diff --git "a/med-Llava3-instruct2/trainer_state.json" "b/med-Llava3-instruct2/trainer_state.json" new file mode 100644--- /dev/null +++ "b/med-Llava3-instruct2/trainer_state.json" @@ -0,0 +1,100191 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 3.0, + "eval_steps": 500, + "global_step": 14307, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00020968756552736424, + "grad_norm": 44.581535249025066, + "learning_rate": 4.6511627906976744e-08, + "loss": 5.2578, + "step": 1 + }, + { + "epoch": 0.0004193751310547285, + "grad_norm": 55.97575030060379, + "learning_rate": 9.302325581395349e-08, + "loss": 5.1115, + "step": 2 + }, + { + "epoch": 0.0006290626965820927, + "grad_norm": 45.6807323797568, + "learning_rate": 1.3953488372093024e-07, + "loss": 5.5506, + "step": 3 + }, + { + "epoch": 0.000838750262109457, + "grad_norm": 44.58814275824773, + "learning_rate": 1.8604651162790698e-07, + "loss": 5.3761, + "step": 4 + }, + { + "epoch": 0.0010484378276368212, + "grad_norm": 45.86331102678746, + "learning_rate": 2.3255813953488374e-07, + "loss": 5.4749, + "step": 5 + }, + { + "epoch": 0.0012581253931641854, + "grad_norm": 50.1396833416372, + "learning_rate": 2.790697674418605e-07, + "loss": 5.4041, + "step": 6 + }, + { + "epoch": 0.0014678129586915495, + "grad_norm": 43.36356048315389, + "learning_rate": 3.2558139534883724e-07, + "loss": 5.3444, + "step": 7 + }, + { + "epoch": 0.001677500524218914, + "grad_norm": 41.237362194380275, + "learning_rate": 3.7209302325581396e-07, + "loss": 5.0789, + "step": 8 + }, + { + "epoch": 0.001887188089746278, + "grad_norm": 44.6268570337259, + "learning_rate": 4.186046511627907e-07, + "loss": 5.225, + "step": 9 + }, + { + "epoch": 0.0020968756552736424, + "grad_norm": 41.93516500698109, + "learning_rate": 4.651162790697675e-07, + "loss": 5.251, + "step": 10 + }, + { + "epoch": 0.0023065632208010066, + "grad_norm": 45.66797322674655, + "learning_rate": 5.116279069767442e-07, + "loss": 5.2581, + "step": 11 + }, + { + "epoch": 0.0025162507863283707, + "grad_norm": 45.92240921350971, + "learning_rate": 5.58139534883721e-07, + "loss": 5.2577, + "step": 12 + }, + { + "epoch": 0.002725938351855735, + "grad_norm": 40.551544063218365, + "learning_rate": 6.046511627906977e-07, + "loss": 5.4346, + "step": 13 + }, + { + "epoch": 0.002935625917383099, + "grad_norm": 38.95477715120408, + "learning_rate": 6.511627906976745e-07, + "loss": 4.9173, + "step": 14 + }, + { + "epoch": 0.003145313482910463, + "grad_norm": 42.505450995173696, + "learning_rate": 6.976744186046513e-07, + "loss": 5.184, + "step": 15 + }, + { + "epoch": 0.003355001048437828, + "grad_norm": 46.532900317909046, + "learning_rate": 7.441860465116279e-07, + "loss": 5.206, + "step": 16 + }, + { + "epoch": 0.003564688613965192, + "grad_norm": 35.2464767755706, + "learning_rate": 7.906976744186047e-07, + "loss": 4.9951, + "step": 17 + }, + { + "epoch": 0.003774376179492556, + "grad_norm": 38.354300316731944, + "learning_rate": 8.372093023255814e-07, + "loss": 5.0014, + "step": 18 + }, + { + "epoch": 0.00398406374501992, + "grad_norm": 39.309026236411896, + "learning_rate": 8.837209302325582e-07, + "loss": 4.9677, + "step": 19 + }, + { + "epoch": 0.004193751310547285, + "grad_norm": 30.962161376619925, + "learning_rate": 9.30232558139535e-07, + "loss": 4.6954, + "step": 20 + }, + { + "epoch": 0.0044034388760746486, + "grad_norm": 31.716183882767574, + "learning_rate": 9.767441860465117e-07, + "loss": 4.9276, + "step": 21 + }, + { + "epoch": 0.004613126441602013, + "grad_norm": 28.029713344025122, + "learning_rate": 1.0232558139534884e-06, + "loss": 4.4202, + "step": 22 + }, + { + "epoch": 0.004822814007129377, + "grad_norm": 29.411298964471545, + "learning_rate": 1.0697674418604653e-06, + "loss": 4.2975, + "step": 23 + }, + { + "epoch": 0.0050325015726567415, + "grad_norm": 31.397717047382393, + "learning_rate": 1.116279069767442e-06, + "loss": 4.7745, + "step": 24 + }, + { + "epoch": 0.005242189138184106, + "grad_norm": 30.17266535106565, + "learning_rate": 1.1627906976744188e-06, + "loss": 4.4112, + "step": 25 + }, + { + "epoch": 0.00545187670371147, + "grad_norm": 28.864912852486626, + "learning_rate": 1.2093023255813954e-06, + "loss": 4.6884, + "step": 26 + }, + { + "epoch": 0.005661564269238834, + "grad_norm": 27.584213780426506, + "learning_rate": 1.2558139534883723e-06, + "loss": 4.287, + "step": 27 + }, + { + "epoch": 0.005871251834766198, + "grad_norm": 32.516567618012104, + "learning_rate": 1.302325581395349e-06, + "loss": 4.7153, + "step": 28 + }, + { + "epoch": 0.006080939400293563, + "grad_norm": 28.518458947751625, + "learning_rate": 1.3488372093023258e-06, + "loss": 4.2179, + "step": 29 + }, + { + "epoch": 0.006290626965820926, + "grad_norm": 32.01356840067489, + "learning_rate": 1.3953488372093025e-06, + "loss": 4.0782, + "step": 30 + }, + { + "epoch": 0.006500314531348291, + "grad_norm": 30.306645519671417, + "learning_rate": 1.4418604651162794e-06, + "loss": 4.2055, + "step": 31 + }, + { + "epoch": 0.006710002096875656, + "grad_norm": 21.555136777198957, + "learning_rate": 1.4883720930232558e-06, + "loss": 3.7999, + "step": 32 + }, + { + "epoch": 0.006919689662403019, + "grad_norm": 16.01301527346332, + "learning_rate": 1.534883720930233e-06, + "loss": 3.8126, + "step": 33 + }, + { + "epoch": 0.007129377227930384, + "grad_norm": 17.677733048222027, + "learning_rate": 1.5813953488372093e-06, + "loss": 3.6181, + "step": 34 + }, + { + "epoch": 0.007339064793457748, + "grad_norm": 22.56973075579131, + "learning_rate": 1.6279069767441862e-06, + "loss": 4.0219, + "step": 35 + }, + { + "epoch": 0.007548752358985112, + "grad_norm": 19.22863542781956, + "learning_rate": 1.6744186046511629e-06, + "loss": 3.684, + "step": 36 + }, + { + "epoch": 0.007758439924512477, + "grad_norm": 41.136305365288855, + "learning_rate": 1.7209302325581397e-06, + "loss": 3.8011, + "step": 37 + }, + { + "epoch": 0.00796812749003984, + "grad_norm": 26.64658586090878, + "learning_rate": 1.7674418604651164e-06, + "loss": 3.7584, + "step": 38 + }, + { + "epoch": 0.008177815055567205, + "grad_norm": 18.393267879776605, + "learning_rate": 1.8139534883720933e-06, + "loss": 3.9381, + "step": 39 + }, + { + "epoch": 0.00838750262109457, + "grad_norm": 14.15050940717148, + "learning_rate": 1.86046511627907e-06, + "loss": 3.3514, + "step": 40 + }, + { + "epoch": 0.008597190186621933, + "grad_norm": 17.425435117299486, + "learning_rate": 1.9069767441860468e-06, + "loss": 3.4709, + "step": 41 + }, + { + "epoch": 0.008806877752149297, + "grad_norm": 20.13210149604207, + "learning_rate": 1.9534883720930235e-06, + "loss": 3.5707, + "step": 42 + }, + { + "epoch": 0.009016565317676662, + "grad_norm": 14.29131526927446, + "learning_rate": 2.0000000000000003e-06, + "loss": 3.4287, + "step": 43 + }, + { + "epoch": 0.009226252883204026, + "grad_norm": 17.350463360046255, + "learning_rate": 2.0465116279069768e-06, + "loss": 3.601, + "step": 44 + }, + { + "epoch": 0.009435940448731391, + "grad_norm": 12.955524208006738, + "learning_rate": 2.0930232558139536e-06, + "loss": 3.6587, + "step": 45 + }, + { + "epoch": 0.009645628014258754, + "grad_norm": 11.725416839586316, + "learning_rate": 2.1395348837209305e-06, + "loss": 3.5439, + "step": 46 + }, + { + "epoch": 0.009855315579786118, + "grad_norm": 12.251306517358946, + "learning_rate": 2.1860465116279074e-06, + "loss": 3.5524, + "step": 47 + }, + { + "epoch": 0.010065003145313483, + "grad_norm": 12.012362016650496, + "learning_rate": 2.232558139534884e-06, + "loss": 3.4082, + "step": 48 + }, + { + "epoch": 0.010274690710840848, + "grad_norm": 13.29171390002502, + "learning_rate": 2.2790697674418607e-06, + "loss": 3.5815, + "step": 49 + }, + { + "epoch": 0.010484378276368212, + "grad_norm": 9.96477777092122, + "learning_rate": 2.3255813953488376e-06, + "loss": 3.3192, + "step": 50 + }, + { + "epoch": 0.010694065841895575, + "grad_norm": 9.602017188577209, + "learning_rate": 2.3720930232558144e-06, + "loss": 3.4421, + "step": 51 + }, + { + "epoch": 0.01090375340742294, + "grad_norm": 10.66111999130771, + "learning_rate": 2.418604651162791e-06, + "loss": 3.5158, + "step": 52 + }, + { + "epoch": 0.011113440972950304, + "grad_norm": 10.848278733863214, + "learning_rate": 2.4651162790697678e-06, + "loss": 3.4538, + "step": 53 + }, + { + "epoch": 0.011323128538477669, + "grad_norm": 10.385904100498394, + "learning_rate": 2.5116279069767446e-06, + "loss": 3.4165, + "step": 54 + }, + { + "epoch": 0.011532816104005033, + "grad_norm": 9.79849245706584, + "learning_rate": 2.558139534883721e-06, + "loss": 3.3661, + "step": 55 + }, + { + "epoch": 0.011742503669532396, + "grad_norm": 10.477143016621993, + "learning_rate": 2.604651162790698e-06, + "loss": 3.5001, + "step": 56 + }, + { + "epoch": 0.01195219123505976, + "grad_norm": 11.313800543376221, + "learning_rate": 2.6511627906976744e-06, + "loss": 3.3892, + "step": 57 + }, + { + "epoch": 0.012161878800587125, + "grad_norm": 10.142488883977713, + "learning_rate": 2.6976744186046517e-06, + "loss": 3.5476, + "step": 58 + }, + { + "epoch": 0.01237156636611449, + "grad_norm": 9.658386187848318, + "learning_rate": 2.744186046511628e-06, + "loss": 3.6256, + "step": 59 + }, + { + "epoch": 0.012581253931641853, + "grad_norm": 9.716817971019752, + "learning_rate": 2.790697674418605e-06, + "loss": 3.1563, + "step": 60 + }, + { + "epoch": 0.012790941497169217, + "grad_norm": 10.080634586272732, + "learning_rate": 2.8372093023255815e-06, + "loss": 3.0472, + "step": 61 + }, + { + "epoch": 0.013000629062696582, + "grad_norm": 8.278465166902485, + "learning_rate": 2.8837209302325587e-06, + "loss": 3.3681, + "step": 62 + }, + { + "epoch": 0.013210316628223947, + "grad_norm": 9.487536575067205, + "learning_rate": 2.930232558139535e-06, + "loss": 3.0451, + "step": 63 + }, + { + "epoch": 0.013420004193751311, + "grad_norm": 8.906592766331341, + "learning_rate": 2.9767441860465116e-06, + "loss": 2.9602, + "step": 64 + }, + { + "epoch": 0.013629691759278674, + "grad_norm": 10.169538227681718, + "learning_rate": 3.0232558139534885e-06, + "loss": 3.1515, + "step": 65 + }, + { + "epoch": 0.013839379324806039, + "grad_norm": 9.69295874549237, + "learning_rate": 3.069767441860466e-06, + "loss": 3.0275, + "step": 66 + }, + { + "epoch": 0.014049066890333403, + "grad_norm": 8.417351385432722, + "learning_rate": 3.1162790697674423e-06, + "loss": 3.1228, + "step": 67 + }, + { + "epoch": 0.014258754455860768, + "grad_norm": 8.359946297125934, + "learning_rate": 3.1627906976744187e-06, + "loss": 3.1764, + "step": 68 + }, + { + "epoch": 0.014468442021388132, + "grad_norm": 8.586224181948893, + "learning_rate": 3.2093023255813956e-06, + "loss": 3.35, + "step": 69 + }, + { + "epoch": 0.014678129586915495, + "grad_norm": 7.775800455437223, + "learning_rate": 3.2558139534883724e-06, + "loss": 3.3618, + "step": 70 + }, + { + "epoch": 0.01488781715244286, + "grad_norm": 8.009191898773206, + "learning_rate": 3.3023255813953493e-06, + "loss": 3.3176, + "step": 71 + }, + { + "epoch": 0.015097504717970224, + "grad_norm": 8.690797306545317, + "learning_rate": 3.3488372093023258e-06, + "loss": 3.4012, + "step": 72 + }, + { + "epoch": 0.015307192283497589, + "grad_norm": 6.759160640370161, + "learning_rate": 3.3953488372093026e-06, + "loss": 3.379, + "step": 73 + }, + { + "epoch": 0.015516879849024954, + "grad_norm": 8.063201572053002, + "learning_rate": 3.4418604651162795e-06, + "loss": 3.1205, + "step": 74 + }, + { + "epoch": 0.015726567414552316, + "grad_norm": 8.382017181883159, + "learning_rate": 3.4883720930232564e-06, + "loss": 3.0159, + "step": 75 + }, + { + "epoch": 0.01593625498007968, + "grad_norm": 7.381364777071937, + "learning_rate": 3.534883720930233e-06, + "loss": 3.0113, + "step": 76 + }, + { + "epoch": 0.016145942545607046, + "grad_norm": 7.181591548601563, + "learning_rate": 3.5813953488372093e-06, + "loss": 3.2055, + "step": 77 + }, + { + "epoch": 0.01635563011113441, + "grad_norm": 8.067824850658068, + "learning_rate": 3.6279069767441866e-06, + "loss": 3.1375, + "step": 78 + }, + { + "epoch": 0.016565317676661775, + "grad_norm": 8.90045917292892, + "learning_rate": 3.674418604651163e-06, + "loss": 3.143, + "step": 79 + }, + { + "epoch": 0.01677500524218914, + "grad_norm": 7.428311648268624, + "learning_rate": 3.72093023255814e-06, + "loss": 3.0114, + "step": 80 + }, + { + "epoch": 0.016984692807716504, + "grad_norm": 8.18283591778806, + "learning_rate": 3.7674418604651163e-06, + "loss": 3.2888, + "step": 81 + }, + { + "epoch": 0.017194380373243865, + "grad_norm": 7.144937554022569, + "learning_rate": 3.8139534883720936e-06, + "loss": 3.0265, + "step": 82 + }, + { + "epoch": 0.01740406793877123, + "grad_norm": 7.809265858472188, + "learning_rate": 3.86046511627907e-06, + "loss": 2.8489, + "step": 83 + }, + { + "epoch": 0.017613755504298594, + "grad_norm": 7.86043319583929, + "learning_rate": 3.906976744186047e-06, + "loss": 3.0097, + "step": 84 + }, + { + "epoch": 0.01782344306982596, + "grad_norm": 8.358068065440838, + "learning_rate": 3.953488372093024e-06, + "loss": 3.1544, + "step": 85 + }, + { + "epoch": 0.018033130635353323, + "grad_norm": 8.32027460266491, + "learning_rate": 4.000000000000001e-06, + "loss": 3.0539, + "step": 86 + }, + { + "epoch": 0.018242818200880688, + "grad_norm": 8.361733270607763, + "learning_rate": 4.0465116279069775e-06, + "loss": 3.0993, + "step": 87 + }, + { + "epoch": 0.018452505766408053, + "grad_norm": 8.244772580707666, + "learning_rate": 4.0930232558139536e-06, + "loss": 2.9546, + "step": 88 + }, + { + "epoch": 0.018662193331935417, + "grad_norm": 7.606218973040087, + "learning_rate": 4.1395348837209304e-06, + "loss": 3.1352, + "step": 89 + }, + { + "epoch": 0.018871880897462782, + "grad_norm": 7.1326650064224895, + "learning_rate": 4.186046511627907e-06, + "loss": 3.0812, + "step": 90 + }, + { + "epoch": 0.019081568462990146, + "grad_norm": 6.9993119864944635, + "learning_rate": 4.232558139534884e-06, + "loss": 3.0158, + "step": 91 + }, + { + "epoch": 0.019291256028517507, + "grad_norm": 6.679637342259109, + "learning_rate": 4.279069767441861e-06, + "loss": 3.0143, + "step": 92 + }, + { + "epoch": 0.019500943594044872, + "grad_norm": 7.0219081176472145, + "learning_rate": 4.325581395348837e-06, + "loss": 3.1109, + "step": 93 + }, + { + "epoch": 0.019710631159572237, + "grad_norm": 7.399432103121332, + "learning_rate": 4.372093023255815e-06, + "loss": 3.1689, + "step": 94 + }, + { + "epoch": 0.0199203187250996, + "grad_norm": 7.67344627391521, + "learning_rate": 4.418604651162791e-06, + "loss": 3.1581, + "step": 95 + }, + { + "epoch": 0.020130006290626966, + "grad_norm": 7.451527879617144, + "learning_rate": 4.465116279069768e-06, + "loss": 2.7311, + "step": 96 + }, + { + "epoch": 0.02033969385615433, + "grad_norm": 8.827861596076998, + "learning_rate": 4.5116279069767445e-06, + "loss": 3.1583, + "step": 97 + }, + { + "epoch": 0.020549381421681695, + "grad_norm": 8.341343814476492, + "learning_rate": 4.558139534883721e-06, + "loss": 3.0021, + "step": 98 + }, + { + "epoch": 0.02075906898720906, + "grad_norm": 7.259040254403374, + "learning_rate": 4.604651162790698e-06, + "loss": 2.7542, + "step": 99 + }, + { + "epoch": 0.020968756552736424, + "grad_norm": 7.265031203493301, + "learning_rate": 4.651162790697675e-06, + "loss": 3.175, + "step": 100 + }, + { + "epoch": 0.021178444118263785, + "grad_norm": 6.62490543346344, + "learning_rate": 4.697674418604651e-06, + "loss": 2.9096, + "step": 101 + }, + { + "epoch": 0.02138813168379115, + "grad_norm": 6.665691734747577, + "learning_rate": 4.744186046511629e-06, + "loss": 3.0775, + "step": 102 + }, + { + "epoch": 0.021597819249318514, + "grad_norm": 7.073363576835186, + "learning_rate": 4.790697674418605e-06, + "loss": 3.0203, + "step": 103 + }, + { + "epoch": 0.02180750681484588, + "grad_norm": 7.337835231721183, + "learning_rate": 4.837209302325582e-06, + "loss": 3.1326, + "step": 104 + }, + { + "epoch": 0.022017194380373244, + "grad_norm": 9.292910347987931, + "learning_rate": 4.883720930232559e-06, + "loss": 2.8739, + "step": 105 + }, + { + "epoch": 0.022226881945900608, + "grad_norm": 6.884255701484647, + "learning_rate": 4.9302325581395355e-06, + "loss": 2.9463, + "step": 106 + }, + { + "epoch": 0.022436569511427973, + "grad_norm": 6.8751041618497695, + "learning_rate": 4.976744186046512e-06, + "loss": 2.9997, + "step": 107 + }, + { + "epoch": 0.022646257076955337, + "grad_norm": 6.371828526115297, + "learning_rate": 5.023255813953489e-06, + "loss": 2.8357, + "step": 108 + }, + { + "epoch": 0.022855944642482702, + "grad_norm": 7.441845262648583, + "learning_rate": 5.069767441860466e-06, + "loss": 3.0136, + "step": 109 + }, + { + "epoch": 0.023065632208010067, + "grad_norm": 6.973505599652898, + "learning_rate": 5.116279069767442e-06, + "loss": 3.0189, + "step": 110 + }, + { + "epoch": 0.023275319773537428, + "grad_norm": 8.031376102956665, + "learning_rate": 5.162790697674419e-06, + "loss": 3.1962, + "step": 111 + }, + { + "epoch": 0.023485007339064792, + "grad_norm": 6.405612779391031, + "learning_rate": 5.209302325581396e-06, + "loss": 2.7497, + "step": 112 + }, + { + "epoch": 0.023694694904592157, + "grad_norm": 7.762453896092673, + "learning_rate": 5.255813953488372e-06, + "loss": 2.836, + "step": 113 + }, + { + "epoch": 0.02390438247011952, + "grad_norm": 6.567603453861468, + "learning_rate": 5.302325581395349e-06, + "loss": 3.0374, + "step": 114 + }, + { + "epoch": 0.024114070035646886, + "grad_norm": 6.363662318269975, + "learning_rate": 5.348837209302326e-06, + "loss": 2.8409, + "step": 115 + }, + { + "epoch": 0.02432375760117425, + "grad_norm": 6.344329633093975, + "learning_rate": 5.395348837209303e-06, + "loss": 2.954, + "step": 116 + }, + { + "epoch": 0.024533445166701615, + "grad_norm": 7.114931878155349, + "learning_rate": 5.44186046511628e-06, + "loss": 2.7483, + "step": 117 + }, + { + "epoch": 0.02474313273222898, + "grad_norm": 6.97592248549356, + "learning_rate": 5.488372093023256e-06, + "loss": 2.8308, + "step": 118 + }, + { + "epoch": 0.024952820297756344, + "grad_norm": 6.749563673827139, + "learning_rate": 5.534883720930233e-06, + "loss": 3.0565, + "step": 119 + }, + { + "epoch": 0.025162507863283706, + "grad_norm": 6.753963206841443, + "learning_rate": 5.58139534883721e-06, + "loss": 2.9406, + "step": 120 + }, + { + "epoch": 0.02537219542881107, + "grad_norm": 7.029681435834868, + "learning_rate": 5.627906976744186e-06, + "loss": 3.0193, + "step": 121 + }, + { + "epoch": 0.025581882994338435, + "grad_norm": 7.302060631425022, + "learning_rate": 5.674418604651163e-06, + "loss": 2.9422, + "step": 122 + }, + { + "epoch": 0.0257915705598658, + "grad_norm": 6.279412677284418, + "learning_rate": 5.72093023255814e-06, + "loss": 3.1161, + "step": 123 + }, + { + "epoch": 0.026001258125393164, + "grad_norm": 6.705253568776916, + "learning_rate": 5.7674418604651175e-06, + "loss": 2.968, + "step": 124 + }, + { + "epoch": 0.02621094569092053, + "grad_norm": 7.307523262657048, + "learning_rate": 5.8139534883720935e-06, + "loss": 2.9159, + "step": 125 + }, + { + "epoch": 0.026420633256447893, + "grad_norm": 7.026743804667868, + "learning_rate": 5.86046511627907e-06, + "loss": 3.0181, + "step": 126 + }, + { + "epoch": 0.026630320821975258, + "grad_norm": 6.2928015872009935, + "learning_rate": 5.906976744186047e-06, + "loss": 2.7893, + "step": 127 + }, + { + "epoch": 0.026840008387502622, + "grad_norm": 6.924752440313978, + "learning_rate": 5.953488372093023e-06, + "loss": 2.8785, + "step": 128 + }, + { + "epoch": 0.027049695953029987, + "grad_norm": 7.591356882506217, + "learning_rate": 6e-06, + "loss": 2.7213, + "step": 129 + }, + { + "epoch": 0.027259383518557348, + "grad_norm": 7.315022847584364, + "learning_rate": 6.046511627906977e-06, + "loss": 2.955, + "step": 130 + }, + { + "epoch": 0.027469071084084713, + "grad_norm": 7.933064282967163, + "learning_rate": 6.093023255813954e-06, + "loss": 3.1728, + "step": 131 + }, + { + "epoch": 0.027678758649612077, + "grad_norm": 6.795893029140771, + "learning_rate": 6.139534883720932e-06, + "loss": 3.042, + "step": 132 + }, + { + "epoch": 0.027888446215139442, + "grad_norm": 7.206132100003922, + "learning_rate": 6.186046511627908e-06, + "loss": 2.9944, + "step": 133 + }, + { + "epoch": 0.028098133780666806, + "grad_norm": 6.8474156142385745, + "learning_rate": 6.2325581395348845e-06, + "loss": 2.8163, + "step": 134 + }, + { + "epoch": 0.02830782134619417, + "grad_norm": 7.976594233426404, + "learning_rate": 6.279069767441861e-06, + "loss": 2.8262, + "step": 135 + }, + { + "epoch": 0.028517508911721536, + "grad_norm": 7.832725676171021, + "learning_rate": 6.325581395348837e-06, + "loss": 2.9799, + "step": 136 + }, + { + "epoch": 0.0287271964772489, + "grad_norm": 7.5292709231393244, + "learning_rate": 6.372093023255814e-06, + "loss": 3.036, + "step": 137 + }, + { + "epoch": 0.028936884042776265, + "grad_norm": 6.940805943484372, + "learning_rate": 6.418604651162791e-06, + "loss": 2.8853, + "step": 138 + }, + { + "epoch": 0.02914657160830363, + "grad_norm": 6.746971856261738, + "learning_rate": 6.465116279069767e-06, + "loss": 3.0046, + "step": 139 + }, + { + "epoch": 0.02935625917383099, + "grad_norm": 6.2941878059733956, + "learning_rate": 6.511627906976745e-06, + "loss": 3.0217, + "step": 140 + }, + { + "epoch": 0.029565946739358355, + "grad_norm": 6.701930124928513, + "learning_rate": 6.558139534883722e-06, + "loss": 3.2442, + "step": 141 + }, + { + "epoch": 0.02977563430488572, + "grad_norm": 7.288812866234423, + "learning_rate": 6.604651162790699e-06, + "loss": 2.7264, + "step": 142 + }, + { + "epoch": 0.029985321870413084, + "grad_norm": 8.024095492047232, + "learning_rate": 6.651162790697675e-06, + "loss": 2.9517, + "step": 143 + }, + { + "epoch": 0.03019500943594045, + "grad_norm": 6.646462356240576, + "learning_rate": 6.6976744186046515e-06, + "loss": 2.8886, + "step": 144 + }, + { + "epoch": 0.030404697001467813, + "grad_norm": 7.4303408597697835, + "learning_rate": 6.744186046511628e-06, + "loss": 2.6785, + "step": 145 + }, + { + "epoch": 0.030614384566995178, + "grad_norm": 7.469936851895788, + "learning_rate": 6.790697674418605e-06, + "loss": 3.0306, + "step": 146 + }, + { + "epoch": 0.030824072132522543, + "grad_norm": 5.665115038975986, + "learning_rate": 6.837209302325581e-06, + "loss": 2.6079, + "step": 147 + }, + { + "epoch": 0.031033759698049907, + "grad_norm": 6.743437262628165, + "learning_rate": 6.883720930232559e-06, + "loss": 2.8679, + "step": 148 + }, + { + "epoch": 0.031243447263577268, + "grad_norm": 6.244858976220253, + "learning_rate": 6.930232558139536e-06, + "loss": 2.7485, + "step": 149 + }, + { + "epoch": 0.03145313482910463, + "grad_norm": 7.46763190974299, + "learning_rate": 6.976744186046513e-06, + "loss": 2.7331, + "step": 150 + }, + { + "epoch": 0.031662822394632, + "grad_norm": 7.4836444766998245, + "learning_rate": 7.023255813953489e-06, + "loss": 2.9845, + "step": 151 + }, + { + "epoch": 0.03187250996015936, + "grad_norm": 7.568525369731155, + "learning_rate": 7.069767441860466e-06, + "loss": 2.8402, + "step": 152 + }, + { + "epoch": 0.03208219752568673, + "grad_norm": 6.391957644079795, + "learning_rate": 7.1162790697674425e-06, + "loss": 2.7253, + "step": 153 + }, + { + "epoch": 0.03229188509121409, + "grad_norm": 6.204025705373054, + "learning_rate": 7.1627906976744185e-06, + "loss": 2.7282, + "step": 154 + }, + { + "epoch": 0.03250157265674145, + "grad_norm": 7.185918901911493, + "learning_rate": 7.209302325581395e-06, + "loss": 3.0344, + "step": 155 + }, + { + "epoch": 0.03271126022226882, + "grad_norm": 6.416435208819019, + "learning_rate": 7.255813953488373e-06, + "loss": 2.8597, + "step": 156 + }, + { + "epoch": 0.03292094778779618, + "grad_norm": 6.524052329984861, + "learning_rate": 7.30232558139535e-06, + "loss": 2.8257, + "step": 157 + }, + { + "epoch": 0.03313063535332355, + "grad_norm": 7.2112682560885775, + "learning_rate": 7.348837209302326e-06, + "loss": 3.0819, + "step": 158 + }, + { + "epoch": 0.03334032291885091, + "grad_norm": 6.061057043931095, + "learning_rate": 7.395348837209303e-06, + "loss": 2.9576, + "step": 159 + }, + { + "epoch": 0.03355001048437828, + "grad_norm": 6.998430546185198, + "learning_rate": 7.44186046511628e-06, + "loss": 2.8175, + "step": 160 + }, + { + "epoch": 0.03375969804990564, + "grad_norm": 7.069055966691737, + "learning_rate": 7.488372093023256e-06, + "loss": 3.0261, + "step": 161 + }, + { + "epoch": 0.03396938561543301, + "grad_norm": 6.627895182582551, + "learning_rate": 7.534883720930233e-06, + "loss": 2.8868, + "step": 162 + }, + { + "epoch": 0.03417907318096037, + "grad_norm": 5.983232879869772, + "learning_rate": 7.5813953488372095e-06, + "loss": 2.7741, + "step": 163 + }, + { + "epoch": 0.03438876074648773, + "grad_norm": 7.561401934252884, + "learning_rate": 7.627906976744187e-06, + "loss": 2.9465, + "step": 164 + }, + { + "epoch": 0.0345984483120151, + "grad_norm": 5.630058872583283, + "learning_rate": 7.674418604651164e-06, + "loss": 2.7352, + "step": 165 + }, + { + "epoch": 0.03480813587754246, + "grad_norm": 5.364526316403802, + "learning_rate": 7.72093023255814e-06, + "loss": 2.838, + "step": 166 + }, + { + "epoch": 0.03501782344306983, + "grad_norm": 5.478025763331473, + "learning_rate": 7.767441860465116e-06, + "loss": 2.7412, + "step": 167 + }, + { + "epoch": 0.03522751100859719, + "grad_norm": 5.845870353207118, + "learning_rate": 7.813953488372094e-06, + "loss": 2.805, + "step": 168 + }, + { + "epoch": 0.03543719857412456, + "grad_norm": 6.813137869162234, + "learning_rate": 7.86046511627907e-06, + "loss": 3.0986, + "step": 169 + }, + { + "epoch": 0.03564688613965192, + "grad_norm": 6.922835858133902, + "learning_rate": 7.906976744186048e-06, + "loss": 2.6966, + "step": 170 + }, + { + "epoch": 0.035856573705179286, + "grad_norm": 6.3040691173508145, + "learning_rate": 7.953488372093024e-06, + "loss": 2.7955, + "step": 171 + }, + { + "epoch": 0.03606626127070665, + "grad_norm": 6.476982121799255, + "learning_rate": 8.000000000000001e-06, + "loss": 2.9843, + "step": 172 + }, + { + "epoch": 0.03627594883623401, + "grad_norm": 6.250873244793059, + "learning_rate": 8.046511627906977e-06, + "loss": 2.6941, + "step": 173 + }, + { + "epoch": 0.036485636401761376, + "grad_norm": 6.142394880645442, + "learning_rate": 8.093023255813955e-06, + "loss": 3.0228, + "step": 174 + }, + { + "epoch": 0.03669532396728874, + "grad_norm": 6.455941279610695, + "learning_rate": 8.139534883720931e-06, + "loss": 3.0294, + "step": 175 + }, + { + "epoch": 0.036905011532816105, + "grad_norm": 7.036015991390055, + "learning_rate": 8.186046511627907e-06, + "loss": 2.9305, + "step": 176 + }, + { + "epoch": 0.037114699098343466, + "grad_norm": 6.930087474195064, + "learning_rate": 8.232558139534885e-06, + "loss": 2.8207, + "step": 177 + }, + { + "epoch": 0.037324386663870834, + "grad_norm": 5.994329531574434, + "learning_rate": 8.279069767441861e-06, + "loss": 2.8105, + "step": 178 + }, + { + "epoch": 0.037534074229398195, + "grad_norm": 6.315983955737537, + "learning_rate": 8.325581395348837e-06, + "loss": 2.7665, + "step": 179 + }, + { + "epoch": 0.037743761794925564, + "grad_norm": 6.264585173042879, + "learning_rate": 8.372093023255815e-06, + "loss": 2.873, + "step": 180 + }, + { + "epoch": 0.037953449360452925, + "grad_norm": 6.263543731620491, + "learning_rate": 8.418604651162792e-06, + "loss": 2.7261, + "step": 181 + }, + { + "epoch": 0.03816313692598029, + "grad_norm": 6.910327208966558, + "learning_rate": 8.465116279069768e-06, + "loss": 2.6568, + "step": 182 + }, + { + "epoch": 0.038372824491507654, + "grad_norm": 7.188497048654426, + "learning_rate": 8.511627906976744e-06, + "loss": 3.0504, + "step": 183 + }, + { + "epoch": 0.038582512057035015, + "grad_norm": 5.911916724608364, + "learning_rate": 8.558139534883722e-06, + "loss": 2.7741, + "step": 184 + }, + { + "epoch": 0.03879219962256238, + "grad_norm": 6.213909048610596, + "learning_rate": 8.604651162790698e-06, + "loss": 2.8767, + "step": 185 + }, + { + "epoch": 0.039001887188089744, + "grad_norm": 6.601058676711457, + "learning_rate": 8.651162790697674e-06, + "loss": 2.8202, + "step": 186 + }, + { + "epoch": 0.03921157475361711, + "grad_norm": 6.579091805768915, + "learning_rate": 8.697674418604652e-06, + "loss": 2.7949, + "step": 187 + }, + { + "epoch": 0.03942126231914447, + "grad_norm": 6.225942052043143, + "learning_rate": 8.74418604651163e-06, + "loss": 2.7729, + "step": 188 + }, + { + "epoch": 0.03963094988467184, + "grad_norm": 6.645340141686921, + "learning_rate": 8.790697674418606e-06, + "loss": 2.6869, + "step": 189 + }, + { + "epoch": 0.0398406374501992, + "grad_norm": 6.342066424715, + "learning_rate": 8.837209302325582e-06, + "loss": 2.7648, + "step": 190 + }, + { + "epoch": 0.04005032501572657, + "grad_norm": 6.942050391424554, + "learning_rate": 8.88372093023256e-06, + "loss": 2.44, + "step": 191 + }, + { + "epoch": 0.04026001258125393, + "grad_norm": 6.879752276057378, + "learning_rate": 8.930232558139535e-06, + "loss": 2.6724, + "step": 192 + }, + { + "epoch": 0.04046970014678129, + "grad_norm": 6.557447416503833, + "learning_rate": 8.976744186046511e-06, + "loss": 2.896, + "step": 193 + }, + { + "epoch": 0.04067938771230866, + "grad_norm": 6.7668012971340055, + "learning_rate": 9.023255813953489e-06, + "loss": 2.8112, + "step": 194 + }, + { + "epoch": 0.04088907527783602, + "grad_norm": 6.149504065706865, + "learning_rate": 9.069767441860465e-06, + "loss": 2.6075, + "step": 195 + }, + { + "epoch": 0.04109876284336339, + "grad_norm": 5.771621281957409, + "learning_rate": 9.116279069767443e-06, + "loss": 2.6253, + "step": 196 + }, + { + "epoch": 0.04130845040889075, + "grad_norm": 6.225253601777948, + "learning_rate": 9.162790697674419e-06, + "loss": 3.0023, + "step": 197 + }, + { + "epoch": 0.04151813797441812, + "grad_norm": 6.672574003306967, + "learning_rate": 9.209302325581397e-06, + "loss": 2.6836, + "step": 198 + }, + { + "epoch": 0.04172782553994548, + "grad_norm": 5.512500309008176, + "learning_rate": 9.255813953488373e-06, + "loss": 2.5928, + "step": 199 + }, + { + "epoch": 0.04193751310547285, + "grad_norm": 5.5206322292409356, + "learning_rate": 9.30232558139535e-06, + "loss": 2.8166, + "step": 200 + }, + { + "epoch": 0.04214720067100021, + "grad_norm": 6.493257736910734, + "learning_rate": 9.348837209302326e-06, + "loss": 2.6397, + "step": 201 + }, + { + "epoch": 0.04235688823652757, + "grad_norm": 5.561370011709336, + "learning_rate": 9.395348837209302e-06, + "loss": 2.7628, + "step": 202 + }, + { + "epoch": 0.04256657580205494, + "grad_norm": 6.4883724309013635, + "learning_rate": 9.44186046511628e-06, + "loss": 2.8979, + "step": 203 + }, + { + "epoch": 0.0427762633675823, + "grad_norm": 6.221200515288901, + "learning_rate": 9.488372093023258e-06, + "loss": 3.0369, + "step": 204 + }, + { + "epoch": 0.04298595093310967, + "grad_norm": 6.039474974421818, + "learning_rate": 9.534883720930234e-06, + "loss": 2.7035, + "step": 205 + }, + { + "epoch": 0.04319563849863703, + "grad_norm": 5.525237525788932, + "learning_rate": 9.58139534883721e-06, + "loss": 2.8386, + "step": 206 + }, + { + "epoch": 0.0434053260641644, + "grad_norm": 5.976020120493838, + "learning_rate": 9.627906976744188e-06, + "loss": 2.8688, + "step": 207 + }, + { + "epoch": 0.04361501362969176, + "grad_norm": 6.3776814779411675, + "learning_rate": 9.674418604651164e-06, + "loss": 2.7952, + "step": 208 + }, + { + "epoch": 0.043824701195219126, + "grad_norm": 6.965685190392773, + "learning_rate": 9.72093023255814e-06, + "loss": 2.7989, + "step": 209 + }, + { + "epoch": 0.04403438876074649, + "grad_norm": 5.971415176834104, + "learning_rate": 9.767441860465117e-06, + "loss": 2.9307, + "step": 210 + }, + { + "epoch": 0.044244076326273855, + "grad_norm": 6.085648978675961, + "learning_rate": 9.813953488372093e-06, + "loss": 2.8971, + "step": 211 + }, + { + "epoch": 0.044453763891801217, + "grad_norm": 6.762147772482043, + "learning_rate": 9.860465116279071e-06, + "loss": 2.7088, + "step": 212 + }, + { + "epoch": 0.04466345145732858, + "grad_norm": 6.088741491621783, + "learning_rate": 9.906976744186047e-06, + "loss": 2.6899, + "step": 213 + }, + { + "epoch": 0.044873139022855946, + "grad_norm": 6.3357003480094285, + "learning_rate": 9.953488372093025e-06, + "loss": 2.9618, + "step": 214 + }, + { + "epoch": 0.04508282658838331, + "grad_norm": 5.800128149585371, + "learning_rate": 1e-05, + "loss": 2.8261, + "step": 215 + }, + { + "epoch": 0.045292514153910675, + "grad_norm": 5.7763607505575445, + "learning_rate": 1.0046511627906979e-05, + "loss": 2.7597, + "step": 216 + }, + { + "epoch": 0.045502201719438036, + "grad_norm": 5.4007140383531285, + "learning_rate": 1.0093023255813955e-05, + "loss": 2.782, + "step": 217 + }, + { + "epoch": 0.045711889284965404, + "grad_norm": 5.5562906602558755, + "learning_rate": 1.0139534883720932e-05, + "loss": 2.7545, + "step": 218 + }, + { + "epoch": 0.045921576850492765, + "grad_norm": 5.160675907752467, + "learning_rate": 1.0186046511627907e-05, + "loss": 2.6036, + "step": 219 + }, + { + "epoch": 0.04613126441602013, + "grad_norm": 6.482894622193402, + "learning_rate": 1.0232558139534884e-05, + "loss": 2.6946, + "step": 220 + }, + { + "epoch": 0.046340951981547494, + "grad_norm": 7.134759447652219, + "learning_rate": 1.027906976744186e-05, + "loss": 2.8082, + "step": 221 + }, + { + "epoch": 0.046550639547074855, + "grad_norm": 6.712574801259102, + "learning_rate": 1.0325581395348838e-05, + "loss": 2.8778, + "step": 222 + }, + { + "epoch": 0.046760327112602224, + "grad_norm": 6.142230410657661, + "learning_rate": 1.0372093023255816e-05, + "loss": 2.8448, + "step": 223 + }, + { + "epoch": 0.046970014678129585, + "grad_norm": 5.792950804593206, + "learning_rate": 1.0418604651162792e-05, + "loss": 2.5658, + "step": 224 + }, + { + "epoch": 0.04717970224365695, + "grad_norm": 6.252937108520555, + "learning_rate": 1.046511627906977e-05, + "loss": 2.7542, + "step": 225 + }, + { + "epoch": 0.047389389809184314, + "grad_norm": 6.903655994729894, + "learning_rate": 1.0511627906976744e-05, + "loss": 2.8082, + "step": 226 + }, + { + "epoch": 0.04759907737471168, + "grad_norm": 5.967108627828422, + "learning_rate": 1.0558139534883722e-05, + "loss": 2.6372, + "step": 227 + }, + { + "epoch": 0.04780876494023904, + "grad_norm": 5.785176436017485, + "learning_rate": 1.0604651162790698e-05, + "loss": 2.7638, + "step": 228 + }, + { + "epoch": 0.04801845250576641, + "grad_norm": 5.911200548984772, + "learning_rate": 1.0651162790697675e-05, + "loss": 2.7544, + "step": 229 + }, + { + "epoch": 0.04822814007129377, + "grad_norm": 6.285402909258177, + "learning_rate": 1.0697674418604651e-05, + "loss": 2.9015, + "step": 230 + }, + { + "epoch": 0.04843782763682113, + "grad_norm": 5.535772285457667, + "learning_rate": 1.0744186046511629e-05, + "loss": 2.8226, + "step": 231 + }, + { + "epoch": 0.0486475152023485, + "grad_norm": 5.891636230427452, + "learning_rate": 1.0790697674418607e-05, + "loss": 2.7513, + "step": 232 + }, + { + "epoch": 0.04885720276787586, + "grad_norm": 5.576360166046296, + "learning_rate": 1.0837209302325583e-05, + "loss": 2.7401, + "step": 233 + }, + { + "epoch": 0.04906689033340323, + "grad_norm": 5.054243993611499, + "learning_rate": 1.088372093023256e-05, + "loss": 2.8342, + "step": 234 + }, + { + "epoch": 0.04927657789893059, + "grad_norm": 7.005846836843509, + "learning_rate": 1.0930232558139535e-05, + "loss": 2.837, + "step": 235 + }, + { + "epoch": 0.04948626546445796, + "grad_norm": 6.15741902291841, + "learning_rate": 1.0976744186046513e-05, + "loss": 2.8706, + "step": 236 + }, + { + "epoch": 0.04969595302998532, + "grad_norm": 5.913666044172258, + "learning_rate": 1.1023255813953489e-05, + "loss": 2.6429, + "step": 237 + }, + { + "epoch": 0.04990564059551269, + "grad_norm": 5.950289705708864, + "learning_rate": 1.1069767441860466e-05, + "loss": 2.8165, + "step": 238 + }, + { + "epoch": 0.05011532816104005, + "grad_norm": 5.274677163222211, + "learning_rate": 1.1116279069767444e-05, + "loss": 2.6037, + "step": 239 + }, + { + "epoch": 0.05032501572656741, + "grad_norm": 5.348982008009826, + "learning_rate": 1.116279069767442e-05, + "loss": 2.469, + "step": 240 + }, + { + "epoch": 0.05053470329209478, + "grad_norm": 5.478702212197505, + "learning_rate": 1.1209302325581398e-05, + "loss": 2.6849, + "step": 241 + }, + { + "epoch": 0.05074439085762214, + "grad_norm": 6.806770523232891, + "learning_rate": 1.1255813953488372e-05, + "loss": 2.9733, + "step": 242 + }, + { + "epoch": 0.05095407842314951, + "grad_norm": 5.274062061765246, + "learning_rate": 1.130232558139535e-05, + "loss": 2.7172, + "step": 243 + }, + { + "epoch": 0.05116376598867687, + "grad_norm": 5.92775882781891, + "learning_rate": 1.1348837209302326e-05, + "loss": 2.686, + "step": 244 + }, + { + "epoch": 0.05137345355420424, + "grad_norm": 5.422156852561135, + "learning_rate": 1.1395348837209304e-05, + "loss": 2.8409, + "step": 245 + }, + { + "epoch": 0.0515831411197316, + "grad_norm": 5.47457712009494, + "learning_rate": 1.144186046511628e-05, + "loss": 2.6795, + "step": 246 + }, + { + "epoch": 0.05179282868525897, + "grad_norm": 6.5079623854230215, + "learning_rate": 1.1488372093023257e-05, + "loss": 2.6468, + "step": 247 + }, + { + "epoch": 0.05200251625078633, + "grad_norm": 6.052979203695229, + "learning_rate": 1.1534883720930235e-05, + "loss": 2.7911, + "step": 248 + }, + { + "epoch": 0.052212203816313696, + "grad_norm": 6.2211508374215905, + "learning_rate": 1.158139534883721e-05, + "loss": 2.646, + "step": 249 + }, + { + "epoch": 0.05242189138184106, + "grad_norm": 6.100754247474486, + "learning_rate": 1.1627906976744187e-05, + "loss": 2.7139, + "step": 250 + }, + { + "epoch": 0.05263157894736842, + "grad_norm": 6.641575016899216, + "learning_rate": 1.1674418604651163e-05, + "loss": 2.8107, + "step": 251 + }, + { + "epoch": 0.052841266512895786, + "grad_norm": 7.103962964668787, + "learning_rate": 1.172093023255814e-05, + "loss": 3.0844, + "step": 252 + }, + { + "epoch": 0.05305095407842315, + "grad_norm": 6.158991274742413, + "learning_rate": 1.1767441860465117e-05, + "loss": 2.6862, + "step": 253 + }, + { + "epoch": 0.053260641643950515, + "grad_norm": 5.493349088138242, + "learning_rate": 1.1813953488372095e-05, + "loss": 2.72, + "step": 254 + }, + { + "epoch": 0.053470329209477876, + "grad_norm": 6.94758241674908, + "learning_rate": 1.1860465116279072e-05, + "loss": 2.9364, + "step": 255 + }, + { + "epoch": 0.053680016775005245, + "grad_norm": 6.477111019252701, + "learning_rate": 1.1906976744186047e-05, + "loss": 2.9216, + "step": 256 + }, + { + "epoch": 0.053889704340532606, + "grad_norm": 6.713537137328682, + "learning_rate": 1.1953488372093024e-05, + "loss": 2.7497, + "step": 257 + }, + { + "epoch": 0.054099391906059974, + "grad_norm": 5.457202553876317, + "learning_rate": 1.2e-05, + "loss": 2.686, + "step": 258 + }, + { + "epoch": 0.054309079471587335, + "grad_norm": 5.928454809699452, + "learning_rate": 1.2046511627906978e-05, + "loss": 2.4659, + "step": 259 + }, + { + "epoch": 0.054518767037114696, + "grad_norm": 6.572174351073993, + "learning_rate": 1.2093023255813954e-05, + "loss": 2.7379, + "step": 260 + }, + { + "epoch": 0.054728454602642064, + "grad_norm": 6.014485618362067, + "learning_rate": 1.2139534883720932e-05, + "loss": 2.6661, + "step": 261 + }, + { + "epoch": 0.054938142168169425, + "grad_norm": 6.828827666462322, + "learning_rate": 1.2186046511627908e-05, + "loss": 2.7704, + "step": 262 + }, + { + "epoch": 0.05514782973369679, + "grad_norm": 5.5016885224260825, + "learning_rate": 1.2232558139534886e-05, + "loss": 2.7052, + "step": 263 + }, + { + "epoch": 0.055357517299224154, + "grad_norm": 5.856462966659524, + "learning_rate": 1.2279069767441863e-05, + "loss": 3.0589, + "step": 264 + }, + { + "epoch": 0.05556720486475152, + "grad_norm": 5.913114702326443, + "learning_rate": 1.2325581395348838e-05, + "loss": 2.6912, + "step": 265 + }, + { + "epoch": 0.055776892430278883, + "grad_norm": 5.879154738611566, + "learning_rate": 1.2372093023255815e-05, + "loss": 2.7632, + "step": 266 + }, + { + "epoch": 0.05598657999580625, + "grad_norm": 5.980828430907205, + "learning_rate": 1.2418604651162791e-05, + "loss": 2.7148, + "step": 267 + }, + { + "epoch": 0.05619626756133361, + "grad_norm": 5.886866577946389, + "learning_rate": 1.2465116279069769e-05, + "loss": 2.7365, + "step": 268 + }, + { + "epoch": 0.056405955126860974, + "grad_norm": 5.930187686501461, + "learning_rate": 1.2511627906976745e-05, + "loss": 2.709, + "step": 269 + }, + { + "epoch": 0.05661564269238834, + "grad_norm": 6.380802844368986, + "learning_rate": 1.2558139534883723e-05, + "loss": 2.7103, + "step": 270 + }, + { + "epoch": 0.0568253302579157, + "grad_norm": 5.708232470252725, + "learning_rate": 1.26046511627907e-05, + "loss": 2.6117, + "step": 271 + }, + { + "epoch": 0.05703501782344307, + "grad_norm": 7.25828101588181, + "learning_rate": 1.2651162790697675e-05, + "loss": 2.662, + "step": 272 + }, + { + "epoch": 0.05724470538897043, + "grad_norm": 6.696050405178439, + "learning_rate": 1.2697674418604653e-05, + "loss": 2.7142, + "step": 273 + }, + { + "epoch": 0.0574543929544978, + "grad_norm": 5.94183434264606, + "learning_rate": 1.2744186046511629e-05, + "loss": 2.629, + "step": 274 + }, + { + "epoch": 0.05766408052002516, + "grad_norm": 6.277838004747979, + "learning_rate": 1.2790697674418606e-05, + "loss": 2.6538, + "step": 275 + }, + { + "epoch": 0.05787376808555253, + "grad_norm": 6.172170849994645, + "learning_rate": 1.2837209302325582e-05, + "loss": 2.6504, + "step": 276 + }, + { + "epoch": 0.05808345565107989, + "grad_norm": 5.709414960334309, + "learning_rate": 1.288372093023256e-05, + "loss": 2.7447, + "step": 277 + }, + { + "epoch": 0.05829314321660726, + "grad_norm": 5.920686961359142, + "learning_rate": 1.2930232558139534e-05, + "loss": 2.877, + "step": 278 + }, + { + "epoch": 0.05850283078213462, + "grad_norm": 6.123953867392272, + "learning_rate": 1.2976744186046512e-05, + "loss": 2.9756, + "step": 279 + }, + { + "epoch": 0.05871251834766198, + "grad_norm": 7.748672546393689, + "learning_rate": 1.302325581395349e-05, + "loss": 3.1345, + "step": 280 + }, + { + "epoch": 0.05892220591318935, + "grad_norm": 5.1763281843620765, + "learning_rate": 1.3069767441860466e-05, + "loss": 2.552, + "step": 281 + }, + { + "epoch": 0.05913189347871671, + "grad_norm": 5.420225289246078, + "learning_rate": 1.3116279069767443e-05, + "loss": 2.6374, + "step": 282 + }, + { + "epoch": 0.05934158104424408, + "grad_norm": 5.837474880142761, + "learning_rate": 1.316279069767442e-05, + "loss": 2.5835, + "step": 283 + }, + { + "epoch": 0.05955126860977144, + "grad_norm": 6.150600393884646, + "learning_rate": 1.3209302325581397e-05, + "loss": 2.902, + "step": 284 + }, + { + "epoch": 0.05976095617529881, + "grad_norm": 7.2554870088917855, + "learning_rate": 1.3255813953488372e-05, + "loss": 2.7621, + "step": 285 + }, + { + "epoch": 0.05997064374082617, + "grad_norm": 6.463778666206971, + "learning_rate": 1.330232558139535e-05, + "loss": 2.874, + "step": 286 + }, + { + "epoch": 0.060180331306353536, + "grad_norm": 6.452281179169747, + "learning_rate": 1.3348837209302327e-05, + "loss": 3.0631, + "step": 287 + }, + { + "epoch": 0.0603900188718809, + "grad_norm": 5.779936757298579, + "learning_rate": 1.3395348837209303e-05, + "loss": 2.7905, + "step": 288 + }, + { + "epoch": 0.06059970643740826, + "grad_norm": 6.306669741091084, + "learning_rate": 1.344186046511628e-05, + "loss": 2.8452, + "step": 289 + }, + { + "epoch": 0.06080939400293563, + "grad_norm": 5.906054386458197, + "learning_rate": 1.3488372093023257e-05, + "loss": 2.8006, + "step": 290 + }, + { + "epoch": 0.06101908156846299, + "grad_norm": 5.533973914178475, + "learning_rate": 1.3534883720930234e-05, + "loss": 2.7965, + "step": 291 + }, + { + "epoch": 0.061228769133990356, + "grad_norm": 5.35877542558296, + "learning_rate": 1.358139534883721e-05, + "loss": 2.7042, + "step": 292 + }, + { + "epoch": 0.06143845669951772, + "grad_norm": 5.4596094718721195, + "learning_rate": 1.3627906976744188e-05, + "loss": 2.4664, + "step": 293 + }, + { + "epoch": 0.061648144265045085, + "grad_norm": 5.53754396482365, + "learning_rate": 1.3674418604651163e-05, + "loss": 2.6338, + "step": 294 + }, + { + "epoch": 0.061857831830572446, + "grad_norm": 6.198278716409282, + "learning_rate": 1.372093023255814e-05, + "loss": 2.698, + "step": 295 + }, + { + "epoch": 0.062067519396099814, + "grad_norm": 6.463561706972344, + "learning_rate": 1.3767441860465118e-05, + "loss": 2.6886, + "step": 296 + }, + { + "epoch": 0.062277206961627175, + "grad_norm": 5.259633563530183, + "learning_rate": 1.3813953488372094e-05, + "loss": 2.6933, + "step": 297 + }, + { + "epoch": 0.062486894527154536, + "grad_norm": 5.204962451071752, + "learning_rate": 1.3860465116279072e-05, + "loss": 2.9227, + "step": 298 + }, + { + "epoch": 0.0626965820926819, + "grad_norm": 5.43931196567226, + "learning_rate": 1.3906976744186048e-05, + "loss": 2.573, + "step": 299 + }, + { + "epoch": 0.06290626965820927, + "grad_norm": 5.726436537818361, + "learning_rate": 1.3953488372093025e-05, + "loss": 2.6359, + "step": 300 + }, + { + "epoch": 0.06311595722373663, + "grad_norm": 5.266601831643755, + "learning_rate": 1.4e-05, + "loss": 2.5716, + "step": 301 + }, + { + "epoch": 0.063325644789264, + "grad_norm": 5.869268627325999, + "learning_rate": 1.4046511627906978e-05, + "loss": 2.364, + "step": 302 + }, + { + "epoch": 0.06353533235479136, + "grad_norm": 6.226587082492448, + "learning_rate": 1.4093023255813955e-05, + "loss": 2.6806, + "step": 303 + }, + { + "epoch": 0.06374501992031872, + "grad_norm": 6.030103234425802, + "learning_rate": 1.4139534883720931e-05, + "loss": 2.7833, + "step": 304 + }, + { + "epoch": 0.06395470748584609, + "grad_norm": 6.478761520238879, + "learning_rate": 1.4186046511627909e-05, + "loss": 2.4918, + "step": 305 + }, + { + "epoch": 0.06416439505137346, + "grad_norm": 5.20215793431158, + "learning_rate": 1.4232558139534885e-05, + "loss": 2.7573, + "step": 306 + }, + { + "epoch": 0.06437408261690082, + "grad_norm": 5.7622062877702716, + "learning_rate": 1.4279069767441863e-05, + "loss": 2.5148, + "step": 307 + }, + { + "epoch": 0.06458377018242818, + "grad_norm": 5.956530426109749, + "learning_rate": 1.4325581395348837e-05, + "loss": 2.6827, + "step": 308 + }, + { + "epoch": 0.06479345774795554, + "grad_norm": 6.059123409614743, + "learning_rate": 1.4372093023255815e-05, + "loss": 2.6931, + "step": 309 + }, + { + "epoch": 0.0650031453134829, + "grad_norm": 6.2455309061981685, + "learning_rate": 1.441860465116279e-05, + "loss": 3.0446, + "step": 310 + }, + { + "epoch": 0.06521283287901028, + "grad_norm": 6.613281937512961, + "learning_rate": 1.4465116279069768e-05, + "loss": 2.9515, + "step": 311 + }, + { + "epoch": 0.06542252044453764, + "grad_norm": 5.412172509867639, + "learning_rate": 1.4511627906976746e-05, + "loss": 2.6634, + "step": 312 + }, + { + "epoch": 0.065632208010065, + "grad_norm": 6.567250187892535, + "learning_rate": 1.4558139534883722e-05, + "loss": 3.0405, + "step": 313 + }, + { + "epoch": 0.06584189557559236, + "grad_norm": 5.84932492317624, + "learning_rate": 1.46046511627907e-05, + "loss": 2.7963, + "step": 314 + }, + { + "epoch": 0.06605158314111974, + "grad_norm": 6.602110069465178, + "learning_rate": 1.4651162790697674e-05, + "loss": 2.7062, + "step": 315 + }, + { + "epoch": 0.0662612707066471, + "grad_norm": 5.498020477299307, + "learning_rate": 1.4697674418604652e-05, + "loss": 2.6295, + "step": 316 + }, + { + "epoch": 0.06647095827217446, + "grad_norm": 6.61416538162114, + "learning_rate": 1.4744186046511628e-05, + "loss": 2.7408, + "step": 317 + }, + { + "epoch": 0.06668064583770182, + "grad_norm": 5.826827768593727, + "learning_rate": 1.4790697674418606e-05, + "loss": 2.4461, + "step": 318 + }, + { + "epoch": 0.06689033340322918, + "grad_norm": 6.660216321981914, + "learning_rate": 1.4837209302325583e-05, + "loss": 2.7952, + "step": 319 + }, + { + "epoch": 0.06710002096875656, + "grad_norm": 4.981011650003628, + "learning_rate": 1.488372093023256e-05, + "loss": 2.8064, + "step": 320 + }, + { + "epoch": 0.06730970853428392, + "grad_norm": 5.831091516525705, + "learning_rate": 1.4930232558139537e-05, + "loss": 2.5788, + "step": 321 + }, + { + "epoch": 0.06751939609981128, + "grad_norm": 6.194265249427232, + "learning_rate": 1.4976744186046512e-05, + "loss": 2.7381, + "step": 322 + }, + { + "epoch": 0.06772908366533864, + "grad_norm": 6.170332847848081, + "learning_rate": 1.5023255813953491e-05, + "loss": 2.599, + "step": 323 + }, + { + "epoch": 0.06793877123086602, + "grad_norm": 6.797964060613839, + "learning_rate": 1.5069767441860465e-05, + "loss": 2.7531, + "step": 324 + }, + { + "epoch": 0.06814845879639338, + "grad_norm": 5.443594485202397, + "learning_rate": 1.5116279069767443e-05, + "loss": 2.7429, + "step": 325 + }, + { + "epoch": 0.06835814636192074, + "grad_norm": 6.024837507150557, + "learning_rate": 1.5162790697674419e-05, + "loss": 2.8143, + "step": 326 + }, + { + "epoch": 0.0685678339274481, + "grad_norm": 5.997588507355551, + "learning_rate": 1.5209302325581397e-05, + "loss": 2.8557, + "step": 327 + }, + { + "epoch": 0.06877752149297546, + "grad_norm": 6.172351164496978, + "learning_rate": 1.5255813953488374e-05, + "loss": 2.5614, + "step": 328 + }, + { + "epoch": 0.06898720905850284, + "grad_norm": 5.419519907161505, + "learning_rate": 1.530232558139535e-05, + "loss": 2.8489, + "step": 329 + }, + { + "epoch": 0.0691968966240302, + "grad_norm": 5.278909144433219, + "learning_rate": 1.5348837209302328e-05, + "loss": 2.4734, + "step": 330 + }, + { + "epoch": 0.06940658418955756, + "grad_norm": 6.484816379060119, + "learning_rate": 1.5395348837209303e-05, + "loss": 2.5781, + "step": 331 + }, + { + "epoch": 0.06961627175508492, + "grad_norm": 6.057262863221625, + "learning_rate": 1.544186046511628e-05, + "loss": 2.7545, + "step": 332 + }, + { + "epoch": 0.0698259593206123, + "grad_norm": 5.799531411057827, + "learning_rate": 1.5488372093023255e-05, + "loss": 2.8803, + "step": 333 + }, + { + "epoch": 0.07003564688613965, + "grad_norm": 7.044371563199469, + "learning_rate": 1.5534883720930232e-05, + "loss": 2.6231, + "step": 334 + }, + { + "epoch": 0.07024533445166702, + "grad_norm": 5.974593860547467, + "learning_rate": 1.558139534883721e-05, + "loss": 2.856, + "step": 335 + }, + { + "epoch": 0.07045502201719438, + "grad_norm": 5.407087813680099, + "learning_rate": 1.5627906976744188e-05, + "loss": 2.8117, + "step": 336 + }, + { + "epoch": 0.07066470958272174, + "grad_norm": 5.744303926677864, + "learning_rate": 1.5674418604651165e-05, + "loss": 2.8545, + "step": 337 + }, + { + "epoch": 0.07087439714824911, + "grad_norm": 5.228189399069513, + "learning_rate": 1.572093023255814e-05, + "loss": 2.4886, + "step": 338 + }, + { + "epoch": 0.07108408471377647, + "grad_norm": 6.035957800156823, + "learning_rate": 1.5767441860465117e-05, + "loss": 2.505, + "step": 339 + }, + { + "epoch": 0.07129377227930384, + "grad_norm": 5.719013219290617, + "learning_rate": 1.5813953488372095e-05, + "loss": 2.7319, + "step": 340 + }, + { + "epoch": 0.0715034598448312, + "grad_norm": 6.062696765411479, + "learning_rate": 1.5860465116279073e-05, + "loss": 2.6984, + "step": 341 + }, + { + "epoch": 0.07171314741035857, + "grad_norm": 6.249201740756078, + "learning_rate": 1.5906976744186047e-05, + "loss": 2.658, + "step": 342 + }, + { + "epoch": 0.07192283497588593, + "grad_norm": 5.834905026219851, + "learning_rate": 1.5953488372093025e-05, + "loss": 2.6295, + "step": 343 + }, + { + "epoch": 0.0721325225414133, + "grad_norm": 5.533371756152266, + "learning_rate": 1.6000000000000003e-05, + "loss": 2.7681, + "step": 344 + }, + { + "epoch": 0.07234221010694065, + "grad_norm": 5.690360174250401, + "learning_rate": 1.6046511627906977e-05, + "loss": 2.4966, + "step": 345 + }, + { + "epoch": 0.07255189767246802, + "grad_norm": 5.753144723036307, + "learning_rate": 1.6093023255813955e-05, + "loss": 2.7304, + "step": 346 + }, + { + "epoch": 0.07276158523799539, + "grad_norm": 5.83958583185687, + "learning_rate": 1.6139534883720932e-05, + "loss": 2.6349, + "step": 347 + }, + { + "epoch": 0.07297127280352275, + "grad_norm": 6.4398852781250016, + "learning_rate": 1.618604651162791e-05, + "loss": 2.6976, + "step": 348 + }, + { + "epoch": 0.07318096036905011, + "grad_norm": 5.735850012586663, + "learning_rate": 1.6232558139534884e-05, + "loss": 2.8161, + "step": 349 + }, + { + "epoch": 0.07339064793457747, + "grad_norm": 5.3063108217473935, + "learning_rate": 1.6279069767441862e-05, + "loss": 2.5386, + "step": 350 + }, + { + "epoch": 0.07360033550010485, + "grad_norm": 5.475051540165912, + "learning_rate": 1.632558139534884e-05, + "loss": 2.7839, + "step": 351 + }, + { + "epoch": 0.07381002306563221, + "grad_norm": 5.191799752934426, + "learning_rate": 1.6372093023255814e-05, + "loss": 2.7398, + "step": 352 + }, + { + "epoch": 0.07401971063115957, + "grad_norm": 6.5219365945949885, + "learning_rate": 1.6418604651162792e-05, + "loss": 2.6315, + "step": 353 + }, + { + "epoch": 0.07422939819668693, + "grad_norm": 6.511208446166503, + "learning_rate": 1.646511627906977e-05, + "loss": 2.6904, + "step": 354 + }, + { + "epoch": 0.07443908576221431, + "grad_norm": 5.586213650329757, + "learning_rate": 1.6511627906976747e-05, + "loss": 2.72, + "step": 355 + }, + { + "epoch": 0.07464877332774167, + "grad_norm": 5.77605073598907, + "learning_rate": 1.6558139534883722e-05, + "loss": 2.3914, + "step": 356 + }, + { + "epoch": 0.07485846089326903, + "grad_norm": 5.3201906015908875, + "learning_rate": 1.66046511627907e-05, + "loss": 2.2321, + "step": 357 + }, + { + "epoch": 0.07506814845879639, + "grad_norm": 5.638380031214597, + "learning_rate": 1.6651162790697674e-05, + "loss": 2.649, + "step": 358 + }, + { + "epoch": 0.07527783602432375, + "grad_norm": 6.692083446003021, + "learning_rate": 1.669767441860465e-05, + "loss": 2.9075, + "step": 359 + }, + { + "epoch": 0.07548752358985113, + "grad_norm": 6.270560780931083, + "learning_rate": 1.674418604651163e-05, + "loss": 2.6546, + "step": 360 + }, + { + "epoch": 0.07569721115537849, + "grad_norm": 6.9541218281501385, + "learning_rate": 1.6790697674418607e-05, + "loss": 3.0673, + "step": 361 + }, + { + "epoch": 0.07590689872090585, + "grad_norm": 5.284210590793171, + "learning_rate": 1.6837209302325585e-05, + "loss": 2.6862, + "step": 362 + }, + { + "epoch": 0.07611658628643321, + "grad_norm": 5.858099499474185, + "learning_rate": 1.688372093023256e-05, + "loss": 2.9625, + "step": 363 + }, + { + "epoch": 0.07632627385196059, + "grad_norm": 6.269620854637626, + "learning_rate": 1.6930232558139537e-05, + "loss": 2.6235, + "step": 364 + }, + { + "epoch": 0.07653596141748795, + "grad_norm": 5.629066276393372, + "learning_rate": 1.697674418604651e-05, + "loss": 2.7425, + "step": 365 + }, + { + "epoch": 0.07674564898301531, + "grad_norm": 6.506410310740582, + "learning_rate": 1.702325581395349e-05, + "loss": 2.5681, + "step": 366 + }, + { + "epoch": 0.07695533654854267, + "grad_norm": 5.609809614006658, + "learning_rate": 1.7069767441860466e-05, + "loss": 2.6783, + "step": 367 + }, + { + "epoch": 0.07716502411407003, + "grad_norm": 6.013526338635598, + "learning_rate": 1.7116279069767444e-05, + "loss": 2.3599, + "step": 368 + }, + { + "epoch": 0.0773747116795974, + "grad_norm": 6.935955960212009, + "learning_rate": 1.7162790697674422e-05, + "loss": 2.9178, + "step": 369 + }, + { + "epoch": 0.07758439924512477, + "grad_norm": 6.187788175726091, + "learning_rate": 1.7209302325581396e-05, + "loss": 2.7254, + "step": 370 + }, + { + "epoch": 0.07779408681065213, + "grad_norm": 6.328826494689137, + "learning_rate": 1.7255813953488374e-05, + "loss": 2.5233, + "step": 371 + }, + { + "epoch": 0.07800377437617949, + "grad_norm": 5.518535549260242, + "learning_rate": 1.7302325581395348e-05, + "loss": 2.7285, + "step": 372 + }, + { + "epoch": 0.07821346194170686, + "grad_norm": 5.773188387861025, + "learning_rate": 1.7348837209302326e-05, + "loss": 2.6425, + "step": 373 + }, + { + "epoch": 0.07842314950723422, + "grad_norm": 5.909261393230419, + "learning_rate": 1.7395348837209304e-05, + "loss": 2.6059, + "step": 374 + }, + { + "epoch": 0.07863283707276159, + "grad_norm": 5.5539565046327795, + "learning_rate": 1.744186046511628e-05, + "loss": 2.6028, + "step": 375 + }, + { + "epoch": 0.07884252463828895, + "grad_norm": 6.503018069121384, + "learning_rate": 1.748837209302326e-05, + "loss": 2.9041, + "step": 376 + }, + { + "epoch": 0.07905221220381631, + "grad_norm": 5.25224348650878, + "learning_rate": 1.7534883720930233e-05, + "loss": 2.7723, + "step": 377 + }, + { + "epoch": 0.07926189976934368, + "grad_norm": 6.589235208720605, + "learning_rate": 1.758139534883721e-05, + "loss": 2.8494, + "step": 378 + }, + { + "epoch": 0.07947158733487104, + "grad_norm": 6.6664500399532365, + "learning_rate": 1.7627906976744185e-05, + "loss": 2.8985, + "step": 379 + }, + { + "epoch": 0.0796812749003984, + "grad_norm": 6.6145968018730095, + "learning_rate": 1.7674418604651163e-05, + "loss": 2.7433, + "step": 380 + }, + { + "epoch": 0.07989096246592577, + "grad_norm": 5.9400843764979845, + "learning_rate": 1.772093023255814e-05, + "loss": 2.6454, + "step": 381 + }, + { + "epoch": 0.08010065003145314, + "grad_norm": 5.345420643176109, + "learning_rate": 1.776744186046512e-05, + "loss": 2.6805, + "step": 382 + }, + { + "epoch": 0.0803103375969805, + "grad_norm": 5.434026829259804, + "learning_rate": 1.7813953488372096e-05, + "loss": 2.8746, + "step": 383 + }, + { + "epoch": 0.08052002516250786, + "grad_norm": 6.511662481541995, + "learning_rate": 1.786046511627907e-05, + "loss": 2.7112, + "step": 384 + }, + { + "epoch": 0.08072971272803522, + "grad_norm": 5.102070572258234, + "learning_rate": 1.790697674418605e-05, + "loss": 2.6143, + "step": 385 + }, + { + "epoch": 0.08093940029356259, + "grad_norm": 5.604191365388476, + "learning_rate": 1.7953488372093023e-05, + "loss": 2.6727, + "step": 386 + }, + { + "epoch": 0.08114908785908996, + "grad_norm": 5.459784455774207, + "learning_rate": 1.8e-05, + "loss": 2.4293, + "step": 387 + }, + { + "epoch": 0.08135877542461732, + "grad_norm": 5.3878168111739875, + "learning_rate": 1.8046511627906978e-05, + "loss": 2.6991, + "step": 388 + }, + { + "epoch": 0.08156846299014468, + "grad_norm": 4.948042673945027, + "learning_rate": 1.8093023255813956e-05, + "loss": 2.725, + "step": 389 + }, + { + "epoch": 0.08177815055567204, + "grad_norm": 6.510208918388066, + "learning_rate": 1.813953488372093e-05, + "loss": 2.6172, + "step": 390 + }, + { + "epoch": 0.08198783812119942, + "grad_norm": 6.386512430603227, + "learning_rate": 1.8186046511627908e-05, + "loss": 2.8203, + "step": 391 + }, + { + "epoch": 0.08219752568672678, + "grad_norm": 4.695139056701934, + "learning_rate": 1.8232558139534886e-05, + "loss": 2.6786, + "step": 392 + }, + { + "epoch": 0.08240721325225414, + "grad_norm": 4.6082432474810195, + "learning_rate": 1.827906976744186e-05, + "loss": 2.9227, + "step": 393 + }, + { + "epoch": 0.0826169008177815, + "grad_norm": 6.1298162665772855, + "learning_rate": 1.8325581395348838e-05, + "loss": 2.8086, + "step": 394 + }, + { + "epoch": 0.08282658838330886, + "grad_norm": 5.234715116361226, + "learning_rate": 1.8372093023255815e-05, + "loss": 2.8152, + "step": 395 + }, + { + "epoch": 0.08303627594883624, + "grad_norm": 5.902818586391133, + "learning_rate": 1.8418604651162793e-05, + "loss": 2.7639, + "step": 396 + }, + { + "epoch": 0.0832459635143636, + "grad_norm": 5.927165057197696, + "learning_rate": 1.8465116279069767e-05, + "loss": 2.6674, + "step": 397 + }, + { + "epoch": 0.08345565107989096, + "grad_norm": 6.251341397312741, + "learning_rate": 1.8511627906976745e-05, + "loss": 2.8503, + "step": 398 + }, + { + "epoch": 0.08366533864541832, + "grad_norm": 5.6091230543080215, + "learning_rate": 1.8558139534883723e-05, + "loss": 2.7752, + "step": 399 + }, + { + "epoch": 0.0838750262109457, + "grad_norm": 4.917752911305816, + "learning_rate": 1.86046511627907e-05, + "loss": 2.7482, + "step": 400 + }, + { + "epoch": 0.08408471377647306, + "grad_norm": 5.125438800059427, + "learning_rate": 1.865116279069768e-05, + "loss": 2.7315, + "step": 401 + }, + { + "epoch": 0.08429440134200042, + "grad_norm": 5.286615016197741, + "learning_rate": 1.8697674418604653e-05, + "loss": 2.6158, + "step": 402 + }, + { + "epoch": 0.08450408890752778, + "grad_norm": 5.304510797463541, + "learning_rate": 1.874418604651163e-05, + "loss": 2.5846, + "step": 403 + }, + { + "epoch": 0.08471377647305514, + "grad_norm": 4.792311878902139, + "learning_rate": 1.8790697674418605e-05, + "loss": 2.7005, + "step": 404 + }, + { + "epoch": 0.08492346403858252, + "grad_norm": 5.617132431670244, + "learning_rate": 1.8837209302325582e-05, + "loss": 2.9455, + "step": 405 + }, + { + "epoch": 0.08513315160410988, + "grad_norm": 5.944688564217573, + "learning_rate": 1.888372093023256e-05, + "loss": 2.5341, + "step": 406 + }, + { + "epoch": 0.08534283916963724, + "grad_norm": 5.457418856033933, + "learning_rate": 1.8930232558139538e-05, + "loss": 2.707, + "step": 407 + }, + { + "epoch": 0.0855525267351646, + "grad_norm": 7.0496032230026735, + "learning_rate": 1.8976744186046516e-05, + "loss": 2.9835, + "step": 408 + }, + { + "epoch": 0.08576221430069197, + "grad_norm": 5.520813416086713, + "learning_rate": 1.902325581395349e-05, + "loss": 2.5473, + "step": 409 + }, + { + "epoch": 0.08597190186621934, + "grad_norm": 4.87398199825445, + "learning_rate": 1.9069767441860468e-05, + "loss": 2.4163, + "step": 410 + }, + { + "epoch": 0.0861815894317467, + "grad_norm": 6.251413406652676, + "learning_rate": 1.9116279069767442e-05, + "loss": 2.6383, + "step": 411 + }, + { + "epoch": 0.08639127699727406, + "grad_norm": 5.2699362722532, + "learning_rate": 1.916279069767442e-05, + "loss": 2.4343, + "step": 412 + }, + { + "epoch": 0.08660096456280142, + "grad_norm": 5.107110416988616, + "learning_rate": 1.9209302325581397e-05, + "loss": 2.5345, + "step": 413 + }, + { + "epoch": 0.0868106521283288, + "grad_norm": 6.6423414665446705, + "learning_rate": 1.9255813953488375e-05, + "loss": 2.9373, + "step": 414 + }, + { + "epoch": 0.08702033969385616, + "grad_norm": 5.349537171025931, + "learning_rate": 1.9302325581395353e-05, + "loss": 2.7612, + "step": 415 + }, + { + "epoch": 0.08723002725938352, + "grad_norm": 6.576590652857589, + "learning_rate": 1.9348837209302327e-05, + "loss": 2.7716, + "step": 416 + }, + { + "epoch": 0.08743971482491088, + "grad_norm": 6.006553360145802, + "learning_rate": 1.9395348837209305e-05, + "loss": 2.8141, + "step": 417 + }, + { + "epoch": 0.08764940239043825, + "grad_norm": 5.556741814789794, + "learning_rate": 1.944186046511628e-05, + "loss": 2.6062, + "step": 418 + }, + { + "epoch": 0.08785908995596561, + "grad_norm": 5.407688232080843, + "learning_rate": 1.9488372093023257e-05, + "loss": 2.6978, + "step": 419 + }, + { + "epoch": 0.08806877752149297, + "grad_norm": 5.456501422868993, + "learning_rate": 1.9534883720930235e-05, + "loss": 2.5827, + "step": 420 + }, + { + "epoch": 0.08827846508702034, + "grad_norm": 6.386595479843006, + "learning_rate": 1.9581395348837212e-05, + "loss": 2.8944, + "step": 421 + }, + { + "epoch": 0.08848815265254771, + "grad_norm": 6.692831232313039, + "learning_rate": 1.9627906976744187e-05, + "loss": 2.5248, + "step": 422 + }, + { + "epoch": 0.08869784021807507, + "grad_norm": 5.341320791247144, + "learning_rate": 1.9674418604651164e-05, + "loss": 2.3906, + "step": 423 + }, + { + "epoch": 0.08890752778360243, + "grad_norm": 5.70436415960133, + "learning_rate": 1.9720930232558142e-05, + "loss": 2.575, + "step": 424 + }, + { + "epoch": 0.0891172153491298, + "grad_norm": 5.0890164850281625, + "learning_rate": 1.9767441860465116e-05, + "loss": 2.7715, + "step": 425 + }, + { + "epoch": 0.08932690291465716, + "grad_norm": 5.721675549050049, + "learning_rate": 1.9813953488372094e-05, + "loss": 2.845, + "step": 426 + }, + { + "epoch": 0.08953659048018453, + "grad_norm": 5.857880106965985, + "learning_rate": 1.9860465116279072e-05, + "loss": 2.6757, + "step": 427 + }, + { + "epoch": 0.08974627804571189, + "grad_norm": 5.273886980398418, + "learning_rate": 1.990697674418605e-05, + "loss": 2.705, + "step": 428 + }, + { + "epoch": 0.08995596561123925, + "grad_norm": 5.414827969068109, + "learning_rate": 1.9953488372093024e-05, + "loss": 2.5024, + "step": 429 + }, + { + "epoch": 0.09016565317676661, + "grad_norm": 6.0070052114533325, + "learning_rate": 2e-05, + "loss": 2.53, + "step": 430 + }, + { + "epoch": 0.09037534074229399, + "grad_norm": 5.960425903800965, + "learning_rate": 1.999999974374133e-05, + "loss": 2.6704, + "step": 431 + }, + { + "epoch": 0.09058502830782135, + "grad_norm": 6.41635429498773, + "learning_rate": 1.9999998974965327e-05, + "loss": 2.5043, + "step": 432 + }, + { + "epoch": 0.09079471587334871, + "grad_norm": 5.621730372574778, + "learning_rate": 1.999999769367203e-05, + "loss": 2.7389, + "step": 433 + }, + { + "epoch": 0.09100440343887607, + "grad_norm": 5.884022718989794, + "learning_rate": 1.999999589986151e-05, + "loss": 2.8743, + "step": 434 + }, + { + "epoch": 0.09121409100440343, + "grad_norm": 6.437487178912405, + "learning_rate": 1.9999993593533858e-05, + "loss": 2.6191, + "step": 435 + }, + { + "epoch": 0.09142377856993081, + "grad_norm": 5.472045448097777, + "learning_rate": 1.9999990774689186e-05, + "loss": 2.7156, + "step": 436 + }, + { + "epoch": 0.09163346613545817, + "grad_norm": 5.9600134087330625, + "learning_rate": 1.9999987443327645e-05, + "loss": 2.5117, + "step": 437 + }, + { + "epoch": 0.09184315370098553, + "grad_norm": 5.040329827674821, + "learning_rate": 1.9999983599449403e-05, + "loss": 2.6345, + "step": 438 + }, + { + "epoch": 0.09205284126651289, + "grad_norm": 5.4961592366018674, + "learning_rate": 1.999997924305466e-05, + "loss": 2.6613, + "step": 439 + }, + { + "epoch": 0.09226252883204027, + "grad_norm": 5.7771173965444005, + "learning_rate": 1.9999974374143632e-05, + "loss": 2.8014, + "step": 440 + }, + { + "epoch": 0.09247221639756763, + "grad_norm": 5.71809983482399, + "learning_rate": 1.9999968992716574e-05, + "loss": 2.6042, + "step": 441 + }, + { + "epoch": 0.09268190396309499, + "grad_norm": 5.402741944518257, + "learning_rate": 1.9999963098773763e-05, + "loss": 2.6389, + "step": 442 + }, + { + "epoch": 0.09289159152862235, + "grad_norm": 4.91373081975557, + "learning_rate": 1.9999956692315496e-05, + "loss": 2.5403, + "step": 443 + }, + { + "epoch": 0.09310127909414971, + "grad_norm": 6.049854120263461, + "learning_rate": 1.999994977334211e-05, + "loss": 2.6504, + "step": 444 + }, + { + "epoch": 0.09331096665967709, + "grad_norm": 4.83617941446969, + "learning_rate": 1.999994234185395e-05, + "loss": 2.6666, + "step": 445 + }, + { + "epoch": 0.09352065422520445, + "grad_norm": 5.375626449202231, + "learning_rate": 1.99999343978514e-05, + "loss": 2.7496, + "step": 446 + }, + { + "epoch": 0.09373034179073181, + "grad_norm": 5.80436245690412, + "learning_rate": 1.9999925941334873e-05, + "loss": 2.7277, + "step": 447 + }, + { + "epoch": 0.09394002935625917, + "grad_norm": 5.588237511682181, + "learning_rate": 1.9999916972304795e-05, + "loss": 2.5495, + "step": 448 + }, + { + "epoch": 0.09414971692178654, + "grad_norm": 5.783832722230089, + "learning_rate": 1.9999907490761624e-05, + "loss": 2.7139, + "step": 449 + }, + { + "epoch": 0.0943594044873139, + "grad_norm": 6.0148287387028345, + "learning_rate": 1.9999897496705856e-05, + "loss": 2.8294, + "step": 450 + }, + { + "epoch": 0.09456909205284127, + "grad_norm": 4.823691801875174, + "learning_rate": 1.9999886990137995e-05, + "loss": 2.4045, + "step": 451 + }, + { + "epoch": 0.09477877961836863, + "grad_norm": 5.664666742946905, + "learning_rate": 1.999987597105858e-05, + "loss": 2.8138, + "step": 452 + }, + { + "epoch": 0.09498846718389599, + "grad_norm": 5.175960849309011, + "learning_rate": 1.999986443946818e-05, + "loss": 2.5709, + "step": 453 + }, + { + "epoch": 0.09519815474942336, + "grad_norm": 5.4327949225361065, + "learning_rate": 1.9999852395367384e-05, + "loss": 2.8675, + "step": 454 + }, + { + "epoch": 0.09540784231495072, + "grad_norm": 6.5224952842492305, + "learning_rate": 1.999983983875681e-05, + "loss": 2.7747, + "step": 455 + }, + { + "epoch": 0.09561752988047809, + "grad_norm": 6.785685535621866, + "learning_rate": 1.99998267696371e-05, + "loss": 2.6202, + "step": 456 + }, + { + "epoch": 0.09582721744600545, + "grad_norm": 7.9980755631953535, + "learning_rate": 1.999981318800892e-05, + "loss": 2.8397, + "step": 457 + }, + { + "epoch": 0.09603690501153282, + "grad_norm": 5.624821398469246, + "learning_rate": 1.9999799093872974e-05, + "loss": 2.6769, + "step": 458 + }, + { + "epoch": 0.09624659257706018, + "grad_norm": 5.60366178340573, + "learning_rate": 1.9999784487229978e-05, + "loss": 2.6743, + "step": 459 + }, + { + "epoch": 0.09645628014258754, + "grad_norm": 5.290331678024499, + "learning_rate": 1.9999769368080684e-05, + "loss": 2.3815, + "step": 460 + }, + { + "epoch": 0.0966659677081149, + "grad_norm": 6.199980941313438, + "learning_rate": 1.999975373642587e-05, + "loss": 2.6348, + "step": 461 + }, + { + "epoch": 0.09687565527364227, + "grad_norm": 5.837642116719691, + "learning_rate": 1.999973759226633e-05, + "loss": 2.6432, + "step": 462 + }, + { + "epoch": 0.09708534283916964, + "grad_norm": 5.713414124198615, + "learning_rate": 1.9999720935602896e-05, + "loss": 2.5827, + "step": 463 + }, + { + "epoch": 0.097295030404697, + "grad_norm": 5.634915914795623, + "learning_rate": 1.9999703766436416e-05, + "loss": 2.6695, + "step": 464 + }, + { + "epoch": 0.09750471797022436, + "grad_norm": 5.6630857711538045, + "learning_rate": 1.999968608476778e-05, + "loss": 2.6379, + "step": 465 + }, + { + "epoch": 0.09771440553575172, + "grad_norm": 5.66602732631865, + "learning_rate": 1.9999667890597885e-05, + "loss": 2.7869, + "step": 466 + }, + { + "epoch": 0.0979240931012791, + "grad_norm": 5.308354869384293, + "learning_rate": 1.9999649183927673e-05, + "loss": 2.6358, + "step": 467 + }, + { + "epoch": 0.09813378066680646, + "grad_norm": 5.496323230462743, + "learning_rate": 1.9999629964758093e-05, + "loss": 2.7075, + "step": 468 + }, + { + "epoch": 0.09834346823233382, + "grad_norm": 5.542408691249312, + "learning_rate": 1.999961023309013e-05, + "loss": 2.7467, + "step": 469 + }, + { + "epoch": 0.09855315579786118, + "grad_norm": 5.422128619389672, + "learning_rate": 1.9999589988924804e-05, + "loss": 2.7005, + "step": 470 + }, + { + "epoch": 0.09876284336338854, + "grad_norm": 5.274937203640006, + "learning_rate": 1.9999569232263145e-05, + "loss": 2.6139, + "step": 471 + }, + { + "epoch": 0.09897253092891592, + "grad_norm": 5.9220959491954215, + "learning_rate": 1.999954796310622e-05, + "loss": 2.6594, + "step": 472 + }, + { + "epoch": 0.09918221849444328, + "grad_norm": 5.39217619898896, + "learning_rate": 1.9999526181455116e-05, + "loss": 2.6863, + "step": 473 + }, + { + "epoch": 0.09939190605997064, + "grad_norm": 5.90160115406471, + "learning_rate": 1.999950388731096e-05, + "loss": 2.6227, + "step": 474 + }, + { + "epoch": 0.099601593625498, + "grad_norm": 5.98777534779657, + "learning_rate": 1.999948108067488e-05, + "loss": 2.506, + "step": 475 + }, + { + "epoch": 0.09981128119102538, + "grad_norm": 6.196923671026067, + "learning_rate": 1.9999457761548054e-05, + "loss": 2.7249, + "step": 476 + }, + { + "epoch": 0.10002096875655274, + "grad_norm": 5.470458354976303, + "learning_rate": 1.999943392993167e-05, + "loss": 2.6591, + "step": 477 + }, + { + "epoch": 0.1002306563220801, + "grad_norm": 5.490708029945062, + "learning_rate": 1.999940958582696e-05, + "loss": 2.4634, + "step": 478 + }, + { + "epoch": 0.10044034388760746, + "grad_norm": 4.78497026880034, + "learning_rate": 1.9999384729235164e-05, + "loss": 2.522, + "step": 479 + }, + { + "epoch": 0.10065003145313482, + "grad_norm": 5.86210394443645, + "learning_rate": 1.9999359360157557e-05, + "loss": 2.5902, + "step": 480 + }, + { + "epoch": 0.1008597190186622, + "grad_norm": 5.431408794205825, + "learning_rate": 1.999933347859544e-05, + "loss": 2.5776, + "step": 481 + }, + { + "epoch": 0.10106940658418956, + "grad_norm": 4.772194403747424, + "learning_rate": 1.9999307084550133e-05, + "loss": 2.6012, + "step": 482 + }, + { + "epoch": 0.10127909414971692, + "grad_norm": 5.957703332978828, + "learning_rate": 1.9999280178023004e-05, + "loss": 2.6663, + "step": 483 + }, + { + "epoch": 0.10148878171524428, + "grad_norm": 6.117437833472381, + "learning_rate": 1.9999252759015418e-05, + "loss": 2.8025, + "step": 484 + }, + { + "epoch": 0.10169846928077166, + "grad_norm": 4.853564132624564, + "learning_rate": 1.999922482752879e-05, + "loss": 2.4937, + "step": 485 + }, + { + "epoch": 0.10190815684629902, + "grad_norm": 4.868952464252578, + "learning_rate": 1.9999196383564544e-05, + "loss": 2.4379, + "step": 486 + }, + { + "epoch": 0.10211784441182638, + "grad_norm": 5.414300387662243, + "learning_rate": 1.999916742712414e-05, + "loss": 2.5766, + "step": 487 + }, + { + "epoch": 0.10232753197735374, + "grad_norm": 6.4439519224976145, + "learning_rate": 1.999913795820906e-05, + "loss": 2.6016, + "step": 488 + }, + { + "epoch": 0.10253721954288111, + "grad_norm": 5.535209643558027, + "learning_rate": 1.9999107976820823e-05, + "loss": 2.6869, + "step": 489 + }, + { + "epoch": 0.10274690710840848, + "grad_norm": 5.408600507073794, + "learning_rate": 1.9999077482960955e-05, + "loss": 2.6929, + "step": 490 + }, + { + "epoch": 0.10295659467393584, + "grad_norm": 4.74942703402957, + "learning_rate": 1.9999046476631025e-05, + "loss": 2.6543, + "step": 491 + }, + { + "epoch": 0.1031662822394632, + "grad_norm": 5.623442300379496, + "learning_rate": 1.999901495783262e-05, + "loss": 2.5465, + "step": 492 + }, + { + "epoch": 0.10337596980499056, + "grad_norm": 5.353827901957465, + "learning_rate": 1.999898292656736e-05, + "loss": 2.7119, + "step": 493 + }, + { + "epoch": 0.10358565737051793, + "grad_norm": 5.265793390687354, + "learning_rate": 1.9998950382836875e-05, + "loss": 2.306, + "step": 494 + }, + { + "epoch": 0.1037953449360453, + "grad_norm": 4.601576136085532, + "learning_rate": 1.999891732664285e-05, + "loss": 2.3583, + "step": 495 + }, + { + "epoch": 0.10400503250157266, + "grad_norm": 7.0899812151715675, + "learning_rate": 1.999888375798696e-05, + "loss": 2.5927, + "step": 496 + }, + { + "epoch": 0.10421472006710002, + "grad_norm": 5.770738287851966, + "learning_rate": 1.9998849676870943e-05, + "loss": 2.8128, + "step": 497 + }, + { + "epoch": 0.10442440763262739, + "grad_norm": 5.343882586098956, + "learning_rate": 1.9998815083296536e-05, + "loss": 2.6729, + "step": 498 + }, + { + "epoch": 0.10463409519815475, + "grad_norm": 5.439766649151531, + "learning_rate": 1.999877997726551e-05, + "loss": 2.626, + "step": 499 + }, + { + "epoch": 0.10484378276368211, + "grad_norm": 5.3463005089974365, + "learning_rate": 1.9998744358779672e-05, + "loss": 2.5737, + "step": 500 + }, + { + "epoch": 0.10505347032920948, + "grad_norm": 5.266449331120233, + "learning_rate": 1.9998708227840842e-05, + "loss": 2.5286, + "step": 501 + }, + { + "epoch": 0.10526315789473684, + "grad_norm": 5.794319244925462, + "learning_rate": 1.9998671584450876e-05, + "loss": 2.5409, + "step": 502 + }, + { + "epoch": 0.10547284546026421, + "grad_norm": 5.895220317942617, + "learning_rate": 1.9998634428611646e-05, + "loss": 2.5332, + "step": 503 + }, + { + "epoch": 0.10568253302579157, + "grad_norm": 5.55712877854596, + "learning_rate": 1.999859676032506e-05, + "loss": 2.6529, + "step": 504 + }, + { + "epoch": 0.10589222059131893, + "grad_norm": 5.626329177204718, + "learning_rate": 1.9998558579593052e-05, + "loss": 2.7653, + "step": 505 + }, + { + "epoch": 0.1061019081568463, + "grad_norm": 6.1517665640732435, + "learning_rate": 1.999851988641757e-05, + "loss": 2.6343, + "step": 506 + }, + { + "epoch": 0.10631159572237367, + "grad_norm": 5.528137493345177, + "learning_rate": 1.9998480680800607e-05, + "loss": 2.6106, + "step": 507 + }, + { + "epoch": 0.10652128328790103, + "grad_norm": 5.730074344966369, + "learning_rate": 1.9998440962744164e-05, + "loss": 2.5419, + "step": 508 + }, + { + "epoch": 0.10673097085342839, + "grad_norm": 5.5326329704019415, + "learning_rate": 1.9998400732250284e-05, + "loss": 2.5181, + "step": 509 + }, + { + "epoch": 0.10694065841895575, + "grad_norm": 5.908567066648113, + "learning_rate": 1.999835998932102e-05, + "loss": 2.6386, + "step": 510 + }, + { + "epoch": 0.10715034598448311, + "grad_norm": 6.082411372734441, + "learning_rate": 1.999831873395847e-05, + "loss": 2.3721, + "step": 511 + }, + { + "epoch": 0.10736003355001049, + "grad_norm": 5.978197588276751, + "learning_rate": 1.9998276966164742e-05, + "loss": 2.8777, + "step": 512 + }, + { + "epoch": 0.10756972111553785, + "grad_norm": 5.679058175856486, + "learning_rate": 1.999823468594198e-05, + "loss": 2.6859, + "step": 513 + }, + { + "epoch": 0.10777940868106521, + "grad_norm": 4.853073321879911, + "learning_rate": 1.9998191893292347e-05, + "loss": 2.5453, + "step": 514 + }, + { + "epoch": 0.10798909624659257, + "grad_norm": 5.80047845788725, + "learning_rate": 1.9998148588218038e-05, + "loss": 2.6444, + "step": 515 + }, + { + "epoch": 0.10819878381211995, + "grad_norm": 5.576124317347189, + "learning_rate": 1.9998104770721274e-05, + "loss": 2.7963, + "step": 516 + }, + { + "epoch": 0.10840847137764731, + "grad_norm": 5.515732212854999, + "learning_rate": 1.9998060440804302e-05, + "loss": 2.6533, + "step": 517 + }, + { + "epoch": 0.10861815894317467, + "grad_norm": 5.833174455645435, + "learning_rate": 1.999801559846939e-05, + "loss": 2.4803, + "step": 518 + }, + { + "epoch": 0.10882784650870203, + "grad_norm": 4.788179962873536, + "learning_rate": 1.9997970243718835e-05, + "loss": 2.4361, + "step": 519 + }, + { + "epoch": 0.10903753407422939, + "grad_norm": 7.199003589993666, + "learning_rate": 1.9997924376554967e-05, + "loss": 2.5668, + "step": 520 + }, + { + "epoch": 0.10924722163975677, + "grad_norm": 5.677166471762073, + "learning_rate": 1.9997877996980135e-05, + "loss": 2.5804, + "step": 521 + }, + { + "epoch": 0.10945690920528413, + "grad_norm": 6.667982980235848, + "learning_rate": 1.999783110499671e-05, + "loss": 2.3781, + "step": 522 + }, + { + "epoch": 0.10966659677081149, + "grad_norm": 5.512445878703349, + "learning_rate": 1.9997783700607107e-05, + "loss": 2.5678, + "step": 523 + }, + { + "epoch": 0.10987628433633885, + "grad_norm": 4.780213508580989, + "learning_rate": 1.9997735783813748e-05, + "loss": 2.5523, + "step": 524 + }, + { + "epoch": 0.11008597190186623, + "grad_norm": 5.8430158104820356, + "learning_rate": 1.999768735461909e-05, + "loss": 2.6848, + "step": 525 + }, + { + "epoch": 0.11029565946739359, + "grad_norm": 5.261270035066101, + "learning_rate": 1.9997638413025612e-05, + "loss": 2.7444, + "step": 526 + }, + { + "epoch": 0.11050534703292095, + "grad_norm": 4.937057553288801, + "learning_rate": 1.9997588959035826e-05, + "loss": 2.6237, + "step": 527 + }, + { + "epoch": 0.11071503459844831, + "grad_norm": 5.7624052028360255, + "learning_rate": 1.9997538992652267e-05, + "loss": 2.7303, + "step": 528 + }, + { + "epoch": 0.11092472216397567, + "grad_norm": 5.172071280493205, + "learning_rate": 1.9997488513877498e-05, + "loss": 2.6619, + "step": 529 + }, + { + "epoch": 0.11113440972950304, + "grad_norm": 5.345373830140982, + "learning_rate": 1.99974375227141e-05, + "loss": 2.6435, + "step": 530 + }, + { + "epoch": 0.1113440972950304, + "grad_norm": 4.909306076745643, + "learning_rate": 1.9997386019164687e-05, + "loss": 2.4053, + "step": 531 + }, + { + "epoch": 0.11155378486055777, + "grad_norm": 4.77733420661915, + "learning_rate": 1.9997334003231904e-05, + "loss": 2.4678, + "step": 532 + }, + { + "epoch": 0.11176347242608513, + "grad_norm": 5.16986640342695, + "learning_rate": 1.999728147491841e-05, + "loss": 2.4487, + "step": 533 + }, + { + "epoch": 0.1119731599916125, + "grad_norm": 5.230628640510131, + "learning_rate": 1.9997228434226903e-05, + "loss": 2.6181, + "step": 534 + }, + { + "epoch": 0.11218284755713986, + "grad_norm": 5.853022871575048, + "learning_rate": 1.99971748811601e-05, + "loss": 2.6099, + "step": 535 + }, + { + "epoch": 0.11239253512266723, + "grad_norm": 6.20050163281424, + "learning_rate": 1.9997120815720748e-05, + "loss": 2.7233, + "step": 536 + }, + { + "epoch": 0.11260222268819459, + "grad_norm": 6.214963391517287, + "learning_rate": 1.9997066237911608e-05, + "loss": 2.6142, + "step": 537 + }, + { + "epoch": 0.11281191025372195, + "grad_norm": 5.4798180256015945, + "learning_rate": 1.999701114773549e-05, + "loss": 2.5321, + "step": 538 + }, + { + "epoch": 0.11302159781924932, + "grad_norm": 5.375298864194612, + "learning_rate": 1.999695554519521e-05, + "loss": 2.7338, + "step": 539 + }, + { + "epoch": 0.11323128538477668, + "grad_norm": 5.7725418154852814, + "learning_rate": 1.999689943029362e-05, + "loss": 2.6815, + "step": 540 + }, + { + "epoch": 0.11344097295030404, + "grad_norm": 5.7692346185079515, + "learning_rate": 1.9996842803033597e-05, + "loss": 2.6926, + "step": 541 + }, + { + "epoch": 0.1136506605158314, + "grad_norm": 5.695150210183655, + "learning_rate": 1.9996785663418038e-05, + "loss": 2.7382, + "step": 542 + }, + { + "epoch": 0.11386034808135878, + "grad_norm": 6.273053977740661, + "learning_rate": 1.9996728011449877e-05, + "loss": 2.8509, + "step": 543 + }, + { + "epoch": 0.11407003564688614, + "grad_norm": 5.753481087512013, + "learning_rate": 1.9996669847132067e-05, + "loss": 2.6474, + "step": 544 + }, + { + "epoch": 0.1142797232124135, + "grad_norm": 5.562055661939869, + "learning_rate": 1.9996611170467587e-05, + "loss": 2.4718, + "step": 545 + }, + { + "epoch": 0.11448941077794086, + "grad_norm": 6.269581425546245, + "learning_rate": 1.9996551981459448e-05, + "loss": 2.5538, + "step": 546 + }, + { + "epoch": 0.11469909834346823, + "grad_norm": 4.778157525726834, + "learning_rate": 1.9996492280110684e-05, + "loss": 2.4509, + "step": 547 + }, + { + "epoch": 0.1149087859089956, + "grad_norm": 6.260218852948971, + "learning_rate": 1.9996432066424352e-05, + "loss": 2.6991, + "step": 548 + }, + { + "epoch": 0.11511847347452296, + "grad_norm": 7.078277664927638, + "learning_rate": 1.9996371340403535e-05, + "loss": 2.6498, + "step": 549 + }, + { + "epoch": 0.11532816104005032, + "grad_norm": 5.994450909932426, + "learning_rate": 1.9996310102051354e-05, + "loss": 2.5986, + "step": 550 + }, + { + "epoch": 0.11553784860557768, + "grad_norm": 5.715137939220944, + "learning_rate": 1.999624835137094e-05, + "loss": 2.6444, + "step": 551 + }, + { + "epoch": 0.11574753617110506, + "grad_norm": 5.544481841817648, + "learning_rate": 1.999618608836546e-05, + "loss": 2.5788, + "step": 552 + }, + { + "epoch": 0.11595722373663242, + "grad_norm": 4.784007352390835, + "learning_rate": 1.9996123313038105e-05, + "loss": 2.5964, + "step": 553 + }, + { + "epoch": 0.11616691130215978, + "grad_norm": 4.7725629325319785, + "learning_rate": 1.9996060025392093e-05, + "loss": 2.43, + "step": 554 + }, + { + "epoch": 0.11637659886768714, + "grad_norm": 5.394147929053285, + "learning_rate": 1.999599622543067e-05, + "loss": 2.6378, + "step": 555 + }, + { + "epoch": 0.11658628643321452, + "grad_norm": 4.848390711476544, + "learning_rate": 1.99959319131571e-05, + "loss": 2.5703, + "step": 556 + }, + { + "epoch": 0.11679597399874188, + "grad_norm": 6.326768339397591, + "learning_rate": 1.999586708857468e-05, + "loss": 2.6833, + "step": 557 + }, + { + "epoch": 0.11700566156426924, + "grad_norm": 6.0610930927826265, + "learning_rate": 1.999580175168674e-05, + "loss": 2.3693, + "step": 558 + }, + { + "epoch": 0.1172153491297966, + "grad_norm": 4.661353301262643, + "learning_rate": 1.9995735902496622e-05, + "loss": 2.7019, + "step": 559 + }, + { + "epoch": 0.11742503669532396, + "grad_norm": 5.635216063572443, + "learning_rate": 1.9995669541007703e-05, + "loss": 2.482, + "step": 560 + }, + { + "epoch": 0.11763472426085134, + "grad_norm": 5.214617419063985, + "learning_rate": 1.999560266722338e-05, + "loss": 2.8172, + "step": 561 + }, + { + "epoch": 0.1178444118263787, + "grad_norm": 5.934904408462649, + "learning_rate": 1.999553528114709e-05, + "loss": 2.7383, + "step": 562 + }, + { + "epoch": 0.11805409939190606, + "grad_norm": 5.480689337975177, + "learning_rate": 1.9995467382782278e-05, + "loss": 2.5456, + "step": 563 + }, + { + "epoch": 0.11826378695743342, + "grad_norm": 5.864430945744579, + "learning_rate": 1.9995398972132422e-05, + "loss": 2.4998, + "step": 564 + }, + { + "epoch": 0.1184734745229608, + "grad_norm": 5.572352742451088, + "learning_rate": 1.9995330049201037e-05, + "loss": 2.7039, + "step": 565 + }, + { + "epoch": 0.11868316208848816, + "grad_norm": 5.485299336145905, + "learning_rate": 1.999526061399165e-05, + "loss": 2.595, + "step": 566 + }, + { + "epoch": 0.11889284965401552, + "grad_norm": 5.449914823873228, + "learning_rate": 1.9995190666507818e-05, + "loss": 2.5257, + "step": 567 + }, + { + "epoch": 0.11910253721954288, + "grad_norm": 5.828691865177637, + "learning_rate": 1.9995120206753132e-05, + "loss": 2.4034, + "step": 568 + }, + { + "epoch": 0.11931222478507024, + "grad_norm": 5.456849930767592, + "learning_rate": 1.9995049234731198e-05, + "loss": 2.6921, + "step": 569 + }, + { + "epoch": 0.11952191235059761, + "grad_norm": 5.232394499420569, + "learning_rate": 1.9994977750445653e-05, + "loss": 2.5938, + "step": 570 + }, + { + "epoch": 0.11973159991612498, + "grad_norm": 4.948895008756202, + "learning_rate": 1.9994905753900168e-05, + "loss": 2.6296, + "step": 571 + }, + { + "epoch": 0.11994128748165234, + "grad_norm": 5.679156712807628, + "learning_rate": 1.9994833245098423e-05, + "loss": 2.5833, + "step": 572 + }, + { + "epoch": 0.1201509750471797, + "grad_norm": 5.685530218272313, + "learning_rate": 1.9994760224044142e-05, + "loss": 2.759, + "step": 573 + }, + { + "epoch": 0.12036066261270707, + "grad_norm": 6.043767935873941, + "learning_rate": 1.9994686690741066e-05, + "loss": 2.7435, + "step": 574 + }, + { + "epoch": 0.12057035017823443, + "grad_norm": 5.502999826879568, + "learning_rate": 1.9994612645192957e-05, + "loss": 2.5119, + "step": 575 + }, + { + "epoch": 0.1207800377437618, + "grad_norm": 6.050329567382397, + "learning_rate": 1.9994538087403623e-05, + "loss": 2.4762, + "step": 576 + }, + { + "epoch": 0.12098972530928916, + "grad_norm": 5.008427850487914, + "learning_rate": 1.999446301737687e-05, + "loss": 2.6466, + "step": 577 + }, + { + "epoch": 0.12119941287481652, + "grad_norm": 5.533879498629465, + "learning_rate": 1.9994387435116557e-05, + "loss": 2.5645, + "step": 578 + }, + { + "epoch": 0.12140910044034389, + "grad_norm": 6.042839943721938, + "learning_rate": 1.999431134062655e-05, + "loss": 2.5004, + "step": 579 + }, + { + "epoch": 0.12161878800587125, + "grad_norm": 5.467755065739939, + "learning_rate": 1.9994234733910754e-05, + "loss": 2.3032, + "step": 580 + }, + { + "epoch": 0.12182847557139861, + "grad_norm": 5.63914808406953, + "learning_rate": 1.9994157614973094e-05, + "loss": 2.5145, + "step": 581 + }, + { + "epoch": 0.12203816313692598, + "grad_norm": 6.162111971834968, + "learning_rate": 1.999407998381752e-05, + "loss": 2.7087, + "step": 582 + }, + { + "epoch": 0.12224785070245335, + "grad_norm": 5.225174953294268, + "learning_rate": 1.999400184044802e-05, + "loss": 2.3773, + "step": 583 + }, + { + "epoch": 0.12245753826798071, + "grad_norm": 5.679697001972673, + "learning_rate": 1.9993923184868588e-05, + "loss": 2.2435, + "step": 584 + }, + { + "epoch": 0.12266722583350807, + "grad_norm": 6.070179558093573, + "learning_rate": 1.999384401708326e-05, + "loss": 2.7846, + "step": 585 + }, + { + "epoch": 0.12287691339903543, + "grad_norm": 4.897891464402482, + "learning_rate": 1.999376433709609e-05, + "loss": 2.3082, + "step": 586 + }, + { + "epoch": 0.1230866009645628, + "grad_norm": 5.836124997880415, + "learning_rate": 1.9993684144911167e-05, + "loss": 2.7073, + "step": 587 + }, + { + "epoch": 0.12329628853009017, + "grad_norm": 5.627032650978459, + "learning_rate": 1.9993603440532596e-05, + "loss": 2.5463, + "step": 588 + }, + { + "epoch": 0.12350597609561753, + "grad_norm": 6.265836229503592, + "learning_rate": 1.9993522223964518e-05, + "loss": 2.6539, + "step": 589 + }, + { + "epoch": 0.12371566366114489, + "grad_norm": 5.781070108933454, + "learning_rate": 1.999344049521109e-05, + "loss": 2.6672, + "step": 590 + }, + { + "epoch": 0.12392535122667225, + "grad_norm": 4.983512777309647, + "learning_rate": 1.9993358254276508e-05, + "loss": 2.6151, + "step": 591 + }, + { + "epoch": 0.12413503879219963, + "grad_norm": 5.321385752891051, + "learning_rate": 1.999327550116498e-05, + "loss": 2.5062, + "step": 592 + }, + { + "epoch": 0.12434472635772699, + "grad_norm": 5.398206354535945, + "learning_rate": 1.999319223588075e-05, + "loss": 2.3726, + "step": 593 + }, + { + "epoch": 0.12455441392325435, + "grad_norm": 5.114306534153933, + "learning_rate": 1.9993108458428087e-05, + "loss": 2.523, + "step": 594 + }, + { + "epoch": 0.12476410148878171, + "grad_norm": 5.170649912916199, + "learning_rate": 1.999302416881128e-05, + "loss": 2.5864, + "step": 595 + }, + { + "epoch": 0.12497378905430907, + "grad_norm": 5.6163494832277845, + "learning_rate": 1.999293936703466e-05, + "loss": 2.5966, + "step": 596 + }, + { + "epoch": 0.12518347661983645, + "grad_norm": 5.456525534616905, + "learning_rate": 1.9992854053102556e-05, + "loss": 2.7464, + "step": 597 + }, + { + "epoch": 0.1253931641853638, + "grad_norm": 6.050680813839292, + "learning_rate": 1.9992768227019357e-05, + "loss": 2.5624, + "step": 598 + }, + { + "epoch": 0.12560285175089117, + "grad_norm": 5.283074996332358, + "learning_rate": 1.9992681888789453e-05, + "loss": 2.6451, + "step": 599 + }, + { + "epoch": 0.12581253931641853, + "grad_norm": 5.367844491072011, + "learning_rate": 1.999259503841727e-05, + "loss": 2.4811, + "step": 600 + }, + { + "epoch": 0.1260222268819459, + "grad_norm": 6.248909724328069, + "learning_rate": 1.999250767590726e-05, + "loss": 2.722, + "step": 601 + }, + { + "epoch": 0.12623191444747325, + "grad_norm": 5.552832606107895, + "learning_rate": 1.9992419801263902e-05, + "loss": 2.7244, + "step": 602 + }, + { + "epoch": 0.12644160201300064, + "grad_norm": 5.252252969697778, + "learning_rate": 1.9992331414491697e-05, + "loss": 2.688, + "step": 603 + }, + { + "epoch": 0.126651289578528, + "grad_norm": 6.123895763343174, + "learning_rate": 1.9992242515595175e-05, + "loss": 2.7344, + "step": 604 + }, + { + "epoch": 0.12686097714405536, + "grad_norm": 5.302942206769204, + "learning_rate": 1.9992153104578892e-05, + "loss": 2.6337, + "step": 605 + }, + { + "epoch": 0.12707066470958273, + "grad_norm": 5.914274981623615, + "learning_rate": 1.9992063181447437e-05, + "loss": 2.732, + "step": 606 + }, + { + "epoch": 0.1272803522751101, + "grad_norm": 6.142168819060165, + "learning_rate": 1.999197274620541e-05, + "loss": 2.6441, + "step": 607 + }, + { + "epoch": 0.12749003984063745, + "grad_norm": 4.973284653106787, + "learning_rate": 1.9991881798857453e-05, + "loss": 2.611, + "step": 608 + }, + { + "epoch": 0.1276997274061648, + "grad_norm": 5.264163042836751, + "learning_rate": 1.999179033940822e-05, + "loss": 2.6344, + "step": 609 + }, + { + "epoch": 0.12790941497169217, + "grad_norm": 4.9408790866947845, + "learning_rate": 1.9991698367862402e-05, + "loss": 2.553, + "step": 610 + }, + { + "epoch": 0.12811910253721953, + "grad_norm": 5.91128114331133, + "learning_rate": 1.9991605884224718e-05, + "loss": 2.5662, + "step": 611 + }, + { + "epoch": 0.12832879010274692, + "grad_norm": 6.278252666825148, + "learning_rate": 1.99915128884999e-05, + "loss": 2.6621, + "step": 612 + }, + { + "epoch": 0.12853847766827428, + "grad_norm": 6.73800212945431, + "learning_rate": 1.9991419380692712e-05, + "loss": 2.5745, + "step": 613 + }, + { + "epoch": 0.12874816523380164, + "grad_norm": 5.702948291536183, + "learning_rate": 1.999132536080796e-05, + "loss": 2.524, + "step": 614 + }, + { + "epoch": 0.128957852799329, + "grad_norm": 5.48889743512358, + "learning_rate": 1.9991230828850448e-05, + "loss": 2.5485, + "step": 615 + }, + { + "epoch": 0.12916754036485636, + "grad_norm": 5.739661561642836, + "learning_rate": 1.999113578482503e-05, + "loss": 2.4045, + "step": 616 + }, + { + "epoch": 0.12937722793038373, + "grad_norm": 5.238697864654882, + "learning_rate": 1.9991040228736573e-05, + "loss": 2.5972, + "step": 617 + }, + { + "epoch": 0.1295869154959111, + "grad_norm": 6.376141328533404, + "learning_rate": 1.9990944160589975e-05, + "loss": 2.7524, + "step": 618 + }, + { + "epoch": 0.12979660306143845, + "grad_norm": 5.536560112421979, + "learning_rate": 1.9990847580390164e-05, + "loss": 2.6407, + "step": 619 + }, + { + "epoch": 0.1300062906269658, + "grad_norm": 6.489063070000069, + "learning_rate": 1.9990750488142084e-05, + "loss": 2.7913, + "step": 620 + }, + { + "epoch": 0.1302159781924932, + "grad_norm": 5.734717822189088, + "learning_rate": 1.9990652883850712e-05, + "loss": 2.6737, + "step": 621 + }, + { + "epoch": 0.13042566575802056, + "grad_norm": 5.4608587689447745, + "learning_rate": 1.9990554767521052e-05, + "loss": 2.419, + "step": 622 + }, + { + "epoch": 0.13063535332354792, + "grad_norm": 4.682214953059233, + "learning_rate": 1.9990456139158136e-05, + "loss": 2.4798, + "step": 623 + }, + { + "epoch": 0.13084504088907528, + "grad_norm": 5.643331962031569, + "learning_rate": 1.999035699876701e-05, + "loss": 2.684, + "step": 624 + }, + { + "epoch": 0.13105472845460264, + "grad_norm": 5.776023410811839, + "learning_rate": 1.9990257346352765e-05, + "loss": 2.59, + "step": 625 + }, + { + "epoch": 0.13126441602013, + "grad_norm": 5.600144616010864, + "learning_rate": 1.99901571819205e-05, + "loss": 2.681, + "step": 626 + }, + { + "epoch": 0.13147410358565736, + "grad_norm": 5.933435259956886, + "learning_rate": 1.9990056505475354e-05, + "loss": 2.7423, + "step": 627 + }, + { + "epoch": 0.13168379115118473, + "grad_norm": 5.136463019492448, + "learning_rate": 1.9989955317022488e-05, + "loss": 2.6539, + "step": 628 + }, + { + "epoch": 0.1318934787167121, + "grad_norm": 5.634815883081778, + "learning_rate": 1.998985361656708e-05, + "loss": 2.5686, + "step": 629 + }, + { + "epoch": 0.13210316628223948, + "grad_norm": 6.1728856249187665, + "learning_rate": 1.9989751404114353e-05, + "loss": 2.6307, + "step": 630 + }, + { + "epoch": 0.13231285384776684, + "grad_norm": 5.617788962890585, + "learning_rate": 1.9989648679669534e-05, + "loss": 2.8053, + "step": 631 + }, + { + "epoch": 0.1325225414132942, + "grad_norm": 5.242809592919124, + "learning_rate": 1.99895454432379e-05, + "loss": 2.6759, + "step": 632 + }, + { + "epoch": 0.13273222897882156, + "grad_norm": 4.697066641496849, + "learning_rate": 1.9989441694824733e-05, + "loss": 2.5688, + "step": 633 + }, + { + "epoch": 0.13294191654434892, + "grad_norm": 5.023310617072543, + "learning_rate": 1.9989337434435355e-05, + "loss": 2.7492, + "step": 634 + }, + { + "epoch": 0.13315160410987628, + "grad_norm": 4.861011579498409, + "learning_rate": 1.998923266207511e-05, + "loss": 2.4153, + "step": 635 + }, + { + "epoch": 0.13336129167540364, + "grad_norm": 4.876867865694473, + "learning_rate": 1.998912737774936e-05, + "loss": 2.4357, + "step": 636 + }, + { + "epoch": 0.133570979240931, + "grad_norm": 5.72943740501241, + "learning_rate": 1.998902158146351e-05, + "loss": 2.3662, + "step": 637 + }, + { + "epoch": 0.13378066680645836, + "grad_norm": 6.564313478184733, + "learning_rate": 1.9988915273222978e-05, + "loss": 2.7265, + "step": 638 + }, + { + "epoch": 0.13399035437198575, + "grad_norm": 6.022619420405473, + "learning_rate": 1.9988808453033213e-05, + "loss": 2.6599, + "step": 639 + }, + { + "epoch": 0.13420004193751311, + "grad_norm": 5.575552897830298, + "learning_rate": 1.9988701120899693e-05, + "loss": 2.6047, + "step": 640 + }, + { + "epoch": 0.13440972950304048, + "grad_norm": 5.512229435159064, + "learning_rate": 1.9988593276827912e-05, + "loss": 2.5313, + "step": 641 + }, + { + "epoch": 0.13461941706856784, + "grad_norm": 5.014022192346478, + "learning_rate": 1.99884849208234e-05, + "loss": 2.5894, + "step": 642 + }, + { + "epoch": 0.1348291046340952, + "grad_norm": 6.3607236447902835, + "learning_rate": 1.9988376052891716e-05, + "loss": 2.6781, + "step": 643 + }, + { + "epoch": 0.13503879219962256, + "grad_norm": 5.40513632427288, + "learning_rate": 1.998826667303843e-05, + "loss": 2.3419, + "step": 644 + }, + { + "epoch": 0.13524847976514992, + "grad_norm": 5.837611823987948, + "learning_rate": 1.9988156781269163e-05, + "loss": 2.6317, + "step": 645 + }, + { + "epoch": 0.13545816733067728, + "grad_norm": 6.721032888952139, + "learning_rate": 1.998804637758953e-05, + "loss": 2.7753, + "step": 646 + }, + { + "epoch": 0.13566785489620464, + "grad_norm": 6.083190464187132, + "learning_rate": 1.9987935462005198e-05, + "loss": 2.5566, + "step": 647 + }, + { + "epoch": 0.13587754246173203, + "grad_norm": 6.077949283580154, + "learning_rate": 1.998782403452185e-05, + "loss": 2.6409, + "step": 648 + }, + { + "epoch": 0.1360872300272594, + "grad_norm": 5.624873727632852, + "learning_rate": 1.9987712095145196e-05, + "loss": 2.5109, + "step": 649 + }, + { + "epoch": 0.13629691759278675, + "grad_norm": 5.149021463965709, + "learning_rate": 1.9987599643880977e-05, + "loss": 2.5359, + "step": 650 + }, + { + "epoch": 0.13650660515831411, + "grad_norm": 4.489888875038243, + "learning_rate": 1.9987486680734955e-05, + "loss": 2.355, + "step": 651 + }, + { + "epoch": 0.13671629272384148, + "grad_norm": 5.294389301140457, + "learning_rate": 1.9987373205712912e-05, + "loss": 2.6437, + "step": 652 + }, + { + "epoch": 0.13692598028936884, + "grad_norm": 6.211927293086614, + "learning_rate": 1.9987259218820678e-05, + "loss": 2.7953, + "step": 653 + }, + { + "epoch": 0.1371356678548962, + "grad_norm": 5.83721495598414, + "learning_rate": 1.998714472006408e-05, + "loss": 2.416, + "step": 654 + }, + { + "epoch": 0.13734535542042356, + "grad_norm": 5.206161014749805, + "learning_rate": 1.9987029709448996e-05, + "loss": 2.6799, + "step": 655 + }, + { + "epoch": 0.13755504298595092, + "grad_norm": 5.5512605404149795, + "learning_rate": 1.998691418698132e-05, + "loss": 2.8845, + "step": 656 + }, + { + "epoch": 0.1377647305514783, + "grad_norm": 5.931845394998795, + "learning_rate": 1.9986798152666966e-05, + "loss": 2.6356, + "step": 657 + }, + { + "epoch": 0.13797441811700567, + "grad_norm": 5.622068071687936, + "learning_rate": 1.9986681606511888e-05, + "loss": 2.4955, + "step": 658 + }, + { + "epoch": 0.13818410568253303, + "grad_norm": 5.2450933279206104, + "learning_rate": 1.9986564548522058e-05, + "loss": 2.5056, + "step": 659 + }, + { + "epoch": 0.1383937932480604, + "grad_norm": 5.213714379448667, + "learning_rate": 1.998644697870347e-05, + "loss": 2.4313, + "step": 660 + }, + { + "epoch": 0.13860348081358775, + "grad_norm": 4.962907536927089, + "learning_rate": 1.9986328897062157e-05, + "loss": 2.5469, + "step": 661 + }, + { + "epoch": 0.13881316837911511, + "grad_norm": 5.448549286590173, + "learning_rate": 1.9986210303604167e-05, + "loss": 2.7182, + "step": 662 + }, + { + "epoch": 0.13902285594464248, + "grad_norm": 5.653507502704269, + "learning_rate": 1.998609119833558e-05, + "loss": 2.7621, + "step": 663 + }, + { + "epoch": 0.13923254351016984, + "grad_norm": 5.958774481371084, + "learning_rate": 1.9985971581262498e-05, + "loss": 2.5839, + "step": 664 + }, + { + "epoch": 0.1394422310756972, + "grad_norm": 5.784459568602851, + "learning_rate": 1.998585145239105e-05, + "loss": 2.458, + "step": 665 + }, + { + "epoch": 0.1396519186412246, + "grad_norm": 5.530312020168745, + "learning_rate": 1.99857308117274e-05, + "loss": 2.5512, + "step": 666 + }, + { + "epoch": 0.13986160620675195, + "grad_norm": 6.1344985417981395, + "learning_rate": 1.9985609659277725e-05, + "loss": 2.4614, + "step": 667 + }, + { + "epoch": 0.1400712937722793, + "grad_norm": 5.7035173869033935, + "learning_rate": 1.9985487995048233e-05, + "loss": 2.523, + "step": 668 + }, + { + "epoch": 0.14028098133780667, + "grad_norm": 6.239261865278675, + "learning_rate": 1.9985365819045167e-05, + "loss": 2.8481, + "step": 669 + }, + { + "epoch": 0.14049066890333403, + "grad_norm": 5.324058490245803, + "learning_rate": 1.998524313127478e-05, + "loss": 2.4229, + "step": 670 + }, + { + "epoch": 0.1407003564688614, + "grad_norm": 4.71799679414778, + "learning_rate": 1.9985119931743365e-05, + "loss": 2.6776, + "step": 671 + }, + { + "epoch": 0.14091004403438875, + "grad_norm": 5.691016122633507, + "learning_rate": 1.9984996220457236e-05, + "loss": 2.5361, + "step": 672 + }, + { + "epoch": 0.14111973159991612, + "grad_norm": 4.249981795124725, + "learning_rate": 1.9984871997422732e-05, + "loss": 2.5796, + "step": 673 + }, + { + "epoch": 0.14132941916544348, + "grad_norm": 5.693901831661372, + "learning_rate": 1.9984747262646222e-05, + "loss": 2.5433, + "step": 674 + }, + { + "epoch": 0.14153910673097087, + "grad_norm": 5.954077652552696, + "learning_rate": 1.9984622016134096e-05, + "loss": 2.5969, + "step": 675 + }, + { + "epoch": 0.14174879429649823, + "grad_norm": 5.428001411309745, + "learning_rate": 1.9984496257892775e-05, + "loss": 2.5192, + "step": 676 + }, + { + "epoch": 0.1419584818620256, + "grad_norm": 5.0598543050965885, + "learning_rate": 1.99843699879287e-05, + "loss": 2.3452, + "step": 677 + }, + { + "epoch": 0.14216816942755295, + "grad_norm": 5.0767043318078064, + "learning_rate": 1.9984243206248347e-05, + "loss": 2.6712, + "step": 678 + }, + { + "epoch": 0.1423778569930803, + "grad_norm": 4.601010244095382, + "learning_rate": 1.9984115912858217e-05, + "loss": 2.5923, + "step": 679 + }, + { + "epoch": 0.14258754455860767, + "grad_norm": 5.424086112836491, + "learning_rate": 1.9983988107764828e-05, + "loss": 2.5772, + "step": 680 + }, + { + "epoch": 0.14279723212413503, + "grad_norm": 5.252267616555135, + "learning_rate": 1.998385979097473e-05, + "loss": 2.6972, + "step": 681 + }, + { + "epoch": 0.1430069196896624, + "grad_norm": 5.463460655872463, + "learning_rate": 1.9983730962494504e-05, + "loss": 2.4517, + "step": 682 + }, + { + "epoch": 0.14321660725518975, + "grad_norm": 5.118115869057551, + "learning_rate": 1.998360162233075e-05, + "loss": 2.5055, + "step": 683 + }, + { + "epoch": 0.14342629482071714, + "grad_norm": 5.744180697178427, + "learning_rate": 1.9983471770490093e-05, + "loss": 2.6948, + "step": 684 + }, + { + "epoch": 0.1436359823862445, + "grad_norm": 5.230441377908594, + "learning_rate": 1.99833414069792e-05, + "loss": 2.6347, + "step": 685 + }, + { + "epoch": 0.14384566995177187, + "grad_norm": 5.406240431378833, + "learning_rate": 1.9983210531804736e-05, + "loss": 2.3523, + "step": 686 + }, + { + "epoch": 0.14405535751729923, + "grad_norm": 5.1111112729171575, + "learning_rate": 1.9983079144973424e-05, + "loss": 2.4532, + "step": 687 + }, + { + "epoch": 0.1442650450828266, + "grad_norm": 6.229726640457781, + "learning_rate": 1.998294724649199e-05, + "loss": 2.6002, + "step": 688 + }, + { + "epoch": 0.14447473264835395, + "grad_norm": 4.842476793201206, + "learning_rate": 1.9982814836367193e-05, + "loss": 2.5075, + "step": 689 + }, + { + "epoch": 0.1446844202138813, + "grad_norm": 5.199600262154254, + "learning_rate": 1.9982681914605823e-05, + "loss": 2.3085, + "step": 690 + }, + { + "epoch": 0.14489410777940867, + "grad_norm": 6.1062450571318365, + "learning_rate": 1.998254848121469e-05, + "loss": 2.8684, + "step": 691 + }, + { + "epoch": 0.14510379534493603, + "grad_norm": 5.772115902147379, + "learning_rate": 1.9982414536200633e-05, + "loss": 2.5755, + "step": 692 + }, + { + "epoch": 0.14531348291046342, + "grad_norm": 5.5088024507795055, + "learning_rate": 1.9982280079570518e-05, + "loss": 2.5159, + "step": 693 + }, + { + "epoch": 0.14552317047599078, + "grad_norm": 5.495057670409247, + "learning_rate": 1.9982145111331235e-05, + "loss": 2.5204, + "step": 694 + }, + { + "epoch": 0.14573285804151814, + "grad_norm": 5.663383224467504, + "learning_rate": 1.9982009631489704e-05, + "loss": 2.4353, + "step": 695 + }, + { + "epoch": 0.1459425456070455, + "grad_norm": 5.466758344070865, + "learning_rate": 1.9981873640052863e-05, + "loss": 2.6035, + "step": 696 + }, + { + "epoch": 0.14615223317257287, + "grad_norm": 6.078622694955836, + "learning_rate": 1.9981737137027684e-05, + "loss": 2.6876, + "step": 697 + }, + { + "epoch": 0.14636192073810023, + "grad_norm": 4.9166578396991385, + "learning_rate": 1.9981600122421172e-05, + "loss": 2.6663, + "step": 698 + }, + { + "epoch": 0.1465716083036276, + "grad_norm": 6.131284466589723, + "learning_rate": 1.9981462596240337e-05, + "loss": 2.6511, + "step": 699 + }, + { + "epoch": 0.14678129586915495, + "grad_norm": 4.979786024235127, + "learning_rate": 1.998132455849223e-05, + "loss": 2.528, + "step": 700 + }, + { + "epoch": 0.1469909834346823, + "grad_norm": 5.057122135958919, + "learning_rate": 1.9981186009183927e-05, + "loss": 2.4776, + "step": 701 + }, + { + "epoch": 0.1472006710002097, + "grad_norm": 5.773137815876697, + "learning_rate": 1.9981046948322533e-05, + "loss": 2.6508, + "step": 702 + }, + { + "epoch": 0.14741035856573706, + "grad_norm": 4.9536915294921435, + "learning_rate": 1.9980907375915173e-05, + "loss": 2.5322, + "step": 703 + }, + { + "epoch": 0.14762004613126442, + "grad_norm": 5.340343694368092, + "learning_rate": 1.9980767291968997e-05, + "loss": 2.4569, + "step": 704 + }, + { + "epoch": 0.14782973369679178, + "grad_norm": 4.839941696325849, + "learning_rate": 1.9980626696491185e-05, + "loss": 2.4397, + "step": 705 + }, + { + "epoch": 0.14803942126231914, + "grad_norm": 5.732583335149285, + "learning_rate": 1.9980485589488946e-05, + "loss": 2.587, + "step": 706 + }, + { + "epoch": 0.1482491088278465, + "grad_norm": 5.105108668151097, + "learning_rate": 1.998034397096951e-05, + "loss": 2.4192, + "step": 707 + }, + { + "epoch": 0.14845879639337387, + "grad_norm": 5.326723333192696, + "learning_rate": 1.9980201840940136e-05, + "loss": 2.5819, + "step": 708 + }, + { + "epoch": 0.14866848395890123, + "grad_norm": 5.663789248616749, + "learning_rate": 1.9980059199408103e-05, + "loss": 2.6536, + "step": 709 + }, + { + "epoch": 0.14887817152442862, + "grad_norm": 5.970658029480189, + "learning_rate": 1.9979916046380734e-05, + "loss": 2.5629, + "step": 710 + }, + { + "epoch": 0.14908785908995598, + "grad_norm": 5.74965717949682, + "learning_rate": 1.997977238186535e-05, + "loss": 2.5661, + "step": 711 + }, + { + "epoch": 0.14929754665548334, + "grad_norm": 6.02301762189861, + "learning_rate": 1.997962820586933e-05, + "loss": 2.7692, + "step": 712 + }, + { + "epoch": 0.1495072342210107, + "grad_norm": 4.579567037465211, + "learning_rate": 1.9979483518400055e-05, + "loss": 2.5482, + "step": 713 + }, + { + "epoch": 0.14971692178653806, + "grad_norm": 6.2559796630509075, + "learning_rate": 1.9979338319464938e-05, + "loss": 2.7388, + "step": 714 + }, + { + "epoch": 0.14992660935206542, + "grad_norm": 5.0962604475265145, + "learning_rate": 1.9979192609071426e-05, + "loss": 2.5554, + "step": 715 + }, + { + "epoch": 0.15013629691759278, + "grad_norm": 5.658627453339846, + "learning_rate": 1.9979046387226982e-05, + "loss": 2.6371, + "step": 716 + }, + { + "epoch": 0.15034598448312014, + "grad_norm": 5.592230022061712, + "learning_rate": 1.9978899653939105e-05, + "loss": 2.7505, + "step": 717 + }, + { + "epoch": 0.1505556720486475, + "grad_norm": 5.052591492697424, + "learning_rate": 1.9978752409215313e-05, + "loss": 2.4662, + "step": 718 + }, + { + "epoch": 0.1507653596141749, + "grad_norm": 4.63635384173006, + "learning_rate": 1.997860465306315e-05, + "loss": 2.4323, + "step": 719 + }, + { + "epoch": 0.15097504717970225, + "grad_norm": 5.119397576289821, + "learning_rate": 1.9978456385490195e-05, + "loss": 2.6242, + "step": 720 + }, + { + "epoch": 0.15118473474522962, + "grad_norm": 5.048631016593217, + "learning_rate": 1.9978307606504044e-05, + "loss": 2.5475, + "step": 721 + }, + { + "epoch": 0.15139442231075698, + "grad_norm": 6.32075870701772, + "learning_rate": 1.997815831611232e-05, + "loss": 2.568, + "step": 722 + }, + { + "epoch": 0.15160410987628434, + "grad_norm": 5.4321602770296, + "learning_rate": 1.9978008514322677e-05, + "loss": 2.6301, + "step": 723 + }, + { + "epoch": 0.1518137974418117, + "grad_norm": 5.605113035140233, + "learning_rate": 1.997785820114279e-05, + "loss": 2.7584, + "step": 724 + }, + { + "epoch": 0.15202348500733906, + "grad_norm": 6.450430833124885, + "learning_rate": 1.9977707376580365e-05, + "loss": 2.6253, + "step": 725 + }, + { + "epoch": 0.15223317257286642, + "grad_norm": 5.982470803305132, + "learning_rate": 1.997755604064313e-05, + "loss": 2.6636, + "step": 726 + }, + { + "epoch": 0.15244286013839378, + "grad_norm": 4.9337984946437, + "learning_rate": 1.997740419333885e-05, + "loss": 2.378, + "step": 727 + }, + { + "epoch": 0.15265254770392117, + "grad_norm": 4.8238368123520186, + "learning_rate": 1.9977251834675294e-05, + "loss": 2.4251, + "step": 728 + }, + { + "epoch": 0.15286223526944853, + "grad_norm": 5.785317364947026, + "learning_rate": 1.9977098964660274e-05, + "loss": 2.8368, + "step": 729 + }, + { + "epoch": 0.1530719228349759, + "grad_norm": 5.711242094605837, + "learning_rate": 1.997694558330163e-05, + "loss": 2.8107, + "step": 730 + }, + { + "epoch": 0.15328161040050325, + "grad_norm": 5.552121805556651, + "learning_rate": 1.9976791690607226e-05, + "loss": 2.5561, + "step": 731 + }, + { + "epoch": 0.15349129796603062, + "grad_norm": 5.049171025402422, + "learning_rate": 1.997663728658494e-05, + "loss": 2.5043, + "step": 732 + }, + { + "epoch": 0.15370098553155798, + "grad_norm": 5.697573904387599, + "learning_rate": 1.9976482371242687e-05, + "loss": 2.3494, + "step": 733 + }, + { + "epoch": 0.15391067309708534, + "grad_norm": 5.641369840965116, + "learning_rate": 1.997632694458841e-05, + "loss": 2.3422, + "step": 734 + }, + { + "epoch": 0.1541203606626127, + "grad_norm": 5.431146595293463, + "learning_rate": 1.9976171006630077e-05, + "loss": 2.3649, + "step": 735 + }, + { + "epoch": 0.15433004822814006, + "grad_norm": 5.101927798951413, + "learning_rate": 1.9976014557375677e-05, + "loss": 2.4986, + "step": 736 + }, + { + "epoch": 0.15453973579366745, + "grad_norm": 4.923014747073183, + "learning_rate": 1.9975857596833227e-05, + "loss": 2.5067, + "step": 737 + }, + { + "epoch": 0.1547494233591948, + "grad_norm": 4.848774676538115, + "learning_rate": 1.9975700125010773e-05, + "loss": 2.5402, + "step": 738 + }, + { + "epoch": 0.15495911092472217, + "grad_norm": 5.115305746956279, + "learning_rate": 1.9975542141916383e-05, + "loss": 2.6393, + "step": 739 + }, + { + "epoch": 0.15516879849024953, + "grad_norm": 5.349510685162554, + "learning_rate": 1.997538364755816e-05, + "loss": 2.7126, + "step": 740 + }, + { + "epoch": 0.1553784860557769, + "grad_norm": 4.967054520234963, + "learning_rate": 1.9975224641944225e-05, + "loss": 2.5267, + "step": 741 + }, + { + "epoch": 0.15558817362130425, + "grad_norm": 5.263228945948572, + "learning_rate": 1.9975065125082723e-05, + "loss": 2.5071, + "step": 742 + }, + { + "epoch": 0.15579786118683162, + "grad_norm": 4.83917189253647, + "learning_rate": 1.9974905096981835e-05, + "loss": 2.3575, + "step": 743 + }, + { + "epoch": 0.15600754875235898, + "grad_norm": 5.3961003506352085, + "learning_rate": 1.9974744557649756e-05, + "loss": 2.6504, + "step": 744 + }, + { + "epoch": 0.15621723631788634, + "grad_norm": 4.98844314553697, + "learning_rate": 1.9974583507094723e-05, + "loss": 2.6481, + "step": 745 + }, + { + "epoch": 0.15642692388341373, + "grad_norm": 5.809041861839085, + "learning_rate": 1.9974421945324982e-05, + "loss": 2.5328, + "step": 746 + }, + { + "epoch": 0.1566366114489411, + "grad_norm": 5.706972443697373, + "learning_rate": 1.9974259872348815e-05, + "loss": 2.6119, + "step": 747 + }, + { + "epoch": 0.15684629901446845, + "grad_norm": 5.2228963489309095, + "learning_rate": 1.9974097288174535e-05, + "loss": 2.4938, + "step": 748 + }, + { + "epoch": 0.1570559865799958, + "grad_norm": 5.5058972830007, + "learning_rate": 1.9973934192810467e-05, + "loss": 2.665, + "step": 749 + }, + { + "epoch": 0.15726567414552317, + "grad_norm": 6.3055086397332625, + "learning_rate": 1.997377058626497e-05, + "loss": 2.5374, + "step": 750 + }, + { + "epoch": 0.15747536171105053, + "grad_norm": 5.2272593588909, + "learning_rate": 1.9973606468546437e-05, + "loss": 2.5509, + "step": 751 + }, + { + "epoch": 0.1576850492765779, + "grad_norm": 5.2389144271210055, + "learning_rate": 1.997344183966327e-05, + "loss": 2.588, + "step": 752 + }, + { + "epoch": 0.15789473684210525, + "grad_norm": 5.170506045786776, + "learning_rate": 1.9973276699623914e-05, + "loss": 2.3833, + "step": 753 + }, + { + "epoch": 0.15810442440763262, + "grad_norm": 6.349661608621654, + "learning_rate": 1.9973111048436827e-05, + "loss": 2.5797, + "step": 754 + }, + { + "epoch": 0.15831411197316, + "grad_norm": 5.467234597564719, + "learning_rate": 1.99729448861105e-05, + "loss": 2.5061, + "step": 755 + }, + { + "epoch": 0.15852379953868737, + "grad_norm": 5.788314282563359, + "learning_rate": 1.9972778212653455e-05, + "loss": 2.5789, + "step": 756 + }, + { + "epoch": 0.15873348710421473, + "grad_norm": 5.766248131111242, + "learning_rate": 1.9972611028074225e-05, + "loss": 2.6302, + "step": 757 + }, + { + "epoch": 0.1589431746697421, + "grad_norm": 5.851124746955546, + "learning_rate": 1.997244333238139e-05, + "loss": 2.5716, + "step": 758 + }, + { + "epoch": 0.15915286223526945, + "grad_norm": 5.647230932231138, + "learning_rate": 1.9972275125583527e-05, + "loss": 2.4891, + "step": 759 + }, + { + "epoch": 0.1593625498007968, + "grad_norm": 5.284928570683543, + "learning_rate": 1.9972106407689278e-05, + "loss": 2.3549, + "step": 760 + }, + { + "epoch": 0.15957223736632417, + "grad_norm": 6.82976176233091, + "learning_rate": 1.9971937178707278e-05, + "loss": 2.451, + "step": 761 + }, + { + "epoch": 0.15978192493185153, + "grad_norm": 5.7689969603794085, + "learning_rate": 1.99717674386462e-05, + "loss": 2.2127, + "step": 762 + }, + { + "epoch": 0.1599916124973789, + "grad_norm": 5.3713345347098755, + "learning_rate": 1.9971597187514743e-05, + "loss": 2.4216, + "step": 763 + }, + { + "epoch": 0.16020130006290628, + "grad_norm": 5.121552622009936, + "learning_rate": 1.9971426425321637e-05, + "loss": 2.4033, + "step": 764 + }, + { + "epoch": 0.16041098762843364, + "grad_norm": 5.1771052992673425, + "learning_rate": 1.9971255152075635e-05, + "loss": 2.2227, + "step": 765 + }, + { + "epoch": 0.160620675193961, + "grad_norm": 5.751104264210732, + "learning_rate": 1.997108336778551e-05, + "loss": 2.562, + "step": 766 + }, + { + "epoch": 0.16083036275948837, + "grad_norm": 6.41324399439957, + "learning_rate": 1.997091107246007e-05, + "loss": 2.8024, + "step": 767 + }, + { + "epoch": 0.16104005032501573, + "grad_norm": 6.246885326610967, + "learning_rate": 1.9970738266108142e-05, + "loss": 2.7224, + "step": 768 + }, + { + "epoch": 0.1612497378905431, + "grad_norm": 5.5788167559955495, + "learning_rate": 1.9970564948738585e-05, + "loss": 2.5663, + "step": 769 + }, + { + "epoch": 0.16145942545607045, + "grad_norm": 6.103869201794778, + "learning_rate": 1.997039112036028e-05, + "loss": 2.6032, + "step": 770 + }, + { + "epoch": 0.1616691130215978, + "grad_norm": 5.093834157772151, + "learning_rate": 1.997021678098214e-05, + "loss": 2.3786, + "step": 771 + }, + { + "epoch": 0.16187880058712517, + "grad_norm": 6.025378098467104, + "learning_rate": 1.9970041930613096e-05, + "loss": 2.4202, + "step": 772 + }, + { + "epoch": 0.16208848815265256, + "grad_norm": 4.9446767239003275, + "learning_rate": 1.996986656926211e-05, + "loss": 2.6345, + "step": 773 + }, + { + "epoch": 0.16229817571817992, + "grad_norm": 5.151429866391372, + "learning_rate": 1.9969690696938173e-05, + "loss": 2.6276, + "step": 774 + }, + { + "epoch": 0.16250786328370728, + "grad_norm": 4.6507668918727285, + "learning_rate": 1.9969514313650294e-05, + "loss": 2.5375, + "step": 775 + }, + { + "epoch": 0.16271755084923464, + "grad_norm": 5.479062357904265, + "learning_rate": 1.9969337419407517e-05, + "loss": 2.6203, + "step": 776 + }, + { + "epoch": 0.162927238414762, + "grad_norm": 5.5366540558457285, + "learning_rate": 1.9969160014218904e-05, + "loss": 2.8279, + "step": 777 + }, + { + "epoch": 0.16313692598028937, + "grad_norm": 6.455368237461392, + "learning_rate": 1.9968982098093553e-05, + "loss": 2.7342, + "step": 778 + }, + { + "epoch": 0.16334661354581673, + "grad_norm": 6.114025923129414, + "learning_rate": 1.9968803671040574e-05, + "loss": 2.7242, + "step": 779 + }, + { + "epoch": 0.1635563011113441, + "grad_norm": 5.165162546701589, + "learning_rate": 1.9968624733069122e-05, + "loss": 2.7608, + "step": 780 + }, + { + "epoch": 0.16376598867687145, + "grad_norm": 5.183951403439922, + "learning_rate": 1.996844528418836e-05, + "loss": 2.6442, + "step": 781 + }, + { + "epoch": 0.16397567624239884, + "grad_norm": 4.932360432093259, + "learning_rate": 1.996826532440749e-05, + "loss": 2.3996, + "step": 782 + }, + { + "epoch": 0.1641853638079262, + "grad_norm": 6.427461911630745, + "learning_rate": 1.9968084853735728e-05, + "loss": 2.6695, + "step": 783 + }, + { + "epoch": 0.16439505137345356, + "grad_norm": 5.839031395817843, + "learning_rate": 1.9967903872182335e-05, + "loss": 2.8268, + "step": 784 + }, + { + "epoch": 0.16460473893898092, + "grad_norm": 4.316046016427989, + "learning_rate": 1.9967722379756575e-05, + "loss": 2.4485, + "step": 785 + }, + { + "epoch": 0.16481442650450828, + "grad_norm": 5.17689136727919, + "learning_rate": 1.996754037646776e-05, + "loss": 2.4977, + "step": 786 + }, + { + "epoch": 0.16502411407003564, + "grad_norm": 4.984608923869862, + "learning_rate": 1.9967357862325206e-05, + "loss": 2.2972, + "step": 787 + }, + { + "epoch": 0.165233801635563, + "grad_norm": 5.764262154632816, + "learning_rate": 1.9967174837338282e-05, + "loss": 2.6403, + "step": 788 + }, + { + "epoch": 0.16544348920109037, + "grad_norm": 5.531201758921919, + "learning_rate": 1.9966991301516356e-05, + "loss": 2.6415, + "step": 789 + }, + { + "epoch": 0.16565317676661773, + "grad_norm": 5.9032219473923835, + "learning_rate": 1.9966807254868835e-05, + "loss": 2.3596, + "step": 790 + }, + { + "epoch": 0.16586286433214512, + "grad_norm": 5.2307080735103195, + "learning_rate": 1.9966622697405163e-05, + "loss": 2.4864, + "step": 791 + }, + { + "epoch": 0.16607255189767248, + "grad_norm": 5.3824415473570175, + "learning_rate": 1.9966437629134786e-05, + "loss": 2.6149, + "step": 792 + }, + { + "epoch": 0.16628223946319984, + "grad_norm": 5.589694336777668, + "learning_rate": 1.9966252050067197e-05, + "loss": 2.5896, + "step": 793 + }, + { + "epoch": 0.1664919270287272, + "grad_norm": 6.493288853507978, + "learning_rate": 1.9966065960211907e-05, + "loss": 2.4996, + "step": 794 + }, + { + "epoch": 0.16670161459425456, + "grad_norm": 5.89449672328628, + "learning_rate": 1.996587935957845e-05, + "loss": 2.376, + "step": 795 + }, + { + "epoch": 0.16691130215978192, + "grad_norm": 5.178448822931242, + "learning_rate": 1.996569224817639e-05, + "loss": 2.5124, + "step": 796 + }, + { + "epoch": 0.16712098972530928, + "grad_norm": 5.314420049466998, + "learning_rate": 1.9965504626015314e-05, + "loss": 2.5139, + "step": 797 + }, + { + "epoch": 0.16733067729083664, + "grad_norm": 4.965540255561048, + "learning_rate": 1.9965316493104845e-05, + "loss": 2.4432, + "step": 798 + }, + { + "epoch": 0.167540364856364, + "grad_norm": 5.9976226159107116, + "learning_rate": 1.9965127849454625e-05, + "loss": 2.6128, + "step": 799 + }, + { + "epoch": 0.1677500524218914, + "grad_norm": 5.6868698522966, + "learning_rate": 1.9964938695074315e-05, + "loss": 2.3999, + "step": 800 + }, + { + "epoch": 0.16795973998741875, + "grad_norm": 5.068783438771577, + "learning_rate": 1.9964749029973616e-05, + "loss": 2.6127, + "step": 801 + }, + { + "epoch": 0.16816942755294612, + "grad_norm": 6.136687509953793, + "learning_rate": 1.9964558854162244e-05, + "loss": 2.8011, + "step": 802 + }, + { + "epoch": 0.16837911511847348, + "grad_norm": 4.800340421160969, + "learning_rate": 1.996436816764995e-05, + "loss": 2.5399, + "step": 803 + }, + { + "epoch": 0.16858880268400084, + "grad_norm": 5.5470642039105975, + "learning_rate": 1.9964176970446504e-05, + "loss": 2.8207, + "step": 804 + }, + { + "epoch": 0.1687984902495282, + "grad_norm": 6.641243790799172, + "learning_rate": 1.9963985262561704e-05, + "loss": 3.0149, + "step": 805 + }, + { + "epoch": 0.16900817781505556, + "grad_norm": 5.2599864963628535, + "learning_rate": 1.9963793044005382e-05, + "loss": 2.7838, + "step": 806 + }, + { + "epoch": 0.16921786538058292, + "grad_norm": 5.551203360839708, + "learning_rate": 1.9963600314787384e-05, + "loss": 2.3039, + "step": 807 + }, + { + "epoch": 0.16942755294611028, + "grad_norm": 5.005656047264504, + "learning_rate": 1.9963407074917586e-05, + "loss": 2.3646, + "step": 808 + }, + { + "epoch": 0.16963724051163767, + "grad_norm": 5.096147192741019, + "learning_rate": 1.99632133244059e-05, + "loss": 2.5875, + "step": 809 + }, + { + "epoch": 0.16984692807716503, + "grad_norm": 6.304971806711211, + "learning_rate": 1.9963019063262245e-05, + "loss": 2.3647, + "step": 810 + }, + { + "epoch": 0.1700566156426924, + "grad_norm": 4.689418131258043, + "learning_rate": 1.9962824291496584e-05, + "loss": 2.4458, + "step": 811 + }, + { + "epoch": 0.17026630320821975, + "grad_norm": 5.679181828413589, + "learning_rate": 1.9962629009118903e-05, + "loss": 2.4693, + "step": 812 + }, + { + "epoch": 0.17047599077374712, + "grad_norm": 5.756522383170872, + "learning_rate": 1.9962433216139204e-05, + "loss": 2.4949, + "step": 813 + }, + { + "epoch": 0.17068567833927448, + "grad_norm": 5.962087484363417, + "learning_rate": 1.9962236912567522e-05, + "loss": 2.3043, + "step": 814 + }, + { + "epoch": 0.17089536590480184, + "grad_norm": 5.160052152709169, + "learning_rate": 1.9962040098413918e-05, + "loss": 2.428, + "step": 815 + }, + { + "epoch": 0.1711050534703292, + "grad_norm": 5.406280541609342, + "learning_rate": 1.9961842773688484e-05, + "loss": 2.4791, + "step": 816 + }, + { + "epoch": 0.17131474103585656, + "grad_norm": 4.734733525477293, + "learning_rate": 1.996164493840133e-05, + "loss": 2.5014, + "step": 817 + }, + { + "epoch": 0.17152442860138395, + "grad_norm": 4.943764561837859, + "learning_rate": 1.9961446592562594e-05, + "loss": 2.3089, + "step": 818 + }, + { + "epoch": 0.1717341161669113, + "grad_norm": 5.5751813653506535, + "learning_rate": 1.9961247736182443e-05, + "loss": 2.5656, + "step": 819 + }, + { + "epoch": 0.17194380373243867, + "grad_norm": 5.13687644705553, + "learning_rate": 1.9961048369271066e-05, + "loss": 2.4772, + "step": 820 + }, + { + "epoch": 0.17215349129796603, + "grad_norm": 6.675941699590338, + "learning_rate": 1.9960848491838688e-05, + "loss": 2.7695, + "step": 821 + }, + { + "epoch": 0.1723631788634934, + "grad_norm": 4.883406365901591, + "learning_rate": 1.9960648103895548e-05, + "loss": 2.3557, + "step": 822 + }, + { + "epoch": 0.17257286642902075, + "grad_norm": 4.94442909495027, + "learning_rate": 1.9960447205451915e-05, + "loss": 2.4459, + "step": 823 + }, + { + "epoch": 0.17278255399454812, + "grad_norm": 5.323802519087413, + "learning_rate": 1.9960245796518086e-05, + "loss": 2.4145, + "step": 824 + }, + { + "epoch": 0.17299224156007548, + "grad_norm": 4.997228864163945, + "learning_rate": 1.996004387710439e-05, + "loss": 2.4581, + "step": 825 + }, + { + "epoch": 0.17320192912560284, + "grad_norm": 5.283336329058954, + "learning_rate": 1.9959841447221165e-05, + "loss": 2.7089, + "step": 826 + }, + { + "epoch": 0.17341161669113023, + "grad_norm": 5.237184579351533, + "learning_rate": 1.9959638506878795e-05, + "loss": 2.5239, + "step": 827 + }, + { + "epoch": 0.1736213042566576, + "grad_norm": 5.65780162637725, + "learning_rate": 1.9959435056087676e-05, + "loss": 2.3824, + "step": 828 + }, + { + "epoch": 0.17383099182218495, + "grad_norm": 4.827351960582513, + "learning_rate": 1.9959231094858236e-05, + "loss": 2.4209, + "step": 829 + }, + { + "epoch": 0.1740406793877123, + "grad_norm": 5.59764664902699, + "learning_rate": 1.995902662320093e-05, + "loss": 2.5681, + "step": 830 + }, + { + "epoch": 0.17425036695323967, + "grad_norm": 5.565930028253618, + "learning_rate": 1.9958821641126234e-05, + "loss": 2.6214, + "step": 831 + }, + { + "epoch": 0.17446005451876703, + "grad_norm": 5.1865682653861205, + "learning_rate": 1.995861614864466e-05, + "loss": 2.3355, + "step": 832 + }, + { + "epoch": 0.1746697420842944, + "grad_norm": 5.587511254245669, + "learning_rate": 1.9958410145766733e-05, + "loss": 2.5303, + "step": 833 + }, + { + "epoch": 0.17487942964982175, + "grad_norm": 6.255446060394876, + "learning_rate": 1.995820363250302e-05, + "loss": 2.4386, + "step": 834 + }, + { + "epoch": 0.17508911721534912, + "grad_norm": 5.5566095073076704, + "learning_rate": 1.995799660886409e-05, + "loss": 2.3336, + "step": 835 + }, + { + "epoch": 0.1752988047808765, + "grad_norm": 5.312228334366878, + "learning_rate": 1.9957789074860566e-05, + "loss": 2.7252, + "step": 836 + }, + { + "epoch": 0.17550849234640387, + "grad_norm": 5.209814225019706, + "learning_rate": 1.9957581030503084e-05, + "loss": 2.2494, + "step": 837 + }, + { + "epoch": 0.17571817991193123, + "grad_norm": 5.809157066825737, + "learning_rate": 1.9957372475802303e-05, + "loss": 2.3917, + "step": 838 + }, + { + "epoch": 0.1759278674774586, + "grad_norm": 4.600111546205428, + "learning_rate": 1.995716341076891e-05, + "loss": 2.4483, + "step": 839 + }, + { + "epoch": 0.17613755504298595, + "grad_norm": 4.928057485084255, + "learning_rate": 1.9956953835413624e-05, + "loss": 2.5772, + "step": 840 + }, + { + "epoch": 0.1763472426085133, + "grad_norm": 5.624418635591188, + "learning_rate": 1.9956743749747184e-05, + "loss": 2.5422, + "step": 841 + }, + { + "epoch": 0.17655693017404067, + "grad_norm": 5.0933223470612665, + "learning_rate": 1.9956533153780358e-05, + "loss": 2.2941, + "step": 842 + }, + { + "epoch": 0.17676661773956803, + "grad_norm": 6.536798514231347, + "learning_rate": 1.9956322047523936e-05, + "loss": 2.6009, + "step": 843 + }, + { + "epoch": 0.17697630530509542, + "grad_norm": 5.450621074969645, + "learning_rate": 1.9956110430988745e-05, + "loss": 2.4368, + "step": 844 + }, + { + "epoch": 0.17718599287062278, + "grad_norm": 6.719090216084803, + "learning_rate": 1.9955898304185623e-05, + "loss": 2.676, + "step": 845 + }, + { + "epoch": 0.17739568043615014, + "grad_norm": 5.468604176484468, + "learning_rate": 1.9955685667125448e-05, + "loss": 2.5359, + "step": 846 + }, + { + "epoch": 0.1776053680016775, + "grad_norm": 5.162680171464789, + "learning_rate": 1.9955472519819114e-05, + "loss": 2.3793, + "step": 847 + }, + { + "epoch": 0.17781505556720487, + "grad_norm": 4.985326564587904, + "learning_rate": 1.995525886227755e-05, + "loss": 2.4282, + "step": 848 + }, + { + "epoch": 0.17802474313273223, + "grad_norm": 5.204117524216867, + "learning_rate": 1.9955044694511697e-05, + "loss": 2.5916, + "step": 849 + }, + { + "epoch": 0.1782344306982596, + "grad_norm": 4.989224401120582, + "learning_rate": 1.995483001653254e-05, + "loss": 2.6525, + "step": 850 + }, + { + "epoch": 0.17844411826378695, + "grad_norm": 5.6958176671058665, + "learning_rate": 1.995461482835108e-05, + "loss": 2.6081, + "step": 851 + }, + { + "epoch": 0.1786538058293143, + "grad_norm": 5.030585029222048, + "learning_rate": 1.995439912997834e-05, + "loss": 2.6039, + "step": 852 + }, + { + "epoch": 0.1788634933948417, + "grad_norm": 4.448185895967703, + "learning_rate": 1.9954182921425388e-05, + "loss": 2.1762, + "step": 853 + }, + { + "epoch": 0.17907318096036906, + "grad_norm": 5.383344612894652, + "learning_rate": 1.995396620270329e-05, + "loss": 2.5764, + "step": 854 + }, + { + "epoch": 0.17928286852589642, + "grad_norm": 5.031615038551207, + "learning_rate": 1.995374897382316e-05, + "loss": 2.2982, + "step": 855 + }, + { + "epoch": 0.17949255609142378, + "grad_norm": 5.18290878826544, + "learning_rate": 1.9953531234796133e-05, + "loss": 2.1095, + "step": 856 + }, + { + "epoch": 0.17970224365695114, + "grad_norm": 5.386646785836298, + "learning_rate": 1.9953312985633366e-05, + "loss": 2.3745, + "step": 857 + }, + { + "epoch": 0.1799119312224785, + "grad_norm": 5.608239066675773, + "learning_rate": 1.9953094226346046e-05, + "loss": 2.4223, + "step": 858 + }, + { + "epoch": 0.18012161878800587, + "grad_norm": 5.738383305715215, + "learning_rate": 1.995287495694538e-05, + "loss": 2.6354, + "step": 859 + }, + { + "epoch": 0.18033130635353323, + "grad_norm": 5.297752061779667, + "learning_rate": 1.9952655177442618e-05, + "loss": 2.4378, + "step": 860 + }, + { + "epoch": 0.1805409939190606, + "grad_norm": 5.815973839977065, + "learning_rate": 1.9952434887849014e-05, + "loss": 2.4679, + "step": 861 + }, + { + "epoch": 0.18075068148458798, + "grad_norm": 5.8559471541561345, + "learning_rate": 1.9952214088175855e-05, + "loss": 2.823, + "step": 862 + }, + { + "epoch": 0.18096036905011534, + "grad_norm": 6.104384028798239, + "learning_rate": 1.995199277843447e-05, + "loss": 2.5181, + "step": 863 + }, + { + "epoch": 0.1811700566156427, + "grad_norm": 4.549457994128656, + "learning_rate": 1.995177095863619e-05, + "loss": 2.3462, + "step": 864 + }, + { + "epoch": 0.18137974418117006, + "grad_norm": 5.436138788899165, + "learning_rate": 1.995154862879239e-05, + "loss": 2.755, + "step": 865 + }, + { + "epoch": 0.18158943174669742, + "grad_norm": 5.472479469383391, + "learning_rate": 1.9951325788914457e-05, + "loss": 2.5968, + "step": 866 + }, + { + "epoch": 0.18179911931222478, + "grad_norm": 4.8993628934332785, + "learning_rate": 1.9951102439013824e-05, + "loss": 2.6024, + "step": 867 + }, + { + "epoch": 0.18200880687775214, + "grad_norm": 5.601454966702046, + "learning_rate": 1.9950878579101927e-05, + "loss": 2.5266, + "step": 868 + }, + { + "epoch": 0.1822184944432795, + "grad_norm": 5.54926763211863, + "learning_rate": 1.9950654209190246e-05, + "loss": 2.6516, + "step": 869 + }, + { + "epoch": 0.18242818200880687, + "grad_norm": 4.975406708451378, + "learning_rate": 1.995042932929028e-05, + "loss": 2.3917, + "step": 870 + }, + { + "epoch": 0.18263786957433426, + "grad_norm": 5.909001873186139, + "learning_rate": 1.995020393941355e-05, + "loss": 2.5046, + "step": 871 + }, + { + "epoch": 0.18284755713986162, + "grad_norm": 5.104492440651537, + "learning_rate": 1.994997803957161e-05, + "loss": 2.4308, + "step": 872 + }, + { + "epoch": 0.18305724470538898, + "grad_norm": 6.2011163694045255, + "learning_rate": 1.9949751629776043e-05, + "loss": 2.7716, + "step": 873 + }, + { + "epoch": 0.18326693227091634, + "grad_norm": 4.739716649909645, + "learning_rate": 1.9949524710038445e-05, + "loss": 2.4716, + "step": 874 + }, + { + "epoch": 0.1834766198364437, + "grad_norm": 6.1425757658306495, + "learning_rate": 1.9949297280370444e-05, + "loss": 2.4702, + "step": 875 + }, + { + "epoch": 0.18368630740197106, + "grad_norm": 5.5573572440882755, + "learning_rate": 1.994906934078371e-05, + "loss": 2.4642, + "step": 876 + }, + { + "epoch": 0.18389599496749842, + "grad_norm": 4.828397823101376, + "learning_rate": 1.994884089128991e-05, + "loss": 2.4723, + "step": 877 + }, + { + "epoch": 0.18410568253302578, + "grad_norm": 5.632723011140619, + "learning_rate": 1.9948611931900764e-05, + "loss": 2.4549, + "step": 878 + }, + { + "epoch": 0.18431537009855314, + "grad_norm": 5.915021700626544, + "learning_rate": 1.9948382462627997e-05, + "loss": 2.4802, + "step": 879 + }, + { + "epoch": 0.18452505766408053, + "grad_norm": 5.192333046317903, + "learning_rate": 1.9948152483483376e-05, + "loss": 2.3059, + "step": 880 + }, + { + "epoch": 0.1847347452296079, + "grad_norm": 4.854540141867747, + "learning_rate": 1.9947921994478686e-05, + "loss": 2.4742, + "step": 881 + }, + { + "epoch": 0.18494443279513526, + "grad_norm": 4.829050196113135, + "learning_rate": 1.994769099562574e-05, + "loss": 2.4339, + "step": 882 + }, + { + "epoch": 0.18515412036066262, + "grad_norm": 6.015103895046883, + "learning_rate": 1.9947459486936378e-05, + "loss": 2.728, + "step": 883 + }, + { + "epoch": 0.18536380792618998, + "grad_norm": 4.768495634323849, + "learning_rate": 1.9947227468422462e-05, + "loss": 2.4327, + "step": 884 + }, + { + "epoch": 0.18557349549171734, + "grad_norm": 6.47748341817548, + "learning_rate": 1.994699494009589e-05, + "loss": 2.8254, + "step": 885 + }, + { + "epoch": 0.1857831830572447, + "grad_norm": 5.674268388327525, + "learning_rate": 1.994676190196857e-05, + "loss": 2.6363, + "step": 886 + }, + { + "epoch": 0.18599287062277206, + "grad_norm": 4.93142945507908, + "learning_rate": 1.9946528354052453e-05, + "loss": 2.3969, + "step": 887 + }, + { + "epoch": 0.18620255818829942, + "grad_norm": 5.745896749665753, + "learning_rate": 1.9946294296359507e-05, + "loss": 2.5042, + "step": 888 + }, + { + "epoch": 0.1864122457538268, + "grad_norm": 4.910569525334854, + "learning_rate": 1.9946059728901724e-05, + "loss": 2.4622, + "step": 889 + }, + { + "epoch": 0.18662193331935417, + "grad_norm": 6.586576062035691, + "learning_rate": 1.9945824651691134e-05, + "loss": 2.5314, + "step": 890 + }, + { + "epoch": 0.18683162088488153, + "grad_norm": 5.079637117467489, + "learning_rate": 1.994558906473978e-05, + "loss": 2.3744, + "step": 891 + }, + { + "epoch": 0.1870413084504089, + "grad_norm": 5.100999598825927, + "learning_rate": 1.9945352968059735e-05, + "loss": 2.3114, + "step": 892 + }, + { + "epoch": 0.18725099601593626, + "grad_norm": 5.522657260137167, + "learning_rate": 1.9945116361663102e-05, + "loss": 2.5035, + "step": 893 + }, + { + "epoch": 0.18746068358146362, + "grad_norm": 5.004913680748999, + "learning_rate": 1.9944879245562006e-05, + "loss": 2.4789, + "step": 894 + }, + { + "epoch": 0.18767037114699098, + "grad_norm": 5.080222853740604, + "learning_rate": 1.9944641619768598e-05, + "loss": 2.483, + "step": 895 + }, + { + "epoch": 0.18788005871251834, + "grad_norm": 5.887346327906223, + "learning_rate": 1.9944403484295062e-05, + "loss": 2.4791, + "step": 896 + }, + { + "epoch": 0.1880897462780457, + "grad_norm": 5.448747530767844, + "learning_rate": 1.9944164839153595e-05, + "loss": 2.5861, + "step": 897 + }, + { + "epoch": 0.1882994338435731, + "grad_norm": 5.149747557197856, + "learning_rate": 1.994392568435644e-05, + "loss": 2.616, + "step": 898 + }, + { + "epoch": 0.18850912140910045, + "grad_norm": 5.4248840812071775, + "learning_rate": 1.9943686019915844e-05, + "loss": 2.5144, + "step": 899 + }, + { + "epoch": 0.1887188089746278, + "grad_norm": 5.019211833713546, + "learning_rate": 1.994344584584409e-05, + "loss": 2.4441, + "step": 900 + }, + { + "epoch": 0.18892849654015517, + "grad_norm": 5.18434435897594, + "learning_rate": 1.9943205162153493e-05, + "loss": 2.5536, + "step": 901 + }, + { + "epoch": 0.18913818410568253, + "grad_norm": 4.8831296119713326, + "learning_rate": 1.9942963968856384e-05, + "loss": 2.3812, + "step": 902 + }, + { + "epoch": 0.1893478716712099, + "grad_norm": 5.684986224455571, + "learning_rate": 1.9942722265965128e-05, + "loss": 2.5858, + "step": 903 + }, + { + "epoch": 0.18955755923673726, + "grad_norm": 5.700698450628685, + "learning_rate": 1.9942480053492112e-05, + "loss": 2.4186, + "step": 904 + }, + { + "epoch": 0.18976724680226462, + "grad_norm": 4.547966659110109, + "learning_rate": 1.9942237331449747e-05, + "loss": 2.2892, + "step": 905 + }, + { + "epoch": 0.18997693436779198, + "grad_norm": 4.521301414957022, + "learning_rate": 1.9941994099850473e-05, + "loss": 2.4115, + "step": 906 + }, + { + "epoch": 0.19018662193331937, + "grad_norm": 4.796428680042201, + "learning_rate": 1.9941750358706763e-05, + "loss": 2.3414, + "step": 907 + }, + { + "epoch": 0.19039630949884673, + "grad_norm": 6.0685827688665235, + "learning_rate": 1.9941506108031102e-05, + "loss": 2.4847, + "step": 908 + }, + { + "epoch": 0.1906059970643741, + "grad_norm": 5.1457261008257404, + "learning_rate": 1.994126134783601e-05, + "loss": 2.6251, + "step": 909 + }, + { + "epoch": 0.19081568462990145, + "grad_norm": 5.239952560216145, + "learning_rate": 1.994101607813403e-05, + "loss": 2.4506, + "step": 910 + }, + { + "epoch": 0.1910253721954288, + "grad_norm": 6.112448725212283, + "learning_rate": 1.9940770298937736e-05, + "loss": 2.5548, + "step": 911 + }, + { + "epoch": 0.19123505976095617, + "grad_norm": 5.191368930630846, + "learning_rate": 1.994052401025972e-05, + "loss": 2.2849, + "step": 912 + }, + { + "epoch": 0.19144474732648353, + "grad_norm": 5.377069654303603, + "learning_rate": 1.994027721211261e-05, + "loss": 2.2073, + "step": 913 + }, + { + "epoch": 0.1916544348920109, + "grad_norm": 5.269656183035728, + "learning_rate": 1.9940029904509054e-05, + "loss": 2.5896, + "step": 914 + }, + { + "epoch": 0.19186412245753826, + "grad_norm": 4.582472713273976, + "learning_rate": 1.9939782087461724e-05, + "loss": 2.3606, + "step": 915 + }, + { + "epoch": 0.19207381002306564, + "grad_norm": 6.011572748683017, + "learning_rate": 1.9939533760983322e-05, + "loss": 2.3621, + "step": 916 + }, + { + "epoch": 0.192283497588593, + "grad_norm": 4.860293072710145, + "learning_rate": 1.9939284925086576e-05, + "loss": 2.4226, + "step": 917 + }, + { + "epoch": 0.19249318515412037, + "grad_norm": 6.224550963894682, + "learning_rate": 1.9939035579784236e-05, + "loss": 2.4929, + "step": 918 + }, + { + "epoch": 0.19270287271964773, + "grad_norm": 4.734580779678761, + "learning_rate": 1.993878572508909e-05, + "loss": 2.2896, + "step": 919 + }, + { + "epoch": 0.1929125602851751, + "grad_norm": 5.821405318413932, + "learning_rate": 1.993853536101393e-05, + "loss": 2.631, + "step": 920 + }, + { + "epoch": 0.19312224785070245, + "grad_norm": 5.630603319863494, + "learning_rate": 1.9938284487571604e-05, + "loss": 2.3614, + "step": 921 + }, + { + "epoch": 0.1933319354162298, + "grad_norm": 5.0940768877114895, + "learning_rate": 1.993803310477495e-05, + "loss": 2.204, + "step": 922 + }, + { + "epoch": 0.19354162298175717, + "grad_norm": 4.922087843773558, + "learning_rate": 1.993778121263687e-05, + "loss": 2.6305, + "step": 923 + }, + { + "epoch": 0.19375131054728453, + "grad_norm": 5.201118970695738, + "learning_rate": 1.9937528811170267e-05, + "loss": 2.6807, + "step": 924 + }, + { + "epoch": 0.19396099811281192, + "grad_norm": 5.338934484350384, + "learning_rate": 1.9937275900388076e-05, + "loss": 2.6133, + "step": 925 + }, + { + "epoch": 0.19417068567833928, + "grad_norm": 5.395068409146181, + "learning_rate": 1.9937022480303256e-05, + "loss": 2.5248, + "step": 926 + }, + { + "epoch": 0.19438037324386664, + "grad_norm": 5.4413205997191705, + "learning_rate": 1.9936768550928804e-05, + "loss": 2.3095, + "step": 927 + }, + { + "epoch": 0.194590060809394, + "grad_norm": 4.60626361060752, + "learning_rate": 1.9936514112277727e-05, + "loss": 2.4582, + "step": 928 + }, + { + "epoch": 0.19479974837492137, + "grad_norm": 4.91685168909383, + "learning_rate": 1.9936259164363066e-05, + "loss": 2.6594, + "step": 929 + }, + { + "epoch": 0.19500943594044873, + "grad_norm": 4.854375241178789, + "learning_rate": 1.9936003707197893e-05, + "loss": 2.116, + "step": 930 + }, + { + "epoch": 0.1952191235059761, + "grad_norm": 5.3419706462791074, + "learning_rate": 1.9935747740795295e-05, + "loss": 2.2324, + "step": 931 + }, + { + "epoch": 0.19542881107150345, + "grad_norm": 5.256376420547868, + "learning_rate": 1.9935491265168392e-05, + "loss": 2.5482, + "step": 932 + }, + { + "epoch": 0.1956384986370308, + "grad_norm": 4.901918345493721, + "learning_rate": 1.9935234280330333e-05, + "loss": 2.3966, + "step": 933 + }, + { + "epoch": 0.1958481862025582, + "grad_norm": 5.317936673143983, + "learning_rate": 1.993497678629428e-05, + "loss": 2.5429, + "step": 934 + }, + { + "epoch": 0.19605787376808556, + "grad_norm": 5.315767067422777, + "learning_rate": 1.993471878307344e-05, + "loss": 2.1342, + "step": 935 + }, + { + "epoch": 0.19626756133361292, + "grad_norm": 5.876888515032318, + "learning_rate": 1.9934460270681028e-05, + "loss": 2.1805, + "step": 936 + }, + { + "epoch": 0.19647724889914028, + "grad_norm": 6.06976386362381, + "learning_rate": 1.99342012491303e-05, + "loss": 2.7713, + "step": 937 + }, + { + "epoch": 0.19668693646466764, + "grad_norm": 5.042156333179805, + "learning_rate": 1.9933941718434523e-05, + "loss": 2.6455, + "step": 938 + }, + { + "epoch": 0.196896624030195, + "grad_norm": 5.373087123649382, + "learning_rate": 1.9933681678607005e-05, + "loss": 2.4848, + "step": 939 + }, + { + "epoch": 0.19710631159572237, + "grad_norm": 4.8283300788311205, + "learning_rate": 1.9933421129661075e-05, + "loss": 2.5254, + "step": 940 + }, + { + "epoch": 0.19731599916124973, + "grad_norm": 5.615108274453707, + "learning_rate": 1.993316007161008e-05, + "loss": 2.6872, + "step": 941 + }, + { + "epoch": 0.1975256867267771, + "grad_norm": 5.632846177763661, + "learning_rate": 1.9932898504467407e-05, + "loss": 2.2772, + "step": 942 + }, + { + "epoch": 0.19773537429230448, + "grad_norm": 4.9900070371932355, + "learning_rate": 1.9932636428246456e-05, + "loss": 2.3392, + "step": 943 + }, + { + "epoch": 0.19794506185783184, + "grad_norm": 5.332409359202788, + "learning_rate": 1.993237384296066e-05, + "loss": 2.5759, + "step": 944 + }, + { + "epoch": 0.1981547494233592, + "grad_norm": 5.618227414596692, + "learning_rate": 1.9932110748623476e-05, + "loss": 2.4152, + "step": 945 + }, + { + "epoch": 0.19836443698888656, + "grad_norm": 5.2710668257843665, + "learning_rate": 1.9931847145248393e-05, + "loss": 2.5688, + "step": 946 + }, + { + "epoch": 0.19857412455441392, + "grad_norm": 4.813334400142518, + "learning_rate": 1.9931583032848918e-05, + "loss": 2.3557, + "step": 947 + }, + { + "epoch": 0.19878381211994128, + "grad_norm": 5.216202596722822, + "learning_rate": 1.9931318411438586e-05, + "loss": 2.4534, + "step": 948 + }, + { + "epoch": 0.19899349968546864, + "grad_norm": 5.175036990798288, + "learning_rate": 1.9931053281030957e-05, + "loss": 2.3328, + "step": 949 + }, + { + "epoch": 0.199203187250996, + "grad_norm": 5.29469629139264, + "learning_rate": 1.9930787641639628e-05, + "loss": 2.4553, + "step": 950 + }, + { + "epoch": 0.19941287481652337, + "grad_norm": 4.6110997078811735, + "learning_rate": 1.9930521493278205e-05, + "loss": 2.4536, + "step": 951 + }, + { + "epoch": 0.19962256238205076, + "grad_norm": 5.529826725346696, + "learning_rate": 1.9930254835960335e-05, + "loss": 2.3811, + "step": 952 + }, + { + "epoch": 0.19983224994757812, + "grad_norm": 5.170745811515311, + "learning_rate": 1.992998766969968e-05, + "loss": 2.4378, + "step": 953 + }, + { + "epoch": 0.20004193751310548, + "grad_norm": 4.459697190515876, + "learning_rate": 1.9929719994509934e-05, + "loss": 2.2087, + "step": 954 + }, + { + "epoch": 0.20025162507863284, + "grad_norm": 5.618558737353918, + "learning_rate": 1.9929451810404816e-05, + "loss": 2.3138, + "step": 955 + }, + { + "epoch": 0.2004613126441602, + "grad_norm": 5.088797994398359, + "learning_rate": 1.992918311739807e-05, + "loss": 2.4355, + "step": 956 + }, + { + "epoch": 0.20067100020968756, + "grad_norm": 5.130162001818529, + "learning_rate": 1.9928913915503473e-05, + "loss": 1.9568, + "step": 957 + }, + { + "epoch": 0.20088068777521492, + "grad_norm": 6.77220147315865, + "learning_rate": 1.992864420473481e-05, + "loss": 2.1486, + "step": 958 + }, + { + "epoch": 0.20109037534074228, + "grad_norm": 5.489317132045333, + "learning_rate": 1.9928373985105914e-05, + "loss": 2.4805, + "step": 959 + }, + { + "epoch": 0.20130006290626964, + "grad_norm": 5.65637728609791, + "learning_rate": 1.9928103256630632e-05, + "loss": 2.7277, + "step": 960 + }, + { + "epoch": 0.20150975047179703, + "grad_norm": 5.604981170941377, + "learning_rate": 1.9927832019322838e-05, + "loss": 2.3853, + "step": 961 + }, + { + "epoch": 0.2017194380373244, + "grad_norm": 6.045773713913119, + "learning_rate": 1.9927560273196436e-05, + "loss": 2.4956, + "step": 962 + }, + { + "epoch": 0.20192912560285176, + "grad_norm": 5.032703277139353, + "learning_rate": 1.9927288018265346e-05, + "loss": 2.4831, + "step": 963 + }, + { + "epoch": 0.20213881316837912, + "grad_norm": 5.870126491171813, + "learning_rate": 1.9927015254543533e-05, + "loss": 2.6288, + "step": 964 + }, + { + "epoch": 0.20234850073390648, + "grad_norm": 4.842696739861761, + "learning_rate": 1.9926741982044967e-05, + "loss": 2.3388, + "step": 965 + }, + { + "epoch": 0.20255818829943384, + "grad_norm": 5.156002782003122, + "learning_rate": 1.992646820078366e-05, + "loss": 2.427, + "step": 966 + }, + { + "epoch": 0.2027678758649612, + "grad_norm": 5.932095944618149, + "learning_rate": 1.992619391077364e-05, + "loss": 2.3469, + "step": 967 + }, + { + "epoch": 0.20297756343048856, + "grad_norm": 6.844332611420199, + "learning_rate": 1.9925919112028968e-05, + "loss": 2.4075, + "step": 968 + }, + { + "epoch": 0.20318725099601595, + "grad_norm": 5.4721063254887214, + "learning_rate": 1.992564380456372e-05, + "loss": 2.4533, + "step": 969 + }, + { + "epoch": 0.2033969385615433, + "grad_norm": 5.689594420915548, + "learning_rate": 1.992536798839202e-05, + "loss": 2.5801, + "step": 970 + }, + { + "epoch": 0.20360662612707067, + "grad_norm": 5.303921923868487, + "learning_rate": 1.9925091663527988e-05, + "loss": 2.4352, + "step": 971 + }, + { + "epoch": 0.20381631369259803, + "grad_norm": 6.102298979386814, + "learning_rate": 1.99248148299858e-05, + "loss": 2.2759, + "step": 972 + }, + { + "epoch": 0.2040260012581254, + "grad_norm": 5.10592501269951, + "learning_rate": 1.9924537487779637e-05, + "loss": 2.4277, + "step": 973 + }, + { + "epoch": 0.20423568882365276, + "grad_norm": 5.44142907343923, + "learning_rate": 1.992425963692371e-05, + "loss": 2.3429, + "step": 974 + }, + { + "epoch": 0.20444537638918012, + "grad_norm": 5.018214761519992, + "learning_rate": 1.9923981277432268e-05, + "loss": 2.4903, + "step": 975 + }, + { + "epoch": 0.20465506395470748, + "grad_norm": 5.1264355271054605, + "learning_rate": 1.9923702409319574e-05, + "loss": 2.4897, + "step": 976 + }, + { + "epoch": 0.20486475152023484, + "grad_norm": 5.616942379959226, + "learning_rate": 1.992342303259992e-05, + "loss": 2.4862, + "step": 977 + }, + { + "epoch": 0.20507443908576223, + "grad_norm": 6.721297698976598, + "learning_rate": 1.9923143147287622e-05, + "loss": 2.5596, + "step": 978 + }, + { + "epoch": 0.2052841266512896, + "grad_norm": 5.94528255614887, + "learning_rate": 1.9922862753397024e-05, + "loss": 2.3787, + "step": 979 + }, + { + "epoch": 0.20549381421681695, + "grad_norm": 5.3440472880217165, + "learning_rate": 1.9922581850942505e-05, + "loss": 2.6357, + "step": 980 + }, + { + "epoch": 0.2057035017823443, + "grad_norm": 5.226061565508953, + "learning_rate": 1.9922300439938454e-05, + "loss": 2.3324, + "step": 981 + }, + { + "epoch": 0.20591318934787167, + "grad_norm": 5.866211480937162, + "learning_rate": 1.9922018520399295e-05, + "loss": 2.3069, + "step": 982 + }, + { + "epoch": 0.20612287691339903, + "grad_norm": 5.194306073907077, + "learning_rate": 1.992173609233948e-05, + "loss": 2.5197, + "step": 983 + }, + { + "epoch": 0.2063325644789264, + "grad_norm": 5.893924553144362, + "learning_rate": 1.9921453155773483e-05, + "loss": 2.2331, + "step": 984 + }, + { + "epoch": 0.20654225204445376, + "grad_norm": 6.2542895594839925, + "learning_rate": 1.9921169710715798e-05, + "loss": 2.7147, + "step": 985 + }, + { + "epoch": 0.20675193960998112, + "grad_norm": 5.412600130038455, + "learning_rate": 1.9920885757180963e-05, + "loss": 2.5721, + "step": 986 + }, + { + "epoch": 0.2069616271755085, + "grad_norm": 5.249332644559337, + "learning_rate": 1.9920601295183525e-05, + "loss": 2.5272, + "step": 987 + }, + { + "epoch": 0.20717131474103587, + "grad_norm": 5.310076871818566, + "learning_rate": 1.9920316324738066e-05, + "loss": 2.3711, + "step": 988 + }, + { + "epoch": 0.20738100230656323, + "grad_norm": 6.120712586398684, + "learning_rate": 1.9920030845859188e-05, + "loss": 2.5763, + "step": 989 + }, + { + "epoch": 0.2075906898720906, + "grad_norm": 5.84672651053916, + "learning_rate": 1.9919744858561522e-05, + "loss": 2.417, + "step": 990 + }, + { + "epoch": 0.20780037743761795, + "grad_norm": 6.859888984723797, + "learning_rate": 1.991945836285973e-05, + "loss": 2.609, + "step": 991 + }, + { + "epoch": 0.2080100650031453, + "grad_norm": 5.198070603555374, + "learning_rate": 1.991917135876849e-05, + "loss": 2.3804, + "step": 992 + }, + { + "epoch": 0.20821975256867267, + "grad_norm": 5.384222366960085, + "learning_rate": 1.9918883846302514e-05, + "loss": 2.4152, + "step": 993 + }, + { + "epoch": 0.20842944013420003, + "grad_norm": 5.68017806619844, + "learning_rate": 1.991859582547654e-05, + "loss": 2.3473, + "step": 994 + }, + { + "epoch": 0.2086391276997274, + "grad_norm": 5.304686557561822, + "learning_rate": 1.9918307296305323e-05, + "loss": 2.562, + "step": 995 + }, + { + "epoch": 0.20884881526525478, + "grad_norm": 5.467332092462531, + "learning_rate": 1.9918018258803657e-05, + "loss": 2.4205, + "step": 996 + }, + { + "epoch": 0.20905850283078214, + "grad_norm": 4.769902789990944, + "learning_rate": 1.9917728712986355e-05, + "loss": 2.2886, + "step": 997 + }, + { + "epoch": 0.2092681903963095, + "grad_norm": 6.794217532194218, + "learning_rate": 1.9917438658868255e-05, + "loss": 2.6853, + "step": 998 + }, + { + "epoch": 0.20947787796183687, + "grad_norm": 5.454906469970778, + "learning_rate": 1.991714809646422e-05, + "loss": 2.2207, + "step": 999 + }, + { + "epoch": 0.20968756552736423, + "grad_norm": 6.298957621995344, + "learning_rate": 1.9916857025789146e-05, + "loss": 2.6666, + "step": 1000 + }, + { + "epoch": 0.2098972530928916, + "grad_norm": 5.521826411995796, + "learning_rate": 1.9916565446857946e-05, + "loss": 2.7545, + "step": 1001 + }, + { + "epoch": 0.21010694065841895, + "grad_norm": 5.757874625011884, + "learning_rate": 1.9916273359685576e-05, + "loss": 2.5922, + "step": 1002 + }, + { + "epoch": 0.2103166282239463, + "grad_norm": 5.27902083093314, + "learning_rate": 1.991598076428699e-05, + "loss": 2.2553, + "step": 1003 + }, + { + "epoch": 0.21052631578947367, + "grad_norm": 5.498085477839137, + "learning_rate": 1.9915687660677192e-05, + "loss": 2.6383, + "step": 1004 + }, + { + "epoch": 0.21073600335500106, + "grad_norm": 5.113704941603106, + "learning_rate": 1.9915394048871205e-05, + "loss": 2.4952, + "step": 1005 + }, + { + "epoch": 0.21094569092052842, + "grad_norm": 5.072259979149427, + "learning_rate": 1.9915099928884078e-05, + "loss": 2.4678, + "step": 1006 + }, + { + "epoch": 0.21115537848605578, + "grad_norm": 4.884467389239115, + "learning_rate": 1.991480530073088e-05, + "loss": 2.4964, + "step": 1007 + }, + { + "epoch": 0.21136506605158314, + "grad_norm": 4.836556286618052, + "learning_rate": 1.9914510164426715e-05, + "loss": 2.5517, + "step": 1008 + }, + { + "epoch": 0.2115747536171105, + "grad_norm": 4.7968385662798285, + "learning_rate": 1.9914214519986704e-05, + "loss": 2.6311, + "step": 1009 + }, + { + "epoch": 0.21178444118263787, + "grad_norm": 5.082890017057909, + "learning_rate": 1.9913918367426008e-05, + "loss": 2.485, + "step": 1010 + }, + { + "epoch": 0.21199412874816523, + "grad_norm": 5.643927808534393, + "learning_rate": 1.99136217067598e-05, + "loss": 2.5945, + "step": 1011 + }, + { + "epoch": 0.2122038163136926, + "grad_norm": 5.84315180999882, + "learning_rate": 1.9913324538003287e-05, + "loss": 2.4666, + "step": 1012 + }, + { + "epoch": 0.21241350387921995, + "grad_norm": 4.746601787110454, + "learning_rate": 1.9913026861171697e-05, + "loss": 2.2797, + "step": 1013 + }, + { + "epoch": 0.21262319144474734, + "grad_norm": 5.932084622281802, + "learning_rate": 1.9912728676280283e-05, + "loss": 2.659, + "step": 1014 + }, + { + "epoch": 0.2128328790102747, + "grad_norm": 6.344257633623337, + "learning_rate": 1.991242998334434e-05, + "loss": 2.6808, + "step": 1015 + }, + { + "epoch": 0.21304256657580206, + "grad_norm": 4.894288146442785, + "learning_rate": 1.991213078237916e-05, + "loss": 2.3413, + "step": 1016 + }, + { + "epoch": 0.21325225414132942, + "grad_norm": 5.068525800195981, + "learning_rate": 1.9911831073400087e-05, + "loss": 2.1377, + "step": 1017 + }, + { + "epoch": 0.21346194170685678, + "grad_norm": 5.214427846582535, + "learning_rate": 1.9911530856422483e-05, + "loss": 2.1781, + "step": 1018 + }, + { + "epoch": 0.21367162927238414, + "grad_norm": 5.408891363357328, + "learning_rate": 1.9911230131461728e-05, + "loss": 2.4674, + "step": 1019 + }, + { + "epoch": 0.2138813168379115, + "grad_norm": 5.8909554529528645, + "learning_rate": 1.991092889853324e-05, + "loss": 2.795, + "step": 1020 + }, + { + "epoch": 0.21409100440343887, + "grad_norm": 5.385242429979247, + "learning_rate": 1.991062715765246e-05, + "loss": 2.4117, + "step": 1021 + }, + { + "epoch": 0.21430069196896623, + "grad_norm": 6.775863301077143, + "learning_rate": 1.9910324908834847e-05, + "loss": 2.6616, + "step": 1022 + }, + { + "epoch": 0.21451037953449362, + "grad_norm": 5.6366557314701415, + "learning_rate": 1.9910022152095893e-05, + "loss": 2.6525, + "step": 1023 + }, + { + "epoch": 0.21472006710002098, + "grad_norm": 5.928054903499012, + "learning_rate": 1.9909718887451116e-05, + "loss": 2.4277, + "step": 1024 + }, + { + "epoch": 0.21492975466554834, + "grad_norm": 5.6545899514519675, + "learning_rate": 1.990941511491606e-05, + "loss": 2.5467, + "step": 1025 + }, + { + "epoch": 0.2151394422310757, + "grad_norm": 5.915372301589685, + "learning_rate": 1.990911083450629e-05, + "loss": 2.6486, + "step": 1026 + }, + { + "epoch": 0.21534912979660306, + "grad_norm": 4.910707852843951, + "learning_rate": 1.9908806046237402e-05, + "loss": 2.4959, + "step": 1027 + }, + { + "epoch": 0.21555881736213042, + "grad_norm": 4.7234529764044595, + "learning_rate": 1.990850075012502e-05, + "loss": 2.5617, + "step": 1028 + }, + { + "epoch": 0.21576850492765778, + "grad_norm": 4.807494910872457, + "learning_rate": 1.990819494618479e-05, + "loss": 2.4391, + "step": 1029 + }, + { + "epoch": 0.21597819249318514, + "grad_norm": 4.977852264588288, + "learning_rate": 1.9907888634432387e-05, + "loss": 2.3121, + "step": 1030 + }, + { + "epoch": 0.2161878800587125, + "grad_norm": 4.752309202791071, + "learning_rate": 1.9907581814883504e-05, + "loss": 2.4005, + "step": 1031 + }, + { + "epoch": 0.2163975676242399, + "grad_norm": 5.340413713610933, + "learning_rate": 1.9907274487553866e-05, + "loss": 2.1291, + "step": 1032 + }, + { + "epoch": 0.21660725518976726, + "grad_norm": 4.8443014728879845, + "learning_rate": 1.990696665245923e-05, + "loss": 2.2789, + "step": 1033 + }, + { + "epoch": 0.21681694275529462, + "grad_norm": 5.018147690157275, + "learning_rate": 1.990665830961537e-05, + "loss": 2.5965, + "step": 1034 + }, + { + "epoch": 0.21702663032082198, + "grad_norm": 5.450430771157589, + "learning_rate": 1.990634945903809e-05, + "loss": 2.382, + "step": 1035 + }, + { + "epoch": 0.21723631788634934, + "grad_norm": 5.629604652990529, + "learning_rate": 1.9906040100743217e-05, + "loss": 2.4539, + "step": 1036 + }, + { + "epoch": 0.2174460054518767, + "grad_norm": 5.258814896738817, + "learning_rate": 1.990573023474661e-05, + "loss": 2.2414, + "step": 1037 + }, + { + "epoch": 0.21765569301740406, + "grad_norm": 5.374829820500496, + "learning_rate": 1.990541986106415e-05, + "loss": 2.6661, + "step": 1038 + }, + { + "epoch": 0.21786538058293142, + "grad_norm": 5.300374494757614, + "learning_rate": 1.9905108979711736e-05, + "loss": 2.2204, + "step": 1039 + }, + { + "epoch": 0.21807506814845878, + "grad_norm": 4.9667622078988165, + "learning_rate": 1.9904797590705312e-05, + "loss": 2.4237, + "step": 1040 + }, + { + "epoch": 0.21828475571398617, + "grad_norm": 5.3780683496462895, + "learning_rate": 1.990448569406083e-05, + "loss": 2.5063, + "step": 1041 + }, + { + "epoch": 0.21849444327951353, + "grad_norm": 5.407520357169475, + "learning_rate": 1.990417328979428e-05, + "loss": 2.2237, + "step": 1042 + }, + { + "epoch": 0.2187041308450409, + "grad_norm": 5.134979884921702, + "learning_rate": 1.9903860377921667e-05, + "loss": 2.2159, + "step": 1043 + }, + { + "epoch": 0.21891381841056826, + "grad_norm": 5.185963481755813, + "learning_rate": 1.9903546958459033e-05, + "loss": 2.5709, + "step": 1044 + }, + { + "epoch": 0.21912350597609562, + "grad_norm": 6.417992882326232, + "learning_rate": 1.9903233031422445e-05, + "loss": 2.5104, + "step": 1045 + }, + { + "epoch": 0.21933319354162298, + "grad_norm": 7.015523503573994, + "learning_rate": 1.9902918596827985e-05, + "loss": 2.7682, + "step": 1046 + }, + { + "epoch": 0.21954288110715034, + "grad_norm": 5.238605408520289, + "learning_rate": 1.990260365469177e-05, + "loss": 2.685, + "step": 1047 + }, + { + "epoch": 0.2197525686726777, + "grad_norm": 5.2584784374393685, + "learning_rate": 1.9902288205029944e-05, + "loss": 2.5484, + "step": 1048 + }, + { + "epoch": 0.21996225623820506, + "grad_norm": 5.017678731434627, + "learning_rate": 1.9901972247858673e-05, + "loss": 2.1928, + "step": 1049 + }, + { + "epoch": 0.22017194380373245, + "grad_norm": 6.009470586969942, + "learning_rate": 1.990165578319415e-05, + "loss": 2.3864, + "step": 1050 + }, + { + "epoch": 0.2203816313692598, + "grad_norm": 5.444071213140894, + "learning_rate": 1.9901338811052593e-05, + "loss": 2.2969, + "step": 1051 + }, + { + "epoch": 0.22059131893478717, + "grad_norm": 5.615432192759512, + "learning_rate": 1.990102133145025e-05, + "loss": 2.5327, + "step": 1052 + }, + { + "epoch": 0.22080100650031453, + "grad_norm": 5.255182920505702, + "learning_rate": 1.9900703344403392e-05, + "loss": 2.3714, + "step": 1053 + }, + { + "epoch": 0.2210106940658419, + "grad_norm": 5.413996056497158, + "learning_rate": 1.9900384849928316e-05, + "loss": 2.5327, + "step": 1054 + }, + { + "epoch": 0.22122038163136926, + "grad_norm": 6.066541748074999, + "learning_rate": 1.9900065848041344e-05, + "loss": 2.4643, + "step": 1055 + }, + { + "epoch": 0.22143006919689662, + "grad_norm": 5.326762180328677, + "learning_rate": 1.9899746338758824e-05, + "loss": 2.3716, + "step": 1056 + }, + { + "epoch": 0.22163975676242398, + "grad_norm": 5.802056536507036, + "learning_rate": 1.989942632209714e-05, + "loss": 2.5801, + "step": 1057 + }, + { + "epoch": 0.22184944432795134, + "grad_norm": 5.151459060862107, + "learning_rate": 1.9899105798072682e-05, + "loss": 2.3164, + "step": 1058 + }, + { + "epoch": 0.22205913189347873, + "grad_norm": 6.18651319712188, + "learning_rate": 1.9898784766701884e-05, + "loss": 2.5993, + "step": 1059 + }, + { + "epoch": 0.2222688194590061, + "grad_norm": 4.982734095422201, + "learning_rate": 1.9898463228001203e-05, + "loss": 2.5444, + "step": 1060 + }, + { + "epoch": 0.22247850702453345, + "grad_norm": 4.850342014040275, + "learning_rate": 1.9898141181987106e-05, + "loss": 2.4433, + "step": 1061 + }, + { + "epoch": 0.2226881945900608, + "grad_norm": 6.324858012002737, + "learning_rate": 1.989781862867611e-05, + "loss": 2.4208, + "step": 1062 + }, + { + "epoch": 0.22289788215558817, + "grad_norm": 5.616595221771015, + "learning_rate": 1.989749556808474e-05, + "loss": 2.4745, + "step": 1063 + }, + { + "epoch": 0.22310756972111553, + "grad_norm": 5.380188697827271, + "learning_rate": 1.989717200022956e-05, + "loss": 2.1131, + "step": 1064 + }, + { + "epoch": 0.2233172572866429, + "grad_norm": 5.450797630408192, + "learning_rate": 1.9896847925127146e-05, + "loss": 2.3037, + "step": 1065 + }, + { + "epoch": 0.22352694485217026, + "grad_norm": 5.457348114452025, + "learning_rate": 1.989652334279411e-05, + "loss": 2.1687, + "step": 1066 + }, + { + "epoch": 0.22373663241769762, + "grad_norm": 6.046739047053696, + "learning_rate": 1.989619825324709e-05, + "loss": 2.4515, + "step": 1067 + }, + { + "epoch": 0.223946319983225, + "grad_norm": 5.711326213587446, + "learning_rate": 1.9895872656502746e-05, + "loss": 2.6118, + "step": 1068 + }, + { + "epoch": 0.22415600754875237, + "grad_norm": 5.008363290924573, + "learning_rate": 1.989554655257776e-05, + "loss": 2.2721, + "step": 1069 + }, + { + "epoch": 0.22436569511427973, + "grad_norm": 5.674063647957833, + "learning_rate": 1.9895219941488858e-05, + "loss": 2.4694, + "step": 1070 + }, + { + "epoch": 0.2245753826798071, + "grad_norm": 5.015264460456114, + "learning_rate": 1.9894892823252765e-05, + "loss": 2.4531, + "step": 1071 + }, + { + "epoch": 0.22478507024533445, + "grad_norm": 5.46331736480284, + "learning_rate": 1.9894565197886258e-05, + "loss": 2.4283, + "step": 1072 + }, + { + "epoch": 0.2249947578108618, + "grad_norm": 5.329346418566365, + "learning_rate": 1.989423706540612e-05, + "loss": 2.4312, + "step": 1073 + }, + { + "epoch": 0.22520444537638917, + "grad_norm": 6.0640851345305755, + "learning_rate": 1.9893908425829173e-05, + "loss": 2.4175, + "step": 1074 + }, + { + "epoch": 0.22541413294191653, + "grad_norm": 5.615423527558611, + "learning_rate": 1.989357927917226e-05, + "loss": 2.4979, + "step": 1075 + }, + { + "epoch": 0.2256238205074439, + "grad_norm": 5.795191410457589, + "learning_rate": 1.9893249625452244e-05, + "loss": 2.283, + "step": 1076 + }, + { + "epoch": 0.22583350807297128, + "grad_norm": 5.402480276096938, + "learning_rate": 1.989291946468603e-05, + "loss": 2.3371, + "step": 1077 + }, + { + "epoch": 0.22604319563849865, + "grad_norm": 6.238517874384495, + "learning_rate": 1.989258879689054e-05, + "loss": 2.315, + "step": 1078 + }, + { + "epoch": 0.226252883204026, + "grad_norm": 5.274754239395247, + "learning_rate": 1.9892257622082706e-05, + "loss": 2.4908, + "step": 1079 + }, + { + "epoch": 0.22646257076955337, + "grad_norm": 5.864569440688154, + "learning_rate": 1.9891925940279518e-05, + "loss": 2.1208, + "step": 1080 + }, + { + "epoch": 0.22667225833508073, + "grad_norm": 5.378517977996407, + "learning_rate": 1.9891593751497967e-05, + "loss": 2.554, + "step": 1081 + }, + { + "epoch": 0.2268819459006081, + "grad_norm": 5.439393863834567, + "learning_rate": 1.9891261055755084e-05, + "loss": 2.3881, + "step": 1082 + }, + { + "epoch": 0.22709163346613545, + "grad_norm": 5.511717855127642, + "learning_rate": 1.989092785306791e-05, + "loss": 2.3734, + "step": 1083 + }, + { + "epoch": 0.2273013210316628, + "grad_norm": 4.680462502286909, + "learning_rate": 1.9890594143453533e-05, + "loss": 2.4086, + "step": 1084 + }, + { + "epoch": 0.22751100859719017, + "grad_norm": 5.5267447915425825, + "learning_rate": 1.9890259926929048e-05, + "loss": 2.4705, + "step": 1085 + }, + { + "epoch": 0.22772069616271756, + "grad_norm": 6.274532441087503, + "learning_rate": 1.9889925203511594e-05, + "loss": 2.8928, + "step": 1086 + }, + { + "epoch": 0.22793038372824492, + "grad_norm": 5.796756491259425, + "learning_rate": 1.9889589973218315e-05, + "loss": 2.0499, + "step": 1087 + }, + { + "epoch": 0.22814007129377228, + "grad_norm": 5.498948459659436, + "learning_rate": 1.98892542360664e-05, + "loss": 2.5408, + "step": 1088 + }, + { + "epoch": 0.22834975885929965, + "grad_norm": 6.00016785085112, + "learning_rate": 1.988891799207305e-05, + "loss": 2.4997, + "step": 1089 + }, + { + "epoch": 0.228559446424827, + "grad_norm": 5.387316315703425, + "learning_rate": 1.9888581241255505e-05, + "loss": 2.4935, + "step": 1090 + }, + { + "epoch": 0.22876913399035437, + "grad_norm": 5.2308199820132995, + "learning_rate": 1.9888243983631018e-05, + "loss": 2.6946, + "step": 1091 + }, + { + "epoch": 0.22897882155588173, + "grad_norm": 5.243182635403196, + "learning_rate": 1.988790621921688e-05, + "loss": 2.3642, + "step": 1092 + }, + { + "epoch": 0.2291885091214091, + "grad_norm": 6.810486904566175, + "learning_rate": 1.9887567948030397e-05, + "loss": 2.3179, + "step": 1093 + }, + { + "epoch": 0.22939819668693645, + "grad_norm": 5.156717603786679, + "learning_rate": 1.9887229170088903e-05, + "loss": 2.2929, + "step": 1094 + }, + { + "epoch": 0.22960788425246384, + "grad_norm": 5.589558908681318, + "learning_rate": 1.988688988540977e-05, + "loss": 2.3762, + "step": 1095 + }, + { + "epoch": 0.2298175718179912, + "grad_norm": 5.858290454627939, + "learning_rate": 1.988655009401038e-05, + "loss": 2.5275, + "step": 1096 + }, + { + "epoch": 0.23002725938351856, + "grad_norm": 5.949439503505859, + "learning_rate": 1.9886209795908153e-05, + "loss": 2.3339, + "step": 1097 + }, + { + "epoch": 0.23023694694904592, + "grad_norm": 5.460743579538918, + "learning_rate": 1.9885868991120527e-05, + "loss": 2.3334, + "step": 1098 + }, + { + "epoch": 0.23044663451457328, + "grad_norm": 5.963284293167345, + "learning_rate": 1.9885527679664968e-05, + "loss": 2.3996, + "step": 1099 + }, + { + "epoch": 0.23065632208010065, + "grad_norm": 5.99662899042123, + "learning_rate": 1.9885185861558972e-05, + "loss": 2.3276, + "step": 1100 + }, + { + "epoch": 0.230866009645628, + "grad_norm": 6.246734905317924, + "learning_rate": 1.9884843536820053e-05, + "loss": 2.3025, + "step": 1101 + }, + { + "epoch": 0.23107569721115537, + "grad_norm": 6.629480689689256, + "learning_rate": 1.9884500705465758e-05, + "loss": 2.5753, + "step": 1102 + }, + { + "epoch": 0.23128538477668276, + "grad_norm": 5.588705383470157, + "learning_rate": 1.988415736751366e-05, + "loss": 2.6218, + "step": 1103 + }, + { + "epoch": 0.23149507234221012, + "grad_norm": 6.03275764601439, + "learning_rate": 1.9883813522981354e-05, + "loss": 2.3798, + "step": 1104 + }, + { + "epoch": 0.23170475990773748, + "grad_norm": 5.945400602844596, + "learning_rate": 1.988346917188646e-05, + "loss": 2.5885, + "step": 1105 + }, + { + "epoch": 0.23191444747326484, + "grad_norm": 5.6174688882209205, + "learning_rate": 1.988312431424663e-05, + "loss": 2.1999, + "step": 1106 + }, + { + "epoch": 0.2321241350387922, + "grad_norm": 5.4661783558100785, + "learning_rate": 1.988277895007954e-05, + "loss": 2.275, + "step": 1107 + }, + { + "epoch": 0.23233382260431956, + "grad_norm": 5.0527309781561875, + "learning_rate": 1.9882433079402883e-05, + "loss": 2.097, + "step": 1108 + }, + { + "epoch": 0.23254351016984692, + "grad_norm": 5.272169916430745, + "learning_rate": 1.9882086702234392e-05, + "loss": 2.2594, + "step": 1109 + }, + { + "epoch": 0.23275319773537428, + "grad_norm": 5.638774706877664, + "learning_rate": 1.988173981859182e-05, + "loss": 2.0905, + "step": 1110 + }, + { + "epoch": 0.23296288530090165, + "grad_norm": 5.095121898086692, + "learning_rate": 1.9881392428492942e-05, + "loss": 2.4484, + "step": 1111 + }, + { + "epoch": 0.23317257286642903, + "grad_norm": 5.312797842547137, + "learning_rate": 1.9881044531955565e-05, + "loss": 2.203, + "step": 1112 + }, + { + "epoch": 0.2333822604319564, + "grad_norm": 6.2745768185297965, + "learning_rate": 1.9880696128997517e-05, + "loss": 2.5075, + "step": 1113 + }, + { + "epoch": 0.23359194799748376, + "grad_norm": 5.5122408090759984, + "learning_rate": 1.988034721963665e-05, + "loss": 2.2074, + "step": 1114 + }, + { + "epoch": 0.23380163556301112, + "grad_norm": 6.0371519336847905, + "learning_rate": 1.987999780389086e-05, + "loss": 2.2155, + "step": 1115 + }, + { + "epoch": 0.23401132312853848, + "grad_norm": 6.317337843256815, + "learning_rate": 1.9879647881778043e-05, + "loss": 2.4839, + "step": 1116 + }, + { + "epoch": 0.23422101069406584, + "grad_norm": 6.208222649238208, + "learning_rate": 1.9879297453316136e-05, + "loss": 2.533, + "step": 1117 + }, + { + "epoch": 0.2344306982595932, + "grad_norm": 6.2122656465287625, + "learning_rate": 1.9878946518523104e-05, + "loss": 2.5666, + "step": 1118 + }, + { + "epoch": 0.23464038582512056, + "grad_norm": 7.153919656271692, + "learning_rate": 1.9878595077416924e-05, + "loss": 2.7303, + "step": 1119 + }, + { + "epoch": 0.23485007339064792, + "grad_norm": 4.792929169826411, + "learning_rate": 1.9878243130015613e-05, + "loss": 2.486, + "step": 1120 + }, + { + "epoch": 0.2350597609561753, + "grad_norm": 5.41165104539102, + "learning_rate": 1.987789067633721e-05, + "loss": 2.2663, + "step": 1121 + }, + { + "epoch": 0.23526944852170267, + "grad_norm": 5.223460485397531, + "learning_rate": 1.9877537716399774e-05, + "loss": 2.478, + "step": 1122 + }, + { + "epoch": 0.23547913608723003, + "grad_norm": 5.729524446536079, + "learning_rate": 1.9877184250221403e-05, + "loss": 2.5641, + "step": 1123 + }, + { + "epoch": 0.2356888236527574, + "grad_norm": 5.502387336399626, + "learning_rate": 1.9876830277820208e-05, + "loss": 2.2782, + "step": 1124 + }, + { + "epoch": 0.23589851121828476, + "grad_norm": 6.15686932207332, + "learning_rate": 1.987647579921433e-05, + "loss": 2.4573, + "step": 1125 + }, + { + "epoch": 0.23610819878381212, + "grad_norm": 4.567698842648741, + "learning_rate": 1.9876120814421934e-05, + "loss": 2.1701, + "step": 1126 + }, + { + "epoch": 0.23631788634933948, + "grad_norm": 6.289496308078909, + "learning_rate": 1.987576532346122e-05, + "loss": 2.2438, + "step": 1127 + }, + { + "epoch": 0.23652757391486684, + "grad_norm": 5.452527689465011, + "learning_rate": 1.9875409326350404e-05, + "loss": 2.3893, + "step": 1128 + }, + { + "epoch": 0.2367372614803942, + "grad_norm": 5.708237255565833, + "learning_rate": 1.9875052823107733e-05, + "loss": 2.2636, + "step": 1129 + }, + { + "epoch": 0.2369469490459216, + "grad_norm": 5.504553428880151, + "learning_rate": 1.9874695813751475e-05, + "loss": 2.1595, + "step": 1130 + }, + { + "epoch": 0.23715663661144895, + "grad_norm": 5.605450790489481, + "learning_rate": 1.9874338298299933e-05, + "loss": 2.753, + "step": 1131 + }, + { + "epoch": 0.2373663241769763, + "grad_norm": 5.367842452443219, + "learning_rate": 1.9873980276771426e-05, + "loss": 2.316, + "step": 1132 + }, + { + "epoch": 0.23757601174250367, + "grad_norm": 5.517744481439908, + "learning_rate": 1.98736217491843e-05, + "loss": 2.4459, + "step": 1133 + }, + { + "epoch": 0.23778569930803103, + "grad_norm": 5.3123463391071875, + "learning_rate": 1.987326271555694e-05, + "loss": 2.4406, + "step": 1134 + }, + { + "epoch": 0.2379953868735584, + "grad_norm": 6.113216911836581, + "learning_rate": 1.987290317590774e-05, + "loss": 2.1981, + "step": 1135 + }, + { + "epoch": 0.23820507443908576, + "grad_norm": 5.499884763202649, + "learning_rate": 1.9872543130255128e-05, + "loss": 2.6627, + "step": 1136 + }, + { + "epoch": 0.23841476200461312, + "grad_norm": 5.274258796880021, + "learning_rate": 1.9872182578617558e-05, + "loss": 2.204, + "step": 1137 + }, + { + "epoch": 0.23862444957014048, + "grad_norm": 5.028270158500906, + "learning_rate": 1.987182152101351e-05, + "loss": 2.4063, + "step": 1138 + }, + { + "epoch": 0.23883413713566787, + "grad_norm": 4.953113826278815, + "learning_rate": 1.9871459957461482e-05, + "loss": 2.2782, + "step": 1139 + }, + { + "epoch": 0.23904382470119523, + "grad_norm": 4.955249712641666, + "learning_rate": 1.9871097887980016e-05, + "loss": 2.5487, + "step": 1140 + }, + { + "epoch": 0.2392535122667226, + "grad_norm": 5.997073655406304, + "learning_rate": 1.9870735312587658e-05, + "loss": 2.2155, + "step": 1141 + }, + { + "epoch": 0.23946319983224995, + "grad_norm": 5.856533857079827, + "learning_rate": 1.9870372231302996e-05, + "loss": 2.3308, + "step": 1142 + }, + { + "epoch": 0.2396728873977773, + "grad_norm": 5.44256584284782, + "learning_rate": 1.987000864414464e-05, + "loss": 2.1604, + "step": 1143 + }, + { + "epoch": 0.23988257496330467, + "grad_norm": 5.495169713569408, + "learning_rate": 1.986964455113122e-05, + "loss": 2.2905, + "step": 1144 + }, + { + "epoch": 0.24009226252883203, + "grad_norm": 6.117013045146031, + "learning_rate": 1.98692799522814e-05, + "loss": 2.4982, + "step": 1145 + }, + { + "epoch": 0.2403019500943594, + "grad_norm": 6.494218601389671, + "learning_rate": 1.9868914847613867e-05, + "loss": 2.5917, + "step": 1146 + }, + { + "epoch": 0.24051163765988676, + "grad_norm": 5.347069593173759, + "learning_rate": 1.986854923714733e-05, + "loss": 2.1281, + "step": 1147 + }, + { + "epoch": 0.24072132522541415, + "grad_norm": 7.068169917508162, + "learning_rate": 1.9868183120900527e-05, + "loss": 2.3845, + "step": 1148 + }, + { + "epoch": 0.2409310127909415, + "grad_norm": 6.15783802810978, + "learning_rate": 1.9867816498892228e-05, + "loss": 2.4638, + "step": 1149 + }, + { + "epoch": 0.24114070035646887, + "grad_norm": 8.244312691717331, + "learning_rate": 1.9867449371141214e-05, + "loss": 2.6692, + "step": 1150 + }, + { + "epoch": 0.24135038792199623, + "grad_norm": 5.68659057279483, + "learning_rate": 1.9867081737666306e-05, + "loss": 2.374, + "step": 1151 + }, + { + "epoch": 0.2415600754875236, + "grad_norm": 5.929399972630398, + "learning_rate": 1.9866713598486345e-05, + "loss": 2.3455, + "step": 1152 + }, + { + "epoch": 0.24176976305305095, + "grad_norm": 5.787811580602739, + "learning_rate": 1.98663449536202e-05, + "loss": 2.2595, + "step": 1153 + }, + { + "epoch": 0.2419794506185783, + "grad_norm": 5.848226090100022, + "learning_rate": 1.9865975803086765e-05, + "loss": 2.4017, + "step": 1154 + }, + { + "epoch": 0.24218913818410567, + "grad_norm": 6.082148243489996, + "learning_rate": 1.9865606146904957e-05, + "loss": 2.5488, + "step": 1155 + }, + { + "epoch": 0.24239882574963303, + "grad_norm": 7.3803871732556185, + "learning_rate": 1.9865235985093724e-05, + "loss": 2.6939, + "step": 1156 + }, + { + "epoch": 0.24260851331516042, + "grad_norm": 5.28203187147514, + "learning_rate": 1.9864865317672034e-05, + "loss": 2.257, + "step": 1157 + }, + { + "epoch": 0.24281820088068778, + "grad_norm": 5.423373224566631, + "learning_rate": 1.986449414465889e-05, + "loss": 2.6038, + "step": 1158 + }, + { + "epoch": 0.24302788844621515, + "grad_norm": 5.801273387364715, + "learning_rate": 1.9864122466073308e-05, + "loss": 2.5051, + "step": 1159 + }, + { + "epoch": 0.2432375760117425, + "grad_norm": 5.379339333505808, + "learning_rate": 1.9863750281934345e-05, + "loss": 2.3499, + "step": 1160 + }, + { + "epoch": 0.24344726357726987, + "grad_norm": 5.362395437761733, + "learning_rate": 1.986337759226107e-05, + "loss": 2.4641, + "step": 1161 + }, + { + "epoch": 0.24365695114279723, + "grad_norm": 5.560355618138492, + "learning_rate": 1.9863004397072586e-05, + "loss": 2.0889, + "step": 1162 + }, + { + "epoch": 0.2438666387083246, + "grad_norm": 5.137534652210816, + "learning_rate": 1.9862630696388023e-05, + "loss": 2.387, + "step": 1163 + }, + { + "epoch": 0.24407632627385195, + "grad_norm": 5.117646251200553, + "learning_rate": 1.986225649022653e-05, + "loss": 2.2911, + "step": 1164 + }, + { + "epoch": 0.2442860138393793, + "grad_norm": 4.929722910793912, + "learning_rate": 1.9861881778607283e-05, + "loss": 2.112, + "step": 1165 + }, + { + "epoch": 0.2444957014049067, + "grad_norm": 4.924007160384863, + "learning_rate": 1.9861506561549494e-05, + "loss": 2.4487, + "step": 1166 + }, + { + "epoch": 0.24470538897043406, + "grad_norm": 5.2735753090830375, + "learning_rate": 1.986113083907239e-05, + "loss": 2.1672, + "step": 1167 + }, + { + "epoch": 0.24491507653596142, + "grad_norm": 6.608704847772229, + "learning_rate": 1.9860754611195226e-05, + "loss": 2.3328, + "step": 1168 + }, + { + "epoch": 0.24512476410148878, + "grad_norm": 5.1966977918464865, + "learning_rate": 1.9860377877937287e-05, + "loss": 2.3562, + "step": 1169 + }, + { + "epoch": 0.24533445166701615, + "grad_norm": 4.948102841994688, + "learning_rate": 1.986000063931788e-05, + "loss": 2.3373, + "step": 1170 + }, + { + "epoch": 0.2455441392325435, + "grad_norm": 5.2097377981363895, + "learning_rate": 1.9859622895356335e-05, + "loss": 2.393, + "step": 1171 + }, + { + "epoch": 0.24575382679807087, + "grad_norm": 5.941671184348702, + "learning_rate": 1.9859244646072017e-05, + "loss": 2.5533, + "step": 1172 + }, + { + "epoch": 0.24596351436359823, + "grad_norm": 5.190811034661953, + "learning_rate": 1.9858865891484313e-05, + "loss": 2.5423, + "step": 1173 + }, + { + "epoch": 0.2461732019291256, + "grad_norm": 5.241397669095447, + "learning_rate": 1.9858486631612633e-05, + "loss": 2.309, + "step": 1174 + }, + { + "epoch": 0.24638288949465298, + "grad_norm": 6.071879430448749, + "learning_rate": 1.985810686647641e-05, + "loss": 2.1952, + "step": 1175 + }, + { + "epoch": 0.24659257706018034, + "grad_norm": 5.415756123932901, + "learning_rate": 1.985772659609512e-05, + "loss": 2.2649, + "step": 1176 + }, + { + "epoch": 0.2468022646257077, + "grad_norm": 5.591952572905724, + "learning_rate": 1.985734582048824e-05, + "loss": 2.5287, + "step": 1177 + }, + { + "epoch": 0.24701195219123506, + "grad_norm": 6.186606283769733, + "learning_rate": 1.9856964539675287e-05, + "loss": 2.4627, + "step": 1178 + }, + { + "epoch": 0.24722163975676242, + "grad_norm": 5.648617180496344, + "learning_rate": 1.9856582753675808e-05, + "loss": 2.4207, + "step": 1179 + }, + { + "epoch": 0.24743132732228978, + "grad_norm": 5.721334461306422, + "learning_rate": 1.985620046250937e-05, + "loss": 2.4322, + "step": 1180 + }, + { + "epoch": 0.24764101488781715, + "grad_norm": 5.8129998608713755, + "learning_rate": 1.985581766619556e-05, + "loss": 2.5754, + "step": 1181 + }, + { + "epoch": 0.2478507024533445, + "grad_norm": 5.6367792272000035, + "learning_rate": 1.9855434364754003e-05, + "loss": 2.3282, + "step": 1182 + }, + { + "epoch": 0.24806039001887187, + "grad_norm": 7.218380738491035, + "learning_rate": 1.985505055820434e-05, + "loss": 2.4668, + "step": 1183 + }, + { + "epoch": 0.24827007758439926, + "grad_norm": 5.177707975673539, + "learning_rate": 1.9854666246566244e-05, + "loss": 2.2937, + "step": 1184 + }, + { + "epoch": 0.24847976514992662, + "grad_norm": 5.572773400010812, + "learning_rate": 1.985428142985941e-05, + "loss": 2.4316, + "step": 1185 + }, + { + "epoch": 0.24868945271545398, + "grad_norm": 5.404683091715184, + "learning_rate": 1.9853896108103565e-05, + "loss": 2.4322, + "step": 1186 + }, + { + "epoch": 0.24889914028098134, + "grad_norm": 5.4586373140361, + "learning_rate": 1.9853510281318446e-05, + "loss": 2.3618, + "step": 1187 + }, + { + "epoch": 0.2491088278465087, + "grad_norm": 5.712605667669462, + "learning_rate": 1.9853123949523843e-05, + "loss": 2.1057, + "step": 1188 + }, + { + "epoch": 0.24931851541203606, + "grad_norm": 5.840087749174537, + "learning_rate": 1.9852737112739545e-05, + "loss": 2.4032, + "step": 1189 + }, + { + "epoch": 0.24952820297756342, + "grad_norm": 5.38535609309615, + "learning_rate": 1.985234977098538e-05, + "loss": 2.2993, + "step": 1190 + }, + { + "epoch": 0.24973789054309078, + "grad_norm": 5.348506150096231, + "learning_rate": 1.98519619242812e-05, + "loss": 2.4278, + "step": 1191 + }, + { + "epoch": 0.24994757810861815, + "grad_norm": 5.583716426867162, + "learning_rate": 1.9851573572646888e-05, + "loss": 2.1251, + "step": 1192 + }, + { + "epoch": 0.25015726567414553, + "grad_norm": 5.956034460626029, + "learning_rate": 1.9851184716102342e-05, + "loss": 2.5705, + "step": 1193 + }, + { + "epoch": 0.2503669532396729, + "grad_norm": 6.671375126114211, + "learning_rate": 1.985079535466749e-05, + "loss": 2.319, + "step": 1194 + }, + { + "epoch": 0.25057664080520026, + "grad_norm": 5.926760972556911, + "learning_rate": 1.9850405488362296e-05, + "loss": 2.5532, + "step": 1195 + }, + { + "epoch": 0.2507863283707276, + "grad_norm": 5.020992536505515, + "learning_rate": 1.9850015117206736e-05, + "loss": 2.6251, + "step": 1196 + }, + { + "epoch": 0.250996015936255, + "grad_norm": 5.177380738976003, + "learning_rate": 1.9849624241220814e-05, + "loss": 2.3906, + "step": 1197 + }, + { + "epoch": 0.25120570350178234, + "grad_norm": 5.439352919616085, + "learning_rate": 1.984923286042457e-05, + "loss": 2.3691, + "step": 1198 + }, + { + "epoch": 0.2514153910673097, + "grad_norm": 5.064151809568982, + "learning_rate": 1.9848840974838054e-05, + "loss": 2.1341, + "step": 1199 + }, + { + "epoch": 0.25162507863283706, + "grad_norm": 5.837547826130716, + "learning_rate": 1.984844858448136e-05, + "loss": 2.5517, + "step": 1200 + }, + { + "epoch": 0.2518347661983644, + "grad_norm": 5.41671745046474, + "learning_rate": 1.9848055689374594e-05, + "loss": 2.139, + "step": 1201 + }, + { + "epoch": 0.2520444537638918, + "grad_norm": 5.5614150019148045, + "learning_rate": 1.9847662289537893e-05, + "loss": 2.1161, + "step": 1202 + }, + { + "epoch": 0.25225414132941915, + "grad_norm": 4.887067996724922, + "learning_rate": 1.984726838499142e-05, + "loss": 2.4905, + "step": 1203 + }, + { + "epoch": 0.2524638288949465, + "grad_norm": 5.266908558874635, + "learning_rate": 1.9846873975755363e-05, + "loss": 2.2074, + "step": 1204 + }, + { + "epoch": 0.25267351646047387, + "grad_norm": 5.700358033778161, + "learning_rate": 1.9846479061849936e-05, + "loss": 2.3268, + "step": 1205 + }, + { + "epoch": 0.2528832040260013, + "grad_norm": 5.033509154854469, + "learning_rate": 1.984608364329538e-05, + "loss": 2.3735, + "step": 1206 + }, + { + "epoch": 0.25309289159152865, + "grad_norm": 6.0183316637739335, + "learning_rate": 1.9845687720111958e-05, + "loss": 2.4834, + "step": 1207 + }, + { + "epoch": 0.253302579157056, + "grad_norm": 5.190534882363171, + "learning_rate": 1.9845291292319967e-05, + "loss": 2.2527, + "step": 1208 + }, + { + "epoch": 0.25351226672258337, + "grad_norm": 5.543751449817277, + "learning_rate": 1.9844894359939716e-05, + "loss": 2.6291, + "step": 1209 + }, + { + "epoch": 0.25372195428811073, + "grad_norm": 5.435129714033043, + "learning_rate": 1.984449692299156e-05, + "loss": 2.3272, + "step": 1210 + }, + { + "epoch": 0.2539316418536381, + "grad_norm": 4.831754248434429, + "learning_rate": 1.984409898149586e-05, + "loss": 2.4798, + "step": 1211 + }, + { + "epoch": 0.25414132941916545, + "grad_norm": 6.5695152689212835, + "learning_rate": 1.984370053547301e-05, + "loss": 2.3515, + "step": 1212 + }, + { + "epoch": 0.2543510169846928, + "grad_norm": 5.897750792729059, + "learning_rate": 1.984330158494344e-05, + "loss": 2.4192, + "step": 1213 + }, + { + "epoch": 0.2545607045502202, + "grad_norm": 5.942370235426671, + "learning_rate": 1.984290212992759e-05, + "loss": 2.3976, + "step": 1214 + }, + { + "epoch": 0.25477039211574753, + "grad_norm": 6.193064285016414, + "learning_rate": 1.9842502170445933e-05, + "loss": 2.4846, + "step": 1215 + }, + { + "epoch": 0.2549800796812749, + "grad_norm": 5.876130410960825, + "learning_rate": 1.9842101706518974e-05, + "loss": 2.3774, + "step": 1216 + }, + { + "epoch": 0.25518976724680226, + "grad_norm": 5.719231147536945, + "learning_rate": 1.9841700738167225e-05, + "loss": 2.5691, + "step": 1217 + }, + { + "epoch": 0.2553994548123296, + "grad_norm": 5.595717421108977, + "learning_rate": 1.984129926541125e-05, + "loss": 2.1098, + "step": 1218 + }, + { + "epoch": 0.255609142377857, + "grad_norm": 6.179802493545814, + "learning_rate": 1.984089728827161e-05, + "loss": 2.3443, + "step": 1219 + }, + { + "epoch": 0.25581882994338434, + "grad_norm": 4.68647201531337, + "learning_rate": 1.9840494806768922e-05, + "loss": 2.2537, + "step": 1220 + }, + { + "epoch": 0.2560285175089117, + "grad_norm": 5.582522111083153, + "learning_rate": 1.984009182092381e-05, + "loss": 2.3954, + "step": 1221 + }, + { + "epoch": 0.25623820507443906, + "grad_norm": 6.359146106136524, + "learning_rate": 1.983968833075692e-05, + "loss": 1.7707, + "step": 1222 + }, + { + "epoch": 0.2564478926399664, + "grad_norm": 6.3285506849987785, + "learning_rate": 1.9839284336288937e-05, + "loss": 2.5248, + "step": 1223 + }, + { + "epoch": 0.25665758020549384, + "grad_norm": 6.151720134525134, + "learning_rate": 1.9838879837540567e-05, + "loss": 2.288, + "step": 1224 + }, + { + "epoch": 0.2568672677710212, + "grad_norm": 5.621075483528397, + "learning_rate": 1.9838474834532546e-05, + "loss": 2.5878, + "step": 1225 + }, + { + "epoch": 0.25707695533654856, + "grad_norm": 5.215333990206144, + "learning_rate": 1.983806932728562e-05, + "loss": 2.4143, + "step": 1226 + }, + { + "epoch": 0.2572866429020759, + "grad_norm": 4.954922257318685, + "learning_rate": 1.9837663315820578e-05, + "loss": 2.4693, + "step": 1227 + }, + { + "epoch": 0.2574963304676033, + "grad_norm": 5.326597655728064, + "learning_rate": 1.9837256800158232e-05, + "loss": 2.3845, + "step": 1228 + }, + { + "epoch": 0.25770601803313065, + "grad_norm": 5.266881696393896, + "learning_rate": 1.9836849780319413e-05, + "loss": 2.1453, + "step": 1229 + }, + { + "epoch": 0.257915705598658, + "grad_norm": 5.151379809070329, + "learning_rate": 1.9836442256324977e-05, + "loss": 2.3946, + "step": 1230 + }, + { + "epoch": 0.25812539316418537, + "grad_norm": 5.5708711932153765, + "learning_rate": 1.9836034228195817e-05, + "loss": 2.6474, + "step": 1231 + }, + { + "epoch": 0.25833508072971273, + "grad_norm": 5.4759966606263575, + "learning_rate": 1.9835625695952846e-05, + "loss": 2.3367, + "step": 1232 + }, + { + "epoch": 0.2585447682952401, + "grad_norm": 4.919886897100189, + "learning_rate": 1.9835216659616997e-05, + "loss": 2.5269, + "step": 1233 + }, + { + "epoch": 0.25875445586076745, + "grad_norm": 6.027754044675275, + "learning_rate": 1.9834807119209233e-05, + "loss": 2.5167, + "step": 1234 + }, + { + "epoch": 0.2589641434262948, + "grad_norm": 4.84284369619074, + "learning_rate": 1.983439707475055e-05, + "loss": 2.2038, + "step": 1235 + }, + { + "epoch": 0.2591738309918222, + "grad_norm": 5.097040535475538, + "learning_rate": 1.9833986526261958e-05, + "loss": 2.3842, + "step": 1236 + }, + { + "epoch": 0.25938351855734953, + "grad_norm": 5.346301675262585, + "learning_rate": 1.9833575473764505e-05, + "loss": 2.5197, + "step": 1237 + }, + { + "epoch": 0.2595932061228769, + "grad_norm": 6.08034660642355, + "learning_rate": 1.9833163917279247e-05, + "loss": 2.3006, + "step": 1238 + }, + { + "epoch": 0.25980289368840426, + "grad_norm": 5.718078908081992, + "learning_rate": 1.9832751856827288e-05, + "loss": 2.4086, + "step": 1239 + }, + { + "epoch": 0.2600125812539316, + "grad_norm": 4.6728544594966, + "learning_rate": 1.983233929242974e-05, + "loss": 2.3108, + "step": 1240 + }, + { + "epoch": 0.260222268819459, + "grad_norm": 5.1393477710538145, + "learning_rate": 1.983192622410775e-05, + "loss": 2.2975, + "step": 1241 + }, + { + "epoch": 0.2604319563849864, + "grad_norm": 5.366751190147697, + "learning_rate": 1.9831512651882493e-05, + "loss": 2.2786, + "step": 1242 + }, + { + "epoch": 0.26064164395051376, + "grad_norm": 5.330911591631136, + "learning_rate": 1.9831098575775157e-05, + "loss": 2.0923, + "step": 1243 + }, + { + "epoch": 0.2608513315160411, + "grad_norm": 5.584312081436149, + "learning_rate": 1.9830683995806966e-05, + "loss": 2.4355, + "step": 1244 + }, + { + "epoch": 0.2610610190815685, + "grad_norm": 5.320522162777732, + "learning_rate": 1.983026891199917e-05, + "loss": 2.3724, + "step": 1245 + }, + { + "epoch": 0.26127070664709584, + "grad_norm": 5.35383690653178, + "learning_rate": 1.9829853324373045e-05, + "loss": 2.3875, + "step": 1246 + }, + { + "epoch": 0.2614803942126232, + "grad_norm": 6.08381273936834, + "learning_rate": 1.9829437232949888e-05, + "loss": 2.2762, + "step": 1247 + }, + { + "epoch": 0.26169008177815056, + "grad_norm": 5.679294095719883, + "learning_rate": 1.9829020637751022e-05, + "loss": 2.3114, + "step": 1248 + }, + { + "epoch": 0.2618997693436779, + "grad_norm": 6.360855883203795, + "learning_rate": 1.9828603538797804e-05, + "loss": 2.7159, + "step": 1249 + }, + { + "epoch": 0.2621094569092053, + "grad_norm": 5.385428800829001, + "learning_rate": 1.9828185936111604e-05, + "loss": 2.2438, + "step": 1250 + }, + { + "epoch": 0.26231914447473265, + "grad_norm": 6.279653511183181, + "learning_rate": 1.9827767829713834e-05, + "loss": 2.32, + "step": 1251 + }, + { + "epoch": 0.26252883204026, + "grad_norm": 6.096860363715155, + "learning_rate": 1.9827349219625912e-05, + "loss": 2.5817, + "step": 1252 + }, + { + "epoch": 0.26273851960578737, + "grad_norm": 6.0660160469466735, + "learning_rate": 1.98269301058693e-05, + "loss": 2.5179, + "step": 1253 + }, + { + "epoch": 0.26294820717131473, + "grad_norm": 5.6939034161962745, + "learning_rate": 1.9826510488465477e-05, + "loss": 1.7898, + "step": 1254 + }, + { + "epoch": 0.2631578947368421, + "grad_norm": 6.157895608774056, + "learning_rate": 1.9826090367435946e-05, + "loss": 2.2776, + "step": 1255 + }, + { + "epoch": 0.26336758230236945, + "grad_norm": 5.111559533735894, + "learning_rate": 1.9825669742802244e-05, + "loss": 2.4436, + "step": 1256 + }, + { + "epoch": 0.2635772698678968, + "grad_norm": 5.657181849764752, + "learning_rate": 1.9825248614585923e-05, + "loss": 2.3923, + "step": 1257 + }, + { + "epoch": 0.2637869574334242, + "grad_norm": 5.932671455166495, + "learning_rate": 1.982482698280857e-05, + "loss": 2.621, + "step": 1258 + }, + { + "epoch": 0.26399664499895154, + "grad_norm": 5.265155833769554, + "learning_rate": 1.9824404847491796e-05, + "loss": 2.3326, + "step": 1259 + }, + { + "epoch": 0.26420633256447895, + "grad_norm": 6.437380784687752, + "learning_rate": 1.982398220865723e-05, + "loss": 2.6448, + "step": 1260 + }, + { + "epoch": 0.2644160201300063, + "grad_norm": 5.617705409317104, + "learning_rate": 1.982355906632654e-05, + "loss": 2.5222, + "step": 1261 + }, + { + "epoch": 0.2646257076955337, + "grad_norm": 5.895967274786925, + "learning_rate": 1.982313542052141e-05, + "loss": 2.3592, + "step": 1262 + }, + { + "epoch": 0.26483539526106104, + "grad_norm": 5.72419044529412, + "learning_rate": 1.9822711271263554e-05, + "loss": 2.4786, + "step": 1263 + }, + { + "epoch": 0.2650450828265884, + "grad_norm": 5.591249549080251, + "learning_rate": 1.9822286618574704e-05, + "loss": 2.2958, + "step": 1264 + }, + { + "epoch": 0.26525477039211576, + "grad_norm": 6.832124623565176, + "learning_rate": 1.9821861462476636e-05, + "loss": 2.1979, + "step": 1265 + }, + { + "epoch": 0.2654644579576431, + "grad_norm": 4.79977835068797, + "learning_rate": 1.9821435802991126e-05, + "loss": 2.453, + "step": 1266 + }, + { + "epoch": 0.2656741455231705, + "grad_norm": 4.640605793316747, + "learning_rate": 1.982100964014e-05, + "loss": 2.274, + "step": 1267 + }, + { + "epoch": 0.26588383308869784, + "grad_norm": 5.093378486241985, + "learning_rate": 1.9820582973945095e-05, + "loss": 2.5203, + "step": 1268 + }, + { + "epoch": 0.2660935206542252, + "grad_norm": 6.586494062817634, + "learning_rate": 1.9820155804428278e-05, + "loss": 2.7188, + "step": 1269 + }, + { + "epoch": 0.26630320821975256, + "grad_norm": 5.253871062667889, + "learning_rate": 1.9819728131611447e-05, + "loss": 2.46, + "step": 1270 + }, + { + "epoch": 0.2665128957852799, + "grad_norm": 6.115208491414063, + "learning_rate": 1.9819299955516515e-05, + "loss": 2.1893, + "step": 1271 + }, + { + "epoch": 0.2667225833508073, + "grad_norm": 6.137026561760813, + "learning_rate": 1.981887127616543e-05, + "loss": 2.5991, + "step": 1272 + }, + { + "epoch": 0.26693227091633465, + "grad_norm": 5.452575290487976, + "learning_rate": 1.9818442093580164e-05, + "loss": 2.5141, + "step": 1273 + }, + { + "epoch": 0.267141958481862, + "grad_norm": 6.113570782322421, + "learning_rate": 1.9818012407782713e-05, + "loss": 2.4623, + "step": 1274 + }, + { + "epoch": 0.26735164604738937, + "grad_norm": 4.749434782189213, + "learning_rate": 1.9817582218795094e-05, + "loss": 2.4898, + "step": 1275 + }, + { + "epoch": 0.26756133361291673, + "grad_norm": 6.4045008635310205, + "learning_rate": 1.981715152663936e-05, + "loss": 2.6487, + "step": 1276 + }, + { + "epoch": 0.2677710211784441, + "grad_norm": 5.03275443583643, + "learning_rate": 1.9816720331337582e-05, + "loss": 2.3014, + "step": 1277 + }, + { + "epoch": 0.2679807087439715, + "grad_norm": 5.073883475047928, + "learning_rate": 1.9816288632911864e-05, + "loss": 2.2536, + "step": 1278 + }, + { + "epoch": 0.26819039630949887, + "grad_norm": 5.937809635860017, + "learning_rate": 1.9815856431384323e-05, + "loss": 2.4236, + "step": 1279 + }, + { + "epoch": 0.26840008387502623, + "grad_norm": 6.636803411490492, + "learning_rate": 1.981542372677712e-05, + "loss": 2.3035, + "step": 1280 + }, + { + "epoch": 0.2686097714405536, + "grad_norm": 5.197509258886102, + "learning_rate": 1.9814990519112424e-05, + "loss": 2.3859, + "step": 1281 + }, + { + "epoch": 0.26881945900608095, + "grad_norm": 5.997065176229385, + "learning_rate": 1.9814556808412444e-05, + "loss": 2.3246, + "step": 1282 + }, + { + "epoch": 0.2690291465716083, + "grad_norm": 5.838210602998649, + "learning_rate": 1.98141225946994e-05, + "loss": 2.635, + "step": 1283 + }, + { + "epoch": 0.2692388341371357, + "grad_norm": 7.133731490626786, + "learning_rate": 1.9813687877995556e-05, + "loss": 2.5888, + "step": 1284 + }, + { + "epoch": 0.26944852170266304, + "grad_norm": 6.212250850778944, + "learning_rate": 1.9813252658323186e-05, + "loss": 1.9985, + "step": 1285 + }, + { + "epoch": 0.2696582092681904, + "grad_norm": 5.386211045262769, + "learning_rate": 1.9812816935704596e-05, + "loss": 2.1854, + "step": 1286 + }, + { + "epoch": 0.26986789683371776, + "grad_norm": 6.303945142855519, + "learning_rate": 1.981238071016212e-05, + "loss": 2.6064, + "step": 1287 + }, + { + "epoch": 0.2700775843992451, + "grad_norm": 5.3162656673676585, + "learning_rate": 1.981194398171811e-05, + "loss": 2.2494, + "step": 1288 + }, + { + "epoch": 0.2702872719647725, + "grad_norm": 5.020536335740002, + "learning_rate": 1.981150675039496e-05, + "loss": 2.4484, + "step": 1289 + }, + { + "epoch": 0.27049695953029984, + "grad_norm": 6.1420272246550915, + "learning_rate": 1.9811069016215066e-05, + "loss": 2.1911, + "step": 1290 + }, + { + "epoch": 0.2707066470958272, + "grad_norm": 5.87394210429339, + "learning_rate": 1.981063077920087e-05, + "loss": 2.2371, + "step": 1291 + }, + { + "epoch": 0.27091633466135456, + "grad_norm": 6.244711660978756, + "learning_rate": 1.981019203937483e-05, + "loss": 2.4066, + "step": 1292 + }, + { + "epoch": 0.2711260222268819, + "grad_norm": 6.165865589693254, + "learning_rate": 1.9809752796759434e-05, + "loss": 2.5739, + "step": 1293 + }, + { + "epoch": 0.2713357097924093, + "grad_norm": 6.88735453147843, + "learning_rate": 1.9809313051377196e-05, + "loss": 2.4636, + "step": 1294 + }, + { + "epoch": 0.2715453973579367, + "grad_norm": 5.810667529576484, + "learning_rate": 1.9808872803250647e-05, + "loss": 2.5927, + "step": 1295 + }, + { + "epoch": 0.27175508492346406, + "grad_norm": 5.0610716016591635, + "learning_rate": 1.9808432052402356e-05, + "loss": 2.2261, + "step": 1296 + }, + { + "epoch": 0.2719647724889914, + "grad_norm": 5.488967514234256, + "learning_rate": 1.980799079885491e-05, + "loss": 2.3501, + "step": 1297 + }, + { + "epoch": 0.2721744600545188, + "grad_norm": 5.810833357151884, + "learning_rate": 1.980754904263092e-05, + "loss": 2.3486, + "step": 1298 + }, + { + "epoch": 0.27238414762004615, + "grad_norm": 5.096805833624817, + "learning_rate": 1.980710678375304e-05, + "loss": 2.1315, + "step": 1299 + }, + { + "epoch": 0.2725938351855735, + "grad_norm": 6.130130225751224, + "learning_rate": 1.980666402224392e-05, + "loss": 2.4333, + "step": 1300 + }, + { + "epoch": 0.27280352275110087, + "grad_norm": 5.476543252682569, + "learning_rate": 1.9806220758126263e-05, + "loss": 2.2098, + "step": 1301 + }, + { + "epoch": 0.27301321031662823, + "grad_norm": 5.8557984383039665, + "learning_rate": 1.9805776991422788e-05, + "loss": 2.3436, + "step": 1302 + }, + { + "epoch": 0.2732228978821556, + "grad_norm": 4.345406756206948, + "learning_rate": 1.980533272215623e-05, + "loss": 2.281, + "step": 1303 + }, + { + "epoch": 0.27343258544768295, + "grad_norm": 6.569083837710685, + "learning_rate": 1.9804887950349366e-05, + "loss": 2.6885, + "step": 1304 + }, + { + "epoch": 0.2736422730132103, + "grad_norm": 4.824719829443905, + "learning_rate": 1.9804442676024986e-05, + "loss": 2.2335, + "step": 1305 + }, + { + "epoch": 0.2738519605787377, + "grad_norm": 5.610564787907202, + "learning_rate": 1.9803996899205916e-05, + "loss": 2.4043, + "step": 1306 + }, + { + "epoch": 0.27406164814426504, + "grad_norm": 5.262691627518141, + "learning_rate": 1.9803550619915e-05, + "loss": 2.3232, + "step": 1307 + }, + { + "epoch": 0.2742713357097924, + "grad_norm": 5.963194042368016, + "learning_rate": 1.9803103838175112e-05, + "loss": 2.2867, + "step": 1308 + }, + { + "epoch": 0.27448102327531976, + "grad_norm": 5.317531294803183, + "learning_rate": 1.980265655400915e-05, + "loss": 2.299, + "step": 1309 + }, + { + "epoch": 0.2746907108408471, + "grad_norm": 5.461498148664218, + "learning_rate": 1.9802208767440033e-05, + "loss": 2.2299, + "step": 1310 + }, + { + "epoch": 0.2749003984063745, + "grad_norm": 5.339881268528538, + "learning_rate": 1.9801760478490722e-05, + "loss": 2.5616, + "step": 1311 + }, + { + "epoch": 0.27511008597190184, + "grad_norm": 5.073466426714328, + "learning_rate": 1.9801311687184182e-05, + "loss": 2.3906, + "step": 1312 + }, + { + "epoch": 0.27531977353742926, + "grad_norm": 5.755291216600122, + "learning_rate": 1.980086239354342e-05, + "loss": 2.4414, + "step": 1313 + }, + { + "epoch": 0.2755294611029566, + "grad_norm": 5.095370180899617, + "learning_rate": 1.9800412597591462e-05, + "loss": 2.4344, + "step": 1314 + }, + { + "epoch": 0.275739148668484, + "grad_norm": 5.8522030111540575, + "learning_rate": 1.9799962299351364e-05, + "loss": 2.3958, + "step": 1315 + }, + { + "epoch": 0.27594883623401134, + "grad_norm": 5.027363512027019, + "learning_rate": 1.9799511498846193e-05, + "loss": 2.2256, + "step": 1316 + }, + { + "epoch": 0.2761585237995387, + "grad_norm": 4.66448394610109, + "learning_rate": 1.9799060196099066e-05, + "loss": 2.2794, + "step": 1317 + }, + { + "epoch": 0.27636821136506606, + "grad_norm": 5.49077036740618, + "learning_rate": 1.979860839113311e-05, + "loss": 2.386, + "step": 1318 + }, + { + "epoch": 0.2765778989305934, + "grad_norm": 6.534615289486747, + "learning_rate": 1.9798156083971474e-05, + "loss": 2.3167, + "step": 1319 + }, + { + "epoch": 0.2767875864961208, + "grad_norm": 4.908776706378267, + "learning_rate": 1.9797703274637348e-05, + "loss": 2.2464, + "step": 1320 + }, + { + "epoch": 0.27699727406164815, + "grad_norm": 5.154659062106493, + "learning_rate": 1.9797249963153936e-05, + "loss": 2.4426, + "step": 1321 + }, + { + "epoch": 0.2772069616271755, + "grad_norm": 5.135014652381726, + "learning_rate": 1.9796796149544472e-05, + "loss": 2.4183, + "step": 1322 + }, + { + "epoch": 0.27741664919270287, + "grad_norm": 5.512941845904786, + "learning_rate": 1.979634183383221e-05, + "loss": 2.2915, + "step": 1323 + }, + { + "epoch": 0.27762633675823023, + "grad_norm": 5.316121042051358, + "learning_rate": 1.979588701604044e-05, + "loss": 2.5303, + "step": 1324 + }, + { + "epoch": 0.2778360243237576, + "grad_norm": 5.832968606183224, + "learning_rate": 1.9795431696192475e-05, + "loss": 2.3015, + "step": 1325 + }, + { + "epoch": 0.27804571188928495, + "grad_norm": 5.184487745454246, + "learning_rate": 1.979497587431164e-05, + "loss": 2.3555, + "step": 1326 + }, + { + "epoch": 0.2782553994548123, + "grad_norm": 5.232871375189829, + "learning_rate": 1.9794519550421305e-05, + "loss": 2.2517, + "step": 1327 + }, + { + "epoch": 0.2784650870203397, + "grad_norm": 5.5910032617212, + "learning_rate": 1.9794062724544855e-05, + "loss": 2.0663, + "step": 1328 + }, + { + "epoch": 0.27867477458586704, + "grad_norm": 5.753176372874745, + "learning_rate": 1.9793605396705706e-05, + "loss": 2.4483, + "step": 1329 + }, + { + "epoch": 0.2788844621513944, + "grad_norm": 6.848631648726224, + "learning_rate": 1.979314756692729e-05, + "loss": 2.5071, + "step": 1330 + }, + { + "epoch": 0.2790941497169218, + "grad_norm": 5.706418835675507, + "learning_rate": 1.9792689235233075e-05, + "loss": 2.1219, + "step": 1331 + }, + { + "epoch": 0.2793038372824492, + "grad_norm": 5.833301559624962, + "learning_rate": 1.9792230401646556e-05, + "loss": 2.4436, + "step": 1332 + }, + { + "epoch": 0.27951352484797654, + "grad_norm": 5.532161417154169, + "learning_rate": 1.9791771066191244e-05, + "loss": 2.4784, + "step": 1333 + }, + { + "epoch": 0.2797232124135039, + "grad_norm": 5.951944046329034, + "learning_rate": 1.9791311228890682e-05, + "loss": 2.3727, + "step": 1334 + }, + { + "epoch": 0.27993289997903126, + "grad_norm": 5.629058746810909, + "learning_rate": 1.9790850889768436e-05, + "loss": 2.635, + "step": 1335 + }, + { + "epoch": 0.2801425875445586, + "grad_norm": 6.3772229118586505, + "learning_rate": 1.9790390048848104e-05, + "loss": 2.4768, + "step": 1336 + }, + { + "epoch": 0.280352275110086, + "grad_norm": 6.9499510630635974, + "learning_rate": 1.9789928706153297e-05, + "loss": 2.392, + "step": 1337 + }, + { + "epoch": 0.28056196267561334, + "grad_norm": 5.175022399981712, + "learning_rate": 1.9789466861707666e-05, + "loss": 2.5344, + "step": 1338 + }, + { + "epoch": 0.2807716502411407, + "grad_norm": 5.267937924448901, + "learning_rate": 1.978900451553488e-05, + "loss": 2.3307, + "step": 1339 + }, + { + "epoch": 0.28098133780666806, + "grad_norm": 5.876642090105805, + "learning_rate": 1.9788541667658638e-05, + "loss": 2.4386, + "step": 1340 + }, + { + "epoch": 0.2811910253721954, + "grad_norm": 6.409002022908709, + "learning_rate": 1.9788078318102652e-05, + "loss": 2.3743, + "step": 1341 + }, + { + "epoch": 0.2814007129377228, + "grad_norm": 6.38620051212082, + "learning_rate": 1.9787614466890678e-05, + "loss": 2.7258, + "step": 1342 + }, + { + "epoch": 0.28161040050325015, + "grad_norm": 5.79571050887188, + "learning_rate": 1.9787150114046487e-05, + "loss": 2.4201, + "step": 1343 + }, + { + "epoch": 0.2818200880687775, + "grad_norm": 6.101610185361124, + "learning_rate": 1.9786685259593877e-05, + "loss": 2.3834, + "step": 1344 + }, + { + "epoch": 0.28202977563430487, + "grad_norm": 5.823266309508754, + "learning_rate": 1.9786219903556675e-05, + "loss": 2.3987, + "step": 1345 + }, + { + "epoch": 0.28223946319983223, + "grad_norm": 5.769019618170738, + "learning_rate": 1.9785754045958727e-05, + "loss": 2.3386, + "step": 1346 + }, + { + "epoch": 0.2824491507653596, + "grad_norm": 5.496938248960908, + "learning_rate": 1.9785287686823912e-05, + "loss": 2.27, + "step": 1347 + }, + { + "epoch": 0.28265883833088695, + "grad_norm": 7.040263575470214, + "learning_rate": 1.9784820826176132e-05, + "loss": 2.7362, + "step": 1348 + }, + { + "epoch": 0.28286852589641437, + "grad_norm": 5.237667922569154, + "learning_rate": 1.9784353464039314e-05, + "loss": 2.3929, + "step": 1349 + }, + { + "epoch": 0.28307821346194173, + "grad_norm": 6.12180871962915, + "learning_rate": 1.9783885600437412e-05, + "loss": 2.4483, + "step": 1350 + }, + { + "epoch": 0.2832879010274691, + "grad_norm": 5.841388344868005, + "learning_rate": 1.97834172353944e-05, + "loss": 2.4532, + "step": 1351 + }, + { + "epoch": 0.28349758859299645, + "grad_norm": 6.830689426606595, + "learning_rate": 1.978294836893429e-05, + "loss": 2.5169, + "step": 1352 + }, + { + "epoch": 0.2837072761585238, + "grad_norm": 5.406962651623831, + "learning_rate": 1.9782479001081107e-05, + "loss": 2.4342, + "step": 1353 + }, + { + "epoch": 0.2839169637240512, + "grad_norm": 5.287942940600494, + "learning_rate": 1.978200913185891e-05, + "loss": 2.1069, + "step": 1354 + }, + { + "epoch": 0.28412665128957854, + "grad_norm": 5.480016943633358, + "learning_rate": 1.9781538761291777e-05, + "loss": 2.5337, + "step": 1355 + }, + { + "epoch": 0.2843363388551059, + "grad_norm": 5.063895214655192, + "learning_rate": 1.9781067889403815e-05, + "loss": 2.154, + "step": 1356 + }, + { + "epoch": 0.28454602642063326, + "grad_norm": 5.439974558336936, + "learning_rate": 1.9780596516219165e-05, + "loss": 2.4046, + "step": 1357 + }, + { + "epoch": 0.2847557139861606, + "grad_norm": 5.823023622451933, + "learning_rate": 1.9780124641761976e-05, + "loss": 2.3336, + "step": 1358 + }, + { + "epoch": 0.284965401551688, + "grad_norm": 5.030387248568188, + "learning_rate": 1.9779652266056435e-05, + "loss": 2.1636, + "step": 1359 + }, + { + "epoch": 0.28517508911721534, + "grad_norm": 5.725925329443499, + "learning_rate": 1.9779179389126758e-05, + "loss": 2.3717, + "step": 1360 + }, + { + "epoch": 0.2853847766827427, + "grad_norm": 5.757928748102482, + "learning_rate": 1.977870601099717e-05, + "loss": 2.4127, + "step": 1361 + }, + { + "epoch": 0.28559446424827006, + "grad_norm": 5.375997260570959, + "learning_rate": 1.9778232131691947e-05, + "loss": 2.4381, + "step": 1362 + }, + { + "epoch": 0.2858041518137974, + "grad_norm": 5.5983631364185875, + "learning_rate": 1.977775775123536e-05, + "loss": 2.3248, + "step": 1363 + }, + { + "epoch": 0.2860138393793248, + "grad_norm": 5.207252969281748, + "learning_rate": 1.9777282869651734e-05, + "loss": 2.3333, + "step": 1364 + }, + { + "epoch": 0.28622352694485215, + "grad_norm": 5.679604890805039, + "learning_rate": 1.9776807486965404e-05, + "loss": 2.5376, + "step": 1365 + }, + { + "epoch": 0.2864332145103795, + "grad_norm": 5.16206243131396, + "learning_rate": 1.9776331603200728e-05, + "loss": 2.1735, + "step": 1366 + }, + { + "epoch": 0.2866429020759069, + "grad_norm": 4.679696428349341, + "learning_rate": 1.9775855218382104e-05, + "loss": 2.3827, + "step": 1367 + }, + { + "epoch": 0.2868525896414343, + "grad_norm": 5.840075579797279, + "learning_rate": 1.9775378332533948e-05, + "loss": 2.2297, + "step": 1368 + }, + { + "epoch": 0.28706227720696165, + "grad_norm": 6.893413814083934, + "learning_rate": 1.9774900945680694e-05, + "loss": 2.0948, + "step": 1369 + }, + { + "epoch": 0.287271964772489, + "grad_norm": 5.296062084816024, + "learning_rate": 1.9774423057846812e-05, + "loss": 2.35, + "step": 1370 + }, + { + "epoch": 0.28748165233801637, + "grad_norm": 5.75696701292084, + "learning_rate": 1.97739446690568e-05, + "loss": 2.524, + "step": 1371 + }, + { + "epoch": 0.28769133990354373, + "grad_norm": 6.774835984463509, + "learning_rate": 1.9773465779335168e-05, + "loss": 2.251, + "step": 1372 + }, + { + "epoch": 0.2879010274690711, + "grad_norm": 5.918354258554908, + "learning_rate": 1.9772986388706466e-05, + "loss": 2.6306, + "step": 1373 + }, + { + "epoch": 0.28811071503459845, + "grad_norm": 5.484237689896832, + "learning_rate": 1.9772506497195256e-05, + "loss": 2.3988, + "step": 1374 + }, + { + "epoch": 0.2883204026001258, + "grad_norm": 6.18271467119657, + "learning_rate": 1.9772026104826146e-05, + "loss": 2.269, + "step": 1375 + }, + { + "epoch": 0.2885300901656532, + "grad_norm": 5.073049190896248, + "learning_rate": 1.9771545211623743e-05, + "loss": 2.3343, + "step": 1376 + }, + { + "epoch": 0.28873977773118054, + "grad_norm": 5.521289749311444, + "learning_rate": 1.9771063817612704e-05, + "loss": 2.4625, + "step": 1377 + }, + { + "epoch": 0.2889494652967079, + "grad_norm": 5.010207279435596, + "learning_rate": 1.9770581922817695e-05, + "loss": 2.2395, + "step": 1378 + }, + { + "epoch": 0.28915915286223526, + "grad_norm": 5.677074069196393, + "learning_rate": 1.9770099527263417e-05, + "loss": 2.5001, + "step": 1379 + }, + { + "epoch": 0.2893688404277626, + "grad_norm": 6.312993090262843, + "learning_rate": 1.9769616630974594e-05, + "loss": 2.3592, + "step": 1380 + }, + { + "epoch": 0.28957852799329, + "grad_norm": 5.276873511139441, + "learning_rate": 1.976913323397597e-05, + "loss": 2.2, + "step": 1381 + }, + { + "epoch": 0.28978821555881734, + "grad_norm": 5.188193178326891, + "learning_rate": 1.9768649336292327e-05, + "loss": 2.5627, + "step": 1382 + }, + { + "epoch": 0.2899979031243447, + "grad_norm": 5.077044232057226, + "learning_rate": 1.9768164937948463e-05, + "loss": 2.2167, + "step": 1383 + }, + { + "epoch": 0.29020759068987206, + "grad_norm": 5.304612498918937, + "learning_rate": 1.9767680038969207e-05, + "loss": 2.3809, + "step": 1384 + }, + { + "epoch": 0.2904172782553995, + "grad_norm": 5.075595825045641, + "learning_rate": 1.9767194639379405e-05, + "loss": 2.2904, + "step": 1385 + }, + { + "epoch": 0.29062696582092684, + "grad_norm": 5.545721663532519, + "learning_rate": 1.9766708739203935e-05, + "loss": 2.4336, + "step": 1386 + }, + { + "epoch": 0.2908366533864542, + "grad_norm": 5.264407927054516, + "learning_rate": 1.9766222338467705e-05, + "loss": 2.3763, + "step": 1387 + }, + { + "epoch": 0.29104634095198156, + "grad_norm": 5.080513593407458, + "learning_rate": 1.9765735437195638e-05, + "loss": 2.2608, + "step": 1388 + }, + { + "epoch": 0.2912560285175089, + "grad_norm": 5.1577538582083635, + "learning_rate": 1.9765248035412696e-05, + "loss": 2.3994, + "step": 1389 + }, + { + "epoch": 0.2914657160830363, + "grad_norm": 5.187958189933125, + "learning_rate": 1.9764760133143857e-05, + "loss": 2.2776, + "step": 1390 + }, + { + "epoch": 0.29167540364856365, + "grad_norm": 7.465843409221242, + "learning_rate": 1.9764271730414122e-05, + "loss": 2.8008, + "step": 1391 + }, + { + "epoch": 0.291885091214091, + "grad_norm": 5.463369237727748, + "learning_rate": 1.9763782827248527e-05, + "loss": 2.4756, + "step": 1392 + }, + { + "epoch": 0.29209477877961837, + "grad_norm": 5.702296479020473, + "learning_rate": 1.9763293423672123e-05, + "loss": 2.3126, + "step": 1393 + }, + { + "epoch": 0.29230446634514573, + "grad_norm": 5.464472570622636, + "learning_rate": 1.9762803519710002e-05, + "loss": 2.4072, + "step": 1394 + }, + { + "epoch": 0.2925141539106731, + "grad_norm": 5.334102500625808, + "learning_rate": 1.9762313115387268e-05, + "loss": 2.53, + "step": 1395 + }, + { + "epoch": 0.29272384147620045, + "grad_norm": 5.471228197868595, + "learning_rate": 1.9761822210729052e-05, + "loss": 2.1569, + "step": 1396 + }, + { + "epoch": 0.2929335290417278, + "grad_norm": 4.937720654859672, + "learning_rate": 1.976133080576052e-05, + "loss": 2.1694, + "step": 1397 + }, + { + "epoch": 0.2931432166072552, + "grad_norm": 4.893923206418719, + "learning_rate": 1.976083890050685e-05, + "loss": 2.3735, + "step": 1398 + }, + { + "epoch": 0.29335290417278254, + "grad_norm": 6.8811178992586175, + "learning_rate": 1.976034649499326e-05, + "loss": 2.5713, + "step": 1399 + }, + { + "epoch": 0.2935625917383099, + "grad_norm": 5.492288795082419, + "learning_rate": 1.9759853589244984e-05, + "loss": 2.3401, + "step": 1400 + }, + { + "epoch": 0.29377227930383726, + "grad_norm": 5.821463970397368, + "learning_rate": 1.9759360183287284e-05, + "loss": 2.2034, + "step": 1401 + }, + { + "epoch": 0.2939819668693646, + "grad_norm": 5.366478947306584, + "learning_rate": 1.9758866277145445e-05, + "loss": 2.0132, + "step": 1402 + }, + { + "epoch": 0.29419165443489204, + "grad_norm": 5.246015318044939, + "learning_rate": 1.9758371870844786e-05, + "loss": 2.435, + "step": 1403 + }, + { + "epoch": 0.2944013420004194, + "grad_norm": 5.864002220307482, + "learning_rate": 1.9757876964410643e-05, + "loss": 2.3031, + "step": 1404 + }, + { + "epoch": 0.29461102956594676, + "grad_norm": 6.511298799388658, + "learning_rate": 1.9757381557868384e-05, + "loss": 2.4059, + "step": 1405 + }, + { + "epoch": 0.2948207171314741, + "grad_norm": 5.624179605917954, + "learning_rate": 1.9756885651243395e-05, + "loss": 2.4428, + "step": 1406 + }, + { + "epoch": 0.2950304046970015, + "grad_norm": 5.547381298034921, + "learning_rate": 1.975638924456109e-05, + "loss": 2.3485, + "step": 1407 + }, + { + "epoch": 0.29524009226252884, + "grad_norm": 6.130195742780277, + "learning_rate": 1.975589233784692e-05, + "loss": 2.5537, + "step": 1408 + }, + { + "epoch": 0.2954497798280562, + "grad_norm": 5.924402322305737, + "learning_rate": 1.9755394931126345e-05, + "loss": 1.9947, + "step": 1409 + }, + { + "epoch": 0.29565946739358356, + "grad_norm": 5.112370655468548, + "learning_rate": 1.9754897024424857e-05, + "loss": 2.2103, + "step": 1410 + }, + { + "epoch": 0.2958691549591109, + "grad_norm": 5.9626419276039, + "learning_rate": 1.9754398617767982e-05, + "loss": 2.5006, + "step": 1411 + }, + { + "epoch": 0.2960788425246383, + "grad_norm": 5.970263228310468, + "learning_rate": 1.9753899711181258e-05, + "loss": 2.3722, + "step": 1412 + }, + { + "epoch": 0.29628853009016565, + "grad_norm": 5.7557982346698635, + "learning_rate": 1.9753400304690256e-05, + "loss": 2.4389, + "step": 1413 + }, + { + "epoch": 0.296498217655693, + "grad_norm": 5.3759009942366, + "learning_rate": 1.9752900398320573e-05, + "loss": 2.5187, + "step": 1414 + }, + { + "epoch": 0.29670790522122037, + "grad_norm": 6.178018382512329, + "learning_rate": 1.9752399992097827e-05, + "loss": 2.5569, + "step": 1415 + }, + { + "epoch": 0.29691759278674773, + "grad_norm": 5.11597217931808, + "learning_rate": 1.975189908604767e-05, + "loss": 2.3324, + "step": 1416 + }, + { + "epoch": 0.2971272803522751, + "grad_norm": 5.759278113157731, + "learning_rate": 1.975139768019577e-05, + "loss": 2.3471, + "step": 1417 + }, + { + "epoch": 0.29733696791780245, + "grad_norm": 5.333804587512334, + "learning_rate": 1.9750895774567822e-05, + "loss": 2.3706, + "step": 1418 + }, + { + "epoch": 0.2975466554833298, + "grad_norm": 5.450258079150982, + "learning_rate": 1.9750393369189556e-05, + "loss": 2.3107, + "step": 1419 + }, + { + "epoch": 0.29775634304885723, + "grad_norm": 5.847153736487412, + "learning_rate": 1.974989046408672e-05, + "loss": 2.134, + "step": 1420 + }, + { + "epoch": 0.2979660306143846, + "grad_norm": 5.29119410342301, + "learning_rate": 1.9749387059285085e-05, + "loss": 2.2594, + "step": 1421 + }, + { + "epoch": 0.29817571817991195, + "grad_norm": 5.724515865347008, + "learning_rate": 1.974888315481046e-05, + "loss": 2.2891, + "step": 1422 + }, + { + "epoch": 0.2983854057454393, + "grad_norm": 6.034746892094913, + "learning_rate": 1.9748378750688657e-05, + "loss": 2.3084, + "step": 1423 + }, + { + "epoch": 0.2985950933109667, + "grad_norm": 5.377297012841492, + "learning_rate": 1.9747873846945537e-05, + "loss": 1.9538, + "step": 1424 + }, + { + "epoch": 0.29880478087649404, + "grad_norm": 5.458174296259753, + "learning_rate": 1.974736844360698e-05, + "loss": 2.3644, + "step": 1425 + }, + { + "epoch": 0.2990144684420214, + "grad_norm": 6.1247899748716925, + "learning_rate": 1.9746862540698882e-05, + "loss": 2.1765, + "step": 1426 + }, + { + "epoch": 0.29922415600754876, + "grad_norm": 5.7119420216059185, + "learning_rate": 1.9746356138247172e-05, + "loss": 2.5048, + "step": 1427 + }, + { + "epoch": 0.2994338435730761, + "grad_norm": 6.398661353954476, + "learning_rate": 1.9745849236277805e-05, + "loss": 2.2233, + "step": 1428 + }, + { + "epoch": 0.2996435311386035, + "grad_norm": 5.430975759887312, + "learning_rate": 1.974534183481676e-05, + "loss": 2.3021, + "step": 1429 + }, + { + "epoch": 0.29985321870413084, + "grad_norm": 6.358817732784684, + "learning_rate": 1.9744833933890048e-05, + "loss": 2.2365, + "step": 1430 + }, + { + "epoch": 0.3000629062696582, + "grad_norm": 5.2373038692391924, + "learning_rate": 1.9744325533523697e-05, + "loss": 2.2837, + "step": 1431 + }, + { + "epoch": 0.30027259383518556, + "grad_norm": 5.444389155516945, + "learning_rate": 1.974381663374376e-05, + "loss": 2.311, + "step": 1432 + }, + { + "epoch": 0.3004822814007129, + "grad_norm": 5.230672184753493, + "learning_rate": 1.9743307234576315e-05, + "loss": 2.2068, + "step": 1433 + }, + { + "epoch": 0.3006919689662403, + "grad_norm": 4.93466298469086, + "learning_rate": 1.9742797336047483e-05, + "loss": 2.3903, + "step": 1434 + }, + { + "epoch": 0.30090165653176765, + "grad_norm": 5.512805175024408, + "learning_rate": 1.9742286938183383e-05, + "loss": 2.431, + "step": 1435 + }, + { + "epoch": 0.301111344097295, + "grad_norm": 4.83193287888066, + "learning_rate": 1.9741776041010183e-05, + "loss": 2.1003, + "step": 1436 + }, + { + "epoch": 0.30132103166282237, + "grad_norm": 6.1005931067695585, + "learning_rate": 1.9741264644554067e-05, + "loss": 2.3219, + "step": 1437 + }, + { + "epoch": 0.3015307192283498, + "grad_norm": 5.0592401441334225, + "learning_rate": 1.974075274884124e-05, + "loss": 2.3954, + "step": 1438 + }, + { + "epoch": 0.30174040679387715, + "grad_norm": 4.994985346391976, + "learning_rate": 1.974024035389794e-05, + "loss": 2.4175, + "step": 1439 + }, + { + "epoch": 0.3019500943594045, + "grad_norm": 5.911793106618969, + "learning_rate": 1.973972745975043e-05, + "loss": 2.3551, + "step": 1440 + }, + { + "epoch": 0.30215978192493187, + "grad_norm": 6.133682346007816, + "learning_rate": 1.9739214066424992e-05, + "loss": 2.3435, + "step": 1441 + }, + { + "epoch": 0.30236946949045923, + "grad_norm": 6.083464090655175, + "learning_rate": 1.9738700173947944e-05, + "loss": 2.5624, + "step": 1442 + }, + { + "epoch": 0.3025791570559866, + "grad_norm": 5.770741610649502, + "learning_rate": 1.9738185782345623e-05, + "loss": 2.5611, + "step": 1443 + }, + { + "epoch": 0.30278884462151395, + "grad_norm": 5.183926297317286, + "learning_rate": 1.9737670891644387e-05, + "loss": 2.089, + "step": 1444 + }, + { + "epoch": 0.3029985321870413, + "grad_norm": 5.611065594858339, + "learning_rate": 1.973715550187063e-05, + "loss": 2.1529, + "step": 1445 + }, + { + "epoch": 0.3032082197525687, + "grad_norm": 5.034184357449487, + "learning_rate": 1.973663961305076e-05, + "loss": 2.1034, + "step": 1446 + }, + { + "epoch": 0.30341790731809604, + "grad_norm": 7.041873809217281, + "learning_rate": 1.9736123225211233e-05, + "loss": 2.6767, + "step": 1447 + }, + { + "epoch": 0.3036275948836234, + "grad_norm": 5.409772700612872, + "learning_rate": 1.97356063383785e-05, + "loss": 2.5794, + "step": 1448 + }, + { + "epoch": 0.30383728244915076, + "grad_norm": 5.806629665273312, + "learning_rate": 1.9735088952579054e-05, + "loss": 2.2354, + "step": 1449 + }, + { + "epoch": 0.3040469700146781, + "grad_norm": 6.015174753195954, + "learning_rate": 1.9734571067839417e-05, + "loss": 2.1491, + "step": 1450 + }, + { + "epoch": 0.3042566575802055, + "grad_norm": 5.474443396407675, + "learning_rate": 1.9734052684186123e-05, + "loss": 2.0347, + "step": 1451 + }, + { + "epoch": 0.30446634514573284, + "grad_norm": 6.699735488091636, + "learning_rate": 1.9733533801645754e-05, + "loss": 2.2523, + "step": 1452 + }, + { + "epoch": 0.3046760327112602, + "grad_norm": 5.7365581935845205, + "learning_rate": 1.9733014420244894e-05, + "loss": 2.7728, + "step": 1453 + }, + { + "epoch": 0.30488572027678756, + "grad_norm": 5.34055720734635, + "learning_rate": 1.973249454001016e-05, + "loss": 2.2638, + "step": 1454 + }, + { + "epoch": 0.3050954078423149, + "grad_norm": 5.275955633358409, + "learning_rate": 1.9731974160968206e-05, + "loss": 2.3764, + "step": 1455 + }, + { + "epoch": 0.30530509540784234, + "grad_norm": 5.338190496900929, + "learning_rate": 1.9731453283145693e-05, + "loss": 2.4801, + "step": 1456 + }, + { + "epoch": 0.3055147829733697, + "grad_norm": 5.327515316400625, + "learning_rate": 1.973093190656932e-05, + "loss": 2.2876, + "step": 1457 + }, + { + "epoch": 0.30572447053889706, + "grad_norm": 6.607126768932123, + "learning_rate": 1.9730410031265813e-05, + "loss": 2.1421, + "step": 1458 + }, + { + "epoch": 0.3059341581044244, + "grad_norm": 5.124583826743587, + "learning_rate": 1.9729887657261914e-05, + "loss": 1.9938, + "step": 1459 + }, + { + "epoch": 0.3061438456699518, + "grad_norm": 4.966580818159501, + "learning_rate": 1.9729364784584396e-05, + "loss": 2.5906, + "step": 1460 + }, + { + "epoch": 0.30635353323547915, + "grad_norm": 5.366111213305886, + "learning_rate": 1.972884141326006e-05, + "loss": 2.0271, + "step": 1461 + }, + { + "epoch": 0.3065632208010065, + "grad_norm": 5.5833115204001285, + "learning_rate": 1.9728317543315727e-05, + "loss": 2.4378, + "step": 1462 + }, + { + "epoch": 0.30677290836653387, + "grad_norm": 5.396520349249426, + "learning_rate": 1.9727793174778247e-05, + "loss": 2.3029, + "step": 1463 + }, + { + "epoch": 0.30698259593206123, + "grad_norm": 5.473257724899315, + "learning_rate": 1.9727268307674495e-05, + "loss": 2.405, + "step": 1464 + }, + { + "epoch": 0.3071922834975886, + "grad_norm": 5.979601176544394, + "learning_rate": 1.972674294203137e-05, + "loss": 2.1358, + "step": 1465 + }, + { + "epoch": 0.30740197106311595, + "grad_norm": 5.264328114008282, + "learning_rate": 1.97262170778758e-05, + "loss": 2.4749, + "step": 1466 + }, + { + "epoch": 0.3076116586286433, + "grad_norm": 5.9035118195019125, + "learning_rate": 1.9725690715234735e-05, + "loss": 2.4332, + "step": 1467 + }, + { + "epoch": 0.3078213461941707, + "grad_norm": 5.489031807224302, + "learning_rate": 1.9725163854135153e-05, + "loss": 2.2876, + "step": 1468 + }, + { + "epoch": 0.30803103375969804, + "grad_norm": 5.536941271034893, + "learning_rate": 1.9724636494604055e-05, + "loss": 2.5461, + "step": 1469 + }, + { + "epoch": 0.3082407213252254, + "grad_norm": 5.455906373338647, + "learning_rate": 1.972410863666847e-05, + "loss": 2.3427, + "step": 1470 + }, + { + "epoch": 0.30845040889075276, + "grad_norm": 5.977633632870943, + "learning_rate": 1.9723580280355455e-05, + "loss": 2.3815, + "step": 1471 + }, + { + "epoch": 0.3086600964562801, + "grad_norm": 5.5493686692025275, + "learning_rate": 1.972305142569208e-05, + "loss": 2.3138, + "step": 1472 + }, + { + "epoch": 0.3088697840218075, + "grad_norm": 6.107626901014838, + "learning_rate": 1.972252207270546e-05, + "loss": 2.4634, + "step": 1473 + }, + { + "epoch": 0.3090794715873349, + "grad_norm": 6.255474493494085, + "learning_rate": 1.972199222142272e-05, + "loss": 2.0997, + "step": 1474 + }, + { + "epoch": 0.30928915915286226, + "grad_norm": 6.8162714043097905, + "learning_rate": 1.9721461871871017e-05, + "loss": 2.312, + "step": 1475 + }, + { + "epoch": 0.3094988467183896, + "grad_norm": 5.288022117194409, + "learning_rate": 1.9720931024077533e-05, + "loss": 2.3113, + "step": 1476 + }, + { + "epoch": 0.309708534283917, + "grad_norm": 5.663020110072521, + "learning_rate": 1.9720399678069474e-05, + "loss": 2.4339, + "step": 1477 + }, + { + "epoch": 0.30991822184944434, + "grad_norm": 5.121772898756945, + "learning_rate": 1.971986783387407e-05, + "loss": 2.2952, + "step": 1478 + }, + { + "epoch": 0.3101279094149717, + "grad_norm": 5.347115126345129, + "learning_rate": 1.971933549151858e-05, + "loss": 2.2576, + "step": 1479 + }, + { + "epoch": 0.31033759698049906, + "grad_norm": 5.413061789510903, + "learning_rate": 1.971880265103029e-05, + "loss": 2.4407, + "step": 1480 + }, + { + "epoch": 0.3105472845460264, + "grad_norm": 5.172916408616316, + "learning_rate": 1.971826931243651e-05, + "loss": 2.3881, + "step": 1481 + }, + { + "epoch": 0.3107569721115538, + "grad_norm": 5.384918675099962, + "learning_rate": 1.971773547576457e-05, + "loss": 2.1328, + "step": 1482 + }, + { + "epoch": 0.31096665967708115, + "grad_norm": 4.589607753773831, + "learning_rate": 1.9717201141041835e-05, + "loss": 2.3187, + "step": 1483 + }, + { + "epoch": 0.3111763472426085, + "grad_norm": 4.952098214472724, + "learning_rate": 1.9716666308295687e-05, + "loss": 2.5112, + "step": 1484 + }, + { + "epoch": 0.31138603480813587, + "grad_norm": 6.319063293801881, + "learning_rate": 1.9716130977553536e-05, + "loss": 2.3032, + "step": 1485 + }, + { + "epoch": 0.31159572237366323, + "grad_norm": 6.08533099718583, + "learning_rate": 1.971559514884282e-05, + "loss": 2.5791, + "step": 1486 + }, + { + "epoch": 0.3118054099391906, + "grad_norm": 6.2827158548655655, + "learning_rate": 1.9715058822191008e-05, + "loss": 2.5598, + "step": 1487 + }, + { + "epoch": 0.31201509750471795, + "grad_norm": 5.504699670486899, + "learning_rate": 1.9714521997625576e-05, + "loss": 2.4408, + "step": 1488 + }, + { + "epoch": 0.3122247850702453, + "grad_norm": 5.86624997625924, + "learning_rate": 1.9713984675174042e-05, + "loss": 2.3816, + "step": 1489 + }, + { + "epoch": 0.3124344726357727, + "grad_norm": 6.224865164676491, + "learning_rate": 1.9713446854863946e-05, + "loss": 2.3346, + "step": 1490 + }, + { + "epoch": 0.31264416020130004, + "grad_norm": 5.091888438800018, + "learning_rate": 1.971290853672285e-05, + "loss": 2.4144, + "step": 1491 + }, + { + "epoch": 0.31285384776682745, + "grad_norm": 6.150932119184296, + "learning_rate": 1.971236972077835e-05, + "loss": 2.3949, + "step": 1492 + }, + { + "epoch": 0.3130635353323548, + "grad_norm": 4.846635162830004, + "learning_rate": 1.9711830407058056e-05, + "loss": 2.2806, + "step": 1493 + }, + { + "epoch": 0.3132732228978822, + "grad_norm": 5.25417386387375, + "learning_rate": 1.9711290595589606e-05, + "loss": 2.5214, + "step": 1494 + }, + { + "epoch": 0.31348291046340954, + "grad_norm": 5.794325600518844, + "learning_rate": 1.971075028640067e-05, + "loss": 2.3367, + "step": 1495 + }, + { + "epoch": 0.3136925980289369, + "grad_norm": 4.948303636813416, + "learning_rate": 1.9710209479518942e-05, + "loss": 2.1773, + "step": 1496 + }, + { + "epoch": 0.31390228559446426, + "grad_norm": 5.345897867183844, + "learning_rate": 1.970966817497214e-05, + "loss": 2.1757, + "step": 1497 + }, + { + "epoch": 0.3141119731599916, + "grad_norm": 4.736279500731373, + "learning_rate": 1.9709126372787994e-05, + "loss": 2.3428, + "step": 1498 + }, + { + "epoch": 0.314321660725519, + "grad_norm": 4.9165498979709215, + "learning_rate": 1.9708584072994287e-05, + "loss": 2.2653, + "step": 1499 + }, + { + "epoch": 0.31453134829104634, + "grad_norm": 5.390229038885812, + "learning_rate": 1.970804127561881e-05, + "loss": 2.2994, + "step": 1500 + }, + { + "epoch": 0.3147410358565737, + "grad_norm": 5.9022811350477085, + "learning_rate": 1.9707497980689375e-05, + "loss": 2.4414, + "step": 1501 + }, + { + "epoch": 0.31495072342210106, + "grad_norm": 5.720020625309213, + "learning_rate": 1.9706954188233832e-05, + "loss": 2.1059, + "step": 1502 + }, + { + "epoch": 0.3151604109876284, + "grad_norm": 5.540632883318829, + "learning_rate": 1.9706409898280053e-05, + "loss": 2.3746, + "step": 1503 + }, + { + "epoch": 0.3153700985531558, + "grad_norm": 4.99106647903385, + "learning_rate": 1.9705865110855936e-05, + "loss": 2.3022, + "step": 1504 + }, + { + "epoch": 0.31557978611868315, + "grad_norm": 5.728424190867146, + "learning_rate": 1.970531982598939e-05, + "loss": 2.1741, + "step": 1505 + }, + { + "epoch": 0.3157894736842105, + "grad_norm": 5.641338266472313, + "learning_rate": 1.9704774043708378e-05, + "loss": 2.2423, + "step": 1506 + }, + { + "epoch": 0.31599916124973787, + "grad_norm": 6.360506019842943, + "learning_rate": 1.9704227764040856e-05, + "loss": 2.3389, + "step": 1507 + }, + { + "epoch": 0.31620884881526523, + "grad_norm": 5.77971291857449, + "learning_rate": 1.9703680987014836e-05, + "loss": 2.3239, + "step": 1508 + }, + { + "epoch": 0.3164185363807926, + "grad_norm": 5.688068680117667, + "learning_rate": 1.9703133712658334e-05, + "loss": 1.9481, + "step": 1509 + }, + { + "epoch": 0.31662822394632, + "grad_norm": 5.312862118132038, + "learning_rate": 1.9702585940999397e-05, + "loss": 2.2139, + "step": 1510 + }, + { + "epoch": 0.31683791151184737, + "grad_norm": 7.674997326080807, + "learning_rate": 1.9702037672066105e-05, + "loss": 2.4508, + "step": 1511 + }, + { + "epoch": 0.31704759907737473, + "grad_norm": 5.615429093840941, + "learning_rate": 1.9701488905886553e-05, + "loss": 2.3787, + "step": 1512 + }, + { + "epoch": 0.3172572866429021, + "grad_norm": 5.197316859027535, + "learning_rate": 1.970093964248887e-05, + "loss": 2.0848, + "step": 1513 + }, + { + "epoch": 0.31746697420842945, + "grad_norm": 5.689690188829284, + "learning_rate": 1.9700389881901203e-05, + "loss": 2.1779, + "step": 1514 + }, + { + "epoch": 0.3176766617739568, + "grad_norm": 5.930786295968636, + "learning_rate": 1.9699839624151732e-05, + "loss": 2.6955, + "step": 1515 + }, + { + "epoch": 0.3178863493394842, + "grad_norm": 5.751919466836633, + "learning_rate": 1.9699288869268657e-05, + "loss": 2.3673, + "step": 1516 + }, + { + "epoch": 0.31809603690501154, + "grad_norm": 5.741769380008591, + "learning_rate": 1.96987376172802e-05, + "loss": 2.5176, + "step": 1517 + }, + { + "epoch": 0.3183057244705389, + "grad_norm": 5.053275356963386, + "learning_rate": 1.9698185868214625e-05, + "loss": 2.4545, + "step": 1518 + }, + { + "epoch": 0.31851541203606626, + "grad_norm": 6.164444263708838, + "learning_rate": 1.9697633622100203e-05, + "loss": 2.6943, + "step": 1519 + }, + { + "epoch": 0.3187250996015936, + "grad_norm": 5.535265046041835, + "learning_rate": 1.9697080878965232e-05, + "loss": 2.3771, + "step": 1520 + }, + { + "epoch": 0.318934787167121, + "grad_norm": 5.243331805782426, + "learning_rate": 1.9696527638838053e-05, + "loss": 2.4186, + "step": 1521 + }, + { + "epoch": 0.31914447473264834, + "grad_norm": 5.608816302323981, + "learning_rate": 1.9695973901747015e-05, + "loss": 2.3495, + "step": 1522 + }, + { + "epoch": 0.3193541622981757, + "grad_norm": 5.441417244285983, + "learning_rate": 1.9695419667720497e-05, + "loss": 2.2835, + "step": 1523 + }, + { + "epoch": 0.31956384986370306, + "grad_norm": 4.645072259927482, + "learning_rate": 1.9694864936786905e-05, + "loss": 2.4053, + "step": 1524 + }, + { + "epoch": 0.3197735374292304, + "grad_norm": 5.538116879192616, + "learning_rate": 1.9694309708974664e-05, + "loss": 2.24, + "step": 1525 + }, + { + "epoch": 0.3199832249947578, + "grad_norm": 6.133448749236639, + "learning_rate": 1.9693753984312244e-05, + "loss": 2.2234, + "step": 1526 + }, + { + "epoch": 0.32019291256028515, + "grad_norm": 5.121488847265792, + "learning_rate": 1.9693197762828116e-05, + "loss": 2.3061, + "step": 1527 + }, + { + "epoch": 0.32040260012581256, + "grad_norm": 5.661960536524321, + "learning_rate": 1.9692641044550793e-05, + "loss": 2.3722, + "step": 1528 + }, + { + "epoch": 0.3206122876913399, + "grad_norm": 5.332912685738281, + "learning_rate": 1.9692083829508805e-05, + "loss": 2.3572, + "step": 1529 + }, + { + "epoch": 0.3208219752568673, + "grad_norm": 5.564093842628362, + "learning_rate": 1.969152611773071e-05, + "loss": 2.4022, + "step": 1530 + }, + { + "epoch": 0.32103166282239465, + "grad_norm": 5.753427394116873, + "learning_rate": 1.9690967909245093e-05, + "loss": 2.2846, + "step": 1531 + }, + { + "epoch": 0.321241350387922, + "grad_norm": 5.971292538542521, + "learning_rate": 1.969040920408056e-05, + "loss": 2.1369, + "step": 1532 + }, + { + "epoch": 0.32145103795344937, + "grad_norm": 6.050386304353231, + "learning_rate": 1.9689850002265748e-05, + "loss": 2.3976, + "step": 1533 + }, + { + "epoch": 0.32166072551897673, + "grad_norm": 5.7453821039470565, + "learning_rate": 1.968929030382932e-05, + "loss": 2.1402, + "step": 1534 + }, + { + "epoch": 0.3218704130845041, + "grad_norm": 5.3989866592591955, + "learning_rate": 1.9688730108799958e-05, + "loss": 2.4425, + "step": 1535 + }, + { + "epoch": 0.32208010065003145, + "grad_norm": 6.120850902242426, + "learning_rate": 1.9688169417206374e-05, + "loss": 2.2639, + "step": 1536 + }, + { + "epoch": 0.3222897882155588, + "grad_norm": 6.249207718670505, + "learning_rate": 1.9687608229077302e-05, + "loss": 2.3343, + "step": 1537 + }, + { + "epoch": 0.3224994757810862, + "grad_norm": 6.136818169893906, + "learning_rate": 1.9687046544441508e-05, + "loss": 2.3297, + "step": 1538 + }, + { + "epoch": 0.32270916334661354, + "grad_norm": 5.8256544480599075, + "learning_rate": 1.9686484363327774e-05, + "loss": 2.0781, + "step": 1539 + }, + { + "epoch": 0.3229188509121409, + "grad_norm": 6.106603338544645, + "learning_rate": 1.9685921685764923e-05, + "loss": 2.2792, + "step": 1540 + }, + { + "epoch": 0.32312853847766826, + "grad_norm": 6.391780487074041, + "learning_rate": 1.968535851178178e-05, + "loss": 2.5572, + "step": 1541 + }, + { + "epoch": 0.3233382260431956, + "grad_norm": 5.143244807163231, + "learning_rate": 1.9684794841407217e-05, + "loss": 2.2325, + "step": 1542 + }, + { + "epoch": 0.323547913608723, + "grad_norm": 5.134603730519258, + "learning_rate": 1.968423067467012e-05, + "loss": 2.0837, + "step": 1543 + }, + { + "epoch": 0.32375760117425034, + "grad_norm": 5.049928770051305, + "learning_rate": 1.9683666011599408e-05, + "loss": 2.2789, + "step": 1544 + }, + { + "epoch": 0.3239672887397777, + "grad_norm": 5.681812380984174, + "learning_rate": 1.9683100852224014e-05, + "loss": 2.3705, + "step": 1545 + }, + { + "epoch": 0.3241769763053051, + "grad_norm": 5.977128581745251, + "learning_rate": 1.9682535196572907e-05, + "loss": 2.4287, + "step": 1546 + }, + { + "epoch": 0.3243866638708325, + "grad_norm": 5.6009328101179525, + "learning_rate": 1.9681969044675076e-05, + "loss": 2.655, + "step": 1547 + }, + { + "epoch": 0.32459635143635984, + "grad_norm": 4.910291568982428, + "learning_rate": 1.9681402396559545e-05, + "loss": 2.4707, + "step": 1548 + }, + { + "epoch": 0.3248060390018872, + "grad_norm": 5.5895750952885175, + "learning_rate": 1.9680835252255345e-05, + "loss": 2.1304, + "step": 1549 + }, + { + "epoch": 0.32501572656741456, + "grad_norm": 5.418398241038866, + "learning_rate": 1.9680267611791547e-05, + "loss": 2.3803, + "step": 1550 + }, + { + "epoch": 0.3252254141329419, + "grad_norm": 5.418530758358457, + "learning_rate": 1.9679699475197248e-05, + "loss": 2.4121, + "step": 1551 + }, + { + "epoch": 0.3254351016984693, + "grad_norm": 6.257273767798533, + "learning_rate": 1.9679130842501556e-05, + "loss": 2.313, + "step": 1552 + }, + { + "epoch": 0.32564478926399665, + "grad_norm": 5.732328243238281, + "learning_rate": 1.9678561713733626e-05, + "loss": 2.7302, + "step": 1553 + }, + { + "epoch": 0.325854476829524, + "grad_norm": 5.492290864673131, + "learning_rate": 1.9677992088922622e-05, + "loss": 2.2384, + "step": 1554 + }, + { + "epoch": 0.32606416439505137, + "grad_norm": 5.764705594021949, + "learning_rate": 1.9677421968097737e-05, + "loss": 2.3671, + "step": 1555 + }, + { + "epoch": 0.32627385196057873, + "grad_norm": 5.508376691207428, + "learning_rate": 1.967685135128819e-05, + "loss": 2.2202, + "step": 1556 + }, + { + "epoch": 0.3264835395261061, + "grad_norm": 5.762762534915171, + "learning_rate": 1.967628023852323e-05, + "loss": 2.1665, + "step": 1557 + }, + { + "epoch": 0.32669322709163345, + "grad_norm": 5.162693486956407, + "learning_rate": 1.967570862983212e-05, + "loss": 2.4545, + "step": 1558 + }, + { + "epoch": 0.3269029146571608, + "grad_norm": 5.9500425410853595, + "learning_rate": 1.9675136525244168e-05, + "loss": 2.3419, + "step": 1559 + }, + { + "epoch": 0.3271126022226882, + "grad_norm": 5.466850158694769, + "learning_rate": 1.9674563924788683e-05, + "loss": 1.8989, + "step": 1560 + }, + { + "epoch": 0.32732228978821554, + "grad_norm": 5.294882977362833, + "learning_rate": 1.967399082849502e-05, + "loss": 2.2718, + "step": 1561 + }, + { + "epoch": 0.3275319773537429, + "grad_norm": 5.0070359353971945, + "learning_rate": 1.9673417236392547e-05, + "loss": 2.2493, + "step": 1562 + }, + { + "epoch": 0.3277416649192703, + "grad_norm": 5.991696998198487, + "learning_rate": 1.9672843148510663e-05, + "loss": 2.5096, + "step": 1563 + }, + { + "epoch": 0.3279513524847977, + "grad_norm": 5.539697722314044, + "learning_rate": 1.9672268564878792e-05, + "loss": 2.2199, + "step": 1564 + }, + { + "epoch": 0.32816104005032504, + "grad_norm": 6.1905639514724395, + "learning_rate": 1.967169348552638e-05, + "loss": 2.5273, + "step": 1565 + }, + { + "epoch": 0.3283707276158524, + "grad_norm": 6.724766239438472, + "learning_rate": 1.9671117910482903e-05, + "loss": 2.4645, + "step": 1566 + }, + { + "epoch": 0.32858041518137976, + "grad_norm": 6.011917438780637, + "learning_rate": 1.967054183977786e-05, + "loss": 2.2566, + "step": 1567 + }, + { + "epoch": 0.3287901027469071, + "grad_norm": 7.572410217569467, + "learning_rate": 1.9669965273440776e-05, + "loss": 2.5161, + "step": 1568 + }, + { + "epoch": 0.3289997903124345, + "grad_norm": 5.421217346534776, + "learning_rate": 1.9669388211501195e-05, + "loss": 2.737, + "step": 1569 + }, + { + "epoch": 0.32920947787796184, + "grad_norm": 5.72999759942945, + "learning_rate": 1.9668810653988702e-05, + "loss": 2.5211, + "step": 1570 + }, + { + "epoch": 0.3294191654434892, + "grad_norm": 6.419669988627167, + "learning_rate": 1.966823260093289e-05, + "loss": 2.3514, + "step": 1571 + }, + { + "epoch": 0.32962885300901656, + "grad_norm": 5.5758203723037925, + "learning_rate": 1.9667654052363393e-05, + "loss": 2.393, + "step": 1572 + }, + { + "epoch": 0.3298385405745439, + "grad_norm": 5.66092324490033, + "learning_rate": 1.9667075008309852e-05, + "loss": 2.2449, + "step": 1573 + }, + { + "epoch": 0.3300482281400713, + "grad_norm": 5.591044279551592, + "learning_rate": 1.9666495468801957e-05, + "loss": 2.307, + "step": 1574 + }, + { + "epoch": 0.33025791570559865, + "grad_norm": 6.430045426141139, + "learning_rate": 1.9665915433869397e-05, + "loss": 2.2199, + "step": 1575 + }, + { + "epoch": 0.330467603271126, + "grad_norm": 5.299603723054972, + "learning_rate": 1.9665334903541913e-05, + "loss": 2.4392, + "step": 1576 + }, + { + "epoch": 0.33067729083665337, + "grad_norm": 5.9714112690123615, + "learning_rate": 1.9664753877849246e-05, + "loss": 2.3323, + "step": 1577 + }, + { + "epoch": 0.33088697840218073, + "grad_norm": 6.231667801598009, + "learning_rate": 1.966417235682118e-05, + "loss": 2.2128, + "step": 1578 + }, + { + "epoch": 0.3310966659677081, + "grad_norm": 5.975659435429296, + "learning_rate": 1.9663590340487528e-05, + "loss": 2.4869, + "step": 1579 + }, + { + "epoch": 0.33130635353323545, + "grad_norm": 5.926377874482364, + "learning_rate": 1.96630078288781e-05, + "loss": 2.5833, + "step": 1580 + }, + { + "epoch": 0.33151604109876287, + "grad_norm": 5.412963581577702, + "learning_rate": 1.9662424822022767e-05, + "loss": 2.3962, + "step": 1581 + }, + { + "epoch": 0.33172572866429023, + "grad_norm": 5.351755265084626, + "learning_rate": 1.96618413199514e-05, + "loss": 2.2915, + "step": 1582 + }, + { + "epoch": 0.3319354162298176, + "grad_norm": 5.840228175903158, + "learning_rate": 1.9661257322693908e-05, + "loss": 2.4686, + "step": 1583 + }, + { + "epoch": 0.33214510379534495, + "grad_norm": 5.426385117783255, + "learning_rate": 1.9660672830280224e-05, + "loss": 2.2827, + "step": 1584 + }, + { + "epoch": 0.3323547913608723, + "grad_norm": 5.506850952057615, + "learning_rate": 1.96600878427403e-05, + "loss": 2.4203, + "step": 1585 + }, + { + "epoch": 0.3325644789263997, + "grad_norm": 5.543616564479627, + "learning_rate": 1.965950236010412e-05, + "loss": 2.2861, + "step": 1586 + }, + { + "epoch": 0.33277416649192704, + "grad_norm": 6.055319824117364, + "learning_rate": 1.965891638240169e-05, + "loss": 2.6901, + "step": 1587 + }, + { + "epoch": 0.3329838540574544, + "grad_norm": 5.658463143578527, + "learning_rate": 1.965832990966304e-05, + "loss": 2.3455, + "step": 1588 + }, + { + "epoch": 0.33319354162298176, + "grad_norm": 5.066372843347675, + "learning_rate": 1.9657742941918237e-05, + "loss": 2.2029, + "step": 1589 + }, + { + "epoch": 0.3334032291885091, + "grad_norm": 6.328899485383506, + "learning_rate": 1.9657155479197356e-05, + "loss": 2.1475, + "step": 1590 + }, + { + "epoch": 0.3336129167540365, + "grad_norm": 5.843336215643689, + "learning_rate": 1.9656567521530502e-05, + "loss": 2.3222, + "step": 1591 + }, + { + "epoch": 0.33382260431956384, + "grad_norm": 5.828937114150826, + "learning_rate": 1.965597906894782e-05, + "loss": 2.5918, + "step": 1592 + }, + { + "epoch": 0.3340322918850912, + "grad_norm": 5.479724679731844, + "learning_rate": 1.965539012147946e-05, + "loss": 2.2339, + "step": 1593 + }, + { + "epoch": 0.33424197945061856, + "grad_norm": 5.479146255059593, + "learning_rate": 1.965480067915561e-05, + "loss": 2.2398, + "step": 1594 + }, + { + "epoch": 0.3344516670161459, + "grad_norm": 5.904441620670538, + "learning_rate": 1.9654210742006482e-05, + "loss": 2.295, + "step": 1595 + }, + { + "epoch": 0.3346613545816733, + "grad_norm": 5.552254721346579, + "learning_rate": 1.9653620310062308e-05, + "loss": 2.1975, + "step": 1596 + }, + { + "epoch": 0.33487104214720065, + "grad_norm": 6.427397347049776, + "learning_rate": 1.965302938335335e-05, + "loss": 2.3856, + "step": 1597 + }, + { + "epoch": 0.335080729712728, + "grad_norm": 6.5043489418639995, + "learning_rate": 1.9652437961909893e-05, + "loss": 2.6487, + "step": 1598 + }, + { + "epoch": 0.3352904172782554, + "grad_norm": 5.786232951460627, + "learning_rate": 1.965184604576225e-05, + "loss": 2.1514, + "step": 1599 + }, + { + "epoch": 0.3355001048437828, + "grad_norm": 6.404225458094209, + "learning_rate": 1.9651253634940753e-05, + "loss": 2.3743, + "step": 1600 + }, + { + "epoch": 0.33570979240931015, + "grad_norm": 4.906540186692006, + "learning_rate": 1.965066072947577e-05, + "loss": 2.0767, + "step": 1601 + }, + { + "epoch": 0.3359194799748375, + "grad_norm": 5.777920523296027, + "learning_rate": 1.965006732939769e-05, + "loss": 2.258, + "step": 1602 + }, + { + "epoch": 0.33612916754036487, + "grad_norm": 4.858194864119895, + "learning_rate": 1.9649473434736917e-05, + "loss": 2.2744, + "step": 1603 + }, + { + "epoch": 0.33633885510589223, + "grad_norm": 5.016821726110959, + "learning_rate": 1.96488790455239e-05, + "loss": 2.2743, + "step": 1604 + }, + { + "epoch": 0.3365485426714196, + "grad_norm": 5.27403008151518, + "learning_rate": 1.9648284161789096e-05, + "loss": 2.4679, + "step": 1605 + }, + { + "epoch": 0.33675823023694695, + "grad_norm": 5.670777104096143, + "learning_rate": 1.9647688783562992e-05, + "loss": 2.6915, + "step": 1606 + }, + { + "epoch": 0.3369679178024743, + "grad_norm": 5.513491182680519, + "learning_rate": 1.9647092910876102e-05, + "loss": 2.3359, + "step": 1607 + }, + { + "epoch": 0.3371776053680017, + "grad_norm": 6.182513180992043, + "learning_rate": 1.9646496543758972e-05, + "loss": 2.445, + "step": 1608 + }, + { + "epoch": 0.33738729293352904, + "grad_norm": 5.876352155212737, + "learning_rate": 1.9645899682242164e-05, + "loss": 2.3224, + "step": 1609 + }, + { + "epoch": 0.3375969804990564, + "grad_norm": 5.632639409916287, + "learning_rate": 1.9645302326356267e-05, + "loss": 2.0864, + "step": 1610 + }, + { + "epoch": 0.33780666806458376, + "grad_norm": 5.104252825325276, + "learning_rate": 1.9644704476131896e-05, + "loss": 2.0801, + "step": 1611 + }, + { + "epoch": 0.3380163556301111, + "grad_norm": 5.1521356080030465, + "learning_rate": 1.964410613159969e-05, + "loss": 2.4843, + "step": 1612 + }, + { + "epoch": 0.3382260431956385, + "grad_norm": 5.960694466519882, + "learning_rate": 1.964350729279032e-05, + "loss": 2.0485, + "step": 1613 + }, + { + "epoch": 0.33843573076116584, + "grad_norm": 6.05848934820254, + "learning_rate": 1.964290795973448e-05, + "loss": 1.9623, + "step": 1614 + }, + { + "epoch": 0.3386454183266932, + "grad_norm": 5.658683941597805, + "learning_rate": 1.9642308132462877e-05, + "loss": 2.3733, + "step": 1615 + }, + { + "epoch": 0.33885510589222056, + "grad_norm": 6.276085311129812, + "learning_rate": 1.964170781100626e-05, + "loss": 2.2151, + "step": 1616 + }, + { + "epoch": 0.339064793457748, + "grad_norm": 5.894701877814183, + "learning_rate": 1.9641106995395393e-05, + "loss": 2.1953, + "step": 1617 + }, + { + "epoch": 0.33927448102327534, + "grad_norm": 5.2367358637664845, + "learning_rate": 1.964050568566107e-05, + "loss": 2.2236, + "step": 1618 + }, + { + "epoch": 0.3394841685888027, + "grad_norm": 5.361926510254171, + "learning_rate": 1.9639903881834115e-05, + "loss": 2.4229, + "step": 1619 + }, + { + "epoch": 0.33969385615433006, + "grad_norm": 6.009733127460258, + "learning_rate": 1.9639301583945364e-05, + "loss": 2.2764, + "step": 1620 + }, + { + "epoch": 0.3399035437198574, + "grad_norm": 6.501378521577173, + "learning_rate": 1.963869879202569e-05, + "loss": 2.5485, + "step": 1621 + }, + { + "epoch": 0.3401132312853848, + "grad_norm": 4.934838676195025, + "learning_rate": 1.963809550610598e-05, + "loss": 2.4331, + "step": 1622 + }, + { + "epoch": 0.34032291885091215, + "grad_norm": 5.5369639586136445, + "learning_rate": 1.9637491726217167e-05, + "loss": 2.3229, + "step": 1623 + }, + { + "epoch": 0.3405326064164395, + "grad_norm": 5.38792022680214, + "learning_rate": 1.963688745239018e-05, + "loss": 2.2018, + "step": 1624 + }, + { + "epoch": 0.34074229398196687, + "grad_norm": 6.27070047343671, + "learning_rate": 1.9636282684656e-05, + "loss": 1.7591, + "step": 1625 + }, + { + "epoch": 0.34095198154749423, + "grad_norm": 6.2044132393959055, + "learning_rate": 1.963567742304562e-05, + "loss": 2.1486, + "step": 1626 + }, + { + "epoch": 0.3411616691130216, + "grad_norm": 6.279797936852069, + "learning_rate": 1.963507166759006e-05, + "loss": 2.2836, + "step": 1627 + }, + { + "epoch": 0.34137135667854895, + "grad_norm": 6.451175365308464, + "learning_rate": 1.9634465418320364e-05, + "loss": 2.3515, + "step": 1628 + }, + { + "epoch": 0.3415810442440763, + "grad_norm": 5.752042339164163, + "learning_rate": 1.9633858675267605e-05, + "loss": 2.2526, + "step": 1629 + }, + { + "epoch": 0.3417907318096037, + "grad_norm": 5.267816016992654, + "learning_rate": 1.963325143846288e-05, + "loss": 2.2267, + "step": 1630 + }, + { + "epoch": 0.34200041937513104, + "grad_norm": 6.796672919437261, + "learning_rate": 1.9632643707937313e-05, + "loss": 2.2328, + "step": 1631 + }, + { + "epoch": 0.3422101069406584, + "grad_norm": 5.569296422166354, + "learning_rate": 1.9632035483722048e-05, + "loss": 2.2812, + "step": 1632 + }, + { + "epoch": 0.34241979450618576, + "grad_norm": 5.833129645716108, + "learning_rate": 1.9631426765848255e-05, + "loss": 2.1111, + "step": 1633 + }, + { + "epoch": 0.3426294820717131, + "grad_norm": 6.869886053626056, + "learning_rate": 1.963081755434714e-05, + "loss": 2.336, + "step": 1634 + }, + { + "epoch": 0.34283916963724054, + "grad_norm": 5.2288693017899535, + "learning_rate": 1.963020784924992e-05, + "loss": 2.4505, + "step": 1635 + }, + { + "epoch": 0.3430488572027679, + "grad_norm": 5.5027633364292985, + "learning_rate": 1.9629597650587847e-05, + "loss": 2.0743, + "step": 1636 + }, + { + "epoch": 0.34325854476829526, + "grad_norm": 6.086821112740867, + "learning_rate": 1.962898695839219e-05, + "loss": 2.3118, + "step": 1637 + }, + { + "epoch": 0.3434682323338226, + "grad_norm": 6.1338275972895, + "learning_rate": 1.9628375772694254e-05, + "loss": 2.1605, + "step": 1638 + }, + { + "epoch": 0.34367791989935, + "grad_norm": 5.734923905222298, + "learning_rate": 1.9627764093525353e-05, + "loss": 1.8931, + "step": 1639 + }, + { + "epoch": 0.34388760746487734, + "grad_norm": 5.403554791675531, + "learning_rate": 1.9627151920916853e-05, + "loss": 2.1866, + "step": 1640 + }, + { + "epoch": 0.3440972950304047, + "grad_norm": 8.125251051343719, + "learning_rate": 1.962653925490011e-05, + "loss": 2.146, + "step": 1641 + }, + { + "epoch": 0.34430698259593207, + "grad_norm": 7.3874474581574185, + "learning_rate": 1.962592609550654e-05, + "loss": 2.5429, + "step": 1642 + }, + { + "epoch": 0.3445166701614594, + "grad_norm": 7.161177938723063, + "learning_rate": 1.962531244276756e-05, + "loss": 2.3126, + "step": 1643 + }, + { + "epoch": 0.3447263577269868, + "grad_norm": 6.215777060934349, + "learning_rate": 1.9624698296714625e-05, + "loss": 2.3888, + "step": 1644 + }, + { + "epoch": 0.34493604529251415, + "grad_norm": 5.959337715393357, + "learning_rate": 1.962408365737921e-05, + "loss": 2.2543, + "step": 1645 + }, + { + "epoch": 0.3451457328580415, + "grad_norm": 5.303832614130179, + "learning_rate": 1.9623468524792814e-05, + "loss": 2.1992, + "step": 1646 + }, + { + "epoch": 0.34535542042356887, + "grad_norm": 6.033451549111891, + "learning_rate": 1.962285289898696e-05, + "loss": 2.3358, + "step": 1647 + }, + { + "epoch": 0.34556510798909623, + "grad_norm": 5.392432382614776, + "learning_rate": 1.9622236779993212e-05, + "loss": 2.3739, + "step": 1648 + }, + { + "epoch": 0.3457747955546236, + "grad_norm": 5.608161651093695, + "learning_rate": 1.962162016784314e-05, + "loss": 2.199, + "step": 1649 + }, + { + "epoch": 0.34598448312015095, + "grad_norm": 6.80955398444917, + "learning_rate": 1.962100306256834e-05, + "loss": 2.6845, + "step": 1650 + }, + { + "epoch": 0.3461941706856783, + "grad_norm": 5.650547318967569, + "learning_rate": 1.962038546420045e-05, + "loss": 2.2906, + "step": 1651 + }, + { + "epoch": 0.3464038582512057, + "grad_norm": 5.99594850177852, + "learning_rate": 1.961976737277112e-05, + "loss": 2.5299, + "step": 1652 + }, + { + "epoch": 0.3466135458167331, + "grad_norm": 6.101160751009999, + "learning_rate": 1.9619148788312027e-05, + "loss": 2.5141, + "step": 1653 + }, + { + "epoch": 0.34682323338226045, + "grad_norm": 5.152095935492857, + "learning_rate": 1.9618529710854875e-05, + "loss": 2.2265, + "step": 1654 + }, + { + "epoch": 0.3470329209477878, + "grad_norm": 5.859880899841965, + "learning_rate": 1.9617910140431396e-05, + "loss": 2.4224, + "step": 1655 + }, + { + "epoch": 0.3472426085133152, + "grad_norm": 6.79054628801436, + "learning_rate": 1.9617290077073336e-05, + "loss": 2.285, + "step": 1656 + }, + { + "epoch": 0.34745229607884254, + "grad_norm": 5.308834508709281, + "learning_rate": 1.9616669520812482e-05, + "loss": 2.2054, + "step": 1657 + }, + { + "epoch": 0.3476619836443699, + "grad_norm": 5.640969058226901, + "learning_rate": 1.9616048471680632e-05, + "loss": 2.2392, + "step": 1658 + }, + { + "epoch": 0.34787167120989726, + "grad_norm": 6.039022430432571, + "learning_rate": 1.9615426929709624e-05, + "loss": 1.9644, + "step": 1659 + }, + { + "epoch": 0.3480813587754246, + "grad_norm": 5.330338928975169, + "learning_rate": 1.9614804894931307e-05, + "loss": 2.441, + "step": 1660 + }, + { + "epoch": 0.348291046340952, + "grad_norm": 6.3073108706206895, + "learning_rate": 1.9614182367377563e-05, + "loss": 2.2988, + "step": 1661 + }, + { + "epoch": 0.34850073390647934, + "grad_norm": 5.895407758624858, + "learning_rate": 1.96135593470803e-05, + "loss": 2.2281, + "step": 1662 + }, + { + "epoch": 0.3487104214720067, + "grad_norm": 5.956443281079859, + "learning_rate": 1.9612935834071445e-05, + "loss": 2.389, + "step": 1663 + }, + { + "epoch": 0.34892010903753407, + "grad_norm": 5.774753742817088, + "learning_rate": 1.9612311828382954e-05, + "loss": 2.3765, + "step": 1664 + }, + { + "epoch": 0.3491297966030614, + "grad_norm": 5.124911428731447, + "learning_rate": 1.961168733004681e-05, + "loss": 1.8536, + "step": 1665 + }, + { + "epoch": 0.3493394841685888, + "grad_norm": 5.29897050230378, + "learning_rate": 1.9611062339095023e-05, + "loss": 1.8147, + "step": 1666 + }, + { + "epoch": 0.34954917173411615, + "grad_norm": 5.205653865348943, + "learning_rate": 1.961043685555962e-05, + "loss": 2.2019, + "step": 1667 + }, + { + "epoch": 0.3497588592996435, + "grad_norm": 5.213782365705236, + "learning_rate": 1.960981087947266e-05, + "loss": 2.1829, + "step": 1668 + }, + { + "epoch": 0.34996854686517087, + "grad_norm": 6.395047417916513, + "learning_rate": 1.9609184410866224e-05, + "loss": 2.1343, + "step": 1669 + }, + { + "epoch": 0.35017823443069823, + "grad_norm": 5.759269846466217, + "learning_rate": 1.960855744977242e-05, + "loss": 2.1201, + "step": 1670 + }, + { + "epoch": 0.35038792199622565, + "grad_norm": 5.866259648829047, + "learning_rate": 1.9607929996223383e-05, + "loss": 1.8925, + "step": 1671 + }, + { + "epoch": 0.350597609561753, + "grad_norm": 6.654200167466861, + "learning_rate": 1.960730205025127e-05, + "loss": 2.4153, + "step": 1672 + }, + { + "epoch": 0.35080729712728037, + "grad_norm": 6.163470414170069, + "learning_rate": 1.9606673611888262e-05, + "loss": 2.1775, + "step": 1673 + }, + { + "epoch": 0.35101698469280773, + "grad_norm": 7.09991290533726, + "learning_rate": 1.960604468116657e-05, + "loss": 2.4434, + "step": 1674 + }, + { + "epoch": 0.3512266722583351, + "grad_norm": 5.746313634499447, + "learning_rate": 1.960541525811843e-05, + "loss": 2.1227, + "step": 1675 + }, + { + "epoch": 0.35143635982386245, + "grad_norm": 6.107570967975584, + "learning_rate": 1.9604785342776095e-05, + "loss": 2.1543, + "step": 1676 + }, + { + "epoch": 0.3516460473893898, + "grad_norm": 5.410935218107208, + "learning_rate": 1.9604154935171854e-05, + "loss": 2.094, + "step": 1677 + }, + { + "epoch": 0.3518557349549172, + "grad_norm": 5.167148370893116, + "learning_rate": 1.9603524035338017e-05, + "loss": 2.2303, + "step": 1678 + }, + { + "epoch": 0.35206542252044454, + "grad_norm": 4.927149623795553, + "learning_rate": 1.9602892643306914e-05, + "loss": 2.2601, + "step": 1679 + }, + { + "epoch": 0.3522751100859719, + "grad_norm": 5.499208224841035, + "learning_rate": 1.9602260759110905e-05, + "loss": 2.3447, + "step": 1680 + }, + { + "epoch": 0.35248479765149926, + "grad_norm": 7.292751486524627, + "learning_rate": 1.960162838278238e-05, + "loss": 2.3028, + "step": 1681 + }, + { + "epoch": 0.3526944852170266, + "grad_norm": 6.077661482205115, + "learning_rate": 1.9600995514353748e-05, + "loss": 2.1586, + "step": 1682 + }, + { + "epoch": 0.352904172782554, + "grad_norm": 5.872212024921919, + "learning_rate": 1.9600362153857445e-05, + "loss": 2.422, + "step": 1683 + }, + { + "epoch": 0.35311386034808134, + "grad_norm": 5.408163191936192, + "learning_rate": 1.959972830132593e-05, + "loss": 2.41, + "step": 1684 + }, + { + "epoch": 0.3533235479136087, + "grad_norm": 6.5853438613257795, + "learning_rate": 1.9599093956791688e-05, + "loss": 1.9205, + "step": 1685 + }, + { + "epoch": 0.35353323547913607, + "grad_norm": 5.25300984712405, + "learning_rate": 1.9598459120287234e-05, + "loss": 2.4163, + "step": 1686 + }, + { + "epoch": 0.3537429230446634, + "grad_norm": 5.303118914330665, + "learning_rate": 1.9597823791845102e-05, + "loss": 2.2151, + "step": 1687 + }, + { + "epoch": 0.35395261061019084, + "grad_norm": 6.041393593496545, + "learning_rate": 1.9597187971497853e-05, + "loss": 2.1429, + "step": 1688 + }, + { + "epoch": 0.3541622981757182, + "grad_norm": 5.241675286667038, + "learning_rate": 1.9596551659278076e-05, + "loss": 2.1497, + "step": 1689 + }, + { + "epoch": 0.35437198574124557, + "grad_norm": 5.804696079211872, + "learning_rate": 1.959591485521838e-05, + "loss": 2.3128, + "step": 1690 + }, + { + "epoch": 0.3545816733067729, + "grad_norm": 5.80782204025035, + "learning_rate": 1.959527755935141e-05, + "loss": 2.2326, + "step": 1691 + }, + { + "epoch": 0.3547913608723003, + "grad_norm": 7.3015953273941365, + "learning_rate": 1.9594639771709814e-05, + "loss": 2.3307, + "step": 1692 + }, + { + "epoch": 0.35500104843782765, + "grad_norm": 5.9833597325486, + "learning_rate": 1.9594001492326295e-05, + "loss": 2.1632, + "step": 1693 + }, + { + "epoch": 0.355210736003355, + "grad_norm": 6.581294966020599, + "learning_rate": 1.959336272123356e-05, + "loss": 2.3893, + "step": 1694 + }, + { + "epoch": 0.35542042356888237, + "grad_norm": 5.420440184064418, + "learning_rate": 1.9592723458464344e-05, + "loss": 2.1637, + "step": 1695 + }, + { + "epoch": 0.35563011113440973, + "grad_norm": 5.160742435019318, + "learning_rate": 1.9592083704051416e-05, + "loss": 2.2996, + "step": 1696 + }, + { + "epoch": 0.3558397986999371, + "grad_norm": 6.076157237108552, + "learning_rate": 1.9591443458027558e-05, + "loss": 1.9313, + "step": 1697 + }, + { + "epoch": 0.35604948626546445, + "grad_norm": 5.434482143608857, + "learning_rate": 1.959080272042559e-05, + "loss": 2.4027, + "step": 1698 + }, + { + "epoch": 0.3562591738309918, + "grad_norm": 5.3882051056958336, + "learning_rate": 1.9590161491278345e-05, + "loss": 2.2472, + "step": 1699 + }, + { + "epoch": 0.3564688613965192, + "grad_norm": 6.249506551515541, + "learning_rate": 1.958951977061869e-05, + "loss": 2.2795, + "step": 1700 + }, + { + "epoch": 0.35667854896204654, + "grad_norm": 6.888026663964633, + "learning_rate": 1.9588877558479517e-05, + "loss": 2.4719, + "step": 1701 + }, + { + "epoch": 0.3568882365275739, + "grad_norm": 5.342502986329096, + "learning_rate": 1.958823485489374e-05, + "loss": 2.3874, + "step": 1702 + }, + { + "epoch": 0.35709792409310126, + "grad_norm": 6.564646797762203, + "learning_rate": 1.9587591659894293e-05, + "loss": 2.1616, + "step": 1703 + }, + { + "epoch": 0.3573076116586286, + "grad_norm": 5.3817900141911315, + "learning_rate": 1.958694797351415e-05, + "loss": 2.4482, + "step": 1704 + }, + { + "epoch": 0.357517299224156, + "grad_norm": 5.810909851341291, + "learning_rate": 1.958630379578629e-05, + "loss": 2.1835, + "step": 1705 + }, + { + "epoch": 0.3577269867896834, + "grad_norm": 5.069083654327974, + "learning_rate": 1.9585659126743734e-05, + "loss": 2.205, + "step": 1706 + }, + { + "epoch": 0.35793667435521076, + "grad_norm": 5.678218097735782, + "learning_rate": 1.9585013966419522e-05, + "loss": 2.3414, + "step": 1707 + }, + { + "epoch": 0.3581463619207381, + "grad_norm": 5.932166825456759, + "learning_rate": 1.9584368314846725e-05, + "loss": 2.4403, + "step": 1708 + }, + { + "epoch": 0.3583560494862655, + "grad_norm": 5.200048716721955, + "learning_rate": 1.9583722172058426e-05, + "loss": 2.2058, + "step": 1709 + }, + { + "epoch": 0.35856573705179284, + "grad_norm": 5.605274255382252, + "learning_rate": 1.958307553808774e-05, + "loss": 2.2684, + "step": 1710 + }, + { + "epoch": 0.3587754246173202, + "grad_norm": 5.050702431116143, + "learning_rate": 1.9582428412967814e-05, + "loss": 2.1875, + "step": 1711 + }, + { + "epoch": 0.35898511218284757, + "grad_norm": 5.922291855023995, + "learning_rate": 1.9581780796731812e-05, + "loss": 2.2152, + "step": 1712 + }, + { + "epoch": 0.3591947997483749, + "grad_norm": 5.280457626243012, + "learning_rate": 1.958113268941293e-05, + "loss": 2.2874, + "step": 1713 + }, + { + "epoch": 0.3594044873139023, + "grad_norm": 5.881500523228601, + "learning_rate": 1.9580484091044373e-05, + "loss": 2.1145, + "step": 1714 + }, + { + "epoch": 0.35961417487942965, + "grad_norm": 6.0607582480168105, + "learning_rate": 1.9579835001659396e-05, + "loss": 2.3703, + "step": 1715 + }, + { + "epoch": 0.359823862444957, + "grad_norm": 6.263590217554079, + "learning_rate": 1.9579185421291257e-05, + "loss": 2.3474, + "step": 1716 + }, + { + "epoch": 0.36003355001048437, + "grad_norm": 5.587184617403286, + "learning_rate": 1.957853534997325e-05, + "loss": 2.573, + "step": 1717 + }, + { + "epoch": 0.36024323757601173, + "grad_norm": 6.004519742100833, + "learning_rate": 1.9577884787738693e-05, + "loss": 2.3938, + "step": 1718 + }, + { + "epoch": 0.3604529251415391, + "grad_norm": 6.138337337081805, + "learning_rate": 1.9577233734620928e-05, + "loss": 2.4345, + "step": 1719 + }, + { + "epoch": 0.36066261270706645, + "grad_norm": 5.827310281471527, + "learning_rate": 1.9576582190653326e-05, + "loss": 2.5692, + "step": 1720 + }, + { + "epoch": 0.3608723002725938, + "grad_norm": 6.211644111480424, + "learning_rate": 1.9575930155869277e-05, + "loss": 2.4837, + "step": 1721 + }, + { + "epoch": 0.3610819878381212, + "grad_norm": 5.599209143469495, + "learning_rate": 1.9575277630302193e-05, + "loss": 2.4039, + "step": 1722 + }, + { + "epoch": 0.36129167540364854, + "grad_norm": 4.959977386523943, + "learning_rate": 1.957462461398553e-05, + "loss": 2.2625, + "step": 1723 + }, + { + "epoch": 0.36150136296917595, + "grad_norm": 4.891211376289939, + "learning_rate": 1.9573971106952746e-05, + "loss": 1.9175, + "step": 1724 + }, + { + "epoch": 0.3617110505347033, + "grad_norm": 6.065117789132127, + "learning_rate": 1.9573317109237337e-05, + "loss": 2.2494, + "step": 1725 + }, + { + "epoch": 0.3619207381002307, + "grad_norm": 6.049693301932738, + "learning_rate": 1.9572662620872825e-05, + "loss": 2.5247, + "step": 1726 + }, + { + "epoch": 0.36213042566575804, + "grad_norm": 5.701447641767417, + "learning_rate": 1.957200764189275e-05, + "loss": 2.2293, + "step": 1727 + }, + { + "epoch": 0.3623401132312854, + "grad_norm": 5.413194695023114, + "learning_rate": 1.957135217233068e-05, + "loss": 2.1644, + "step": 1728 + }, + { + "epoch": 0.36254980079681276, + "grad_norm": 5.54205890665771, + "learning_rate": 1.957069621222021e-05, + "loss": 2.1572, + "step": 1729 + }, + { + "epoch": 0.3627594883623401, + "grad_norm": 5.369699570129372, + "learning_rate": 1.957003976159496e-05, + "loss": 2.2254, + "step": 1730 + }, + { + "epoch": 0.3629691759278675, + "grad_norm": 5.606180579482364, + "learning_rate": 1.9569382820488575e-05, + "loss": 2.5863, + "step": 1731 + }, + { + "epoch": 0.36317886349339484, + "grad_norm": 5.983595416644938, + "learning_rate": 1.9568725388934723e-05, + "loss": 2.284, + "step": 1732 + }, + { + "epoch": 0.3633885510589222, + "grad_norm": 5.60807875619355, + "learning_rate": 1.95680674669671e-05, + "loss": 2.2106, + "step": 1733 + }, + { + "epoch": 0.36359823862444957, + "grad_norm": 7.6127692459497265, + "learning_rate": 1.956740905461942e-05, + "loss": 2.5548, + "step": 1734 + }, + { + "epoch": 0.3638079261899769, + "grad_norm": 5.636538864772774, + "learning_rate": 1.9566750151925436e-05, + "loss": 2.3459, + "step": 1735 + }, + { + "epoch": 0.3640176137555043, + "grad_norm": 5.933719806951929, + "learning_rate": 1.9566090758918913e-05, + "loss": 2.3465, + "step": 1736 + }, + { + "epoch": 0.36422730132103165, + "grad_norm": 6.256097942666154, + "learning_rate": 1.9565430875633648e-05, + "loss": 2.3948, + "step": 1737 + }, + { + "epoch": 0.364436988886559, + "grad_norm": 5.66321593403629, + "learning_rate": 1.956477050210346e-05, + "loss": 2.1085, + "step": 1738 + }, + { + "epoch": 0.36464667645208637, + "grad_norm": 5.2921111983457445, + "learning_rate": 1.956410963836219e-05, + "loss": 2.2224, + "step": 1739 + }, + { + "epoch": 0.36485636401761373, + "grad_norm": 5.980122971396161, + "learning_rate": 1.9563448284443717e-05, + "loss": 2.2052, + "step": 1740 + }, + { + "epoch": 0.3650660515831411, + "grad_norm": 5.1069635358808565, + "learning_rate": 1.9562786440381933e-05, + "loss": 2.2934, + "step": 1741 + }, + { + "epoch": 0.3652757391486685, + "grad_norm": 5.318715414569083, + "learning_rate": 1.9562124106210758e-05, + "loss": 2.2583, + "step": 1742 + }, + { + "epoch": 0.36548542671419587, + "grad_norm": 5.721554311240375, + "learning_rate": 1.9561461281964135e-05, + "loss": 2.3655, + "step": 1743 + }, + { + "epoch": 0.36569511427972323, + "grad_norm": 5.73652509627809, + "learning_rate": 1.956079796767604e-05, + "loss": 2.304, + "step": 1744 + }, + { + "epoch": 0.3659048018452506, + "grad_norm": 5.368147113544979, + "learning_rate": 1.9560134163380467e-05, + "loss": 2.4203, + "step": 1745 + }, + { + "epoch": 0.36611448941077795, + "grad_norm": 5.272285917075877, + "learning_rate": 1.9559469869111437e-05, + "loss": 2.0666, + "step": 1746 + }, + { + "epoch": 0.3663241769763053, + "grad_norm": 4.8188749260014365, + "learning_rate": 1.9558805084902995e-05, + "loss": 2.2457, + "step": 1747 + }, + { + "epoch": 0.3665338645418327, + "grad_norm": 6.257958190807909, + "learning_rate": 1.9558139810789216e-05, + "loss": 2.4562, + "step": 1748 + }, + { + "epoch": 0.36674355210736004, + "grad_norm": 5.104900912262871, + "learning_rate": 1.9557474046804194e-05, + "loss": 2.1576, + "step": 1749 + }, + { + "epoch": 0.3669532396728874, + "grad_norm": 6.226248081374414, + "learning_rate": 1.9556807792982045e-05, + "loss": 2.3186, + "step": 1750 + }, + { + "epoch": 0.36716292723841476, + "grad_norm": 5.040256406768721, + "learning_rate": 1.9556141049356924e-05, + "loss": 2.0172, + "step": 1751 + }, + { + "epoch": 0.3673726148039421, + "grad_norm": 5.1157351849864865, + "learning_rate": 1.9555473815963002e-05, + "loss": 2.325, + "step": 1752 + }, + { + "epoch": 0.3675823023694695, + "grad_norm": 5.085508325886163, + "learning_rate": 1.9554806092834473e-05, + "loss": 2.3395, + "step": 1753 + }, + { + "epoch": 0.36779198993499684, + "grad_norm": 5.8794078042652185, + "learning_rate": 1.955413788000556e-05, + "loss": 2.4353, + "step": 1754 + }, + { + "epoch": 0.3680016775005242, + "grad_norm": 6.237286463140929, + "learning_rate": 1.955346917751051e-05, + "loss": 2.335, + "step": 1755 + }, + { + "epoch": 0.36821136506605157, + "grad_norm": 5.882972795049092, + "learning_rate": 1.9552799985383593e-05, + "loss": 2.4365, + "step": 1756 + }, + { + "epoch": 0.3684210526315789, + "grad_norm": 5.5896404710564465, + "learning_rate": 1.955213030365911e-05, + "loss": 2.1801, + "step": 1757 + }, + { + "epoch": 0.3686307401971063, + "grad_norm": 5.111875555213847, + "learning_rate": 1.955146013237138e-05, + "loss": 2.3078, + "step": 1758 + }, + { + "epoch": 0.36884042776263365, + "grad_norm": 5.421058354005871, + "learning_rate": 1.9550789471554754e-05, + "loss": 2.3057, + "step": 1759 + }, + { + "epoch": 0.36905011532816107, + "grad_norm": 5.821862040099351, + "learning_rate": 1.95501183212436e-05, + "loss": 2.4892, + "step": 1760 + }, + { + "epoch": 0.3692598028936884, + "grad_norm": 6.018041114057587, + "learning_rate": 1.9549446681472323e-05, + "loss": 2.0957, + "step": 1761 + }, + { + "epoch": 0.3694694904592158, + "grad_norm": 5.52649957626701, + "learning_rate": 1.9548774552275335e-05, + "loss": 2.3012, + "step": 1762 + }, + { + "epoch": 0.36967917802474315, + "grad_norm": 5.614782683905725, + "learning_rate": 1.9548101933687094e-05, + "loss": 2.3079, + "step": 1763 + }, + { + "epoch": 0.3698888655902705, + "grad_norm": 5.84946680341017, + "learning_rate": 1.9547428825742064e-05, + "loss": 2.3786, + "step": 1764 + }, + { + "epoch": 0.37009855315579787, + "grad_norm": 5.44972348490551, + "learning_rate": 1.954675522847475e-05, + "loss": 2.031, + "step": 1765 + }, + { + "epoch": 0.37030824072132523, + "grad_norm": 5.388306391552122, + "learning_rate": 1.9546081141919676e-05, + "loss": 2.1767, + "step": 1766 + }, + { + "epoch": 0.3705179282868526, + "grad_norm": 5.859141628704206, + "learning_rate": 1.954540656611138e-05, + "loss": 2.2229, + "step": 1767 + }, + { + "epoch": 0.37072761585237995, + "grad_norm": 5.864720920219384, + "learning_rate": 1.9544731501084445e-05, + "loss": 2.1253, + "step": 1768 + }, + { + "epoch": 0.3709373034179073, + "grad_norm": 5.604754448946263, + "learning_rate": 1.954405594687347e-05, + "loss": 2.5683, + "step": 1769 + }, + { + "epoch": 0.3711469909834347, + "grad_norm": 6.398507553923447, + "learning_rate": 1.954337990351307e-05, + "loss": 2.4562, + "step": 1770 + }, + { + "epoch": 0.37135667854896204, + "grad_norm": 5.855825785013247, + "learning_rate": 1.95427033710379e-05, + "loss": 2.3962, + "step": 1771 + }, + { + "epoch": 0.3715663661144894, + "grad_norm": 5.4992735136383235, + "learning_rate": 1.954202634948263e-05, + "loss": 2.5828, + "step": 1772 + }, + { + "epoch": 0.37177605368001676, + "grad_norm": 5.266256005586283, + "learning_rate": 1.954134883888196e-05, + "loss": 2.4694, + "step": 1773 + }, + { + "epoch": 0.3719857412455441, + "grad_norm": 5.803277768654528, + "learning_rate": 1.9540670839270614e-05, + "loss": 2.2878, + "step": 1774 + }, + { + "epoch": 0.3721954288110715, + "grad_norm": 6.420229155901658, + "learning_rate": 1.9539992350683337e-05, + "loss": 2.359, + "step": 1775 + }, + { + "epoch": 0.37240511637659884, + "grad_norm": 5.946736918596325, + "learning_rate": 1.9539313373154907e-05, + "loss": 2.2908, + "step": 1776 + }, + { + "epoch": 0.3726148039421262, + "grad_norm": 5.897266705001163, + "learning_rate": 1.9538633906720123e-05, + "loss": 2.186, + "step": 1777 + }, + { + "epoch": 0.3728244915076536, + "grad_norm": 6.227492316729951, + "learning_rate": 1.9537953951413806e-05, + "loss": 2.3604, + "step": 1778 + }, + { + "epoch": 0.373034179073181, + "grad_norm": 5.260347230947926, + "learning_rate": 1.953727350727081e-05, + "loss": 2.1722, + "step": 1779 + }, + { + "epoch": 0.37324386663870834, + "grad_norm": 5.298609573526748, + "learning_rate": 1.9536592574326e-05, + "loss": 2.2996, + "step": 1780 + }, + { + "epoch": 0.3734535542042357, + "grad_norm": 5.5121981120352075, + "learning_rate": 1.9535911152614284e-05, + "loss": 2.2951, + "step": 1781 + }, + { + "epoch": 0.37366324176976307, + "grad_norm": 5.623201849578347, + "learning_rate": 1.953522924217058e-05, + "loss": 2.2044, + "step": 1782 + }, + { + "epoch": 0.3738729293352904, + "grad_norm": 5.965575327911224, + "learning_rate": 1.953454684302984e-05, + "loss": 2.3269, + "step": 1783 + }, + { + "epoch": 0.3740826169008178, + "grad_norm": 6.06988319820644, + "learning_rate": 1.9533863955227035e-05, + "loss": 2.4885, + "step": 1784 + }, + { + "epoch": 0.37429230446634515, + "grad_norm": 5.9272562588300435, + "learning_rate": 1.953318057879717e-05, + "loss": 1.9623, + "step": 1785 + }, + { + "epoch": 0.3745019920318725, + "grad_norm": 5.209736945066261, + "learning_rate": 1.9532496713775266e-05, + "loss": 2.24, + "step": 1786 + }, + { + "epoch": 0.37471167959739987, + "grad_norm": 5.344081708985399, + "learning_rate": 1.953181236019637e-05, + "loss": 2.066, + "step": 1787 + }, + { + "epoch": 0.37492136716292723, + "grad_norm": 5.605855067993545, + "learning_rate": 1.953112751809556e-05, + "loss": 2.2658, + "step": 1788 + }, + { + "epoch": 0.3751310547284546, + "grad_norm": 5.011748979544954, + "learning_rate": 1.9530442187507934e-05, + "loss": 2.4437, + "step": 1789 + }, + { + "epoch": 0.37534074229398195, + "grad_norm": 5.775688418011658, + "learning_rate": 1.9529756368468616e-05, + "loss": 2.2131, + "step": 1790 + }, + { + "epoch": 0.3755504298595093, + "grad_norm": 6.104004423939276, + "learning_rate": 1.9529070061012753e-05, + "loss": 2.5495, + "step": 1791 + }, + { + "epoch": 0.3757601174250367, + "grad_norm": 5.25476828898554, + "learning_rate": 1.9528383265175527e-05, + "loss": 2.0111, + "step": 1792 + }, + { + "epoch": 0.37596980499056404, + "grad_norm": 6.026721085477506, + "learning_rate": 1.952769598099213e-05, + "loss": 2.3937, + "step": 1793 + }, + { + "epoch": 0.3761794925560914, + "grad_norm": 6.065106070500315, + "learning_rate": 1.9527008208497787e-05, + "loss": 2.3891, + "step": 1794 + }, + { + "epoch": 0.37638918012161876, + "grad_norm": 5.493475065879344, + "learning_rate": 1.952631994772775e-05, + "loss": 2.0691, + "step": 1795 + }, + { + "epoch": 0.3765988676871462, + "grad_norm": 5.413536039386208, + "learning_rate": 1.9525631198717297e-05, + "loss": 2.1441, + "step": 1796 + }, + { + "epoch": 0.37680855525267354, + "grad_norm": 5.343933474412332, + "learning_rate": 1.952494196150172e-05, + "loss": 2.0614, + "step": 1797 + }, + { + "epoch": 0.3770182428182009, + "grad_norm": 4.921739814336073, + "learning_rate": 1.952425223611635e-05, + "loss": 2.2284, + "step": 1798 + }, + { + "epoch": 0.37722793038372826, + "grad_norm": 5.585877886367357, + "learning_rate": 1.952356202259653e-05, + "loss": 2.6193, + "step": 1799 + }, + { + "epoch": 0.3774376179492556, + "grad_norm": 5.1345446354232225, + "learning_rate": 1.952287132097764e-05, + "loss": 2.1897, + "step": 1800 + }, + { + "epoch": 0.377647305514783, + "grad_norm": 5.083000134912092, + "learning_rate": 1.952218013129508e-05, + "loss": 2.3993, + "step": 1801 + }, + { + "epoch": 0.37785699308031034, + "grad_norm": 5.563620657023212, + "learning_rate": 1.9521488453584273e-05, + "loss": 1.9981, + "step": 1802 + }, + { + "epoch": 0.3780666806458377, + "grad_norm": 5.418593657875649, + "learning_rate": 1.9520796287880665e-05, + "loss": 2.2082, + "step": 1803 + }, + { + "epoch": 0.37827636821136507, + "grad_norm": 4.89514384485245, + "learning_rate": 1.952010363421974e-05, + "loss": 2.3457, + "step": 1804 + }, + { + "epoch": 0.3784860557768924, + "grad_norm": 5.911870982962662, + "learning_rate": 1.9519410492636987e-05, + "loss": 2.3468, + "step": 1805 + }, + { + "epoch": 0.3786957433424198, + "grad_norm": 5.190451477521803, + "learning_rate": 1.951871686316794e-05, + "loss": 2.4207, + "step": 1806 + }, + { + "epoch": 0.37890543090794715, + "grad_norm": 6.236609353571022, + "learning_rate": 1.9518022745848138e-05, + "loss": 2.2166, + "step": 1807 + }, + { + "epoch": 0.3791151184734745, + "grad_norm": 5.566966338171751, + "learning_rate": 1.951732814071317e-05, + "loss": 2.1937, + "step": 1808 + }, + { + "epoch": 0.37932480603900187, + "grad_norm": 5.3646977857229015, + "learning_rate": 1.951663304779862e-05, + "loss": 2.3121, + "step": 1809 + }, + { + "epoch": 0.37953449360452923, + "grad_norm": 6.041580362498255, + "learning_rate": 1.9515937467140126e-05, + "loss": 2.3487, + "step": 1810 + }, + { + "epoch": 0.3797441811700566, + "grad_norm": 5.937631434629704, + "learning_rate": 1.951524139877333e-05, + "loss": 2.0437, + "step": 1811 + }, + { + "epoch": 0.37995386873558396, + "grad_norm": 4.702474627762571, + "learning_rate": 1.9514544842733907e-05, + "loss": 2.0293, + "step": 1812 + }, + { + "epoch": 0.38016355630111137, + "grad_norm": 6.2191410024144576, + "learning_rate": 1.9513847799057565e-05, + "loss": 2.0789, + "step": 1813 + }, + { + "epoch": 0.38037324386663873, + "grad_norm": 6.09845026031638, + "learning_rate": 1.9513150267780018e-05, + "loss": 2.2375, + "step": 1814 + }, + { + "epoch": 0.3805829314321661, + "grad_norm": 6.151681530295985, + "learning_rate": 1.9512452248937022e-05, + "loss": 2.4172, + "step": 1815 + }, + { + "epoch": 0.38079261899769346, + "grad_norm": 6.2437661159773254, + "learning_rate": 1.9511753742564345e-05, + "loss": 2.2121, + "step": 1816 + }, + { + "epoch": 0.3810023065632208, + "grad_norm": 5.972548574427595, + "learning_rate": 1.9511054748697797e-05, + "loss": 2.3818, + "step": 1817 + }, + { + "epoch": 0.3812119941287482, + "grad_norm": 5.6958767391655245, + "learning_rate": 1.9510355267373193e-05, + "loss": 2.0622, + "step": 1818 + }, + { + "epoch": 0.38142168169427554, + "grad_norm": 5.925729691011436, + "learning_rate": 1.950965529862639e-05, + "loss": 2.1346, + "step": 1819 + }, + { + "epoch": 0.3816313692598029, + "grad_norm": 6.437704271425044, + "learning_rate": 1.9508954842493257e-05, + "loss": 2.4005, + "step": 1820 + }, + { + "epoch": 0.38184105682533026, + "grad_norm": 5.728483053173975, + "learning_rate": 1.9508253899009697e-05, + "loss": 2.2269, + "step": 1821 + }, + { + "epoch": 0.3820507443908576, + "grad_norm": 5.667349795046705, + "learning_rate": 1.9507552468211632e-05, + "loss": 2.2204, + "step": 1822 + }, + { + "epoch": 0.382260431956385, + "grad_norm": 5.244236341061125, + "learning_rate": 1.9506850550135015e-05, + "loss": 1.8756, + "step": 1823 + }, + { + "epoch": 0.38247011952191234, + "grad_norm": 5.58873787539878, + "learning_rate": 1.9506148144815818e-05, + "loss": 2.1292, + "step": 1824 + }, + { + "epoch": 0.3826798070874397, + "grad_norm": 6.786100292525884, + "learning_rate": 1.950544525229004e-05, + "loss": 2.241, + "step": 1825 + }, + { + "epoch": 0.38288949465296707, + "grad_norm": 6.481006854056999, + "learning_rate": 1.9504741872593707e-05, + "loss": 2.4264, + "step": 1826 + }, + { + "epoch": 0.3830991822184944, + "grad_norm": 6.375598964649223, + "learning_rate": 1.950403800576287e-05, + "loss": 2.306, + "step": 1827 + }, + { + "epoch": 0.3833088697840218, + "grad_norm": 5.141686094663298, + "learning_rate": 1.95033336518336e-05, + "loss": 2.0233, + "step": 1828 + }, + { + "epoch": 0.38351855734954915, + "grad_norm": 5.501768906196965, + "learning_rate": 1.9502628810841996e-05, + "loss": 2.2887, + "step": 1829 + }, + { + "epoch": 0.3837282449150765, + "grad_norm": 6.9227133018860085, + "learning_rate": 1.9501923482824182e-05, + "loss": 2.4476, + "step": 1830 + }, + { + "epoch": 0.3839379324806039, + "grad_norm": 7.5199385007583635, + "learning_rate": 1.9501217667816312e-05, + "loss": 2.2419, + "step": 1831 + }, + { + "epoch": 0.3841476200461313, + "grad_norm": 5.160971370597283, + "learning_rate": 1.9500511365854558e-05, + "loss": 1.8077, + "step": 1832 + }, + { + "epoch": 0.38435730761165865, + "grad_norm": 6.827245150658511, + "learning_rate": 1.9499804576975117e-05, + "loss": 2.2194, + "step": 1833 + }, + { + "epoch": 0.384566995177186, + "grad_norm": 6.14911191175889, + "learning_rate": 1.9499097301214216e-05, + "loss": 2.3399, + "step": 1834 + }, + { + "epoch": 0.38477668274271337, + "grad_norm": 6.141183349714347, + "learning_rate": 1.94983895386081e-05, + "loss": 2.3681, + "step": 1835 + }, + { + "epoch": 0.38498637030824073, + "grad_norm": 5.960549119418551, + "learning_rate": 1.9497681289193053e-05, + "loss": 2.217, + "step": 1836 + }, + { + "epoch": 0.3851960578737681, + "grad_norm": 6.813486625485713, + "learning_rate": 1.9496972553005365e-05, + "loss": 2.0061, + "step": 1837 + }, + { + "epoch": 0.38540574543929546, + "grad_norm": 6.265557676414994, + "learning_rate": 1.9496263330081357e-05, + "loss": 2.1097, + "step": 1838 + }, + { + "epoch": 0.3856154330048228, + "grad_norm": 5.133315055230063, + "learning_rate": 1.9495553620457385e-05, + "loss": 2.0452, + "step": 1839 + }, + { + "epoch": 0.3858251205703502, + "grad_norm": 6.173405561313836, + "learning_rate": 1.9494843424169823e-05, + "loss": 2.3421, + "step": 1840 + }, + { + "epoch": 0.38603480813587754, + "grad_norm": 5.4848310846079755, + "learning_rate": 1.9494132741255067e-05, + "loss": 2.2359, + "step": 1841 + }, + { + "epoch": 0.3862444957014049, + "grad_norm": 6.331471632008918, + "learning_rate": 1.9493421571749543e-05, + "loss": 2.5103, + "step": 1842 + }, + { + "epoch": 0.38645418326693226, + "grad_norm": 5.174664578086117, + "learning_rate": 1.9492709915689693e-05, + "loss": 2.023, + "step": 1843 + }, + { + "epoch": 0.3866638708324596, + "grad_norm": 5.5631940789043295, + "learning_rate": 1.9491997773111997e-05, + "loss": 2.2882, + "step": 1844 + }, + { + "epoch": 0.386873558397987, + "grad_norm": 5.804478665474135, + "learning_rate": 1.9491285144052955e-05, + "loss": 2.2521, + "step": 1845 + }, + { + "epoch": 0.38708324596351434, + "grad_norm": 5.513779357743572, + "learning_rate": 1.9490572028549086e-05, + "loss": 2.2151, + "step": 1846 + }, + { + "epoch": 0.3872929335290417, + "grad_norm": 5.770079966300676, + "learning_rate": 1.948985842663694e-05, + "loss": 2.1534, + "step": 1847 + }, + { + "epoch": 0.38750262109456907, + "grad_norm": 5.732785030799801, + "learning_rate": 1.948914433835309e-05, + "loss": 2.1023, + "step": 1848 + }, + { + "epoch": 0.3877123086600965, + "grad_norm": 6.392519378479588, + "learning_rate": 1.9488429763734132e-05, + "loss": 2.0394, + "step": 1849 + }, + { + "epoch": 0.38792199622562384, + "grad_norm": 4.798667761130312, + "learning_rate": 1.9487714702816696e-05, + "loss": 2.3865, + "step": 1850 + }, + { + "epoch": 0.3881316837911512, + "grad_norm": 5.429639162390453, + "learning_rate": 1.9486999155637424e-05, + "loss": 2.0597, + "step": 1851 + }, + { + "epoch": 0.38834137135667857, + "grad_norm": 5.392244512945511, + "learning_rate": 1.948628312223299e-05, + "loss": 2.2082, + "step": 1852 + }, + { + "epoch": 0.3885510589222059, + "grad_norm": 6.755034231590265, + "learning_rate": 1.9485566602640095e-05, + "loss": 2.4778, + "step": 1853 + }, + { + "epoch": 0.3887607464877333, + "grad_norm": 4.953659009463895, + "learning_rate": 1.948484959689546e-05, + "loss": 2.0787, + "step": 1854 + }, + { + "epoch": 0.38897043405326065, + "grad_norm": 5.807587794949061, + "learning_rate": 1.9484132105035832e-05, + "loss": 2.0223, + "step": 1855 + }, + { + "epoch": 0.389180121618788, + "grad_norm": 5.350079350058516, + "learning_rate": 1.948341412709798e-05, + "loss": 2.5058, + "step": 1856 + }, + { + "epoch": 0.38938980918431537, + "grad_norm": 5.23715133920839, + "learning_rate": 1.948269566311871e-05, + "loss": 1.9987, + "step": 1857 + }, + { + "epoch": 0.38959949674984273, + "grad_norm": 5.593821027473954, + "learning_rate": 1.948197671313484e-05, + "loss": 2.3818, + "step": 1858 + }, + { + "epoch": 0.3898091843153701, + "grad_norm": 5.119605514207335, + "learning_rate": 1.948125727718322e-05, + "loss": 2.3033, + "step": 1859 + }, + { + "epoch": 0.39001887188089746, + "grad_norm": 5.620300100417322, + "learning_rate": 1.9480537355300715e-05, + "loss": 2.3298, + "step": 1860 + }, + { + "epoch": 0.3902285594464248, + "grad_norm": 7.210124954699667, + "learning_rate": 1.947981694752423e-05, + "loss": 2.4181, + "step": 1861 + }, + { + "epoch": 0.3904382470119522, + "grad_norm": 6.222535625493142, + "learning_rate": 1.9479096053890686e-05, + "loss": 2.4226, + "step": 1862 + }, + { + "epoch": 0.39064793457747954, + "grad_norm": 5.411448404953495, + "learning_rate": 1.9478374674437027e-05, + "loss": 2.3347, + "step": 1863 + }, + { + "epoch": 0.3908576221430069, + "grad_norm": 5.912691002972943, + "learning_rate": 1.9477652809200224e-05, + "loss": 2.2209, + "step": 1864 + }, + { + "epoch": 0.39106730970853426, + "grad_norm": 5.801364757590116, + "learning_rate": 1.947693045821728e-05, + "loss": 2.1197, + "step": 1865 + }, + { + "epoch": 0.3912769972740616, + "grad_norm": 6.185036871439721, + "learning_rate": 1.9476207621525215e-05, + "loss": 2.4563, + "step": 1866 + }, + { + "epoch": 0.39148668483958904, + "grad_norm": 5.63604263565064, + "learning_rate": 1.9475484299161066e-05, + "loss": 2.2743, + "step": 1867 + }, + { + "epoch": 0.3916963724051164, + "grad_norm": 5.459923500836046, + "learning_rate": 1.947476049116192e-05, + "loss": 2.0538, + "step": 1868 + }, + { + "epoch": 0.39190605997064376, + "grad_norm": 5.662428623856443, + "learning_rate": 1.9474036197564864e-05, + "loss": 2.0575, + "step": 1869 + }, + { + "epoch": 0.3921157475361711, + "grad_norm": 5.28329185545302, + "learning_rate": 1.9473311418407017e-05, + "loss": 1.9175, + "step": 1870 + }, + { + "epoch": 0.3923254351016985, + "grad_norm": 5.269757960336864, + "learning_rate": 1.9472586153725532e-05, + "loss": 2.4371, + "step": 1871 + }, + { + "epoch": 0.39253512266722584, + "grad_norm": 5.325441176060469, + "learning_rate": 1.947186040355758e-05, + "loss": 2.099, + "step": 1872 + }, + { + "epoch": 0.3927448102327532, + "grad_norm": 6.168658986315181, + "learning_rate": 1.947113416794035e-05, + "loss": 2.0369, + "step": 1873 + }, + { + "epoch": 0.39295449779828057, + "grad_norm": 5.144962237885244, + "learning_rate": 1.9470407446911068e-05, + "loss": 2.1908, + "step": 1874 + }, + { + "epoch": 0.3931641853638079, + "grad_norm": 6.138871633142044, + "learning_rate": 1.9469680240506982e-05, + "loss": 2.5676, + "step": 1875 + }, + { + "epoch": 0.3933738729293353, + "grad_norm": 5.658040325645907, + "learning_rate": 1.9468952548765353e-05, + "loss": 2.002, + "step": 1876 + }, + { + "epoch": 0.39358356049486265, + "grad_norm": 5.370314318369707, + "learning_rate": 1.946822437172349e-05, + "loss": 2.0515, + "step": 1877 + }, + { + "epoch": 0.39379324806039, + "grad_norm": 5.945691469761393, + "learning_rate": 1.94674957094187e-05, + "loss": 2.252, + "step": 1878 + }, + { + "epoch": 0.39400293562591737, + "grad_norm": 5.470246503296191, + "learning_rate": 1.9466766561888337e-05, + "loss": 1.7592, + "step": 1879 + }, + { + "epoch": 0.39421262319144473, + "grad_norm": 5.840271399222884, + "learning_rate": 1.946603692916977e-05, + "loss": 2.4053, + "step": 1880 + }, + { + "epoch": 0.3944223107569721, + "grad_norm": 5.852093743622925, + "learning_rate": 1.946530681130039e-05, + "loss": 2.0543, + "step": 1881 + }, + { + "epoch": 0.39463199832249946, + "grad_norm": 6.135305852856054, + "learning_rate": 1.9464576208317617e-05, + "loss": 2.2118, + "step": 1882 + }, + { + "epoch": 0.3948416858880268, + "grad_norm": 6.140807251464246, + "learning_rate": 1.9463845120258903e-05, + "loss": 2.3049, + "step": 1883 + }, + { + "epoch": 0.3950513734535542, + "grad_norm": 5.422938242095375, + "learning_rate": 1.946311354716171e-05, + "loss": 1.88, + "step": 1884 + }, + { + "epoch": 0.3952610610190816, + "grad_norm": 5.251847127764769, + "learning_rate": 1.9462381489063532e-05, + "loss": 2.0867, + "step": 1885 + }, + { + "epoch": 0.39547074858460896, + "grad_norm": 6.303060814259278, + "learning_rate": 1.9461648946001898e-05, + "loss": 2.3058, + "step": 1886 + }, + { + "epoch": 0.3956804361501363, + "grad_norm": 6.013341286899624, + "learning_rate": 1.946091591801434e-05, + "loss": 2.3494, + "step": 1887 + }, + { + "epoch": 0.3958901237156637, + "grad_norm": 6.4192967776392615, + "learning_rate": 1.946018240513843e-05, + "loss": 2.3006, + "step": 1888 + }, + { + "epoch": 0.39609981128119104, + "grad_norm": 6.281004130476324, + "learning_rate": 1.945944840741177e-05, + "loss": 2.0074, + "step": 1889 + }, + { + "epoch": 0.3963094988467184, + "grad_norm": 5.909475354444969, + "learning_rate": 1.945871392487197e-05, + "loss": 2.3995, + "step": 1890 + }, + { + "epoch": 0.39651918641224576, + "grad_norm": 5.748361824524642, + "learning_rate": 1.9457978957556675e-05, + "loss": 1.9936, + "step": 1891 + }, + { + "epoch": 0.3967288739777731, + "grad_norm": 5.589512218717744, + "learning_rate": 1.9457243505503552e-05, + "loss": 2.1116, + "step": 1892 + }, + { + "epoch": 0.3969385615433005, + "grad_norm": 4.813067779704906, + "learning_rate": 1.9456507568750303e-05, + "loss": 2.0623, + "step": 1893 + }, + { + "epoch": 0.39714824910882784, + "grad_norm": 5.8312918681549775, + "learning_rate": 1.9455771147334634e-05, + "loss": 2.3951, + "step": 1894 + }, + { + "epoch": 0.3973579366743552, + "grad_norm": 5.630188593868802, + "learning_rate": 1.9455034241294298e-05, + "loss": 1.9422, + "step": 1895 + }, + { + "epoch": 0.39756762423988257, + "grad_norm": 5.726957346586343, + "learning_rate": 1.9454296850667058e-05, + "loss": 2.1655, + "step": 1896 + }, + { + "epoch": 0.3977773118054099, + "grad_norm": 5.710283570253907, + "learning_rate": 1.94535589754907e-05, + "loss": 2.0759, + "step": 1897 + }, + { + "epoch": 0.3979869993709373, + "grad_norm": 6.018976066968295, + "learning_rate": 1.9452820615803055e-05, + "loss": 2.5303, + "step": 1898 + }, + { + "epoch": 0.39819668693646465, + "grad_norm": 6.404827914917179, + "learning_rate": 1.9452081771641954e-05, + "loss": 2.1774, + "step": 1899 + }, + { + "epoch": 0.398406374501992, + "grad_norm": 5.989486646424691, + "learning_rate": 1.9451342443045275e-05, + "loss": 2.1306, + "step": 1900 + }, + { + "epoch": 0.39861606206751937, + "grad_norm": 6.162279013491712, + "learning_rate": 1.9450602630050898e-05, + "loss": 2.0919, + "step": 1901 + }, + { + "epoch": 0.39882574963304673, + "grad_norm": 5.882485437451917, + "learning_rate": 1.9449862332696745e-05, + "loss": 2.4434, + "step": 1902 + }, + { + "epoch": 0.39903543719857415, + "grad_norm": 5.618885858075298, + "learning_rate": 1.9449121551020757e-05, + "loss": 2.0226, + "step": 1903 + }, + { + "epoch": 0.3992451247641015, + "grad_norm": 8.12403816550133, + "learning_rate": 1.9448380285060902e-05, + "loss": 2.2967, + "step": 1904 + }, + { + "epoch": 0.39945481232962887, + "grad_norm": 6.0100040213393235, + "learning_rate": 1.944763853485517e-05, + "loss": 1.907, + "step": 1905 + }, + { + "epoch": 0.39966449989515623, + "grad_norm": 5.430596194283071, + "learning_rate": 1.9446896300441578e-05, + "loss": 1.9985, + "step": 1906 + }, + { + "epoch": 0.3998741874606836, + "grad_norm": 5.409521757371335, + "learning_rate": 1.9446153581858165e-05, + "loss": 1.8089, + "step": 1907 + }, + { + "epoch": 0.40008387502621096, + "grad_norm": 6.467306966086017, + "learning_rate": 1.9445410379142995e-05, + "loss": 2.3695, + "step": 1908 + }, + { + "epoch": 0.4002935625917383, + "grad_norm": 5.908066545279196, + "learning_rate": 1.9444666692334164e-05, + "loss": 2.1107, + "step": 1909 + }, + { + "epoch": 0.4005032501572657, + "grad_norm": 5.3699400708350975, + "learning_rate": 1.944392252146978e-05, + "loss": 2.2557, + "step": 1910 + }, + { + "epoch": 0.40071293772279304, + "grad_norm": 5.9692236072071285, + "learning_rate": 1.9443177866587995e-05, + "loss": 2.2331, + "step": 1911 + }, + { + "epoch": 0.4009226252883204, + "grad_norm": 5.663327693465919, + "learning_rate": 1.944243272772696e-05, + "loss": 2.131, + "step": 1912 + }, + { + "epoch": 0.40113231285384776, + "grad_norm": 5.742374662670714, + "learning_rate": 1.944168710492487e-05, + "loss": 2.1401, + "step": 1913 + }, + { + "epoch": 0.4013420004193751, + "grad_norm": 5.540911240883638, + "learning_rate": 1.9440940998219944e-05, + "loss": 2.3733, + "step": 1914 + }, + { + "epoch": 0.4015516879849025, + "grad_norm": 5.787970281732358, + "learning_rate": 1.9440194407650416e-05, + "loss": 1.8939, + "step": 1915 + }, + { + "epoch": 0.40176137555042984, + "grad_norm": 6.5950426253030985, + "learning_rate": 1.943944733325455e-05, + "loss": 2.239, + "step": 1916 + }, + { + "epoch": 0.4019710631159572, + "grad_norm": 5.41092777500551, + "learning_rate": 1.943869977507064e-05, + "loss": 2.094, + "step": 1917 + }, + { + "epoch": 0.40218075068148457, + "grad_norm": 5.275996145545977, + "learning_rate": 1.943795173313699e-05, + "loss": 2.0839, + "step": 1918 + }, + { + "epoch": 0.40239043824701193, + "grad_norm": 5.845261694296803, + "learning_rate": 1.9437203207491948e-05, + "loss": 2.2976, + "step": 1919 + }, + { + "epoch": 0.4026001258125393, + "grad_norm": 5.623578486698355, + "learning_rate": 1.943645419817387e-05, + "loss": 1.96, + "step": 1920 + }, + { + "epoch": 0.4028098133780667, + "grad_norm": 6.04765277627018, + "learning_rate": 1.9435704705221155e-05, + "loss": 2.4378, + "step": 1921 + }, + { + "epoch": 0.40301950094359407, + "grad_norm": 5.141463983765473, + "learning_rate": 1.9434954728672202e-05, + "loss": 2.0341, + "step": 1922 + }, + { + "epoch": 0.40322918850912143, + "grad_norm": 5.801059022026269, + "learning_rate": 1.943420426856546e-05, + "loss": 2.4334, + "step": 1923 + }, + { + "epoch": 0.4034388760746488, + "grad_norm": 6.088325178561093, + "learning_rate": 1.9433453324939383e-05, + "loss": 2.3452, + "step": 1924 + }, + { + "epoch": 0.40364856364017615, + "grad_norm": 5.282094073983423, + "learning_rate": 1.943270189783246e-05, + "loss": 2.15, + "step": 1925 + }, + { + "epoch": 0.4038582512057035, + "grad_norm": 6.54386698151231, + "learning_rate": 1.9431949987283206e-05, + "loss": 2.239, + "step": 1926 + }, + { + "epoch": 0.4040679387712309, + "grad_norm": 5.713375697124021, + "learning_rate": 1.943119759333016e-05, + "loss": 2.2938, + "step": 1927 + }, + { + "epoch": 0.40427762633675823, + "grad_norm": 5.8333048924465185, + "learning_rate": 1.9430444716011875e-05, + "loss": 1.942, + "step": 1928 + }, + { + "epoch": 0.4044873139022856, + "grad_norm": 6.517545971492462, + "learning_rate": 1.9429691355366944e-05, + "loss": 2.2491, + "step": 1929 + }, + { + "epoch": 0.40469700146781296, + "grad_norm": 5.437227110157038, + "learning_rate": 1.9428937511433975e-05, + "loss": 2.3364, + "step": 1930 + }, + { + "epoch": 0.4049066890333403, + "grad_norm": 6.310440477579955, + "learning_rate": 1.9428183184251606e-05, + "loss": 1.8814, + "step": 1931 + }, + { + "epoch": 0.4051163765988677, + "grad_norm": 6.330108793455676, + "learning_rate": 1.9427428373858496e-05, + "loss": 2.7728, + "step": 1932 + }, + { + "epoch": 0.40532606416439504, + "grad_norm": 5.700414575768137, + "learning_rate": 1.942667308029333e-05, + "loss": 2.0245, + "step": 1933 + }, + { + "epoch": 0.4055357517299224, + "grad_norm": 5.972185828031188, + "learning_rate": 1.9425917303594823e-05, + "loss": 2.2138, + "step": 1934 + }, + { + "epoch": 0.40574543929544976, + "grad_norm": 5.827876684316709, + "learning_rate": 1.94251610438017e-05, + "loss": 2.3002, + "step": 1935 + }, + { + "epoch": 0.4059551268609771, + "grad_norm": 6.242194809509504, + "learning_rate": 1.942440430095273e-05, + "loss": 2.3103, + "step": 1936 + }, + { + "epoch": 0.4061648144265045, + "grad_norm": 6.014832790031313, + "learning_rate": 1.9423647075086693e-05, + "loss": 2.0443, + "step": 1937 + }, + { + "epoch": 0.4063745019920319, + "grad_norm": 6.731533423516322, + "learning_rate": 1.94228893662424e-05, + "loss": 2.4112, + "step": 1938 + }, + { + "epoch": 0.40658418955755926, + "grad_norm": 5.100344149858468, + "learning_rate": 1.942213117445868e-05, + "loss": 1.9673, + "step": 1939 + }, + { + "epoch": 0.4067938771230866, + "grad_norm": 5.920650543366484, + "learning_rate": 1.94213724997744e-05, + "loss": 2.2348, + "step": 1940 + }, + { + "epoch": 0.407003564688614, + "grad_norm": 5.745769287027729, + "learning_rate": 1.9420613342228437e-05, + "loss": 2.0051, + "step": 1941 + }, + { + "epoch": 0.40721325225414134, + "grad_norm": 5.25576506290497, + "learning_rate": 1.9419853701859698e-05, + "loss": 2.1942, + "step": 1942 + }, + { + "epoch": 0.4074229398196687, + "grad_norm": 5.602093551048213, + "learning_rate": 1.9419093578707125e-05, + "loss": 2.1569, + "step": 1943 + }, + { + "epoch": 0.40763262738519607, + "grad_norm": 6.038675721324789, + "learning_rate": 1.9418332972809667e-05, + "loss": 2.1447, + "step": 1944 + }, + { + "epoch": 0.40784231495072343, + "grad_norm": 5.568682957563024, + "learning_rate": 1.941757188420631e-05, + "loss": 2.0811, + "step": 1945 + }, + { + "epoch": 0.4080520025162508, + "grad_norm": 6.560601566957782, + "learning_rate": 1.9416810312936056e-05, + "loss": 2.1096, + "step": 1946 + }, + { + "epoch": 0.40826169008177815, + "grad_norm": 5.938554965271541, + "learning_rate": 1.9416048259037942e-05, + "loss": 2.2737, + "step": 1947 + }, + { + "epoch": 0.4084713776473055, + "grad_norm": 5.649517221423893, + "learning_rate": 1.9415285722551026e-05, + "loss": 2.02, + "step": 1948 + }, + { + "epoch": 0.4086810652128329, + "grad_norm": 5.872569513695348, + "learning_rate": 1.9414522703514384e-05, + "loss": 2.1347, + "step": 1949 + }, + { + "epoch": 0.40889075277836023, + "grad_norm": 5.376845176851961, + "learning_rate": 1.9413759201967124e-05, + "loss": 1.8583, + "step": 1950 + }, + { + "epoch": 0.4091004403438876, + "grad_norm": 4.832515896166294, + "learning_rate": 1.941299521794838e-05, + "loss": 1.7673, + "step": 1951 + }, + { + "epoch": 0.40931012790941496, + "grad_norm": 5.6830955250949104, + "learning_rate": 1.9412230751497306e-05, + "loss": 2.1055, + "step": 1952 + }, + { + "epoch": 0.4095198154749423, + "grad_norm": 6.3690993252994135, + "learning_rate": 1.941146580265308e-05, + "loss": 1.9798, + "step": 1953 + }, + { + "epoch": 0.4097295030404697, + "grad_norm": 5.5432250469034985, + "learning_rate": 1.941070037145491e-05, + "loss": 2.092, + "step": 1954 + }, + { + "epoch": 0.40993919060599704, + "grad_norm": 5.002611567338543, + "learning_rate": 1.940993445794202e-05, + "loss": 2.316, + "step": 1955 + }, + { + "epoch": 0.41014887817152446, + "grad_norm": 5.482867095252288, + "learning_rate": 1.9409168062153668e-05, + "loss": 2.2611, + "step": 1956 + }, + { + "epoch": 0.4103585657370518, + "grad_norm": 5.865665536175434, + "learning_rate": 1.9408401184129135e-05, + "loss": 2.3242, + "step": 1957 + }, + { + "epoch": 0.4105682533025792, + "grad_norm": 5.631661599101595, + "learning_rate": 1.9407633823907725e-05, + "loss": 2.3295, + "step": 1958 + }, + { + "epoch": 0.41077794086810654, + "grad_norm": 6.734560103345229, + "learning_rate": 1.9406865981528765e-05, + "loss": 2.0642, + "step": 1959 + }, + { + "epoch": 0.4109876284336339, + "grad_norm": 5.286191746354845, + "learning_rate": 1.9406097657031604e-05, + "loss": 1.9251, + "step": 1960 + }, + { + "epoch": 0.41119731599916126, + "grad_norm": 5.484666641168899, + "learning_rate": 1.940532885045563e-05, + "loss": 2.3152, + "step": 1961 + }, + { + "epoch": 0.4114070035646886, + "grad_norm": 5.510693198200412, + "learning_rate": 1.9404559561840235e-05, + "loss": 2.1775, + "step": 1962 + }, + { + "epoch": 0.411616691130216, + "grad_norm": 6.082396449135801, + "learning_rate": 1.9403789791224855e-05, + "loss": 2.415, + "step": 1963 + }, + { + "epoch": 0.41182637869574334, + "grad_norm": 5.450417707000446, + "learning_rate": 1.940301953864893e-05, + "loss": 2.2307, + "step": 1964 + }, + { + "epoch": 0.4120360662612707, + "grad_norm": 5.605150020655879, + "learning_rate": 1.9402248804151952e-05, + "loss": 2.125, + "step": 1965 + }, + { + "epoch": 0.41224575382679807, + "grad_norm": 4.615372468950363, + "learning_rate": 1.9401477587773415e-05, + "loss": 2.3, + "step": 1966 + }, + { + "epoch": 0.41245544139232543, + "grad_norm": 5.912093454099453, + "learning_rate": 1.9400705889552845e-05, + "loss": 2.3672, + "step": 1967 + }, + { + "epoch": 0.4126651289578528, + "grad_norm": 4.885158072336608, + "learning_rate": 1.9399933709529793e-05, + "loss": 2.1908, + "step": 1968 + }, + { + "epoch": 0.41287481652338015, + "grad_norm": 5.488791015816429, + "learning_rate": 1.9399161047743838e-05, + "loss": 2.4788, + "step": 1969 + }, + { + "epoch": 0.4130845040889075, + "grad_norm": 5.373212277401722, + "learning_rate": 1.939838790423457e-05, + "loss": 2.1084, + "step": 1970 + }, + { + "epoch": 0.4132941916544349, + "grad_norm": 5.147488078337295, + "learning_rate": 1.939761427904163e-05, + "loss": 2.1311, + "step": 1971 + }, + { + "epoch": 0.41350387921996223, + "grad_norm": 5.878288971852596, + "learning_rate": 1.9396840172204654e-05, + "loss": 2.4611, + "step": 1972 + }, + { + "epoch": 0.4137135667854896, + "grad_norm": 5.512726405067035, + "learning_rate": 1.9396065583763323e-05, + "loss": 2.3146, + "step": 1973 + }, + { + "epoch": 0.413923254351017, + "grad_norm": 6.948480560974792, + "learning_rate": 1.939529051375733e-05, + "loss": 2.3548, + "step": 1974 + }, + { + "epoch": 0.4141329419165444, + "grad_norm": 5.760753375338554, + "learning_rate": 1.9394514962226406e-05, + "loss": 2.2447, + "step": 1975 + }, + { + "epoch": 0.41434262948207173, + "grad_norm": 4.665031705307105, + "learning_rate": 1.9393738929210296e-05, + "loss": 2.2618, + "step": 1976 + }, + { + "epoch": 0.4145523170475991, + "grad_norm": 5.662796359610734, + "learning_rate": 1.9392962414748776e-05, + "loss": 2.4672, + "step": 1977 + }, + { + "epoch": 0.41476200461312646, + "grad_norm": 5.603834743461014, + "learning_rate": 1.9392185418881637e-05, + "loss": 2.1682, + "step": 1978 + }, + { + "epoch": 0.4149716921786538, + "grad_norm": 5.761363284897148, + "learning_rate": 1.9391407941648704e-05, + "loss": 2.2492, + "step": 1979 + }, + { + "epoch": 0.4151813797441812, + "grad_norm": 5.55979120910232, + "learning_rate": 1.939062998308983e-05, + "loss": 2.478, + "step": 1980 + }, + { + "epoch": 0.41539106730970854, + "grad_norm": 5.256870641705434, + "learning_rate": 1.938985154324488e-05, + "loss": 2.2195, + "step": 1981 + }, + { + "epoch": 0.4156007548752359, + "grad_norm": 4.598153324651477, + "learning_rate": 1.9389072622153754e-05, + "loss": 2.3643, + "step": 1982 + }, + { + "epoch": 0.41581044244076326, + "grad_norm": 5.478001173789781, + "learning_rate": 1.9388293219856373e-05, + "loss": 2.0816, + "step": 1983 + }, + { + "epoch": 0.4160201300062906, + "grad_norm": 5.228408349610673, + "learning_rate": 1.9387513336392676e-05, + "loss": 1.9742, + "step": 1984 + }, + { + "epoch": 0.416229817571818, + "grad_norm": 4.808294444991597, + "learning_rate": 1.938673297180264e-05, + "loss": 2.2271, + "step": 1985 + }, + { + "epoch": 0.41643950513734534, + "grad_norm": 5.438841530953087, + "learning_rate": 1.938595212612626e-05, + "loss": 2.2377, + "step": 1986 + }, + { + "epoch": 0.4166491927028727, + "grad_norm": 5.375547713308862, + "learning_rate": 1.9385170799403553e-05, + "loss": 2.2204, + "step": 1987 + }, + { + "epoch": 0.41685888026840007, + "grad_norm": 5.790129632703097, + "learning_rate": 1.9384388991674568e-05, + "loss": 2.1856, + "step": 1988 + }, + { + "epoch": 0.41706856783392743, + "grad_norm": 5.9808512888628185, + "learning_rate": 1.938360670297937e-05, + "loss": 2.0137, + "step": 1989 + }, + { + "epoch": 0.4172782553994548, + "grad_norm": 5.359704491298829, + "learning_rate": 1.9382823933358052e-05, + "loss": 1.8506, + "step": 1990 + }, + { + "epoch": 0.41748794296498215, + "grad_norm": 5.746329330604571, + "learning_rate": 1.9382040682850736e-05, + "loss": 2.353, + "step": 1991 + }, + { + "epoch": 0.41769763053050957, + "grad_norm": 6.7324940984961605, + "learning_rate": 1.938125695149756e-05, + "loss": 2.2354, + "step": 1992 + }, + { + "epoch": 0.41790731809603693, + "grad_norm": 5.549159840172159, + "learning_rate": 1.9380472739338697e-05, + "loss": 1.8577, + "step": 1993 + }, + { + "epoch": 0.4181170056615643, + "grad_norm": 6.567428796883728, + "learning_rate": 1.9379688046414333e-05, + "loss": 2.1882, + "step": 1994 + }, + { + "epoch": 0.41832669322709165, + "grad_norm": 6.442302177292334, + "learning_rate": 1.9378902872764687e-05, + "loss": 2.1867, + "step": 1995 + }, + { + "epoch": 0.418536380792619, + "grad_norm": 5.721773668328924, + "learning_rate": 1.9378117218430007e-05, + "loss": 1.9088, + "step": 1996 + }, + { + "epoch": 0.4187460683581464, + "grad_norm": 7.116147412325039, + "learning_rate": 1.9377331083450553e-05, + "loss": 2.4424, + "step": 1997 + }, + { + "epoch": 0.41895575592367373, + "grad_norm": 6.294320289918357, + "learning_rate": 1.9376544467866618e-05, + "loss": 2.4735, + "step": 1998 + }, + { + "epoch": 0.4191654434892011, + "grad_norm": 5.862530198939572, + "learning_rate": 1.9375757371718512e-05, + "loss": 2.1935, + "step": 1999 + }, + { + "epoch": 0.41937513105472846, + "grad_norm": 6.265295477683987, + "learning_rate": 1.937496979504658e-05, + "loss": 2.2382, + "step": 2000 + }, + { + "epoch": 0.4195848186202558, + "grad_norm": 6.060609809494632, + "learning_rate": 1.937418173789119e-05, + "loss": 2.0672, + "step": 2001 + }, + { + "epoch": 0.4197945061857832, + "grad_norm": 5.832967778638011, + "learning_rate": 1.937339320029272e-05, + "loss": 2.1739, + "step": 2002 + }, + { + "epoch": 0.42000419375131054, + "grad_norm": 5.9144866941147, + "learning_rate": 1.93726041822916e-05, + "loss": 2.42, + "step": 2003 + }, + { + "epoch": 0.4202138813168379, + "grad_norm": 5.769492824448438, + "learning_rate": 1.9371814683928254e-05, + "loss": 2.107, + "step": 2004 + }, + { + "epoch": 0.42042356888236526, + "grad_norm": 6.128745604677519, + "learning_rate": 1.9371024705243153e-05, + "loss": 2.2926, + "step": 2005 + }, + { + "epoch": 0.4206332564478926, + "grad_norm": 6.94799669224784, + "learning_rate": 1.937023424627678e-05, + "loss": 2.0711, + "step": 2006 + }, + { + "epoch": 0.42084294401342, + "grad_norm": 6.253542380198567, + "learning_rate": 1.9369443307069652e-05, + "loss": 2.1404, + "step": 2007 + }, + { + "epoch": 0.42105263157894735, + "grad_norm": 5.731252920061354, + "learning_rate": 1.9368651887662305e-05, + "loss": 1.9719, + "step": 2008 + }, + { + "epoch": 0.4212623191444747, + "grad_norm": 5.54887162896101, + "learning_rate": 1.93678599880953e-05, + "loss": 2.1454, + "step": 2009 + }, + { + "epoch": 0.4214720067100021, + "grad_norm": 6.9384749252134865, + "learning_rate": 1.936706760840922e-05, + "loss": 2.3265, + "step": 2010 + }, + { + "epoch": 0.4216816942755295, + "grad_norm": 5.880858818188466, + "learning_rate": 1.936627474864468e-05, + "loss": 2.356, + "step": 2011 + }, + { + "epoch": 0.42189138184105685, + "grad_norm": 5.834522963809873, + "learning_rate": 1.9365481408842316e-05, + "loss": 2.2803, + "step": 2012 + }, + { + "epoch": 0.4221010694065842, + "grad_norm": 5.890280060192829, + "learning_rate": 1.9364687589042784e-05, + "loss": 2.2733, + "step": 2013 + }, + { + "epoch": 0.42231075697211157, + "grad_norm": 6.7025715581683825, + "learning_rate": 1.936389328928677e-05, + "loss": 2.2507, + "step": 2014 + }, + { + "epoch": 0.42252044453763893, + "grad_norm": 5.862394687737632, + "learning_rate": 1.9363098509614986e-05, + "loss": 2.3053, + "step": 2015 + }, + { + "epoch": 0.4227301321031663, + "grad_norm": 6.616369166435254, + "learning_rate": 1.9362303250068166e-05, + "loss": 2.2514, + "step": 2016 + }, + { + "epoch": 0.42293981966869365, + "grad_norm": 5.314512617857007, + "learning_rate": 1.936150751068706e-05, + "loss": 1.8628, + "step": 2017 + }, + { + "epoch": 0.423149507234221, + "grad_norm": 6.13395910387215, + "learning_rate": 1.9360711291512463e-05, + "loss": 2.4728, + "step": 2018 + }, + { + "epoch": 0.4233591947997484, + "grad_norm": 6.6003835680659115, + "learning_rate": 1.9359914592585174e-05, + "loss": 2.0128, + "step": 2019 + }, + { + "epoch": 0.42356888236527573, + "grad_norm": 5.536355983436875, + "learning_rate": 1.935911741394603e-05, + "loss": 2.0144, + "step": 2020 + }, + { + "epoch": 0.4237785699308031, + "grad_norm": 6.770656421196363, + "learning_rate": 1.9358319755635886e-05, + "loss": 2.237, + "step": 2021 + }, + { + "epoch": 0.42398825749633046, + "grad_norm": 5.178819318223887, + "learning_rate": 1.9357521617695622e-05, + "loss": 2.2974, + "step": 2022 + }, + { + "epoch": 0.4241979450618578, + "grad_norm": 5.249639613602149, + "learning_rate": 1.9356723000166144e-05, + "loss": 2.2677, + "step": 2023 + }, + { + "epoch": 0.4244076326273852, + "grad_norm": 6.109689196967141, + "learning_rate": 1.9355923903088387e-05, + "loss": 1.96, + "step": 2024 + }, + { + "epoch": 0.42461732019291254, + "grad_norm": 4.92541630073264, + "learning_rate": 1.93551243265033e-05, + "loss": 2.2351, + "step": 2025 + }, + { + "epoch": 0.4248270077584399, + "grad_norm": 5.602064620773708, + "learning_rate": 1.9354324270451868e-05, + "loss": 2.1125, + "step": 2026 + }, + { + "epoch": 0.42503669532396726, + "grad_norm": 6.330968911541508, + "learning_rate": 1.9353523734975093e-05, + "loss": 2.2718, + "step": 2027 + }, + { + "epoch": 0.4252463828894947, + "grad_norm": 5.135501745144233, + "learning_rate": 1.9352722720114e-05, + "loss": 2.3348, + "step": 2028 + }, + { + "epoch": 0.42545607045502204, + "grad_norm": 5.126734120913945, + "learning_rate": 1.935192122590965e-05, + "loss": 1.8358, + "step": 2029 + }, + { + "epoch": 0.4256657580205494, + "grad_norm": 6.506488498787985, + "learning_rate": 1.9351119252403117e-05, + "loss": 2.07, + "step": 2030 + }, + { + "epoch": 0.42587544558607676, + "grad_norm": 5.996247350017663, + "learning_rate": 1.93503167996355e-05, + "loss": 2.3962, + "step": 2031 + }, + { + "epoch": 0.4260851331516041, + "grad_norm": 6.091321862307605, + "learning_rate": 1.9349513867647934e-05, + "loss": 2.3194, + "step": 2032 + }, + { + "epoch": 0.4262948207171315, + "grad_norm": 6.715491028435017, + "learning_rate": 1.9348710456481567e-05, + "loss": 2.1027, + "step": 2033 + }, + { + "epoch": 0.42650450828265885, + "grad_norm": 5.94320774840823, + "learning_rate": 1.9347906566177572e-05, + "loss": 2.2841, + "step": 2034 + }, + { + "epoch": 0.4267141958481862, + "grad_norm": 5.734420617348297, + "learning_rate": 1.9347102196777153e-05, + "loss": 2.0446, + "step": 2035 + }, + { + "epoch": 0.42692388341371357, + "grad_norm": 5.824742045198555, + "learning_rate": 1.9346297348321534e-05, + "loss": 2.2614, + "step": 2036 + }, + { + "epoch": 0.42713357097924093, + "grad_norm": 6.252753944137568, + "learning_rate": 1.9345492020851967e-05, + "loss": 2.1476, + "step": 2037 + }, + { + "epoch": 0.4273432585447683, + "grad_norm": 6.717286218553124, + "learning_rate": 1.9344686214409724e-05, + "loss": 2.253, + "step": 2038 + }, + { + "epoch": 0.42755294611029565, + "grad_norm": 5.373113340246127, + "learning_rate": 1.9343879929036103e-05, + "loss": 2.1445, + "step": 2039 + }, + { + "epoch": 0.427762633675823, + "grad_norm": 6.3643725544219505, + "learning_rate": 1.9343073164772432e-05, + "loss": 2.1806, + "step": 2040 + }, + { + "epoch": 0.4279723212413504, + "grad_norm": 6.199200701371801, + "learning_rate": 1.9342265921660054e-05, + "loss": 2.143, + "step": 2041 + }, + { + "epoch": 0.42818200880687773, + "grad_norm": 5.66314781975456, + "learning_rate": 1.934145819974035e-05, + "loss": 1.9979, + "step": 2042 + }, + { + "epoch": 0.4283916963724051, + "grad_norm": 6.286261967743655, + "learning_rate": 1.9340649999054707e-05, + "loss": 2.1439, + "step": 2043 + }, + { + "epoch": 0.42860138393793246, + "grad_norm": 5.476334925418123, + "learning_rate": 1.933984131964455e-05, + "loss": 2.3297, + "step": 2044 + }, + { + "epoch": 0.4288110715034598, + "grad_norm": 5.656426868090549, + "learning_rate": 1.9339032161551328e-05, + "loss": 2.2493, + "step": 2045 + }, + { + "epoch": 0.42902075906898723, + "grad_norm": 5.283818501014784, + "learning_rate": 1.933822252481651e-05, + "loss": 2.3128, + "step": 2046 + }, + { + "epoch": 0.4292304466345146, + "grad_norm": 5.579397238106657, + "learning_rate": 1.933741240948159e-05, + "loss": 2.0862, + "step": 2047 + }, + { + "epoch": 0.42944013420004196, + "grad_norm": 5.35244711965899, + "learning_rate": 1.9336601815588092e-05, + "loss": 2.3516, + "step": 2048 + }, + { + "epoch": 0.4296498217655693, + "grad_norm": 5.848286279440088, + "learning_rate": 1.9335790743177555e-05, + "loss": 2.2656, + "step": 2049 + }, + { + "epoch": 0.4298595093310967, + "grad_norm": 5.326979090134375, + "learning_rate": 1.933497919229155e-05, + "loss": 2.0895, + "step": 2050 + }, + { + "epoch": 0.43006919689662404, + "grad_norm": 5.032408912348559, + "learning_rate": 1.9334167162971672e-05, + "loss": 2.1415, + "step": 2051 + }, + { + "epoch": 0.4302788844621514, + "grad_norm": 5.537158925475578, + "learning_rate": 1.9333354655259536e-05, + "loss": 2.3665, + "step": 2052 + }, + { + "epoch": 0.43048857202767876, + "grad_norm": 5.88610943559457, + "learning_rate": 1.9332541669196788e-05, + "loss": 2.1844, + "step": 2053 + }, + { + "epoch": 0.4306982595932061, + "grad_norm": 6.30399584432925, + "learning_rate": 1.9331728204825092e-05, + "loss": 1.8749, + "step": 2054 + }, + { + "epoch": 0.4309079471587335, + "grad_norm": 5.597292511090274, + "learning_rate": 1.933091426218614e-05, + "loss": 2.0928, + "step": 2055 + }, + { + "epoch": 0.43111763472426085, + "grad_norm": 5.638930277821436, + "learning_rate": 1.933009984132165e-05, + "loss": 2.1757, + "step": 2056 + }, + { + "epoch": 0.4313273222897882, + "grad_norm": 5.275767560046034, + "learning_rate": 1.932928494227336e-05, + "loss": 2.0879, + "step": 2057 + }, + { + "epoch": 0.43153700985531557, + "grad_norm": 5.891912118126988, + "learning_rate": 1.9328469565083033e-05, + "loss": 2.2518, + "step": 2058 + }, + { + "epoch": 0.43174669742084293, + "grad_norm": 6.131723412163059, + "learning_rate": 1.9327653709792467e-05, + "loss": 2.0922, + "step": 2059 + }, + { + "epoch": 0.4319563849863703, + "grad_norm": 4.806275608474459, + "learning_rate": 1.9326837376443465e-05, + "loss": 2.0747, + "step": 2060 + }, + { + "epoch": 0.43216607255189765, + "grad_norm": 5.65893793256204, + "learning_rate": 1.9326020565077877e-05, + "loss": 2.1196, + "step": 2061 + }, + { + "epoch": 0.432375760117425, + "grad_norm": 7.573833858858292, + "learning_rate": 1.9325203275737557e-05, + "loss": 2.7746, + "step": 2062 + }, + { + "epoch": 0.43258544768295243, + "grad_norm": 6.734672915264905, + "learning_rate": 1.932438550846439e-05, + "loss": 2.359, + "step": 2063 + }, + { + "epoch": 0.4327951352484798, + "grad_norm": 5.601178993413717, + "learning_rate": 1.93235672633003e-05, + "loss": 1.9529, + "step": 2064 + }, + { + "epoch": 0.43300482281400715, + "grad_norm": 5.49168067409043, + "learning_rate": 1.9322748540287214e-05, + "loss": 2.0043, + "step": 2065 + }, + { + "epoch": 0.4332145103795345, + "grad_norm": 6.622302844621882, + "learning_rate": 1.9321929339467097e-05, + "loss": 2.218, + "step": 2066 + }, + { + "epoch": 0.4334241979450619, + "grad_norm": 5.99192606304044, + "learning_rate": 1.9321109660881934e-05, + "loss": 1.9851, + "step": 2067 + }, + { + "epoch": 0.43363388551058923, + "grad_norm": 6.003115940602221, + "learning_rate": 1.9320289504573734e-05, + "loss": 2.327, + "step": 2068 + }, + { + "epoch": 0.4338435730761166, + "grad_norm": 5.616618635469566, + "learning_rate": 1.9319468870584532e-05, + "loss": 2.4658, + "step": 2069 + }, + { + "epoch": 0.43405326064164396, + "grad_norm": 5.783200780107306, + "learning_rate": 1.9318647758956385e-05, + "loss": 2.0661, + "step": 2070 + }, + { + "epoch": 0.4342629482071713, + "grad_norm": 5.5309950475475524, + "learning_rate": 1.9317826169731377e-05, + "loss": 2.5602, + "step": 2071 + }, + { + "epoch": 0.4344726357726987, + "grad_norm": 5.818403769826662, + "learning_rate": 1.931700410295162e-05, + "loss": 2.0612, + "step": 2072 + }, + { + "epoch": 0.43468232333822604, + "grad_norm": 5.94637769575716, + "learning_rate": 1.9316181558659243e-05, + "loss": 2.4089, + "step": 2073 + }, + { + "epoch": 0.4348920109037534, + "grad_norm": 4.700776204946042, + "learning_rate": 1.9315358536896397e-05, + "loss": 2.0653, + "step": 2074 + }, + { + "epoch": 0.43510169846928076, + "grad_norm": 5.681035212156418, + "learning_rate": 1.9314535037705274e-05, + "loss": 2.1765, + "step": 2075 + }, + { + "epoch": 0.4353113860348081, + "grad_norm": 5.431748804059685, + "learning_rate": 1.9313711061128075e-05, + "loss": 2.1142, + "step": 2076 + }, + { + "epoch": 0.4355210736003355, + "grad_norm": 6.300833276473555, + "learning_rate": 1.931288660720703e-05, + "loss": 2.1151, + "step": 2077 + }, + { + "epoch": 0.43573076116586285, + "grad_norm": 5.635273946679019, + "learning_rate": 1.9312061675984396e-05, + "loss": 2.4139, + "step": 2078 + }, + { + "epoch": 0.4359404487313902, + "grad_norm": 6.235775951649308, + "learning_rate": 1.931123626750245e-05, + "loss": 2.3655, + "step": 2079 + }, + { + "epoch": 0.43615013629691757, + "grad_norm": 5.710130425109169, + "learning_rate": 1.9310410381803494e-05, + "loss": 2.0791, + "step": 2080 + }, + { + "epoch": 0.436359823862445, + "grad_norm": 5.158089292368677, + "learning_rate": 1.9309584018929857e-05, + "loss": 2.32, + "step": 2081 + }, + { + "epoch": 0.43656951142797235, + "grad_norm": 5.378579223958286, + "learning_rate": 1.9308757178923892e-05, + "loss": 1.9725, + "step": 2082 + }, + { + "epoch": 0.4367791989934997, + "grad_norm": 6.345658724832935, + "learning_rate": 1.9307929861827978e-05, + "loss": 2.3201, + "step": 2083 + }, + { + "epoch": 0.43698888655902707, + "grad_norm": 5.32513022247471, + "learning_rate": 1.9307102067684515e-05, + "loss": 2.0496, + "step": 2084 + }, + { + "epoch": 0.43719857412455443, + "grad_norm": 5.187660334016663, + "learning_rate": 1.9306273796535927e-05, + "loss": 1.8756, + "step": 2085 + }, + { + "epoch": 0.4374082616900818, + "grad_norm": 5.696723864006713, + "learning_rate": 1.930544504842467e-05, + "loss": 2.3171, + "step": 2086 + }, + { + "epoch": 0.43761794925560915, + "grad_norm": 5.7490410534754215, + "learning_rate": 1.9304615823393212e-05, + "loss": 2.2151, + "step": 2087 + }, + { + "epoch": 0.4378276368211365, + "grad_norm": 6.042000149835223, + "learning_rate": 1.9303786121484055e-05, + "loss": 2.1976, + "step": 2088 + }, + { + "epoch": 0.4380373243866639, + "grad_norm": 6.244626165358541, + "learning_rate": 1.930295594273972e-05, + "loss": 1.9764, + "step": 2089 + }, + { + "epoch": 0.43824701195219123, + "grad_norm": 6.660763283582525, + "learning_rate": 1.930212528720276e-05, + "loss": 2.4895, + "step": 2090 + }, + { + "epoch": 0.4384566995177186, + "grad_norm": 5.010957273057531, + "learning_rate": 1.9301294154915748e-05, + "loss": 2.0326, + "step": 2091 + }, + { + "epoch": 0.43866638708324596, + "grad_norm": 5.642691429878223, + "learning_rate": 1.9300462545921275e-05, + "loss": 2.0941, + "step": 2092 + }, + { + "epoch": 0.4388760746487733, + "grad_norm": 5.587524958435849, + "learning_rate": 1.9299630460261965e-05, + "loss": 1.7575, + "step": 2093 + }, + { + "epoch": 0.4390857622143007, + "grad_norm": 6.126055005929078, + "learning_rate": 1.929879789798047e-05, + "loss": 2.3099, + "step": 2094 + }, + { + "epoch": 0.43929544977982804, + "grad_norm": 6.3326167998416585, + "learning_rate": 1.9297964859119452e-05, + "loss": 2.3565, + "step": 2095 + }, + { + "epoch": 0.4395051373453554, + "grad_norm": 5.8339022278469805, + "learning_rate": 1.9297131343721606e-05, + "loss": 2.0313, + "step": 2096 + }, + { + "epoch": 0.43971482491088276, + "grad_norm": 5.912496507784254, + "learning_rate": 1.9296297351829655e-05, + "loss": 1.998, + "step": 2097 + }, + { + "epoch": 0.4399245124764101, + "grad_norm": 5.518182485277831, + "learning_rate": 1.9295462883486343e-05, + "loss": 1.719, + "step": 2098 + }, + { + "epoch": 0.44013420004193754, + "grad_norm": 5.225034649020508, + "learning_rate": 1.9294627938734432e-05, + "loss": 2.1939, + "step": 2099 + }, + { + "epoch": 0.4403438876074649, + "grad_norm": 6.3373451638869716, + "learning_rate": 1.929379251761672e-05, + "loss": 2.1479, + "step": 2100 + }, + { + "epoch": 0.44055357517299226, + "grad_norm": 6.676127481438523, + "learning_rate": 1.9292956620176026e-05, + "loss": 1.8878, + "step": 2101 + }, + { + "epoch": 0.4407632627385196, + "grad_norm": 5.469884784332731, + "learning_rate": 1.9292120246455182e-05, + "loss": 2.4373, + "step": 2102 + }, + { + "epoch": 0.440972950304047, + "grad_norm": 5.64658135255303, + "learning_rate": 1.9291283396497064e-05, + "loss": 1.9612, + "step": 2103 + }, + { + "epoch": 0.44118263786957435, + "grad_norm": 6.221941424803897, + "learning_rate": 1.929044607034455e-05, + "loss": 2.3523, + "step": 2104 + }, + { + "epoch": 0.4413923254351017, + "grad_norm": 6.549847020720109, + "learning_rate": 1.928960826804057e-05, + "loss": 2.0915, + "step": 2105 + }, + { + "epoch": 0.44160201300062907, + "grad_norm": 6.46659717443009, + "learning_rate": 1.928876998962805e-05, + "loss": 1.9885, + "step": 2106 + }, + { + "epoch": 0.44181170056615643, + "grad_norm": 7.692818894809973, + "learning_rate": 1.9287931235149953e-05, + "loss": 2.2936, + "step": 2107 + }, + { + "epoch": 0.4420213881316838, + "grad_norm": 6.194831982241896, + "learning_rate": 1.9287092004649278e-05, + "loss": 2.1111, + "step": 2108 + }, + { + "epoch": 0.44223107569721115, + "grad_norm": 5.381950824437453, + "learning_rate": 1.928625229816903e-05, + "loss": 2.1017, + "step": 2109 + }, + { + "epoch": 0.4424407632627385, + "grad_norm": 6.069627677798655, + "learning_rate": 1.928541211575224e-05, + "loss": 2.0528, + "step": 2110 + }, + { + "epoch": 0.4426504508282659, + "grad_norm": 6.196281401837651, + "learning_rate": 1.928457145744198e-05, + "loss": 2.068, + "step": 2111 + }, + { + "epoch": 0.44286013839379323, + "grad_norm": 6.47798398596338, + "learning_rate": 1.9283730323281324e-05, + "loss": 2.1255, + "step": 2112 + }, + { + "epoch": 0.4430698259593206, + "grad_norm": 5.206729071527936, + "learning_rate": 1.928288871331339e-05, + "loss": 2.1531, + "step": 2113 + }, + { + "epoch": 0.44327951352484796, + "grad_norm": 5.936718978146401, + "learning_rate": 1.9282046627581312e-05, + "loss": 2.5002, + "step": 2114 + }, + { + "epoch": 0.4434892010903753, + "grad_norm": 6.1021245039425445, + "learning_rate": 1.9281204066128244e-05, + "loss": 2.332, + "step": 2115 + }, + { + "epoch": 0.4436988886559027, + "grad_norm": 6.406287733699576, + "learning_rate": 1.9280361028997366e-05, + "loss": 2.2513, + "step": 2116 + }, + { + "epoch": 0.4439085762214301, + "grad_norm": 5.330564933676017, + "learning_rate": 1.9279517516231894e-05, + "loss": 2.2691, + "step": 2117 + }, + { + "epoch": 0.44411826378695746, + "grad_norm": 6.514738203209957, + "learning_rate": 1.927867352787505e-05, + "loss": 2.1958, + "step": 2118 + }, + { + "epoch": 0.4443279513524848, + "grad_norm": 6.08413407798268, + "learning_rate": 1.92778290639701e-05, + "loss": 2.0121, + "step": 2119 + }, + { + "epoch": 0.4445376389180122, + "grad_norm": 6.192734970302713, + "learning_rate": 1.9276984124560316e-05, + "loss": 2.4204, + "step": 2120 + }, + { + "epoch": 0.44474732648353954, + "grad_norm": 5.519097254092197, + "learning_rate": 1.927613870968901e-05, + "loss": 2.0228, + "step": 2121 + }, + { + "epoch": 0.4449570140490669, + "grad_norm": 5.802823216417202, + "learning_rate": 1.9275292819399497e-05, + "loss": 2.2928, + "step": 2122 + }, + { + "epoch": 0.44516670161459426, + "grad_norm": 5.758362580275623, + "learning_rate": 1.927444645373515e-05, + "loss": 2.2429, + "step": 2123 + }, + { + "epoch": 0.4453763891801216, + "grad_norm": 5.783980183378622, + "learning_rate": 1.927359961273933e-05, + "loss": 2.1413, + "step": 2124 + }, + { + "epoch": 0.445586076745649, + "grad_norm": 6.518825745108229, + "learning_rate": 1.9272752296455445e-05, + "loss": 2.299, + "step": 2125 + }, + { + "epoch": 0.44579576431117635, + "grad_norm": 5.761284680422236, + "learning_rate": 1.9271904504926926e-05, + "loss": 2.1173, + "step": 2126 + }, + { + "epoch": 0.4460054518767037, + "grad_norm": 5.353737782145257, + "learning_rate": 1.927105623819722e-05, + "loss": 2.0069, + "step": 2127 + }, + { + "epoch": 0.44621513944223107, + "grad_norm": 5.607547248719613, + "learning_rate": 1.9270207496309797e-05, + "loss": 1.6257, + "step": 2128 + }, + { + "epoch": 0.44642482700775843, + "grad_norm": 5.763310483879941, + "learning_rate": 1.9269358279308166e-05, + "loss": 2.3788, + "step": 2129 + }, + { + "epoch": 0.4466345145732858, + "grad_norm": 6.098194229322987, + "learning_rate": 1.9268508587235845e-05, + "loss": 2.3351, + "step": 2130 + }, + { + "epoch": 0.44684420213881315, + "grad_norm": 5.56995439916535, + "learning_rate": 1.9267658420136382e-05, + "loss": 1.9447, + "step": 2131 + }, + { + "epoch": 0.4470538897043405, + "grad_norm": 5.648570887849179, + "learning_rate": 1.9266807778053355e-05, + "loss": 2.0965, + "step": 2132 + }, + { + "epoch": 0.4472635772698679, + "grad_norm": 6.061262143542905, + "learning_rate": 1.9265956661030354e-05, + "loss": 2.3161, + "step": 2133 + }, + { + "epoch": 0.44747326483539523, + "grad_norm": 6.22562323241189, + "learning_rate": 1.9265105069111004e-05, + "loss": 1.9543, + "step": 2134 + }, + { + "epoch": 0.44768295240092265, + "grad_norm": 6.338761354478839, + "learning_rate": 1.926425300233895e-05, + "loss": 2.2737, + "step": 2135 + }, + { + "epoch": 0.44789263996645, + "grad_norm": 5.425742059783069, + "learning_rate": 1.9263400460757862e-05, + "loss": 2.0995, + "step": 2136 + }, + { + "epoch": 0.4481023275319774, + "grad_norm": 5.580454103356368, + "learning_rate": 1.9262547444411434e-05, + "loss": 2.0169, + "step": 2137 + }, + { + "epoch": 0.44831201509750473, + "grad_norm": 5.699511976228824, + "learning_rate": 1.926169395334338e-05, + "loss": 2.4202, + "step": 2138 + }, + { + "epoch": 0.4485217026630321, + "grad_norm": 5.252596678127221, + "learning_rate": 1.9260839987597452e-05, + "loss": 2.0182, + "step": 2139 + }, + { + "epoch": 0.44873139022855946, + "grad_norm": 5.859247990503234, + "learning_rate": 1.925998554721741e-05, + "loss": 2.0346, + "step": 2140 + }, + { + "epoch": 0.4489410777940868, + "grad_norm": 6.153703690951179, + "learning_rate": 1.9259130632247047e-05, + "loss": 2.4379, + "step": 2141 + }, + { + "epoch": 0.4491507653596142, + "grad_norm": 5.954251759696108, + "learning_rate": 1.9258275242730186e-05, + "loss": 2.1387, + "step": 2142 + }, + { + "epoch": 0.44936045292514154, + "grad_norm": 6.412397700313111, + "learning_rate": 1.9257419378710655e-05, + "loss": 2.3891, + "step": 2143 + }, + { + "epoch": 0.4495701404906689, + "grad_norm": 5.275799310870207, + "learning_rate": 1.9256563040232327e-05, + "loss": 2.0199, + "step": 2144 + }, + { + "epoch": 0.44977982805619626, + "grad_norm": 5.788502863681894, + "learning_rate": 1.925570622733909e-05, + "loss": 2.1101, + "step": 2145 + }, + { + "epoch": 0.4499895156217236, + "grad_norm": 6.281865542887358, + "learning_rate": 1.925484894007485e-05, + "loss": 1.9435, + "step": 2146 + }, + { + "epoch": 0.450199203187251, + "grad_norm": 5.029955211994221, + "learning_rate": 1.9253991178483558e-05, + "loss": 1.6969, + "step": 2147 + }, + { + "epoch": 0.45040889075277835, + "grad_norm": 6.226058554567945, + "learning_rate": 1.925313294260916e-05, + "loss": 2.323, + "step": 2148 + }, + { + "epoch": 0.4506185783183057, + "grad_norm": 6.173703084842196, + "learning_rate": 1.9252274232495657e-05, + "loss": 2.1683, + "step": 2149 + }, + { + "epoch": 0.45082826588383307, + "grad_norm": 5.191343436819127, + "learning_rate": 1.9251415048187048e-05, + "loss": 1.894, + "step": 2150 + }, + { + "epoch": 0.45103795344936043, + "grad_norm": 5.883486803450713, + "learning_rate": 1.9250555389727372e-05, + "loss": 2.1668, + "step": 2151 + }, + { + "epoch": 0.4512476410148878, + "grad_norm": 7.076499150662879, + "learning_rate": 1.924969525716069e-05, + "loss": 2.3622, + "step": 2152 + }, + { + "epoch": 0.4514573285804152, + "grad_norm": 6.0310037323600865, + "learning_rate": 1.9248834650531082e-05, + "loss": 2.0964, + "step": 2153 + }, + { + "epoch": 0.45166701614594257, + "grad_norm": 7.175554750332626, + "learning_rate": 1.9247973569882658e-05, + "loss": 2.0424, + "step": 2154 + }, + { + "epoch": 0.45187670371146993, + "grad_norm": 6.105869658042161, + "learning_rate": 1.924711201525955e-05, + "loss": 1.7389, + "step": 2155 + }, + { + "epoch": 0.4520863912769973, + "grad_norm": 5.691638384452805, + "learning_rate": 1.9246249986705908e-05, + "loss": 1.8976, + "step": 2156 + }, + { + "epoch": 0.45229607884252465, + "grad_norm": 5.42756617344828, + "learning_rate": 1.9245387484265922e-05, + "loss": 1.9626, + "step": 2157 + }, + { + "epoch": 0.452505766408052, + "grad_norm": 6.038008616429139, + "learning_rate": 1.924452450798379e-05, + "loss": 2.1422, + "step": 2158 + }, + { + "epoch": 0.4527154539735794, + "grad_norm": 5.704012336271476, + "learning_rate": 1.9243661057903745e-05, + "loss": 1.8792, + "step": 2159 + }, + { + "epoch": 0.45292514153910673, + "grad_norm": 5.8213075043338645, + "learning_rate": 1.924279713407004e-05, + "loss": 2.1707, + "step": 2160 + }, + { + "epoch": 0.4531348291046341, + "grad_norm": 5.93516429303292, + "learning_rate": 1.9241932736526946e-05, + "loss": 2.0073, + "step": 2161 + }, + { + "epoch": 0.45334451667016146, + "grad_norm": 7.380665068681982, + "learning_rate": 1.9241067865318773e-05, + "loss": 2.4388, + "step": 2162 + }, + { + "epoch": 0.4535542042356888, + "grad_norm": 6.302986263222068, + "learning_rate": 1.924020252048984e-05, + "loss": 2.0992, + "step": 2163 + }, + { + "epoch": 0.4537638918012162, + "grad_norm": 5.156662616331745, + "learning_rate": 1.9239336702084504e-05, + "loss": 1.9568, + "step": 2164 + }, + { + "epoch": 0.45397357936674354, + "grad_norm": 6.538949072160556, + "learning_rate": 1.923847041014714e-05, + "loss": 2.328, + "step": 2165 + }, + { + "epoch": 0.4541832669322709, + "grad_norm": 5.785433967010931, + "learning_rate": 1.923760364472214e-05, + "loss": 2.5033, + "step": 2166 + }, + { + "epoch": 0.45439295449779826, + "grad_norm": 5.9903884281129205, + "learning_rate": 1.9236736405853936e-05, + "loss": 1.9587, + "step": 2167 + }, + { + "epoch": 0.4546026420633256, + "grad_norm": 5.515737482442287, + "learning_rate": 1.9235868693586968e-05, + "loss": 2.0225, + "step": 2168 + }, + { + "epoch": 0.454812329628853, + "grad_norm": 6.707401275980805, + "learning_rate": 1.923500050796571e-05, + "loss": 2.3053, + "step": 2169 + }, + { + "epoch": 0.45502201719438035, + "grad_norm": 6.207836087239906, + "learning_rate": 1.923413184903466e-05, + "loss": 2.2179, + "step": 2170 + }, + { + "epoch": 0.45523170475990776, + "grad_norm": 5.888412981526102, + "learning_rate": 1.9233262716838336e-05, + "loss": 2.3701, + "step": 2171 + }, + { + "epoch": 0.4554413923254351, + "grad_norm": 7.4226005879639745, + "learning_rate": 1.9232393111421285e-05, + "loss": 2.2473, + "step": 2172 + }, + { + "epoch": 0.4556510798909625, + "grad_norm": 5.018636253827584, + "learning_rate": 1.9231523032828073e-05, + "loss": 2.167, + "step": 2173 + }, + { + "epoch": 0.45586076745648985, + "grad_norm": 6.685540707178862, + "learning_rate": 1.9230652481103297e-05, + "loss": 2.3116, + "step": 2174 + }, + { + "epoch": 0.4560704550220172, + "grad_norm": 5.402939053047352, + "learning_rate": 1.922978145629157e-05, + "loss": 2.2652, + "step": 2175 + }, + { + "epoch": 0.45628014258754457, + "grad_norm": 5.75831079413424, + "learning_rate": 1.9228909958437532e-05, + "loss": 1.5863, + "step": 2176 + }, + { + "epoch": 0.45648983015307193, + "grad_norm": 5.308506535451557, + "learning_rate": 1.9228037987585855e-05, + "loss": 2.1101, + "step": 2177 + }, + { + "epoch": 0.4566995177185993, + "grad_norm": 5.958508989244659, + "learning_rate": 1.9227165543781226e-05, + "loss": 2.0866, + "step": 2178 + }, + { + "epoch": 0.45690920528412665, + "grad_norm": 5.930466722596854, + "learning_rate": 1.9226292627068358e-05, + "loss": 2.1167, + "step": 2179 + }, + { + "epoch": 0.457118892849654, + "grad_norm": 6.41026921843967, + "learning_rate": 1.9225419237491988e-05, + "loss": 2.6108, + "step": 2180 + }, + { + "epoch": 0.4573285804151814, + "grad_norm": 5.9257686378480585, + "learning_rate": 1.9224545375096887e-05, + "loss": 2.1733, + "step": 2181 + }, + { + "epoch": 0.45753826798070873, + "grad_norm": 5.801829726257146, + "learning_rate": 1.9223671039927833e-05, + "loss": 2.2511, + "step": 2182 + }, + { + "epoch": 0.4577479555462361, + "grad_norm": 5.549752188009174, + "learning_rate": 1.922279623202964e-05, + "loss": 2.1756, + "step": 2183 + }, + { + "epoch": 0.45795764311176346, + "grad_norm": 5.501722891793999, + "learning_rate": 1.9221920951447147e-05, + "loss": 2.0517, + "step": 2184 + }, + { + "epoch": 0.4581673306772908, + "grad_norm": 6.710408918793639, + "learning_rate": 1.922104519822521e-05, + "loss": 2.2381, + "step": 2185 + }, + { + "epoch": 0.4583770182428182, + "grad_norm": 5.72671393575448, + "learning_rate": 1.9220168972408713e-05, + "loss": 2.2189, + "step": 2186 + }, + { + "epoch": 0.45858670580834554, + "grad_norm": 6.36804343786305, + "learning_rate": 1.9219292274042564e-05, + "loss": 2.1695, + "step": 2187 + }, + { + "epoch": 0.4587963933738729, + "grad_norm": 6.132848501859067, + "learning_rate": 1.9218415103171696e-05, + "loss": 2.2488, + "step": 2188 + }, + { + "epoch": 0.4590060809394003, + "grad_norm": 6.421598509999375, + "learning_rate": 1.9217537459841066e-05, + "loss": 2.055, + "step": 2189 + }, + { + "epoch": 0.4592157685049277, + "grad_norm": 5.594864388131729, + "learning_rate": 1.9216659344095652e-05, + "loss": 2.0946, + "step": 2190 + }, + { + "epoch": 0.45942545607045504, + "grad_norm": 6.904996811110198, + "learning_rate": 1.9215780755980465e-05, + "loss": 2.4237, + "step": 2191 + }, + { + "epoch": 0.4596351436359824, + "grad_norm": 6.545390301866025, + "learning_rate": 1.9214901695540526e-05, + "loss": 2.3833, + "step": 2192 + }, + { + "epoch": 0.45984483120150976, + "grad_norm": 6.285725749432022, + "learning_rate": 1.9214022162820894e-05, + "loss": 2.0051, + "step": 2193 + }, + { + "epoch": 0.4600545187670371, + "grad_norm": 5.914363234399766, + "learning_rate": 1.9213142157866645e-05, + "loss": 2.0454, + "step": 2194 + }, + { + "epoch": 0.4602642063325645, + "grad_norm": 6.004414256617915, + "learning_rate": 1.9212261680722882e-05, + "loss": 2.2039, + "step": 2195 + }, + { + "epoch": 0.46047389389809185, + "grad_norm": 5.345417968860954, + "learning_rate": 1.921138073143473e-05, + "loss": 2.2903, + "step": 2196 + }, + { + "epoch": 0.4606835814636192, + "grad_norm": 6.728334912224999, + "learning_rate": 1.921049931004734e-05, + "loss": 2.2343, + "step": 2197 + }, + { + "epoch": 0.46089326902914657, + "grad_norm": 5.979838770395121, + "learning_rate": 1.9209617416605887e-05, + "loss": 2.2212, + "step": 2198 + }, + { + "epoch": 0.46110295659467393, + "grad_norm": 5.2713312350823625, + "learning_rate": 1.9208735051155566e-05, + "loss": 2.2407, + "step": 2199 + }, + { + "epoch": 0.4613126441602013, + "grad_norm": 5.6957515091447, + "learning_rate": 1.92078522137416e-05, + "loss": 2.1771, + "step": 2200 + }, + { + "epoch": 0.46152233172572865, + "grad_norm": 6.2806806757058595, + "learning_rate": 1.920696890440924e-05, + "loss": 2.1071, + "step": 2201 + }, + { + "epoch": 0.461732019291256, + "grad_norm": 6.306937379373556, + "learning_rate": 1.9206085123203752e-05, + "loss": 2.3453, + "step": 2202 + }, + { + "epoch": 0.4619417068567834, + "grad_norm": 6.864168241152256, + "learning_rate": 1.920520087017044e-05, + "loss": 2.3866, + "step": 2203 + }, + { + "epoch": 0.46215139442231074, + "grad_norm": 4.911490014827521, + "learning_rate": 1.9204316145354613e-05, + "loss": 2.0165, + "step": 2204 + }, + { + "epoch": 0.4623610819878381, + "grad_norm": 5.850954038316423, + "learning_rate": 1.9203430948801623e-05, + "loss": 2.0233, + "step": 2205 + }, + { + "epoch": 0.4625707695533655, + "grad_norm": 4.887587741902803, + "learning_rate": 1.9202545280556832e-05, + "loss": 1.9609, + "step": 2206 + }, + { + "epoch": 0.4627804571188929, + "grad_norm": 6.35976619595104, + "learning_rate": 1.9201659140665634e-05, + "loss": 2.5015, + "step": 2207 + }, + { + "epoch": 0.46299014468442024, + "grad_norm": 5.513955918886161, + "learning_rate": 1.9200772529173447e-05, + "loss": 2.0362, + "step": 2208 + }, + { + "epoch": 0.4631998322499476, + "grad_norm": 5.359570745551045, + "learning_rate": 1.919988544612571e-05, + "loss": 2.2955, + "step": 2209 + }, + { + "epoch": 0.46340951981547496, + "grad_norm": 5.435620889286572, + "learning_rate": 1.9198997891567885e-05, + "loss": 2.1157, + "step": 2210 + }, + { + "epoch": 0.4636192073810023, + "grad_norm": 5.132557438764978, + "learning_rate": 1.9198109865545465e-05, + "loss": 2.3518, + "step": 2211 + }, + { + "epoch": 0.4638288949465297, + "grad_norm": 6.216866769826188, + "learning_rate": 1.919722136810396e-05, + "loss": 2.094, + "step": 2212 + }, + { + "epoch": 0.46403858251205704, + "grad_norm": 5.715510949972599, + "learning_rate": 1.919633239928891e-05, + "loss": 2.1414, + "step": 2213 + }, + { + "epoch": 0.4642482700775844, + "grad_norm": 6.05166003185328, + "learning_rate": 1.9195442959145873e-05, + "loss": 2.2933, + "step": 2214 + }, + { + "epoch": 0.46445795764311176, + "grad_norm": 5.795164276777185, + "learning_rate": 1.919455304772044e-05, + "loss": 2.2086, + "step": 2215 + }, + { + "epoch": 0.4646676452086391, + "grad_norm": 5.5968244475981574, + "learning_rate": 1.919366266505821e-05, + "loss": 2.35, + "step": 2216 + }, + { + "epoch": 0.4648773327741665, + "grad_norm": 5.231984952353256, + "learning_rate": 1.9192771811204823e-05, + "loss": 2.0885, + "step": 2217 + }, + { + "epoch": 0.46508702033969385, + "grad_norm": 5.507646860851503, + "learning_rate": 1.919188048620594e-05, + "loss": 1.9552, + "step": 2218 + }, + { + "epoch": 0.4652967079052212, + "grad_norm": 5.342541964172121, + "learning_rate": 1.919098869010724e-05, + "loss": 2.2781, + "step": 2219 + }, + { + "epoch": 0.46550639547074857, + "grad_norm": 5.129520596276407, + "learning_rate": 1.9190096422954427e-05, + "loss": 2.0327, + "step": 2220 + }, + { + "epoch": 0.46571608303627593, + "grad_norm": 6.499360271125452, + "learning_rate": 1.9189203684793233e-05, + "loss": 2.0239, + "step": 2221 + }, + { + "epoch": 0.4659257706018033, + "grad_norm": 5.407031658580738, + "learning_rate": 1.918831047566941e-05, + "loss": 2.126, + "step": 2222 + }, + { + "epoch": 0.46613545816733065, + "grad_norm": 5.8079699961900655, + "learning_rate": 1.9187416795628743e-05, + "loss": 2.1462, + "step": 2223 + }, + { + "epoch": 0.46634514573285807, + "grad_norm": 6.990552814762791, + "learning_rate": 1.9186522644717027e-05, + "loss": 2.24, + "step": 2224 + }, + { + "epoch": 0.46655483329838543, + "grad_norm": 5.912756516457348, + "learning_rate": 1.9185628022980093e-05, + "loss": 1.9577, + "step": 2225 + }, + { + "epoch": 0.4667645208639128, + "grad_norm": 5.535657852727496, + "learning_rate": 1.9184732930463796e-05, + "loss": 2.3345, + "step": 2226 + }, + { + "epoch": 0.46697420842944015, + "grad_norm": 6.433725293753883, + "learning_rate": 1.9183837367214e-05, + "loss": 2.2902, + "step": 2227 + }, + { + "epoch": 0.4671838959949675, + "grad_norm": 7.117438170345291, + "learning_rate": 1.9182941333276614e-05, + "loss": 2.0742, + "step": 2228 + }, + { + "epoch": 0.4673935835604949, + "grad_norm": 5.761837137751725, + "learning_rate": 1.918204482869755e-05, + "loss": 1.9299, + "step": 2229 + }, + { + "epoch": 0.46760327112602224, + "grad_norm": 6.246421759726161, + "learning_rate": 1.918114785352277e-05, + "loss": 2.2229, + "step": 2230 + }, + { + "epoch": 0.4678129586915496, + "grad_norm": 5.404958116350511, + "learning_rate": 1.9180250407798242e-05, + "loss": 2.0502, + "step": 2231 + }, + { + "epoch": 0.46802264625707696, + "grad_norm": 5.85599494447238, + "learning_rate": 1.9179352491569953e-05, + "loss": 2.3847, + "step": 2232 + }, + { + "epoch": 0.4682323338226043, + "grad_norm": 5.191862108195184, + "learning_rate": 1.917845410488393e-05, + "loss": 2.3421, + "step": 2233 + }, + { + "epoch": 0.4684420213881317, + "grad_norm": 5.263309440272586, + "learning_rate": 1.9177555247786212e-05, + "loss": 2.0166, + "step": 2234 + }, + { + "epoch": 0.46865170895365904, + "grad_norm": 5.866289917141282, + "learning_rate": 1.9176655920322872e-05, + "loss": 2.1082, + "step": 2235 + }, + { + "epoch": 0.4688613965191864, + "grad_norm": 5.873472828637023, + "learning_rate": 1.9175756122540002e-05, + "loss": 2.2015, + "step": 2236 + }, + { + "epoch": 0.46907108408471376, + "grad_norm": 5.883851351902903, + "learning_rate": 1.9174855854483715e-05, + "loss": 2.0549, + "step": 2237 + }, + { + "epoch": 0.4692807716502411, + "grad_norm": 5.392928383936007, + "learning_rate": 1.917395511620015e-05, + "loss": 2.1924, + "step": 2238 + }, + { + "epoch": 0.4694904592157685, + "grad_norm": 5.5921929222847, + "learning_rate": 1.9173053907735475e-05, + "loss": 1.8906, + "step": 2239 + }, + { + "epoch": 0.46970014678129585, + "grad_norm": 5.9949091249008335, + "learning_rate": 1.9172152229135878e-05, + "loss": 2.3645, + "step": 2240 + }, + { + "epoch": 0.4699098343468232, + "grad_norm": 5.907666521846647, + "learning_rate": 1.917125008044757e-05, + "loss": 1.8574, + "step": 2241 + }, + { + "epoch": 0.4701195219123506, + "grad_norm": 5.321537405818213, + "learning_rate": 1.917034746171679e-05, + "loss": 2.156, + "step": 2242 + }, + { + "epoch": 0.470329209477878, + "grad_norm": 5.113487971303939, + "learning_rate": 1.91694443729898e-05, + "loss": 2.1939, + "step": 2243 + }, + { + "epoch": 0.47053889704340535, + "grad_norm": 6.015557514778198, + "learning_rate": 1.9168540814312875e-05, + "loss": 2.1233, + "step": 2244 + }, + { + "epoch": 0.4707485846089327, + "grad_norm": 7.187292944994101, + "learning_rate": 1.9167636785732337e-05, + "loss": 2.1076, + "step": 2245 + }, + { + "epoch": 0.47095827217446007, + "grad_norm": 5.731787413492678, + "learning_rate": 1.9166732287294512e-05, + "loss": 2.2406, + "step": 2246 + }, + { + "epoch": 0.47116795973998743, + "grad_norm": 5.844299153066626, + "learning_rate": 1.9165827319045757e-05, + "loss": 1.8746, + "step": 2247 + }, + { + "epoch": 0.4713776473055148, + "grad_norm": 6.6831618795984475, + "learning_rate": 1.9164921881032454e-05, + "loss": 2.4474, + "step": 2248 + }, + { + "epoch": 0.47158733487104215, + "grad_norm": 5.211801386725437, + "learning_rate": 1.9164015973301012e-05, + "loss": 1.8424, + "step": 2249 + }, + { + "epoch": 0.4717970224365695, + "grad_norm": 6.9212471379060085, + "learning_rate": 1.9163109595897854e-05, + "loss": 2.2099, + "step": 2250 + }, + { + "epoch": 0.4720067100020969, + "grad_norm": 6.116129951239323, + "learning_rate": 1.916220274886944e-05, + "loss": 2.1847, + "step": 2251 + }, + { + "epoch": 0.47221639756762424, + "grad_norm": 5.490557160858078, + "learning_rate": 1.9161295432262243e-05, + "loss": 2.3115, + "step": 2252 + }, + { + "epoch": 0.4724260851331516, + "grad_norm": 6.0974434254007255, + "learning_rate": 1.9160387646122765e-05, + "loss": 2.304, + "step": 2253 + }, + { + "epoch": 0.47263577269867896, + "grad_norm": 6.225053555931331, + "learning_rate": 1.9159479390497535e-05, + "loss": 2.1636, + "step": 2254 + }, + { + "epoch": 0.4728454602642063, + "grad_norm": 5.903043682410186, + "learning_rate": 1.9158570665433093e-05, + "loss": 2.2441, + "step": 2255 + }, + { + "epoch": 0.4730551478297337, + "grad_norm": 5.603766050656394, + "learning_rate": 1.9157661470976022e-05, + "loss": 1.9925, + "step": 2256 + }, + { + "epoch": 0.47326483539526104, + "grad_norm": 5.458032051711044, + "learning_rate": 1.915675180717292e-05, + "loss": 2.0725, + "step": 2257 + }, + { + "epoch": 0.4734745229607884, + "grad_norm": 5.405296290141715, + "learning_rate": 1.9155841674070408e-05, + "loss": 1.8653, + "step": 2258 + }, + { + "epoch": 0.47368421052631576, + "grad_norm": 5.724640047374485, + "learning_rate": 1.915493107171513e-05, + "loss": 2.0941, + "step": 2259 + }, + { + "epoch": 0.4738938980918432, + "grad_norm": 6.01612171779029, + "learning_rate": 1.915402000015375e-05, + "loss": 2.2116, + "step": 2260 + }, + { + "epoch": 0.47410358565737054, + "grad_norm": 6.375110819849119, + "learning_rate": 1.915310845943297e-05, + "loss": 1.9799, + "step": 2261 + }, + { + "epoch": 0.4743132732228979, + "grad_norm": 5.713590711588439, + "learning_rate": 1.915219644959951e-05, + "loss": 2.2308, + "step": 2262 + }, + { + "epoch": 0.47452296078842526, + "grad_norm": 5.208890653753317, + "learning_rate": 1.9151283970700107e-05, + "loss": 2.2139, + "step": 2263 + }, + { + "epoch": 0.4747326483539526, + "grad_norm": 6.154497579160804, + "learning_rate": 1.915037102278153e-05, + "loss": 2.0922, + "step": 2264 + }, + { + "epoch": 0.47494233591948, + "grad_norm": 5.844617358520422, + "learning_rate": 1.914945760589056e-05, + "loss": 2.3324, + "step": 2265 + }, + { + "epoch": 0.47515202348500735, + "grad_norm": 5.6421170430866905, + "learning_rate": 1.9148543720074027e-05, + "loss": 2.1882, + "step": 2266 + }, + { + "epoch": 0.4753617110505347, + "grad_norm": 5.589636396631507, + "learning_rate": 1.914762936537876e-05, + "loss": 2.2831, + "step": 2267 + }, + { + "epoch": 0.47557139861606207, + "grad_norm": 5.7891177306391315, + "learning_rate": 1.9146714541851617e-05, + "loss": 2.1765, + "step": 2268 + }, + { + "epoch": 0.47578108618158943, + "grad_norm": 7.833356649071576, + "learning_rate": 1.9145799249539497e-05, + "loss": 2.6146, + "step": 2269 + }, + { + "epoch": 0.4759907737471168, + "grad_norm": 6.164159591348915, + "learning_rate": 1.91448834884893e-05, + "loss": 2.3648, + "step": 2270 + }, + { + "epoch": 0.47620046131264415, + "grad_norm": 5.507057721589435, + "learning_rate": 1.914396725874796e-05, + "loss": 1.8405, + "step": 2271 + }, + { + "epoch": 0.4764101488781715, + "grad_norm": 5.483572197943918, + "learning_rate": 1.914305056036244e-05, + "loss": 2.125, + "step": 2272 + }, + { + "epoch": 0.4766198364436989, + "grad_norm": 6.035955026227375, + "learning_rate": 1.9142133393379724e-05, + "loss": 2.5276, + "step": 2273 + }, + { + "epoch": 0.47682952400922624, + "grad_norm": 5.229814449143713, + "learning_rate": 1.9141215757846812e-05, + "loss": 2.1619, + "step": 2274 + }, + { + "epoch": 0.4770392115747536, + "grad_norm": 5.662789258735806, + "learning_rate": 1.914029765381074e-05, + "loss": 2.1363, + "step": 2275 + }, + { + "epoch": 0.47724889914028096, + "grad_norm": 5.239699392383437, + "learning_rate": 1.913937908131856e-05, + "loss": 2.148, + "step": 2276 + }, + { + "epoch": 0.4774585867058083, + "grad_norm": 5.488282216570894, + "learning_rate": 1.9138460040417346e-05, + "loss": 2.4493, + "step": 2277 + }, + { + "epoch": 0.47766827427133574, + "grad_norm": 6.315343853868752, + "learning_rate": 1.913754053115421e-05, + "loss": 2.3307, + "step": 2278 + }, + { + "epoch": 0.4778779618368631, + "grad_norm": 5.6080925670579465, + "learning_rate": 1.9136620553576272e-05, + "loss": 2.107, + "step": 2279 + }, + { + "epoch": 0.47808764940239046, + "grad_norm": 5.831982304680028, + "learning_rate": 1.9135700107730682e-05, + "loss": 2.2452, + "step": 2280 + }, + { + "epoch": 0.4782973369679178, + "grad_norm": 5.787905327077409, + "learning_rate": 1.913477919366462e-05, + "loss": 2.2639, + "step": 2281 + }, + { + "epoch": 0.4785070245334452, + "grad_norm": 5.442313003708371, + "learning_rate": 1.913385781142528e-05, + "loss": 2.141, + "step": 2282 + }, + { + "epoch": 0.47871671209897254, + "grad_norm": 5.547397250284771, + "learning_rate": 1.9132935961059886e-05, + "loss": 2.0073, + "step": 2283 + }, + { + "epoch": 0.4789263996644999, + "grad_norm": 6.108542161008461, + "learning_rate": 1.9132013642615682e-05, + "loss": 2.09, + "step": 2284 + }, + { + "epoch": 0.47913608723002726, + "grad_norm": 5.398926224578731, + "learning_rate": 1.9131090856139944e-05, + "loss": 2.201, + "step": 2285 + }, + { + "epoch": 0.4793457747955546, + "grad_norm": 5.52596060057175, + "learning_rate": 1.9130167601679956e-05, + "loss": 2.221, + "step": 2286 + }, + { + "epoch": 0.479555462361082, + "grad_norm": 6.949766376300676, + "learning_rate": 1.9129243879283046e-05, + "loss": 2.3306, + "step": 2287 + }, + { + "epoch": 0.47976514992660935, + "grad_norm": 5.886875888686392, + "learning_rate": 1.9128319688996555e-05, + "loss": 2.0702, + "step": 2288 + }, + { + "epoch": 0.4799748374921367, + "grad_norm": 6.6835449840955965, + "learning_rate": 1.9127395030867845e-05, + "loss": 2.3101, + "step": 2289 + }, + { + "epoch": 0.48018452505766407, + "grad_norm": 5.8480708469635685, + "learning_rate": 1.912646990494431e-05, + "loss": 2.0345, + "step": 2290 + }, + { + "epoch": 0.48039421262319143, + "grad_norm": 5.840093448971638, + "learning_rate": 1.912554431127336e-05, + "loss": 2.0652, + "step": 2291 + }, + { + "epoch": 0.4806039001887188, + "grad_norm": 5.739225106424425, + "learning_rate": 1.9124618249902444e-05, + "loss": 2.1156, + "step": 2292 + }, + { + "epoch": 0.48081358775424615, + "grad_norm": 6.867760260883154, + "learning_rate": 1.912369172087901e-05, + "loss": 2.2225, + "step": 2293 + }, + { + "epoch": 0.4810232753197735, + "grad_norm": 6.028072151540249, + "learning_rate": 1.9122764724250555e-05, + "loss": 1.9526, + "step": 2294 + }, + { + "epoch": 0.4812329628853009, + "grad_norm": 5.9129118702387045, + "learning_rate": 1.9121837260064585e-05, + "loss": 2.1773, + "step": 2295 + }, + { + "epoch": 0.4814426504508283, + "grad_norm": 6.4484740762525945, + "learning_rate": 1.9120909328368636e-05, + "loss": 1.8367, + "step": 2296 + }, + { + "epoch": 0.48165233801635565, + "grad_norm": 6.639897003379941, + "learning_rate": 1.911998092921026e-05, + "loss": 2.1958, + "step": 2297 + }, + { + "epoch": 0.481862025581883, + "grad_norm": 6.37943282307499, + "learning_rate": 1.9119052062637046e-05, + "loss": 2.1032, + "step": 2298 + }, + { + "epoch": 0.4820717131474104, + "grad_norm": 6.378506483538332, + "learning_rate": 1.9118122728696596e-05, + "loss": 2.3076, + "step": 2299 + }, + { + "epoch": 0.48228140071293774, + "grad_norm": 7.335221080156312, + "learning_rate": 1.9117192927436546e-05, + "loss": 2.2956, + "step": 2300 + }, + { + "epoch": 0.4824910882784651, + "grad_norm": 5.463400886485292, + "learning_rate": 1.9116262658904542e-05, + "loss": 2.0164, + "step": 2301 + }, + { + "epoch": 0.48270077584399246, + "grad_norm": 5.043648173543413, + "learning_rate": 1.9115331923148264e-05, + "loss": 1.9418, + "step": 2302 + }, + { + "epoch": 0.4829104634095198, + "grad_norm": 6.11473623913284, + "learning_rate": 1.9114400720215416e-05, + "loss": 1.9098, + "step": 2303 + }, + { + "epoch": 0.4831201509750472, + "grad_norm": 6.045354085085585, + "learning_rate": 1.9113469050153727e-05, + "loss": 1.9945, + "step": 2304 + }, + { + "epoch": 0.48332983854057454, + "grad_norm": 5.897279169964477, + "learning_rate": 1.9112536913010938e-05, + "loss": 2.1018, + "step": 2305 + }, + { + "epoch": 0.4835395261061019, + "grad_norm": 7.209973801693477, + "learning_rate": 1.911160430883483e-05, + "loss": 2.4253, + "step": 2306 + }, + { + "epoch": 0.48374921367162926, + "grad_norm": 6.068208849899813, + "learning_rate": 1.9110671237673193e-05, + "loss": 2.1066, + "step": 2307 + }, + { + "epoch": 0.4839589012371566, + "grad_norm": 7.019115825772522, + "learning_rate": 1.9109737699573858e-05, + "loss": 1.9369, + "step": 2308 + }, + { + "epoch": 0.484168588802684, + "grad_norm": 5.980490709073885, + "learning_rate": 1.9108803694584666e-05, + "loss": 2.1301, + "step": 2309 + }, + { + "epoch": 0.48437827636821135, + "grad_norm": 7.093500850835242, + "learning_rate": 1.9107869222753486e-05, + "loss": 2.3909, + "step": 2310 + }, + { + "epoch": 0.4845879639337387, + "grad_norm": 5.809965110153048, + "learning_rate": 1.910693428412821e-05, + "loss": 1.8726, + "step": 2311 + }, + { + "epoch": 0.48479765149926607, + "grad_norm": 5.661632544245621, + "learning_rate": 1.9105998878756758e-05, + "loss": 2.18, + "step": 2312 + }, + { + "epoch": 0.48500733906479343, + "grad_norm": 7.238416599463696, + "learning_rate": 1.910506300668707e-05, + "loss": 2.4346, + "step": 2313 + }, + { + "epoch": 0.48521702663032085, + "grad_norm": 5.474727495118865, + "learning_rate": 1.910412666796711e-05, + "loss": 2.3721, + "step": 2314 + }, + { + "epoch": 0.4854267141958482, + "grad_norm": 6.295115074822989, + "learning_rate": 1.910318986264487e-05, + "loss": 1.8663, + "step": 2315 + }, + { + "epoch": 0.48563640176137557, + "grad_norm": 5.689241407981616, + "learning_rate": 1.9102252590768357e-05, + "loss": 2.1643, + "step": 2316 + }, + { + "epoch": 0.48584608932690293, + "grad_norm": 5.270619351102859, + "learning_rate": 1.9101314852385614e-05, + "loss": 2.3577, + "step": 2317 + }, + { + "epoch": 0.4860557768924303, + "grad_norm": 6.2924365733813445, + "learning_rate": 1.91003766475447e-05, + "loss": 2.1038, + "step": 2318 + }, + { + "epoch": 0.48626546445795765, + "grad_norm": 5.870149865231637, + "learning_rate": 1.90994379762937e-05, + "loss": 2.2398, + "step": 2319 + }, + { + "epoch": 0.486475152023485, + "grad_norm": 5.217401366307402, + "learning_rate": 1.9098498838680716e-05, + "loss": 2.2572, + "step": 2320 + }, + { + "epoch": 0.4866848395890124, + "grad_norm": 4.934154555265756, + "learning_rate": 1.9097559234753893e-05, + "loss": 1.9839, + "step": 2321 + }, + { + "epoch": 0.48689452715453974, + "grad_norm": 6.306084827825201, + "learning_rate": 1.9096619164561376e-05, + "loss": 2.3003, + "step": 2322 + }, + { + "epoch": 0.4871042147200671, + "grad_norm": 5.420472033131771, + "learning_rate": 1.909567862815135e-05, + "loss": 2.1472, + "step": 2323 + }, + { + "epoch": 0.48731390228559446, + "grad_norm": 5.766418672886656, + "learning_rate": 1.9094737625572022e-05, + "loss": 1.9894, + "step": 2324 + }, + { + "epoch": 0.4875235898511218, + "grad_norm": 5.4623096039110495, + "learning_rate": 1.9093796156871612e-05, + "loss": 1.9375, + "step": 2325 + }, + { + "epoch": 0.4877332774166492, + "grad_norm": 5.387138463765321, + "learning_rate": 1.909285422209838e-05, + "loss": 2.1706, + "step": 2326 + }, + { + "epoch": 0.48794296498217654, + "grad_norm": 5.439546977728199, + "learning_rate": 1.9091911821300594e-05, + "loss": 1.9532, + "step": 2327 + }, + { + "epoch": 0.4881526525477039, + "grad_norm": 5.716428250886569, + "learning_rate": 1.9090968954526564e-05, + "loss": 2.2681, + "step": 2328 + }, + { + "epoch": 0.48836234011323126, + "grad_norm": 5.535149121498627, + "learning_rate": 1.9090025621824603e-05, + "loss": 1.9706, + "step": 2329 + }, + { + "epoch": 0.4885720276787586, + "grad_norm": 9.22039450670552, + "learning_rate": 1.9089081823243066e-05, + "loss": 2.1087, + "step": 2330 + }, + { + "epoch": 0.48878171524428604, + "grad_norm": 5.310122222694555, + "learning_rate": 1.9088137558830318e-05, + "loss": 2.056, + "step": 2331 + }, + { + "epoch": 0.4889914028098134, + "grad_norm": 5.295513419045383, + "learning_rate": 1.908719282863476e-05, + "loss": 2.0356, + "step": 2332 + }, + { + "epoch": 0.48920109037534076, + "grad_norm": 5.425010593808835, + "learning_rate": 1.9086247632704808e-05, + "loss": 1.9661, + "step": 2333 + }, + { + "epoch": 0.4894107779408681, + "grad_norm": 6.111997610064877, + "learning_rate": 1.9085301971088906e-05, + "loss": 1.879, + "step": 2334 + }, + { + "epoch": 0.4896204655063955, + "grad_norm": 6.421672708221043, + "learning_rate": 1.9084355843835516e-05, + "loss": 2.2508, + "step": 2335 + }, + { + "epoch": 0.48983015307192285, + "grad_norm": 6.881533398074077, + "learning_rate": 1.908340925099314e-05, + "loss": 2.2109, + "step": 2336 + }, + { + "epoch": 0.4900398406374502, + "grad_norm": 6.518485068188767, + "learning_rate": 1.9082462192610283e-05, + "loss": 2.0763, + "step": 2337 + }, + { + "epoch": 0.49024952820297757, + "grad_norm": 6.819380817404949, + "learning_rate": 1.9081514668735485e-05, + "loss": 2.1932, + "step": 2338 + }, + { + "epoch": 0.49045921576850493, + "grad_norm": 7.3194014231403415, + "learning_rate": 1.9080566679417313e-05, + "loss": 1.9668, + "step": 2339 + }, + { + "epoch": 0.4906689033340323, + "grad_norm": 6.851200789078422, + "learning_rate": 1.9079618224704345e-05, + "loss": 2.3139, + "step": 2340 + }, + { + "epoch": 0.49087859089955965, + "grad_norm": 7.173304358486707, + "learning_rate": 1.907866930464519e-05, + "loss": 1.9716, + "step": 2341 + }, + { + "epoch": 0.491088278465087, + "grad_norm": 6.690516006788188, + "learning_rate": 1.9077719919288497e-05, + "loss": 2.2947, + "step": 2342 + }, + { + "epoch": 0.4912979660306144, + "grad_norm": 6.13828021216144, + "learning_rate": 1.907677006868291e-05, + "loss": 1.8088, + "step": 2343 + }, + { + "epoch": 0.49150765359614174, + "grad_norm": 6.6533509459098985, + "learning_rate": 1.907581975287711e-05, + "loss": 2.3695, + "step": 2344 + }, + { + "epoch": 0.4917173411616691, + "grad_norm": 5.572810442397147, + "learning_rate": 1.9074868971919812e-05, + "loss": 2.258, + "step": 2345 + }, + { + "epoch": 0.49192702872719646, + "grad_norm": 5.695927312028442, + "learning_rate": 1.9073917725859738e-05, + "loss": 2.3008, + "step": 2346 + }, + { + "epoch": 0.4921367162927238, + "grad_norm": 5.562874174324754, + "learning_rate": 1.9072966014745643e-05, + "loss": 1.6022, + "step": 2347 + }, + { + "epoch": 0.4923464038582512, + "grad_norm": 5.487272229550957, + "learning_rate": 1.9072013838626303e-05, + "loss": 2.3415, + "step": 2348 + }, + { + "epoch": 0.4925560914237786, + "grad_norm": 5.074277148633861, + "learning_rate": 1.907106119755052e-05, + "loss": 2.1237, + "step": 2349 + }, + { + "epoch": 0.49276577898930596, + "grad_norm": 6.517231037750821, + "learning_rate": 1.9070108091567115e-05, + "loss": 2.5566, + "step": 2350 + }, + { + "epoch": 0.4929754665548333, + "grad_norm": 5.624852956730777, + "learning_rate": 1.906915452072494e-05, + "loss": 1.9445, + "step": 2351 + }, + { + "epoch": 0.4931851541203607, + "grad_norm": 5.775054823995273, + "learning_rate": 1.9068200485072867e-05, + "loss": 2.1182, + "step": 2352 + }, + { + "epoch": 0.49339484168588804, + "grad_norm": 5.296104029724121, + "learning_rate": 1.906724598465979e-05, + "loss": 2.2116, + "step": 2353 + }, + { + "epoch": 0.4936045292514154, + "grad_norm": 5.025189306268243, + "learning_rate": 1.9066291019534632e-05, + "loss": 1.8211, + "step": 2354 + }, + { + "epoch": 0.49381421681694276, + "grad_norm": 6.545413003683983, + "learning_rate": 1.9065335589746335e-05, + "loss": 2.3161, + "step": 2355 + }, + { + "epoch": 0.4940239043824701, + "grad_norm": 6.415484420931244, + "learning_rate": 1.906437969534386e-05, + "loss": 2.1413, + "step": 2356 + }, + { + "epoch": 0.4942335919479975, + "grad_norm": 5.54241740507023, + "learning_rate": 1.906342333637621e-05, + "loss": 1.9692, + "step": 2357 + }, + { + "epoch": 0.49444327951352485, + "grad_norm": 6.036925989066589, + "learning_rate": 1.9062466512892395e-05, + "loss": 2.2641, + "step": 2358 + }, + { + "epoch": 0.4946529670790522, + "grad_norm": 5.420006516972742, + "learning_rate": 1.906150922494145e-05, + "loss": 2.0105, + "step": 2359 + }, + { + "epoch": 0.49486265464457957, + "grad_norm": 5.864509212812893, + "learning_rate": 1.9060551472572442e-05, + "loss": 1.86, + "step": 2360 + }, + { + "epoch": 0.49507234221010693, + "grad_norm": 7.406032077685681, + "learning_rate": 1.9059593255834457e-05, + "loss": 1.9605, + "step": 2361 + }, + { + "epoch": 0.4952820297756343, + "grad_norm": 6.715096049536995, + "learning_rate": 1.9058634574776606e-05, + "loss": 2.19, + "step": 2362 + }, + { + "epoch": 0.49549171734116165, + "grad_norm": 5.437844276662292, + "learning_rate": 1.9057675429448015e-05, + "loss": 2.019, + "step": 2363 + }, + { + "epoch": 0.495701404906689, + "grad_norm": 6.758132630549353, + "learning_rate": 1.9056715819897855e-05, + "loss": 2.2599, + "step": 2364 + }, + { + "epoch": 0.4959110924722164, + "grad_norm": 5.5133441340501, + "learning_rate": 1.90557557461753e-05, + "loss": 2.04, + "step": 2365 + }, + { + "epoch": 0.49612078003774374, + "grad_norm": 5.7863830826148535, + "learning_rate": 1.9054795208329558e-05, + "loss": 1.7747, + "step": 2366 + }, + { + "epoch": 0.49633046760327115, + "grad_norm": 6.053456590647323, + "learning_rate": 1.9053834206409854e-05, + "loss": 2.272, + "step": 2367 + }, + { + "epoch": 0.4965401551687985, + "grad_norm": 5.503077556820159, + "learning_rate": 1.9052872740465444e-05, + "loss": 2.1596, + "step": 2368 + }, + { + "epoch": 0.4967498427343259, + "grad_norm": 6.214830320920408, + "learning_rate": 1.9051910810545604e-05, + "loss": 2.0359, + "step": 2369 + }, + { + "epoch": 0.49695953029985324, + "grad_norm": 6.70460844242227, + "learning_rate": 1.9050948416699636e-05, + "loss": 2.0042, + "step": 2370 + }, + { + "epoch": 0.4971692178653806, + "grad_norm": 7.981658960225294, + "learning_rate": 1.9049985558976864e-05, + "loss": 1.8125, + "step": 2371 + }, + { + "epoch": 0.49737890543090796, + "grad_norm": 6.424302636271175, + "learning_rate": 1.9049022237426635e-05, + "loss": 2.0033, + "step": 2372 + }, + { + "epoch": 0.4975885929964353, + "grad_norm": 5.483404936748399, + "learning_rate": 1.9048058452098323e-05, + "loss": 2.0798, + "step": 2373 + }, + { + "epoch": 0.4977982805619627, + "grad_norm": 6.178462083181409, + "learning_rate": 1.9047094203041322e-05, + "loss": 2.1715, + "step": 2374 + }, + { + "epoch": 0.49800796812749004, + "grad_norm": 6.252924502536067, + "learning_rate": 1.904612949030505e-05, + "loss": 2.1147, + "step": 2375 + }, + { + "epoch": 0.4982176556930174, + "grad_norm": 5.8091544377416575, + "learning_rate": 1.9045164313938952e-05, + "loss": 2.1987, + "step": 2376 + }, + { + "epoch": 0.49842734325854476, + "grad_norm": 7.532699761994082, + "learning_rate": 1.90441986739925e-05, + "loss": 2.398, + "step": 2377 + }, + { + "epoch": 0.4986370308240721, + "grad_norm": 5.099261833607678, + "learning_rate": 1.9043232570515174e-05, + "loss": 1.9185, + "step": 2378 + }, + { + "epoch": 0.4988467183895995, + "grad_norm": 6.320403373369112, + "learning_rate": 1.9042266003556494e-05, + "loss": 2.0786, + "step": 2379 + }, + { + "epoch": 0.49905640595512685, + "grad_norm": 5.203193542742107, + "learning_rate": 1.9041298973166e-05, + "loss": 2.1385, + "step": 2380 + }, + { + "epoch": 0.4992660935206542, + "grad_norm": 6.249330609505911, + "learning_rate": 1.904033147939325e-05, + "loss": 1.9189, + "step": 2381 + }, + { + "epoch": 0.49947578108618157, + "grad_norm": 5.664877339319939, + "learning_rate": 1.9039363522287835e-05, + "loss": 2.1726, + "step": 2382 + }, + { + "epoch": 0.49968546865170893, + "grad_norm": 5.740542861207297, + "learning_rate": 1.903839510189936e-05, + "loss": 2.0998, + "step": 2383 + }, + { + "epoch": 0.4998951562172363, + "grad_norm": 5.7689742490636355, + "learning_rate": 1.9037426218277463e-05, + "loss": 2.2739, + "step": 2384 + }, + { + "epoch": 0.5001048437827637, + "grad_norm": 6.033687363725323, + "learning_rate": 1.9036456871471792e-05, + "loss": 2.1637, + "step": 2385 + }, + { + "epoch": 0.5003145313482911, + "grad_norm": 8.01699680841515, + "learning_rate": 1.903548706153204e-05, + "loss": 2.2429, + "step": 2386 + }, + { + "epoch": 0.5005242189138184, + "grad_norm": 5.2260689823153275, + "learning_rate": 1.90345167885079e-05, + "loss": 2.0867, + "step": 2387 + }, + { + "epoch": 0.5007339064793458, + "grad_norm": 5.581233110187377, + "learning_rate": 1.9033546052449106e-05, + "loss": 2.185, + "step": 2388 + }, + { + "epoch": 0.5009435940448731, + "grad_norm": 8.356945135235096, + "learning_rate": 1.903257485340541e-05, + "loss": 2.5881, + "step": 2389 + }, + { + "epoch": 0.5011532816104005, + "grad_norm": 5.293277979893477, + "learning_rate": 1.9031603191426587e-05, + "loss": 2.1925, + "step": 2390 + }, + { + "epoch": 0.5013629691759278, + "grad_norm": 5.65899763464518, + "learning_rate": 1.9030631066562436e-05, + "loss": 2.5119, + "step": 2391 + }, + { + "epoch": 0.5015726567414552, + "grad_norm": 5.452565421328935, + "learning_rate": 1.9029658478862776e-05, + "loss": 1.7918, + "step": 2392 + }, + { + "epoch": 0.5017823443069825, + "grad_norm": 9.15098421579578, + "learning_rate": 1.9028685428377462e-05, + "loss": 2.4517, + "step": 2393 + }, + { + "epoch": 0.50199203187251, + "grad_norm": 6.153944910687597, + "learning_rate": 1.902771191515636e-05, + "loss": 2.2359, + "step": 2394 + }, + { + "epoch": 0.5022017194380374, + "grad_norm": 5.667062866273726, + "learning_rate": 1.9026737939249363e-05, + "loss": 1.91, + "step": 2395 + }, + { + "epoch": 0.5024114070035647, + "grad_norm": 5.9757697757745945, + "learning_rate": 1.902576350070639e-05, + "loss": 2.1999, + "step": 2396 + }, + { + "epoch": 0.5026210945690921, + "grad_norm": 5.887686237164335, + "learning_rate": 1.9024788599577383e-05, + "loss": 1.9772, + "step": 2397 + }, + { + "epoch": 0.5028307821346194, + "grad_norm": 5.058824111966105, + "learning_rate": 1.902381323591231e-05, + "loss": 2.0835, + "step": 2398 + }, + { + "epoch": 0.5030404697001468, + "grad_norm": 5.7260635696615285, + "learning_rate": 1.9022837409761155e-05, + "loss": 2.127, + "step": 2399 + }, + { + "epoch": 0.5032501572656741, + "grad_norm": 5.14106195159542, + "learning_rate": 1.9021861121173936e-05, + "loss": 2.1018, + "step": 2400 + }, + { + "epoch": 0.5034598448312015, + "grad_norm": 5.432524328519932, + "learning_rate": 1.9020884370200685e-05, + "loss": 2.1782, + "step": 2401 + }, + { + "epoch": 0.5036695323967288, + "grad_norm": 6.286592021837883, + "learning_rate": 1.9019907156891464e-05, + "loss": 2.4022, + "step": 2402 + }, + { + "epoch": 0.5038792199622563, + "grad_norm": 6.102193936181787, + "learning_rate": 1.901892948129636e-05, + "loss": 2.1732, + "step": 2403 + }, + { + "epoch": 0.5040889075277836, + "grad_norm": 4.753235931711819, + "learning_rate": 1.9017951343465473e-05, + "loss": 2.0187, + "step": 2404 + }, + { + "epoch": 0.504298595093311, + "grad_norm": 5.647004233542891, + "learning_rate": 1.9016972743448944e-05, + "loss": 2.0608, + "step": 2405 + }, + { + "epoch": 0.5045082826588383, + "grad_norm": 5.605049904184601, + "learning_rate": 1.9015993681296917e-05, + "loss": 2.1389, + "step": 2406 + }, + { + "epoch": 0.5047179702243657, + "grad_norm": 6.021952684035758, + "learning_rate": 1.9015014157059578e-05, + "loss": 2.3593, + "step": 2407 + }, + { + "epoch": 0.504927657789893, + "grad_norm": 6.103797458703036, + "learning_rate": 1.9014034170787132e-05, + "loss": 2.0789, + "step": 2408 + }, + { + "epoch": 0.5051373453554204, + "grad_norm": 5.967531630592809, + "learning_rate": 1.9013053722529798e-05, + "loss": 2.0806, + "step": 2409 + }, + { + "epoch": 0.5053470329209477, + "grad_norm": 6.029084380909239, + "learning_rate": 1.9012072812337828e-05, + "loss": 2.261, + "step": 2410 + }, + { + "epoch": 0.5055567204864752, + "grad_norm": 6.012398732609155, + "learning_rate": 1.9011091440261495e-05, + "loss": 1.9123, + "step": 2411 + }, + { + "epoch": 0.5057664080520026, + "grad_norm": 5.350645210145288, + "learning_rate": 1.9010109606351098e-05, + "loss": 2.0402, + "step": 2412 + }, + { + "epoch": 0.5059760956175299, + "grad_norm": 5.633013623478876, + "learning_rate": 1.9009127310656958e-05, + "loss": 2.2955, + "step": 2413 + }, + { + "epoch": 0.5061857831830573, + "grad_norm": 5.8473098450970715, + "learning_rate": 1.9008144553229414e-05, + "loss": 2.1665, + "step": 2414 + }, + { + "epoch": 0.5063954707485846, + "grad_norm": 5.875046523618884, + "learning_rate": 1.900716133411884e-05, + "loss": 2.1546, + "step": 2415 + }, + { + "epoch": 0.506605158314112, + "grad_norm": 5.683185050794553, + "learning_rate": 1.9006177653375623e-05, + "loss": 1.9391, + "step": 2416 + }, + { + "epoch": 0.5068148458796393, + "grad_norm": 5.661859288865748, + "learning_rate": 1.9005193511050183e-05, + "loss": 2.1656, + "step": 2417 + }, + { + "epoch": 0.5070245334451667, + "grad_norm": 6.508985724057258, + "learning_rate": 1.900420890719296e-05, + "loss": 1.9514, + "step": 2418 + }, + { + "epoch": 0.507234221010694, + "grad_norm": 5.453614688749812, + "learning_rate": 1.900322384185441e-05, + "loss": 2.0257, + "step": 2419 + }, + { + "epoch": 0.5074439085762215, + "grad_norm": 6.151785595113499, + "learning_rate": 1.900223831508502e-05, + "loss": 2.2917, + "step": 2420 + }, + { + "epoch": 0.5076535961417488, + "grad_norm": 6.551169074036723, + "learning_rate": 1.9001252326935306e-05, + "loss": 2.5201, + "step": 2421 + }, + { + "epoch": 0.5078632837072762, + "grad_norm": 5.479608564104697, + "learning_rate": 1.90002658774558e-05, + "loss": 2.118, + "step": 2422 + }, + { + "epoch": 0.5080729712728035, + "grad_norm": 5.978904310512989, + "learning_rate": 1.8999278966697055e-05, + "loss": 1.9548, + "step": 2423 + }, + { + "epoch": 0.5082826588383309, + "grad_norm": 5.838716078489018, + "learning_rate": 1.8998291594709655e-05, + "loss": 2.2903, + "step": 2424 + }, + { + "epoch": 0.5084923464038582, + "grad_norm": 6.929947633172149, + "learning_rate": 1.8997303761544206e-05, + "loss": 2.2608, + "step": 2425 + }, + { + "epoch": 0.5087020339693856, + "grad_norm": 6.610662867772675, + "learning_rate": 1.8996315467251333e-05, + "loss": 2.0301, + "step": 2426 + }, + { + "epoch": 0.5089117215349129, + "grad_norm": 5.679471733482516, + "learning_rate": 1.8995326711881687e-05, + "loss": 2.0016, + "step": 2427 + }, + { + "epoch": 0.5091214091004403, + "grad_norm": 6.0816843081496925, + "learning_rate": 1.8994337495485946e-05, + "loss": 1.9341, + "step": 2428 + }, + { + "epoch": 0.5093310966659677, + "grad_norm": 5.475189235774921, + "learning_rate": 1.899334781811481e-05, + "loss": 1.8723, + "step": 2429 + }, + { + "epoch": 0.5095407842314951, + "grad_norm": 4.94909873518066, + "learning_rate": 1.8992357679819e-05, + "loss": 1.8506, + "step": 2430 + }, + { + "epoch": 0.5097504717970225, + "grad_norm": 5.668184705303442, + "learning_rate": 1.8991367080649262e-05, + "loss": 2.3004, + "step": 2431 + }, + { + "epoch": 0.5099601593625498, + "grad_norm": 6.548289442959977, + "learning_rate": 1.8990376020656363e-05, + "loss": 2.0827, + "step": 2432 + }, + { + "epoch": 0.5101698469280772, + "grad_norm": 6.200477921712447, + "learning_rate": 1.8989384499891105e-05, + "loss": 2.0737, + "step": 2433 + }, + { + "epoch": 0.5103795344936045, + "grad_norm": 6.699598810408789, + "learning_rate": 1.89883925184043e-05, + "loss": 1.9239, + "step": 2434 + }, + { + "epoch": 0.5105892220591319, + "grad_norm": 5.933597619317245, + "learning_rate": 1.8987400076246786e-05, + "loss": 2.0603, + "step": 2435 + }, + { + "epoch": 0.5107989096246592, + "grad_norm": 5.594914065145635, + "learning_rate": 1.898640717346943e-05, + "loss": 1.8536, + "step": 2436 + }, + { + "epoch": 0.5110085971901867, + "grad_norm": 6.124003405592209, + "learning_rate": 1.8985413810123122e-05, + "loss": 2.168, + "step": 2437 + }, + { + "epoch": 0.511218284755714, + "grad_norm": 5.117847996774858, + "learning_rate": 1.8984419986258773e-05, + "loss": 2.1751, + "step": 2438 + }, + { + "epoch": 0.5114279723212414, + "grad_norm": 6.496920280151962, + "learning_rate": 1.8983425701927313e-05, + "loss": 2.3464, + "step": 2439 + }, + { + "epoch": 0.5116376598867687, + "grad_norm": 5.446051004489931, + "learning_rate": 1.898243095717971e-05, + "loss": 2.2514, + "step": 2440 + }, + { + "epoch": 0.5118473474522961, + "grad_norm": 6.395967676245346, + "learning_rate": 1.8981435752066937e-05, + "loss": 1.6687, + "step": 2441 + }, + { + "epoch": 0.5120570350178234, + "grad_norm": 5.836971640083478, + "learning_rate": 1.898044008664001e-05, + "loss": 2.2307, + "step": 2442 + }, + { + "epoch": 0.5122667225833508, + "grad_norm": 7.052926053273921, + "learning_rate": 1.8979443960949945e-05, + "loss": 2.0052, + "step": 2443 + }, + { + "epoch": 0.5124764101488781, + "grad_norm": 5.7966212634184044, + "learning_rate": 1.8978447375047805e-05, + "loss": 2.245, + "step": 2444 + }, + { + "epoch": 0.5126860977144055, + "grad_norm": 4.752110915106241, + "learning_rate": 1.897745032898467e-05, + "loss": 2.02, + "step": 2445 + }, + { + "epoch": 0.5128957852799328, + "grad_norm": 6.124714542750896, + "learning_rate": 1.897645282281163e-05, + "loss": 2.3848, + "step": 2446 + }, + { + "epoch": 0.5131054728454603, + "grad_norm": 5.907468372087065, + "learning_rate": 1.8975454856579817e-05, + "loss": 2.1063, + "step": 2447 + }, + { + "epoch": 0.5133151604109877, + "grad_norm": 6.7460353381616125, + "learning_rate": 1.897445643034037e-05, + "loss": 2.2471, + "step": 2448 + }, + { + "epoch": 0.513524847976515, + "grad_norm": 5.256027015938374, + "learning_rate": 1.897345754414447e-05, + "loss": 1.8617, + "step": 2449 + }, + { + "epoch": 0.5137345355420424, + "grad_norm": 6.888810526007073, + "learning_rate": 1.8972458198043306e-05, + "loss": 1.8979, + "step": 2450 + }, + { + "epoch": 0.5139442231075697, + "grad_norm": 6.274816385128185, + "learning_rate": 1.8971458392088097e-05, + "loss": 1.8552, + "step": 2451 + }, + { + "epoch": 0.5141539106730971, + "grad_norm": 5.701180089204313, + "learning_rate": 1.8970458126330088e-05, + "loss": 1.854, + "step": 2452 + }, + { + "epoch": 0.5143635982386244, + "grad_norm": 5.497722221126138, + "learning_rate": 1.896945740082054e-05, + "loss": 2.1337, + "step": 2453 + }, + { + "epoch": 0.5145732858041518, + "grad_norm": 5.539619232366782, + "learning_rate": 1.8968456215610742e-05, + "loss": 2.5457, + "step": 2454 + }, + { + "epoch": 0.5147829733696792, + "grad_norm": 6.165740341382778, + "learning_rate": 1.896745457075201e-05, + "loss": 2.1018, + "step": 2455 + }, + { + "epoch": 0.5149926609352066, + "grad_norm": 5.864414206491105, + "learning_rate": 1.8966452466295674e-05, + "loss": 2.0048, + "step": 2456 + }, + { + "epoch": 0.5152023485007339, + "grad_norm": 5.707363694056109, + "learning_rate": 1.8965449902293103e-05, + "loss": 2.2108, + "step": 2457 + }, + { + "epoch": 0.5154120360662613, + "grad_norm": 5.936006779252624, + "learning_rate": 1.896444687879567e-05, + "loss": 1.9133, + "step": 2458 + }, + { + "epoch": 0.5156217236317886, + "grad_norm": 5.6405877269955775, + "learning_rate": 1.896344339585479e-05, + "loss": 1.6468, + "step": 2459 + }, + { + "epoch": 0.515831411197316, + "grad_norm": 6.633958221080308, + "learning_rate": 1.8962439453521887e-05, + "loss": 2.4828, + "step": 2460 + }, + { + "epoch": 0.5160410987628433, + "grad_norm": 5.422441272269181, + "learning_rate": 1.8961435051848422e-05, + "loss": 1.9458, + "step": 2461 + }, + { + "epoch": 0.5162507863283707, + "grad_norm": 6.4908141697417285, + "learning_rate": 1.8960430190885862e-05, + "loss": 2.3311, + "step": 2462 + }, + { + "epoch": 0.516460473893898, + "grad_norm": 5.626908643083947, + "learning_rate": 1.8959424870685718e-05, + "loss": 1.9031, + "step": 2463 + }, + { + "epoch": 0.5166701614594255, + "grad_norm": 5.645935982009144, + "learning_rate": 1.895841909129951e-05, + "loss": 2.1234, + "step": 2464 + }, + { + "epoch": 0.5168798490249529, + "grad_norm": 6.355549663504592, + "learning_rate": 1.895741285277878e-05, + "loss": 1.8931, + "step": 2465 + }, + { + "epoch": 0.5170895365904802, + "grad_norm": 5.56836460649157, + "learning_rate": 1.895640615517511e-05, + "loss": 2.0302, + "step": 2466 + }, + { + "epoch": 0.5172992241560076, + "grad_norm": 5.867739565581109, + "learning_rate": 1.895539899854009e-05, + "loss": 1.9763, + "step": 2467 + }, + { + "epoch": 0.5175089117215349, + "grad_norm": 6.2409512987061175, + "learning_rate": 1.895439138292534e-05, + "loss": 1.9439, + "step": 2468 + }, + { + "epoch": 0.5177185992870623, + "grad_norm": 5.500422630646335, + "learning_rate": 1.89533833083825e-05, + "loss": 2.1511, + "step": 2469 + }, + { + "epoch": 0.5179282868525896, + "grad_norm": 6.392000632653926, + "learning_rate": 1.8952374774963238e-05, + "loss": 2.0937, + "step": 2470 + }, + { + "epoch": 0.518137974418117, + "grad_norm": 5.866132675037486, + "learning_rate": 1.895136578271924e-05, + "loss": 1.9408, + "step": 2471 + }, + { + "epoch": 0.5183476619836443, + "grad_norm": 5.646877043591221, + "learning_rate": 1.895035633170222e-05, + "loss": 1.7246, + "step": 2472 + }, + { + "epoch": 0.5185573495491718, + "grad_norm": 6.319846113211682, + "learning_rate": 1.8949346421963914e-05, + "loss": 2.0778, + "step": 2473 + }, + { + "epoch": 0.5187670371146991, + "grad_norm": 6.1000946977443755, + "learning_rate": 1.8948336053556085e-05, + "loss": 2.1154, + "step": 2474 + }, + { + "epoch": 0.5189767246802265, + "grad_norm": 6.201465736094627, + "learning_rate": 1.894732522653051e-05, + "loss": 2.2514, + "step": 2475 + }, + { + "epoch": 0.5191864122457538, + "grad_norm": 6.156466409315422, + "learning_rate": 1.8946313940939e-05, + "loss": 2.0332, + "step": 2476 + }, + { + "epoch": 0.5193960998112812, + "grad_norm": 4.851808363217823, + "learning_rate": 1.8945302196833383e-05, + "loss": 1.7619, + "step": 2477 + }, + { + "epoch": 0.5196057873768085, + "grad_norm": 6.3137014719236255, + "learning_rate": 1.8944289994265516e-05, + "loss": 2.2938, + "step": 2478 + }, + { + "epoch": 0.5198154749423359, + "grad_norm": 5.424746886631281, + "learning_rate": 1.8943277333287274e-05, + "loss": 1.8157, + "step": 2479 + }, + { + "epoch": 0.5200251625078632, + "grad_norm": 5.8054338980657745, + "learning_rate": 1.8942264213950553e-05, + "loss": 2.2112, + "step": 2480 + }, + { + "epoch": 0.5202348500733907, + "grad_norm": 5.623690997076992, + "learning_rate": 1.8941250636307285e-05, + "loss": 2.0334, + "step": 2481 + }, + { + "epoch": 0.520444537638918, + "grad_norm": 6.479274120125267, + "learning_rate": 1.8940236600409415e-05, + "loss": 2.2167, + "step": 2482 + }, + { + "epoch": 0.5206542252044454, + "grad_norm": 6.763323296402688, + "learning_rate": 1.893922210630891e-05, + "loss": 2.0633, + "step": 2483 + }, + { + "epoch": 0.5208639127699728, + "grad_norm": 6.345361107743618, + "learning_rate": 1.893820715405777e-05, + "loss": 1.8266, + "step": 2484 + }, + { + "epoch": 0.5210736003355001, + "grad_norm": 5.788027774176796, + "learning_rate": 1.8937191743708007e-05, + "loss": 2.074, + "step": 2485 + }, + { + "epoch": 0.5212832879010275, + "grad_norm": 5.214432151138426, + "learning_rate": 1.893617587531167e-05, + "loss": 2.1817, + "step": 2486 + }, + { + "epoch": 0.5214929754665548, + "grad_norm": 5.912122837167264, + "learning_rate": 1.893515954892082e-05, + "loss": 2.001, + "step": 2487 + }, + { + "epoch": 0.5217026630320822, + "grad_norm": 5.882870610521304, + "learning_rate": 1.8934142764587548e-05, + "loss": 2.025, + "step": 2488 + }, + { + "epoch": 0.5219123505976095, + "grad_norm": 5.990650409757526, + "learning_rate": 1.8933125522363957e-05, + "loss": 2.1127, + "step": 2489 + }, + { + "epoch": 0.522122038163137, + "grad_norm": 5.870296045274698, + "learning_rate": 1.8932107822302194e-05, + "loss": 2.0589, + "step": 2490 + }, + { + "epoch": 0.5223317257286643, + "grad_norm": 6.715016531205387, + "learning_rate": 1.8931089664454415e-05, + "loss": 2.4405, + "step": 2491 + }, + { + "epoch": 0.5225414132941917, + "grad_norm": 5.982506981343005, + "learning_rate": 1.89300710488728e-05, + "loss": 2.1682, + "step": 2492 + }, + { + "epoch": 0.522751100859719, + "grad_norm": 5.131394305041194, + "learning_rate": 1.892905197560955e-05, + "loss": 2.0204, + "step": 2493 + }, + { + "epoch": 0.5229607884252464, + "grad_norm": 6.056956846506437, + "learning_rate": 1.8928032444716905e-05, + "loss": 2.1809, + "step": 2494 + }, + { + "epoch": 0.5231704759907737, + "grad_norm": 5.70724378372189, + "learning_rate": 1.892701245624711e-05, + "loss": 2.4067, + "step": 2495 + }, + { + "epoch": 0.5233801635563011, + "grad_norm": 5.642034277483455, + "learning_rate": 1.8925992010252442e-05, + "loss": 1.9798, + "step": 2496 + }, + { + "epoch": 0.5235898511218284, + "grad_norm": 6.082824402807786, + "learning_rate": 1.8924971106785203e-05, + "loss": 2.2759, + "step": 2497 + }, + { + "epoch": 0.5237995386873558, + "grad_norm": 5.99309358311386, + "learning_rate": 1.892394974589772e-05, + "loss": 2.212, + "step": 2498 + }, + { + "epoch": 0.5240092262528832, + "grad_norm": 5.6779773412157795, + "learning_rate": 1.892292792764233e-05, + "loss": 1.9566, + "step": 2499 + }, + { + "epoch": 0.5242189138184106, + "grad_norm": 6.070708602362544, + "learning_rate": 1.8921905652071407e-05, + "loss": 2.1451, + "step": 2500 + }, + { + "epoch": 0.524428601383938, + "grad_norm": 7.218418417680411, + "learning_rate": 1.892088291923735e-05, + "loss": 2.399, + "step": 2501 + }, + { + "epoch": 0.5246382889494653, + "grad_norm": 5.5088831545755745, + "learning_rate": 1.8919859729192564e-05, + "loss": 2.1479, + "step": 2502 + }, + { + "epoch": 0.5248479765149927, + "grad_norm": 6.992877515385851, + "learning_rate": 1.89188360819895e-05, + "loss": 2.2418, + "step": 2503 + }, + { + "epoch": 0.52505766408052, + "grad_norm": 5.578938026909314, + "learning_rate": 1.891781197768062e-05, + "loss": 2.3097, + "step": 2504 + }, + { + "epoch": 0.5252673516460474, + "grad_norm": 5.614939133387399, + "learning_rate": 1.8916787416318404e-05, + "loss": 2.2699, + "step": 2505 + }, + { + "epoch": 0.5254770392115747, + "grad_norm": 5.583994581207203, + "learning_rate": 1.891576239795537e-05, + "loss": 2.3479, + "step": 2506 + }, + { + "epoch": 0.5256867267771022, + "grad_norm": 5.420076296041309, + "learning_rate": 1.891473692264405e-05, + "loss": 2.26, + "step": 2507 + }, + { + "epoch": 0.5258964143426295, + "grad_norm": 6.786149398370626, + "learning_rate": 1.8913710990437003e-05, + "loss": 2.2815, + "step": 2508 + }, + { + "epoch": 0.5261061019081569, + "grad_norm": 6.437573523645526, + "learning_rate": 1.8912684601386805e-05, + "loss": 2.0489, + "step": 2509 + }, + { + "epoch": 0.5263157894736842, + "grad_norm": 5.257089587666138, + "learning_rate": 1.8911657755546063e-05, + "loss": 2.1594, + "step": 2510 + }, + { + "epoch": 0.5265254770392116, + "grad_norm": 5.785725256332309, + "learning_rate": 1.8910630452967403e-05, + "loss": 2.29, + "step": 2511 + }, + { + "epoch": 0.5267351646047389, + "grad_norm": 5.190953858034122, + "learning_rate": 1.8909602693703482e-05, + "loss": 2.2172, + "step": 2512 + }, + { + "epoch": 0.5269448521702663, + "grad_norm": 5.95180831168956, + "learning_rate": 1.890857447780697e-05, + "loss": 1.9096, + "step": 2513 + }, + { + "epoch": 0.5271545397357936, + "grad_norm": 5.793914220672826, + "learning_rate": 1.890754580533056e-05, + "loss": 1.9626, + "step": 2514 + }, + { + "epoch": 0.527364227301321, + "grad_norm": 5.794222872174821, + "learning_rate": 1.890651667632698e-05, + "loss": 2.1394, + "step": 2515 + }, + { + "epoch": 0.5275739148668483, + "grad_norm": 6.324613866456327, + "learning_rate": 1.890548709084897e-05, + "loss": 2.1301, + "step": 2516 + }, + { + "epoch": 0.5277836024323758, + "grad_norm": 6.13198617558845, + "learning_rate": 1.8904457048949306e-05, + "loss": 2.1908, + "step": 2517 + }, + { + "epoch": 0.5279932899979031, + "grad_norm": 5.349212019707747, + "learning_rate": 1.890342655068077e-05, + "loss": 2.0959, + "step": 2518 + }, + { + "epoch": 0.5282029775634305, + "grad_norm": 7.0716757372954335, + "learning_rate": 1.8902395596096184e-05, + "loss": 2.1723, + "step": 2519 + }, + { + "epoch": 0.5284126651289579, + "grad_norm": 7.486544596485488, + "learning_rate": 1.8901364185248382e-05, + "loss": 2.2201, + "step": 2520 + }, + { + "epoch": 0.5286223526944852, + "grad_norm": 6.06742622916226, + "learning_rate": 1.8900332318190227e-05, + "loss": 2.2142, + "step": 2521 + }, + { + "epoch": 0.5288320402600126, + "grad_norm": 5.625352313015393, + "learning_rate": 1.8899299994974605e-05, + "loss": 2.1964, + "step": 2522 + }, + { + "epoch": 0.5290417278255399, + "grad_norm": 5.602054241009448, + "learning_rate": 1.8898267215654424e-05, + "loss": 2.0153, + "step": 2523 + }, + { + "epoch": 0.5292514153910673, + "grad_norm": 6.299744307991921, + "learning_rate": 1.8897233980282614e-05, + "loss": 2.4547, + "step": 2524 + }, + { + "epoch": 0.5294611029565947, + "grad_norm": 5.140554876015905, + "learning_rate": 1.889620028891213e-05, + "loss": 1.7746, + "step": 2525 + }, + { + "epoch": 0.5296707905221221, + "grad_norm": 6.018089094757581, + "learning_rate": 1.8895166141595952e-05, + "loss": 1.8517, + "step": 2526 + }, + { + "epoch": 0.5298804780876494, + "grad_norm": 5.270471274648795, + "learning_rate": 1.8894131538387078e-05, + "loss": 2.2218, + "step": 2527 + }, + { + "epoch": 0.5300901656531768, + "grad_norm": 5.729871918211228, + "learning_rate": 1.8893096479338543e-05, + "loss": 2.128, + "step": 2528 + }, + { + "epoch": 0.5302998532187041, + "grad_norm": 6.423095525940002, + "learning_rate": 1.8892060964503385e-05, + "loss": 2.3252, + "step": 2529 + }, + { + "epoch": 0.5305095407842315, + "grad_norm": 5.639575883968387, + "learning_rate": 1.889102499393468e-05, + "loss": 2.1844, + "step": 2530 + }, + { + "epoch": 0.5307192283497588, + "grad_norm": 5.914693416864942, + "learning_rate": 1.8889988567685528e-05, + "loss": 1.7898, + "step": 2531 + }, + { + "epoch": 0.5309289159152862, + "grad_norm": 5.907057379136445, + "learning_rate": 1.8888951685809036e-05, + "loss": 1.9841, + "step": 2532 + }, + { + "epoch": 0.5311386034808135, + "grad_norm": 6.005674083766735, + "learning_rate": 1.8887914348358357e-05, + "loss": 1.7599, + "step": 2533 + }, + { + "epoch": 0.531348291046341, + "grad_norm": 6.144101867538987, + "learning_rate": 1.8886876555386654e-05, + "loss": 1.8793, + "step": 2534 + }, + { + "epoch": 0.5315579786118683, + "grad_norm": 5.836397378552884, + "learning_rate": 1.888583830694711e-05, + "loss": 1.8599, + "step": 2535 + }, + { + "epoch": 0.5317676661773957, + "grad_norm": 5.707268278185482, + "learning_rate": 1.8884799603092944e-05, + "loss": 1.9251, + "step": 2536 + }, + { + "epoch": 0.5319773537429231, + "grad_norm": 5.409215441129544, + "learning_rate": 1.8883760443877385e-05, + "loss": 2.0358, + "step": 2537 + }, + { + "epoch": 0.5321870413084504, + "grad_norm": 5.782022069269918, + "learning_rate": 1.88827208293537e-05, + "loss": 2.356, + "step": 2538 + }, + { + "epoch": 0.5323967288739778, + "grad_norm": 6.025708045357777, + "learning_rate": 1.8881680759575164e-05, + "loss": 2.1097, + "step": 2539 + }, + { + "epoch": 0.5326064164395051, + "grad_norm": 6.113074775414307, + "learning_rate": 1.888064023459508e-05, + "loss": 2.0383, + "step": 2540 + }, + { + "epoch": 0.5328161040050325, + "grad_norm": 6.874447921239057, + "learning_rate": 1.887959925446679e-05, + "loss": 1.8064, + "step": 2541 + }, + { + "epoch": 0.5330257915705598, + "grad_norm": 6.974773252832469, + "learning_rate": 1.887855781924363e-05, + "loss": 2.2379, + "step": 2542 + }, + { + "epoch": 0.5332354791360873, + "grad_norm": 7.143128441944039, + "learning_rate": 1.8877515928978988e-05, + "loss": 1.6974, + "step": 2543 + }, + { + "epoch": 0.5334451667016146, + "grad_norm": 5.506020343500608, + "learning_rate": 1.8876473583726255e-05, + "loss": 1.9358, + "step": 2544 + }, + { + "epoch": 0.533654854267142, + "grad_norm": 5.295286436205329, + "learning_rate": 1.8875430783538854e-05, + "loss": 1.4723, + "step": 2545 + }, + { + "epoch": 0.5338645418326693, + "grad_norm": 6.818605675315057, + "learning_rate": 1.887438752847023e-05, + "loss": 2.1726, + "step": 2546 + }, + { + "epoch": 0.5340742293981967, + "grad_norm": 7.437139298754355, + "learning_rate": 1.887334381857386e-05, + "loss": 2.3991, + "step": 2547 + }, + { + "epoch": 0.534283916963724, + "grad_norm": 6.6226961714605705, + "learning_rate": 1.8872299653903224e-05, + "loss": 2.2787, + "step": 2548 + }, + { + "epoch": 0.5344936045292514, + "grad_norm": 5.920636651180079, + "learning_rate": 1.8871255034511843e-05, + "loss": 2.1613, + "step": 2549 + }, + { + "epoch": 0.5347032920947787, + "grad_norm": 5.704353782310128, + "learning_rate": 1.8870209960453257e-05, + "loss": 2.2763, + "step": 2550 + }, + { + "epoch": 0.5349129796603062, + "grad_norm": 6.365168925361749, + "learning_rate": 1.8869164431781025e-05, + "loss": 2.2851, + "step": 2551 + }, + { + "epoch": 0.5351226672258335, + "grad_norm": 6.852482383812198, + "learning_rate": 1.8868118448548732e-05, + "loss": 1.6981, + "step": 2552 + }, + { + "epoch": 0.5353323547913609, + "grad_norm": 5.341167673368236, + "learning_rate": 1.886707201080999e-05, + "loss": 1.9602, + "step": 2553 + }, + { + "epoch": 0.5355420423568882, + "grad_norm": 6.580285107003099, + "learning_rate": 1.8866025118618427e-05, + "loss": 2.1814, + "step": 2554 + }, + { + "epoch": 0.5357517299224156, + "grad_norm": 6.737933866419712, + "learning_rate": 1.88649777720277e-05, + "loss": 1.9446, + "step": 2555 + }, + { + "epoch": 0.535961417487943, + "grad_norm": 6.009354519815355, + "learning_rate": 1.8863929971091483e-05, + "loss": 1.897, + "step": 2556 + }, + { + "epoch": 0.5361711050534703, + "grad_norm": 6.164549947304244, + "learning_rate": 1.8862881715863485e-05, + "loss": 2.2016, + "step": 2557 + }, + { + "epoch": 0.5363807926189977, + "grad_norm": 5.545686855804167, + "learning_rate": 1.8861833006397426e-05, + "loss": 2.0923, + "step": 2558 + }, + { + "epoch": 0.536590480184525, + "grad_norm": 6.488597114632481, + "learning_rate": 1.8860783842747056e-05, + "loss": 2.1825, + "step": 2559 + }, + { + "epoch": 0.5368001677500525, + "grad_norm": 5.813140867138001, + "learning_rate": 1.8859734224966145e-05, + "loss": 2.447, + "step": 2560 + }, + { + "epoch": 0.5370098553155798, + "grad_norm": 6.04901192822802, + "learning_rate": 1.885868415310849e-05, + "loss": 2.2761, + "step": 2561 + }, + { + "epoch": 0.5372195428811072, + "grad_norm": 6.045359823819872, + "learning_rate": 1.8857633627227905e-05, + "loss": 2.1769, + "step": 2562 + }, + { + "epoch": 0.5374292304466345, + "grad_norm": 5.5660647245674095, + "learning_rate": 1.8856582647378238e-05, + "loss": 2.0056, + "step": 2563 + }, + { + "epoch": 0.5376389180121619, + "grad_norm": 6.22577447568348, + "learning_rate": 1.8855531213613348e-05, + "loss": 2.0116, + "step": 2564 + }, + { + "epoch": 0.5378486055776892, + "grad_norm": 5.549439006043563, + "learning_rate": 1.885447932598712e-05, + "loss": 2.0624, + "step": 2565 + }, + { + "epoch": 0.5380582931432166, + "grad_norm": 6.066944574652426, + "learning_rate": 1.8853426984553472e-05, + "loss": 2.0386, + "step": 2566 + }, + { + "epoch": 0.5382679807087439, + "grad_norm": 6.04419168859851, + "learning_rate": 1.8852374189366336e-05, + "loss": 2.0921, + "step": 2567 + }, + { + "epoch": 0.5384776682742713, + "grad_norm": 5.606980320856114, + "learning_rate": 1.885132094047967e-05, + "loss": 2.1022, + "step": 2568 + }, + { + "epoch": 0.5386873558397987, + "grad_norm": 5.996642453563527, + "learning_rate": 1.885026723794745e-05, + "loss": 1.671, + "step": 2569 + }, + { + "epoch": 0.5388970434053261, + "grad_norm": 5.453699393885708, + "learning_rate": 1.884921308182369e-05, + "loss": 1.928, + "step": 2570 + }, + { + "epoch": 0.5391067309708534, + "grad_norm": 6.017831150741545, + "learning_rate": 1.8848158472162408e-05, + "loss": 1.9925, + "step": 2571 + }, + { + "epoch": 0.5393164185363808, + "grad_norm": 5.019441325591909, + "learning_rate": 1.8847103409017658e-05, + "loss": 1.5908, + "step": 2572 + }, + { + "epoch": 0.5395261061019082, + "grad_norm": 7.434357621169412, + "learning_rate": 1.8846047892443513e-05, + "loss": 2.2383, + "step": 2573 + }, + { + "epoch": 0.5397357936674355, + "grad_norm": 5.826370864395905, + "learning_rate": 1.884499192249407e-05, + "loss": 2.0631, + "step": 2574 + }, + { + "epoch": 0.5399454812329629, + "grad_norm": 6.2153723496133235, + "learning_rate": 1.8843935499223455e-05, + "loss": 1.7688, + "step": 2575 + }, + { + "epoch": 0.5401551687984902, + "grad_norm": 6.066741173415851, + "learning_rate": 1.8842878622685804e-05, + "loss": 1.7406, + "step": 2576 + }, + { + "epoch": 0.5403648563640177, + "grad_norm": 5.462624577483679, + "learning_rate": 1.8841821292935286e-05, + "loss": 1.9572, + "step": 2577 + }, + { + "epoch": 0.540574543929545, + "grad_norm": 5.9342953808306165, + "learning_rate": 1.884076351002609e-05, + "loss": 2.0587, + "step": 2578 + }, + { + "epoch": 0.5407842314950724, + "grad_norm": 7.314594636167027, + "learning_rate": 1.8839705274012433e-05, + "loss": 2.2103, + "step": 2579 + }, + { + "epoch": 0.5409939190605997, + "grad_norm": 6.320814660898045, + "learning_rate": 1.8838646584948548e-05, + "loss": 1.9895, + "step": 2580 + }, + { + "epoch": 0.5412036066261271, + "grad_norm": 6.3041781940040185, + "learning_rate": 1.8837587442888695e-05, + "loss": 1.9651, + "step": 2581 + }, + { + "epoch": 0.5414132941916544, + "grad_norm": 6.076140858127466, + "learning_rate": 1.8836527847887155e-05, + "loss": 2.2243, + "step": 2582 + }, + { + "epoch": 0.5416229817571818, + "grad_norm": 6.609548119353039, + "learning_rate": 1.8835467799998244e-05, + "loss": 1.7389, + "step": 2583 + }, + { + "epoch": 0.5418326693227091, + "grad_norm": 5.645032221596158, + "learning_rate": 1.883440729927628e-05, + "loss": 2.2814, + "step": 2584 + }, + { + "epoch": 0.5420423568882365, + "grad_norm": 6.265513245942741, + "learning_rate": 1.8833346345775618e-05, + "loss": 1.6286, + "step": 2585 + }, + { + "epoch": 0.5422520444537638, + "grad_norm": 6.889547940581719, + "learning_rate": 1.8832284939550635e-05, + "loss": 2.1864, + "step": 2586 + }, + { + "epoch": 0.5424617320192913, + "grad_norm": 6.046332804758267, + "learning_rate": 1.883122308065573e-05, + "loss": 1.932, + "step": 2587 + }, + { + "epoch": 0.5426714195848186, + "grad_norm": 7.528004394604743, + "learning_rate": 1.8830160769145325e-05, + "loss": 2.2459, + "step": 2588 + }, + { + "epoch": 0.542881107150346, + "grad_norm": 5.8426351049485525, + "learning_rate": 1.882909800507387e-05, + "loss": 2.312, + "step": 2589 + }, + { + "epoch": 0.5430907947158734, + "grad_norm": 6.574870114972231, + "learning_rate": 1.8828034788495825e-05, + "loss": 2.2739, + "step": 2590 + }, + { + "epoch": 0.5433004822814007, + "grad_norm": 6.476936864156679, + "learning_rate": 1.8826971119465685e-05, + "loss": 2.1414, + "step": 2591 + }, + { + "epoch": 0.5435101698469281, + "grad_norm": 6.444448301870777, + "learning_rate": 1.8825906998037968e-05, + "loss": 1.9769, + "step": 2592 + }, + { + "epoch": 0.5437198574124554, + "grad_norm": 5.944910416967087, + "learning_rate": 1.882484242426721e-05, + "loss": 2.1644, + "step": 2593 + }, + { + "epoch": 0.5439295449779828, + "grad_norm": 6.216191310254095, + "learning_rate": 1.8823777398207967e-05, + "loss": 1.9783, + "step": 2594 + }, + { + "epoch": 0.5441392325435102, + "grad_norm": 6.232777991800609, + "learning_rate": 1.8822711919914832e-05, + "loss": 2.0438, + "step": 2595 + }, + { + "epoch": 0.5443489201090376, + "grad_norm": 6.796762743384618, + "learning_rate": 1.882164598944241e-05, + "loss": 2.323, + "step": 2596 + }, + { + "epoch": 0.5445586076745649, + "grad_norm": 6.0599533602428375, + "learning_rate": 1.882057960684533e-05, + "loss": 1.8952, + "step": 2597 + }, + { + "epoch": 0.5447682952400923, + "grad_norm": 5.54785777187151, + "learning_rate": 1.8819512772178245e-05, + "loss": 1.7238, + "step": 2598 + }, + { + "epoch": 0.5449779828056196, + "grad_norm": 4.915630012202421, + "learning_rate": 1.8818445485495832e-05, + "loss": 1.8832, + "step": 2599 + }, + { + "epoch": 0.545187670371147, + "grad_norm": 7.059105383533619, + "learning_rate": 1.88173777468528e-05, + "loss": 2.1849, + "step": 2600 + }, + { + "epoch": 0.5453973579366743, + "grad_norm": 5.337910178951622, + "learning_rate": 1.881630955630386e-05, + "loss": 2.0503, + "step": 2601 + }, + { + "epoch": 0.5456070455022017, + "grad_norm": 6.126488900739763, + "learning_rate": 1.8815240913903764e-05, + "loss": 1.9927, + "step": 2602 + }, + { + "epoch": 0.545816733067729, + "grad_norm": 6.962111979280045, + "learning_rate": 1.8814171819707282e-05, + "loss": 1.5928, + "step": 2603 + }, + { + "epoch": 0.5460264206332565, + "grad_norm": 6.510655898069793, + "learning_rate": 1.881310227376921e-05, + "loss": 2.0686, + "step": 2604 + }, + { + "epoch": 0.5462361081987838, + "grad_norm": 7.140428743075681, + "learning_rate": 1.881203227614436e-05, + "loss": 2.0624, + "step": 2605 + }, + { + "epoch": 0.5464457957643112, + "grad_norm": 7.032297272042619, + "learning_rate": 1.881096182688757e-05, + "loss": 2.3407, + "step": 2606 + }, + { + "epoch": 0.5466554833298385, + "grad_norm": 6.566799320265648, + "learning_rate": 1.8809890926053705e-05, + "loss": 2.3214, + "step": 2607 + }, + { + "epoch": 0.5468651708953659, + "grad_norm": 5.705523634143453, + "learning_rate": 1.880881957369765e-05, + "loss": 1.997, + "step": 2608 + }, + { + "epoch": 0.5470748584608933, + "grad_norm": 6.197984985801809, + "learning_rate": 1.8807747769874315e-05, + "loss": 1.9006, + "step": 2609 + }, + { + "epoch": 0.5472845460264206, + "grad_norm": 6.155445026431149, + "learning_rate": 1.880667551463863e-05, + "loss": 2.1163, + "step": 2610 + }, + { + "epoch": 0.547494233591948, + "grad_norm": 6.627898538936987, + "learning_rate": 1.880560280804555e-05, + "loss": 2.2132, + "step": 2611 + }, + { + "epoch": 0.5477039211574753, + "grad_norm": 6.441122380762986, + "learning_rate": 1.8804529650150054e-05, + "loss": 2.3559, + "step": 2612 + }, + { + "epoch": 0.5479136087230028, + "grad_norm": 6.523963593210011, + "learning_rate": 1.8803456041007142e-05, + "loss": 2.2058, + "step": 2613 + }, + { + "epoch": 0.5481232962885301, + "grad_norm": 6.090534268779164, + "learning_rate": 1.880238198067184e-05, + "loss": 1.7876, + "step": 2614 + }, + { + "epoch": 0.5483329838540575, + "grad_norm": 5.70892252025577, + "learning_rate": 1.880130746919919e-05, + "loss": 2.3092, + "step": 2615 + }, + { + "epoch": 0.5485426714195848, + "grad_norm": 5.975602913938529, + "learning_rate": 1.8800232506644274e-05, + "loss": 2.0855, + "step": 2616 + }, + { + "epoch": 0.5487523589851122, + "grad_norm": 5.181952961539222, + "learning_rate": 1.879915709306218e-05, + "loss": 2.0363, + "step": 2617 + }, + { + "epoch": 0.5489620465506395, + "grad_norm": 5.957147466276341, + "learning_rate": 1.879808122850802e-05, + "loss": 2.2269, + "step": 2618 + }, + { + "epoch": 0.5491717341161669, + "grad_norm": 5.70177242393609, + "learning_rate": 1.8797004913036937e-05, + "loss": 2.1188, + "step": 2619 + }, + { + "epoch": 0.5493814216816942, + "grad_norm": 6.0469326749966354, + "learning_rate": 1.879592814670409e-05, + "loss": 2.088, + "step": 2620 + }, + { + "epoch": 0.5495911092472217, + "grad_norm": 5.603066859220233, + "learning_rate": 1.879485092956468e-05, + "loss": 2.2123, + "step": 2621 + }, + { + "epoch": 0.549800796812749, + "grad_norm": 6.684463330144917, + "learning_rate": 1.8793773261673898e-05, + "loss": 2.0154, + "step": 2622 + }, + { + "epoch": 0.5500104843782764, + "grad_norm": 6.051459721069936, + "learning_rate": 1.879269514308699e-05, + "loss": 1.9838, + "step": 2623 + }, + { + "epoch": 0.5502201719438037, + "grad_norm": 6.504314365555888, + "learning_rate": 1.87916165738592e-05, + "loss": 1.9246, + "step": 2624 + }, + { + "epoch": 0.5504298595093311, + "grad_norm": 5.335514199646772, + "learning_rate": 1.8790537554045816e-05, + "loss": 1.9697, + "step": 2625 + }, + { + "epoch": 0.5506395470748585, + "grad_norm": 6.466563052930602, + "learning_rate": 1.8789458083702134e-05, + "loss": 2.1781, + "step": 2626 + }, + { + "epoch": 0.5508492346403858, + "grad_norm": 4.991162315360214, + "learning_rate": 1.878837816288348e-05, + "loss": 1.9836, + "step": 2627 + }, + { + "epoch": 0.5510589222059132, + "grad_norm": 5.530390858435121, + "learning_rate": 1.8787297791645206e-05, + "loss": 2.1867, + "step": 2628 + }, + { + "epoch": 0.5512686097714405, + "grad_norm": 6.0951658650466936, + "learning_rate": 1.8786216970042677e-05, + "loss": 2.4288, + "step": 2629 + }, + { + "epoch": 0.551478297336968, + "grad_norm": 6.146414981998222, + "learning_rate": 1.8785135698131287e-05, + "loss": 1.9042, + "step": 2630 + }, + { + "epoch": 0.5516879849024953, + "grad_norm": 6.043900892902211, + "learning_rate": 1.878405397596646e-05, + "loss": 2.0916, + "step": 2631 + }, + { + "epoch": 0.5518976724680227, + "grad_norm": 5.493283793183004, + "learning_rate": 1.878297180360363e-05, + "loss": 1.72, + "step": 2632 + }, + { + "epoch": 0.55210736003355, + "grad_norm": 6.523742918756957, + "learning_rate": 1.878188918109826e-05, + "loss": 2.1661, + "step": 2633 + }, + { + "epoch": 0.5523170475990774, + "grad_norm": 6.268987761913348, + "learning_rate": 1.8780806108505843e-05, + "loss": 1.6766, + "step": 2634 + }, + { + "epoch": 0.5525267351646047, + "grad_norm": 7.081458621037962, + "learning_rate": 1.8779722585881877e-05, + "loss": 2.1916, + "step": 2635 + }, + { + "epoch": 0.5527364227301321, + "grad_norm": 6.985263527659814, + "learning_rate": 1.8778638613281904e-05, + "loss": 1.7591, + "step": 2636 + }, + { + "epoch": 0.5529461102956594, + "grad_norm": 6.450472955554288, + "learning_rate": 1.8777554190761478e-05, + "loss": 1.6982, + "step": 2637 + }, + { + "epoch": 0.5531557978611868, + "grad_norm": 6.544425544827743, + "learning_rate": 1.8776469318376175e-05, + "loss": 2.0519, + "step": 2638 + }, + { + "epoch": 0.5533654854267142, + "grad_norm": 6.621091158480967, + "learning_rate": 1.8775383996181595e-05, + "loss": 1.7643, + "step": 2639 + }, + { + "epoch": 0.5535751729922416, + "grad_norm": 6.6094164115710905, + "learning_rate": 1.8774298224233364e-05, + "loss": 1.9746, + "step": 2640 + }, + { + "epoch": 0.5537848605577689, + "grad_norm": 6.053812782568108, + "learning_rate": 1.8773212002587135e-05, + "loss": 2.044, + "step": 2641 + }, + { + "epoch": 0.5539945481232963, + "grad_norm": 6.983403383375147, + "learning_rate": 1.877212533129857e-05, + "loss": 2.3276, + "step": 2642 + }, + { + "epoch": 0.5542042356888236, + "grad_norm": 6.416429155049442, + "learning_rate": 1.8771038210423375e-05, + "loss": 2.0387, + "step": 2643 + }, + { + "epoch": 0.554413923254351, + "grad_norm": 7.029730292703193, + "learning_rate": 1.8769950640017253e-05, + "loss": 2.0282, + "step": 2644 + }, + { + "epoch": 0.5546236108198784, + "grad_norm": 6.577962749200691, + "learning_rate": 1.8768862620135947e-05, + "loss": 2.1805, + "step": 2645 + }, + { + "epoch": 0.5548332983854057, + "grad_norm": 6.5731343776378095, + "learning_rate": 1.876777415083523e-05, + "loss": 1.6695, + "step": 2646 + }, + { + "epoch": 0.5550429859509332, + "grad_norm": 6.467715596166901, + "learning_rate": 1.8766685232170877e-05, + "loss": 1.8594, + "step": 2647 + }, + { + "epoch": 0.5552526735164605, + "grad_norm": 6.76420042303689, + "learning_rate": 1.8765595864198702e-05, + "loss": 1.8977, + "step": 2648 + }, + { + "epoch": 0.5554623610819879, + "grad_norm": 5.59630298445021, + "learning_rate": 1.8764506046974536e-05, + "loss": 1.6031, + "step": 2649 + }, + { + "epoch": 0.5556720486475152, + "grad_norm": 5.959674279827521, + "learning_rate": 1.8763415780554232e-05, + "loss": 1.9359, + "step": 2650 + }, + { + "epoch": 0.5558817362130426, + "grad_norm": 5.9601162940686265, + "learning_rate": 1.876232506499367e-05, + "loss": 2.3092, + "step": 2651 + }, + { + "epoch": 0.5560914237785699, + "grad_norm": 5.450924639150595, + "learning_rate": 1.8761233900348755e-05, + "loss": 1.8073, + "step": 2652 + }, + { + "epoch": 0.5563011113440973, + "grad_norm": 5.793176530396519, + "learning_rate": 1.8760142286675403e-05, + "loss": 1.9406, + "step": 2653 + }, + { + "epoch": 0.5565107989096246, + "grad_norm": 5.171880563937533, + "learning_rate": 1.8759050224029564e-05, + "loss": 1.9244, + "step": 2654 + }, + { + "epoch": 0.556720486475152, + "grad_norm": 5.788681288637579, + "learning_rate": 1.8757957712467213e-05, + "loss": 2.0788, + "step": 2655 + }, + { + "epoch": 0.5569301740406793, + "grad_norm": 5.525966055569547, + "learning_rate": 1.8756864752044336e-05, + "loss": 1.9819, + "step": 2656 + }, + { + "epoch": 0.5571398616062068, + "grad_norm": 6.493016913575826, + "learning_rate": 1.8755771342816956e-05, + "loss": 2.1539, + "step": 2657 + }, + { + "epoch": 0.5573495491717341, + "grad_norm": 5.680384664958545, + "learning_rate": 1.875467748484111e-05, + "loss": 1.8154, + "step": 2658 + }, + { + "epoch": 0.5575592367372615, + "grad_norm": 7.4916988363603405, + "learning_rate": 1.8753583178172855e-05, + "loss": 2.4587, + "step": 2659 + }, + { + "epoch": 0.5577689243027888, + "grad_norm": 5.500849148239562, + "learning_rate": 1.8752488422868278e-05, + "loss": 2.1669, + "step": 2660 + }, + { + "epoch": 0.5579786118683162, + "grad_norm": 5.9440986408268355, + "learning_rate": 1.875139321898349e-05, + "loss": 2.3973, + "step": 2661 + }, + { + "epoch": 0.5581882994338436, + "grad_norm": 5.933776625062114, + "learning_rate": 1.8750297566574623e-05, + "loss": 2.1191, + "step": 2662 + }, + { + "epoch": 0.5583979869993709, + "grad_norm": 5.432377400318719, + "learning_rate": 1.874920146569783e-05, + "loss": 1.7428, + "step": 2663 + }, + { + "epoch": 0.5586076745648983, + "grad_norm": 5.324312088306347, + "learning_rate": 1.874810491640928e-05, + "loss": 2.0472, + "step": 2664 + }, + { + "epoch": 0.5588173621304257, + "grad_norm": 5.757867124769394, + "learning_rate": 1.874700791876519e-05, + "loss": 2.0221, + "step": 2665 + }, + { + "epoch": 0.5590270496959531, + "grad_norm": 5.659279041224531, + "learning_rate": 1.8745910472821767e-05, + "loss": 2.1093, + "step": 2666 + }, + { + "epoch": 0.5592367372614804, + "grad_norm": 5.8152583903364325, + "learning_rate": 1.8744812578635264e-05, + "loss": 2.3153, + "step": 2667 + }, + { + "epoch": 0.5594464248270078, + "grad_norm": 6.274909912584504, + "learning_rate": 1.874371423626195e-05, + "loss": 2.1075, + "step": 2668 + }, + { + "epoch": 0.5596561123925351, + "grad_norm": 5.7198017657466576, + "learning_rate": 1.874261544575812e-05, + "loss": 2.2077, + "step": 2669 + }, + { + "epoch": 0.5598657999580625, + "grad_norm": 6.208011647448722, + "learning_rate": 1.8741516207180082e-05, + "loss": 2.2804, + "step": 2670 + }, + { + "epoch": 0.5600754875235898, + "grad_norm": 5.1911253368763655, + "learning_rate": 1.8740416520584174e-05, + "loss": 2.1247, + "step": 2671 + }, + { + "epoch": 0.5602851750891172, + "grad_norm": 5.225813556654828, + "learning_rate": 1.8739316386026764e-05, + "loss": 1.9818, + "step": 2672 + }, + { + "epoch": 0.5604948626546445, + "grad_norm": 4.7691457650823, + "learning_rate": 1.873821580356423e-05, + "loss": 2.074, + "step": 2673 + }, + { + "epoch": 0.560704550220172, + "grad_norm": 5.344593547062895, + "learning_rate": 1.873711477325298e-05, + "loss": 1.9696, + "step": 2674 + }, + { + "epoch": 0.5609142377856993, + "grad_norm": 4.95845880405185, + "learning_rate": 1.8736013295149444e-05, + "loss": 1.9341, + "step": 2675 + }, + { + "epoch": 0.5611239253512267, + "grad_norm": 7.361642565632095, + "learning_rate": 1.8734911369310078e-05, + "loss": 2.041, + "step": 2676 + }, + { + "epoch": 0.561333612916754, + "grad_norm": 5.7560843924351035, + "learning_rate": 1.873380899579135e-05, + "loss": 2.0045, + "step": 2677 + }, + { + "epoch": 0.5615433004822814, + "grad_norm": 5.698414140311639, + "learning_rate": 1.8732706174649765e-05, + "loss": 2.3569, + "step": 2678 + }, + { + "epoch": 0.5617529880478087, + "grad_norm": 5.638760562710624, + "learning_rate": 1.8731602905941843e-05, + "loss": 2.0654, + "step": 2679 + }, + { + "epoch": 0.5619626756133361, + "grad_norm": 5.543907889938176, + "learning_rate": 1.8730499189724125e-05, + "loss": 2.0676, + "step": 2680 + }, + { + "epoch": 0.5621723631788635, + "grad_norm": 5.093171829741121, + "learning_rate": 1.8729395026053185e-05, + "loss": 2.0148, + "step": 2681 + }, + { + "epoch": 0.5623820507443908, + "grad_norm": 6.602405268014984, + "learning_rate": 1.872829041498561e-05, + "loss": 1.8978, + "step": 2682 + }, + { + "epoch": 0.5625917383099183, + "grad_norm": 4.917211262729164, + "learning_rate": 1.872718535657801e-05, + "loss": 2.2525, + "step": 2683 + }, + { + "epoch": 0.5628014258754456, + "grad_norm": 6.20822782818195, + "learning_rate": 1.872607985088702e-05, + "loss": 2.2741, + "step": 2684 + }, + { + "epoch": 0.563011113440973, + "grad_norm": 6.286655280455439, + "learning_rate": 1.8724973897969308e-05, + "loss": 2.0052, + "step": 2685 + }, + { + "epoch": 0.5632208010065003, + "grad_norm": 5.087714316119428, + "learning_rate": 1.8723867497881553e-05, + "loss": 1.9857, + "step": 2686 + }, + { + "epoch": 0.5634304885720277, + "grad_norm": 5.203714422900515, + "learning_rate": 1.8722760650680454e-05, + "loss": 2.0874, + "step": 2687 + }, + { + "epoch": 0.563640176137555, + "grad_norm": 6.376718930125289, + "learning_rate": 1.8721653356422745e-05, + "loss": 2.3018, + "step": 2688 + }, + { + "epoch": 0.5638498637030824, + "grad_norm": 5.150186999573697, + "learning_rate": 1.8720545615165174e-05, + "loss": 2.0346, + "step": 2689 + }, + { + "epoch": 0.5640595512686097, + "grad_norm": 5.9658279364258195, + "learning_rate": 1.8719437426964515e-05, + "loss": 2.1122, + "step": 2690 + }, + { + "epoch": 0.5642692388341372, + "grad_norm": 5.787366077157618, + "learning_rate": 1.8718328791877563e-05, + "loss": 1.8914, + "step": 2691 + }, + { + "epoch": 0.5644789263996645, + "grad_norm": 6.903315531310373, + "learning_rate": 1.8717219709961142e-05, + "loss": 2.3797, + "step": 2692 + }, + { + "epoch": 0.5646886139651919, + "grad_norm": 5.767717891657005, + "learning_rate": 1.8716110181272094e-05, + "loss": 1.9054, + "step": 2693 + }, + { + "epoch": 0.5648983015307192, + "grad_norm": 6.021351567504202, + "learning_rate": 1.8715000205867278e-05, + "loss": 2.0147, + "step": 2694 + }, + { + "epoch": 0.5651079890962466, + "grad_norm": 6.017687709569518, + "learning_rate": 1.8713889783803588e-05, + "loss": 2.1029, + "step": 2695 + }, + { + "epoch": 0.5653176766617739, + "grad_norm": 5.446117259902237, + "learning_rate": 1.8712778915137928e-05, + "loss": 1.9414, + "step": 2696 + }, + { + "epoch": 0.5655273642273013, + "grad_norm": 5.298011868684031, + "learning_rate": 1.8711667599927242e-05, + "loss": 1.9641, + "step": 2697 + }, + { + "epoch": 0.5657370517928287, + "grad_norm": 6.206091165472913, + "learning_rate": 1.871055583822848e-05, + "loss": 1.9994, + "step": 2698 + }, + { + "epoch": 0.565946739358356, + "grad_norm": 6.092905590961644, + "learning_rate": 1.8709443630098626e-05, + "loss": 2.0255, + "step": 2699 + }, + { + "epoch": 0.5661564269238835, + "grad_norm": 5.995637053235547, + "learning_rate": 1.8708330975594677e-05, + "loss": 2.1665, + "step": 2700 + }, + { + "epoch": 0.5663661144894108, + "grad_norm": 6.671332802115261, + "learning_rate": 1.8707217874773662e-05, + "loss": 2.1807, + "step": 2701 + }, + { + "epoch": 0.5665758020549382, + "grad_norm": 6.7459237712878215, + "learning_rate": 1.8706104327692634e-05, + "loss": 2.1185, + "step": 2702 + }, + { + "epoch": 0.5667854896204655, + "grad_norm": 7.982511796624487, + "learning_rate": 1.8704990334408656e-05, + "loss": 1.9709, + "step": 2703 + }, + { + "epoch": 0.5669951771859929, + "grad_norm": 6.046707582558813, + "learning_rate": 1.870387589497882e-05, + "loss": 2.0603, + "step": 2704 + }, + { + "epoch": 0.5672048647515202, + "grad_norm": 6.0054835322872915, + "learning_rate": 1.8702761009460257e-05, + "loss": 2.2664, + "step": 2705 + }, + { + "epoch": 0.5674145523170476, + "grad_norm": 6.971973131413373, + "learning_rate": 1.870164567791009e-05, + "loss": 1.9682, + "step": 2706 + }, + { + "epoch": 0.5676242398825749, + "grad_norm": 5.456181558165276, + "learning_rate": 1.8700529900385497e-05, + "loss": 1.8005, + "step": 2707 + }, + { + "epoch": 0.5678339274481023, + "grad_norm": 6.86761808320295, + "learning_rate": 1.869941367694365e-05, + "loss": 2.3656, + "step": 2708 + }, + { + "epoch": 0.5680436150136297, + "grad_norm": 6.944595423798982, + "learning_rate": 1.8698297007641767e-05, + "loss": 1.8246, + "step": 2709 + }, + { + "epoch": 0.5682533025791571, + "grad_norm": 6.0645727191871694, + "learning_rate": 1.869717989253708e-05, + "loss": 1.8002, + "step": 2710 + }, + { + "epoch": 0.5684629901446844, + "grad_norm": 6.188568172044652, + "learning_rate": 1.869606233168683e-05, + "loss": 2.36, + "step": 2711 + }, + { + "epoch": 0.5686726777102118, + "grad_norm": 5.891868649907786, + "learning_rate": 1.869494432514831e-05, + "loss": 2.1397, + "step": 2712 + }, + { + "epoch": 0.5688823652757391, + "grad_norm": 6.347512614165626, + "learning_rate": 1.869382587297881e-05, + "loss": 2.0811, + "step": 2713 + }, + { + "epoch": 0.5690920528412665, + "grad_norm": 6.150558442739384, + "learning_rate": 1.8692706975235657e-05, + "loss": 1.8356, + "step": 2714 + }, + { + "epoch": 0.5693017404067939, + "grad_norm": 6.415217463317695, + "learning_rate": 1.8691587631976195e-05, + "loss": 2.3428, + "step": 2715 + }, + { + "epoch": 0.5695114279723212, + "grad_norm": 6.121405345262252, + "learning_rate": 1.869046784325779e-05, + "loss": 2.142, + "step": 2716 + }, + { + "epoch": 0.5697211155378487, + "grad_norm": 7.433282054083341, + "learning_rate": 1.8689347609137834e-05, + "loss": 2.3011, + "step": 2717 + }, + { + "epoch": 0.569930803103376, + "grad_norm": 6.47311274821471, + "learning_rate": 1.8688226929673747e-05, + "loss": 2.0019, + "step": 2718 + }, + { + "epoch": 0.5701404906689034, + "grad_norm": 6.45377360232322, + "learning_rate": 1.8687105804922955e-05, + "loss": 2.227, + "step": 2719 + }, + { + "epoch": 0.5703501782344307, + "grad_norm": 6.047156282100101, + "learning_rate": 1.8685984234942928e-05, + "loss": 1.9088, + "step": 2720 + }, + { + "epoch": 0.5705598657999581, + "grad_norm": 5.315206635865982, + "learning_rate": 1.8684862219791142e-05, + "loss": 1.9434, + "step": 2721 + }, + { + "epoch": 0.5707695533654854, + "grad_norm": 6.627961023081533, + "learning_rate": 1.8683739759525105e-05, + "loss": 2.1496, + "step": 2722 + }, + { + "epoch": 0.5709792409310128, + "grad_norm": 5.760285731253, + "learning_rate": 1.868261685420234e-05, + "loss": 1.6827, + "step": 2723 + }, + { + "epoch": 0.5711889284965401, + "grad_norm": 6.598516163423714, + "learning_rate": 1.8681493503880406e-05, + "loss": 1.9888, + "step": 2724 + }, + { + "epoch": 0.5713986160620675, + "grad_norm": 6.230349513261235, + "learning_rate": 1.868036970861687e-05, + "loss": 2.2908, + "step": 2725 + }, + { + "epoch": 0.5716083036275948, + "grad_norm": 6.505106907214323, + "learning_rate": 1.8679245468469334e-05, + "loss": 2.1056, + "step": 2726 + }, + { + "epoch": 0.5718179911931223, + "grad_norm": 6.079516966081716, + "learning_rate": 1.8678120783495415e-05, + "loss": 1.8275, + "step": 2727 + }, + { + "epoch": 0.5720276787586496, + "grad_norm": 7.022574927643935, + "learning_rate": 1.8676995653752755e-05, + "loss": 2.0554, + "step": 2728 + }, + { + "epoch": 0.572237366324177, + "grad_norm": 6.411236972209665, + "learning_rate": 1.8675870079299014e-05, + "loss": 1.7686, + "step": 2729 + }, + { + "epoch": 0.5724470538897043, + "grad_norm": 5.677846621554343, + "learning_rate": 1.867474406019189e-05, + "loss": 2.1153, + "step": 2730 + }, + { + "epoch": 0.5726567414552317, + "grad_norm": 6.250652107104516, + "learning_rate": 1.8673617596489082e-05, + "loss": 2.049, + "step": 2731 + }, + { + "epoch": 0.572866429020759, + "grad_norm": 5.549938127805381, + "learning_rate": 1.8672490688248328e-05, + "loss": 2.1606, + "step": 2732 + }, + { + "epoch": 0.5730761165862864, + "grad_norm": 5.665935603894634, + "learning_rate": 1.8671363335527387e-05, + "loss": 1.7843, + "step": 2733 + }, + { + "epoch": 0.5732858041518138, + "grad_norm": 6.207191963647915, + "learning_rate": 1.8670235538384033e-05, + "loss": 2.1177, + "step": 2734 + }, + { + "epoch": 0.5734954917173412, + "grad_norm": 7.1470663510588555, + "learning_rate": 1.8669107296876073e-05, + "loss": 2.1896, + "step": 2735 + }, + { + "epoch": 0.5737051792828686, + "grad_norm": 6.191036326130498, + "learning_rate": 1.866797861106133e-05, + "loss": 1.9805, + "step": 2736 + }, + { + "epoch": 0.5739148668483959, + "grad_norm": 7.008784218917093, + "learning_rate": 1.8666849480997644e-05, + "loss": 1.9644, + "step": 2737 + }, + { + "epoch": 0.5741245544139233, + "grad_norm": 6.68268645709309, + "learning_rate": 1.8665719906742893e-05, + "loss": 2.2133, + "step": 2738 + }, + { + "epoch": 0.5743342419794506, + "grad_norm": 5.9350242021380994, + "learning_rate": 1.8664589888354966e-05, + "loss": 2.1331, + "step": 2739 + }, + { + "epoch": 0.574543929544978, + "grad_norm": 5.75101298241169, + "learning_rate": 1.866345942589178e-05, + "loss": 1.9286, + "step": 2740 + }, + { + "epoch": 0.5747536171105053, + "grad_norm": 7.210009665016464, + "learning_rate": 1.866232851941127e-05, + "loss": 2.0464, + "step": 2741 + }, + { + "epoch": 0.5749633046760327, + "grad_norm": 5.2990713313272595, + "learning_rate": 1.8661197168971403e-05, + "loss": 1.9399, + "step": 2742 + }, + { + "epoch": 0.57517299224156, + "grad_norm": 7.4849935933744955, + "learning_rate": 1.8660065374630155e-05, + "loss": 1.6694, + "step": 2743 + }, + { + "epoch": 0.5753826798070875, + "grad_norm": 6.914434132044269, + "learning_rate": 1.865893313644554e-05, + "loss": 1.849, + "step": 2744 + }, + { + "epoch": 0.5755923673726148, + "grad_norm": 6.278163681295618, + "learning_rate": 1.865780045447558e-05, + "loss": 2.2206, + "step": 2745 + }, + { + "epoch": 0.5758020549381422, + "grad_norm": 5.854642415445766, + "learning_rate": 1.8656667328778336e-05, + "loss": 1.9686, + "step": 2746 + }, + { + "epoch": 0.5760117425036695, + "grad_norm": 6.034483674605276, + "learning_rate": 1.8655533759411876e-05, + "loss": 1.9169, + "step": 2747 + }, + { + "epoch": 0.5762214300691969, + "grad_norm": 5.828105447086947, + "learning_rate": 1.8654399746434296e-05, + "loss": 1.8914, + "step": 2748 + }, + { + "epoch": 0.5764311176347242, + "grad_norm": 7.732193301976145, + "learning_rate": 1.865326528990372e-05, + "loss": 1.561, + "step": 2749 + }, + { + "epoch": 0.5766408052002516, + "grad_norm": 6.977330496882968, + "learning_rate": 1.865213038987829e-05, + "loss": 2.0952, + "step": 2750 + }, + { + "epoch": 0.576850492765779, + "grad_norm": 6.4732742621727475, + "learning_rate": 1.8650995046416167e-05, + "loss": 2.094, + "step": 2751 + }, + { + "epoch": 0.5770601803313063, + "grad_norm": 6.17100823584381, + "learning_rate": 1.8649859259575548e-05, + "loss": 2.1014, + "step": 2752 + }, + { + "epoch": 0.5772698678968338, + "grad_norm": 8.400581570556966, + "learning_rate": 1.8648723029414638e-05, + "loss": 2.239, + "step": 2753 + }, + { + "epoch": 0.5774795554623611, + "grad_norm": 6.358691614627974, + "learning_rate": 1.864758635599167e-05, + "loss": 1.9169, + "step": 2754 + }, + { + "epoch": 0.5776892430278885, + "grad_norm": 6.625139041557192, + "learning_rate": 1.8646449239364906e-05, + "loss": 2.4451, + "step": 2755 + }, + { + "epoch": 0.5778989305934158, + "grad_norm": 5.396835085962694, + "learning_rate": 1.8645311679592622e-05, + "loss": 1.8255, + "step": 2756 + }, + { + "epoch": 0.5781086181589432, + "grad_norm": 6.612106918883937, + "learning_rate": 1.864417367673312e-05, + "loss": 1.9214, + "step": 2757 + }, + { + "epoch": 0.5783183057244705, + "grad_norm": 6.605480488474029, + "learning_rate": 1.8643035230844725e-05, + "loss": 2.1115, + "step": 2758 + }, + { + "epoch": 0.5785279932899979, + "grad_norm": 6.573471124856382, + "learning_rate": 1.864189634198578e-05, + "loss": 2.1502, + "step": 2759 + }, + { + "epoch": 0.5787376808555252, + "grad_norm": 6.23270152363272, + "learning_rate": 1.864075701021466e-05, + "loss": 2.152, + "step": 2760 + }, + { + "epoch": 0.5789473684210527, + "grad_norm": 5.4339113024414365, + "learning_rate": 1.863961723558976e-05, + "loss": 1.9662, + "step": 2761 + }, + { + "epoch": 0.57915705598658, + "grad_norm": 6.039031895674701, + "learning_rate": 1.863847701816949e-05, + "loss": 1.9518, + "step": 2762 + }, + { + "epoch": 0.5793667435521074, + "grad_norm": 5.2411896271802805, + "learning_rate": 1.863733635801229e-05, + "loss": 2.0189, + "step": 2763 + }, + { + "epoch": 0.5795764311176347, + "grad_norm": 5.4320705186084774, + "learning_rate": 1.863619525517662e-05, + "loss": 2.0442, + "step": 2764 + }, + { + "epoch": 0.5797861186831621, + "grad_norm": 6.40665092609313, + "learning_rate": 1.8635053709720963e-05, + "loss": 1.9739, + "step": 2765 + }, + { + "epoch": 0.5799958062486894, + "grad_norm": 6.7146461245807085, + "learning_rate": 1.8633911721703833e-05, + "loss": 2.2687, + "step": 2766 + }, + { + "epoch": 0.5802054938142168, + "grad_norm": 5.53533748368371, + "learning_rate": 1.8632769291183748e-05, + "loss": 1.8952, + "step": 2767 + }, + { + "epoch": 0.5804151813797441, + "grad_norm": 6.197429178287639, + "learning_rate": 1.8631626418219262e-05, + "loss": 2.1775, + "step": 2768 + }, + { + "epoch": 0.5806248689452715, + "grad_norm": 5.6572588882299755, + "learning_rate": 1.8630483102868954e-05, + "loss": 1.7355, + "step": 2769 + }, + { + "epoch": 0.580834556510799, + "grad_norm": 6.667840159438343, + "learning_rate": 1.862933934519142e-05, + "loss": 1.7145, + "step": 2770 + }, + { + "epoch": 0.5810442440763263, + "grad_norm": 5.437844620737426, + "learning_rate": 1.8628195145245275e-05, + "loss": 2.077, + "step": 2771 + }, + { + "epoch": 0.5812539316418537, + "grad_norm": 6.088118114755378, + "learning_rate": 1.8627050503089164e-05, + "loss": 1.872, + "step": 2772 + }, + { + "epoch": 0.581463619207381, + "grad_norm": 6.728400677888094, + "learning_rate": 1.8625905418781754e-05, + "loss": 2.4107, + "step": 2773 + }, + { + "epoch": 0.5816733067729084, + "grad_norm": 6.166774301311825, + "learning_rate": 1.862475989238173e-05, + "loss": 2.1215, + "step": 2774 + }, + { + "epoch": 0.5818829943384357, + "grad_norm": 6.191949331219909, + "learning_rate": 1.8623613923947802e-05, + "loss": 1.9648, + "step": 2775 + }, + { + "epoch": 0.5820926819039631, + "grad_norm": 6.378446343983823, + "learning_rate": 1.86224675135387e-05, + "loss": 2.0417, + "step": 2776 + }, + { + "epoch": 0.5823023694694904, + "grad_norm": 6.2375076055480365, + "learning_rate": 1.8621320661213186e-05, + "loss": 1.8335, + "step": 2777 + }, + { + "epoch": 0.5825120570350178, + "grad_norm": 5.701035106367095, + "learning_rate": 1.8620173367030036e-05, + "loss": 1.9553, + "step": 2778 + }, + { + "epoch": 0.5827217446005452, + "grad_norm": 6.091961221841586, + "learning_rate": 1.8619025631048047e-05, + "loss": 2.0834, + "step": 2779 + }, + { + "epoch": 0.5829314321660726, + "grad_norm": 6.575829598696737, + "learning_rate": 1.861787745332605e-05, + "loss": 1.9819, + "step": 2780 + }, + { + "epoch": 0.5831411197315999, + "grad_norm": 7.304348619015532, + "learning_rate": 1.8616728833922884e-05, + "loss": 2.0994, + "step": 2781 + }, + { + "epoch": 0.5833508072971273, + "grad_norm": 6.937064025578596, + "learning_rate": 1.861557977289742e-05, + "loss": 1.9736, + "step": 2782 + }, + { + "epoch": 0.5835604948626546, + "grad_norm": 5.303755529620455, + "learning_rate": 1.8614430270308554e-05, + "loss": 1.6312, + "step": 2783 + }, + { + "epoch": 0.583770182428182, + "grad_norm": 7.529976332468666, + "learning_rate": 1.8613280326215192e-05, + "loss": 1.9495, + "step": 2784 + }, + { + "epoch": 0.5839798699937093, + "grad_norm": 6.279715443093613, + "learning_rate": 1.8612129940676272e-05, + "loss": 1.8962, + "step": 2785 + }, + { + "epoch": 0.5841895575592367, + "grad_norm": 6.675047264909567, + "learning_rate": 1.861097911375076e-05, + "loss": 2.1242, + "step": 2786 + }, + { + "epoch": 0.5843992451247642, + "grad_norm": 6.348626906787265, + "learning_rate": 1.860982784549763e-05, + "loss": 1.8807, + "step": 2787 + }, + { + "epoch": 0.5846089326902915, + "grad_norm": 6.348096122172748, + "learning_rate": 1.860867613597589e-05, + "loss": 1.733, + "step": 2788 + }, + { + "epoch": 0.5848186202558189, + "grad_norm": 5.942722257261544, + "learning_rate": 1.8607523985244567e-05, + "loss": 2.2468, + "step": 2789 + }, + { + "epoch": 0.5850283078213462, + "grad_norm": 7.243354422238376, + "learning_rate": 1.8606371393362713e-05, + "loss": 2.2528, + "step": 2790 + }, + { + "epoch": 0.5852379953868736, + "grad_norm": 5.524743898035511, + "learning_rate": 1.86052183603894e-05, + "loss": 2.0669, + "step": 2791 + }, + { + "epoch": 0.5854476829524009, + "grad_norm": 5.437122265737213, + "learning_rate": 1.8604064886383718e-05, + "loss": 2.201, + "step": 2792 + }, + { + "epoch": 0.5856573705179283, + "grad_norm": 6.073973431340435, + "learning_rate": 1.8602910971404786e-05, + "loss": 2.2775, + "step": 2793 + }, + { + "epoch": 0.5858670580834556, + "grad_norm": 5.4748252490280365, + "learning_rate": 1.860175661551175e-05, + "loss": 1.9526, + "step": 2794 + }, + { + "epoch": 0.586076745648983, + "grad_norm": 5.889596274557162, + "learning_rate": 1.8600601818763764e-05, + "loss": 1.9138, + "step": 2795 + }, + { + "epoch": 0.5862864332145103, + "grad_norm": 5.818773399302393, + "learning_rate": 1.8599446581220018e-05, + "loss": 1.8794, + "step": 2796 + }, + { + "epoch": 0.5864961207800378, + "grad_norm": 5.126087766732159, + "learning_rate": 1.8598290902939724e-05, + "loss": 2.1673, + "step": 2797 + }, + { + "epoch": 0.5867058083455651, + "grad_norm": 5.9354522772227085, + "learning_rate": 1.8597134783982103e-05, + "loss": 2.4213, + "step": 2798 + }, + { + "epoch": 0.5869154959110925, + "grad_norm": 5.5948818181582896, + "learning_rate": 1.859597822440642e-05, + "loss": 1.9696, + "step": 2799 + }, + { + "epoch": 0.5871251834766198, + "grad_norm": 5.833858255715922, + "learning_rate": 1.8594821224271937e-05, + "loss": 2.1716, + "step": 2800 + }, + { + "epoch": 0.5873348710421472, + "grad_norm": 5.908323227946351, + "learning_rate": 1.8593663783637962e-05, + "loss": 2.3965, + "step": 2801 + }, + { + "epoch": 0.5875445586076745, + "grad_norm": 5.64316681768169, + "learning_rate": 1.8592505902563816e-05, + "loss": 1.8461, + "step": 2802 + }, + { + "epoch": 0.5877542461732019, + "grad_norm": 6.797872025434759, + "learning_rate": 1.8591347581108835e-05, + "loss": 1.9383, + "step": 2803 + }, + { + "epoch": 0.5879639337387292, + "grad_norm": 5.322537800890154, + "learning_rate": 1.8590188819332392e-05, + "loss": 1.992, + "step": 2804 + }, + { + "epoch": 0.5881736213042567, + "grad_norm": 5.524221411534303, + "learning_rate": 1.8589029617293875e-05, + "loss": 2.1636, + "step": 2805 + }, + { + "epoch": 0.5883833088697841, + "grad_norm": 7.135498871497982, + "learning_rate": 1.858786997505269e-05, + "loss": 1.9046, + "step": 2806 + }, + { + "epoch": 0.5885929964353114, + "grad_norm": 5.950978899457029, + "learning_rate": 1.8586709892668274e-05, + "loss": 2.0656, + "step": 2807 + }, + { + "epoch": 0.5888026840008388, + "grad_norm": 6.473838064300838, + "learning_rate": 1.858554937020008e-05, + "loss": 2.0665, + "step": 2808 + }, + { + "epoch": 0.5890123715663661, + "grad_norm": 5.800764570950762, + "learning_rate": 1.8584388407707598e-05, + "loss": 1.9313, + "step": 2809 + }, + { + "epoch": 0.5892220591318935, + "grad_norm": 6.320039863800695, + "learning_rate": 1.8583227005250316e-05, + "loss": 2.2284, + "step": 2810 + }, + { + "epoch": 0.5894317466974208, + "grad_norm": 5.523034828041235, + "learning_rate": 1.8582065162887763e-05, + "loss": 2.3193, + "step": 2811 + }, + { + "epoch": 0.5896414342629482, + "grad_norm": 5.723140928424983, + "learning_rate": 1.8580902880679487e-05, + "loss": 1.9274, + "step": 2812 + }, + { + "epoch": 0.5898511218284755, + "grad_norm": 6.629506154808888, + "learning_rate": 1.8579740158685057e-05, + "loss": 2.282, + "step": 2813 + }, + { + "epoch": 0.590060809394003, + "grad_norm": 6.210511427186196, + "learning_rate": 1.857857699696406e-05, + "loss": 2.0276, + "step": 2814 + }, + { + "epoch": 0.5902704969595303, + "grad_norm": 5.767235434414866, + "learning_rate": 1.8577413395576114e-05, + "loss": 1.9594, + "step": 2815 + }, + { + "epoch": 0.5904801845250577, + "grad_norm": 6.425811354731852, + "learning_rate": 1.8576249354580857e-05, + "loss": 2.0, + "step": 2816 + }, + { + "epoch": 0.590689872090585, + "grad_norm": 7.252463844689402, + "learning_rate": 1.8575084874037944e-05, + "loss": 2.2094, + "step": 2817 + }, + { + "epoch": 0.5908995596561124, + "grad_norm": 5.997443830607522, + "learning_rate": 1.857391995400706e-05, + "loss": 2.1512, + "step": 2818 + }, + { + "epoch": 0.5911092472216397, + "grad_norm": 5.73839274685035, + "learning_rate": 1.8572754594547903e-05, + "loss": 2.1362, + "step": 2819 + }, + { + "epoch": 0.5913189347871671, + "grad_norm": 6.08102002355284, + "learning_rate": 1.8571588795720207e-05, + "loss": 2.1546, + "step": 2820 + }, + { + "epoch": 0.5915286223526944, + "grad_norm": 6.564845547130511, + "learning_rate": 1.857042255758372e-05, + "loss": 2.0197, + "step": 2821 + }, + { + "epoch": 0.5917383099182218, + "grad_norm": 6.253006059024423, + "learning_rate": 1.8569255880198213e-05, + "loss": 2.0026, + "step": 2822 + }, + { + "epoch": 0.5919479974837493, + "grad_norm": 6.043827078687982, + "learning_rate": 1.8568088763623478e-05, + "loss": 1.9969, + "step": 2823 + }, + { + "epoch": 0.5921576850492766, + "grad_norm": 6.426688536911274, + "learning_rate": 1.8566921207919332e-05, + "loss": 2.1987, + "step": 2824 + }, + { + "epoch": 0.592367372614804, + "grad_norm": 5.880424982159923, + "learning_rate": 1.8565753213145615e-05, + "loss": 2.304, + "step": 2825 + }, + { + "epoch": 0.5925770601803313, + "grad_norm": 6.269099861378297, + "learning_rate": 1.8564584779362192e-05, + "loss": 2.2299, + "step": 2826 + }, + { + "epoch": 0.5927867477458587, + "grad_norm": 5.493781815782301, + "learning_rate": 1.8563415906628942e-05, + "loss": 1.9815, + "step": 2827 + }, + { + "epoch": 0.592996435311386, + "grad_norm": 6.482903463790162, + "learning_rate": 1.8562246595005774e-05, + "loss": 1.9065, + "step": 2828 + }, + { + "epoch": 0.5932061228769134, + "grad_norm": 5.4657120503943935, + "learning_rate": 1.856107684455262e-05, + "loss": 2.0115, + "step": 2829 + }, + { + "epoch": 0.5934158104424407, + "grad_norm": 5.881506615973131, + "learning_rate": 1.855990665532943e-05, + "loss": 1.8569, + "step": 2830 + }, + { + "epoch": 0.5936254980079682, + "grad_norm": 6.1975118592375, + "learning_rate": 1.8558736027396177e-05, + "loss": 2.2694, + "step": 2831 + }, + { + "epoch": 0.5938351855734955, + "grad_norm": 5.280387541626495, + "learning_rate": 1.8557564960812855e-05, + "loss": 1.8424, + "step": 2832 + }, + { + "epoch": 0.5940448731390229, + "grad_norm": 5.994526462094391, + "learning_rate": 1.855639345563949e-05, + "loss": 1.9947, + "step": 2833 + }, + { + "epoch": 0.5942545607045502, + "grad_norm": 5.4735144671289175, + "learning_rate": 1.8555221511936118e-05, + "loss": 2.0521, + "step": 2834 + }, + { + "epoch": 0.5944642482700776, + "grad_norm": 5.754751226434579, + "learning_rate": 1.8554049129762804e-05, + "loss": 1.7855, + "step": 2835 + }, + { + "epoch": 0.5946739358356049, + "grad_norm": 5.532918599236389, + "learning_rate": 1.855287630917964e-05, + "loss": 2.0404, + "step": 2836 + }, + { + "epoch": 0.5948836234011323, + "grad_norm": 5.220453338413051, + "learning_rate": 1.855170305024673e-05, + "loss": 1.8221, + "step": 2837 + }, + { + "epoch": 0.5950933109666596, + "grad_norm": 5.7450617480030655, + "learning_rate": 1.8550529353024204e-05, + "loss": 2.0589, + "step": 2838 + }, + { + "epoch": 0.595302998532187, + "grad_norm": 6.363323676364217, + "learning_rate": 1.854935521757222e-05, + "loss": 2.1924, + "step": 2839 + }, + { + "epoch": 0.5955126860977145, + "grad_norm": 6.764205381451113, + "learning_rate": 1.8548180643950955e-05, + "loss": 1.9919, + "step": 2840 + }, + { + "epoch": 0.5957223736632418, + "grad_norm": 6.635936417589668, + "learning_rate": 1.8547005632220602e-05, + "loss": 1.9, + "step": 2841 + }, + { + "epoch": 0.5959320612287692, + "grad_norm": 5.82680697197393, + "learning_rate": 1.8545830182441386e-05, + "loss": 2.2936, + "step": 2842 + }, + { + "epoch": 0.5961417487942965, + "grad_norm": 6.35022104767472, + "learning_rate": 1.8544654294673553e-05, + "loss": 2.0621, + "step": 2843 + }, + { + "epoch": 0.5963514363598239, + "grad_norm": 7.036026652989272, + "learning_rate": 1.854347796897737e-05, + "loss": 2.1278, + "step": 2844 + }, + { + "epoch": 0.5965611239253512, + "grad_norm": 5.847140920811413, + "learning_rate": 1.854230120541312e-05, + "loss": 1.8412, + "step": 2845 + }, + { + "epoch": 0.5967708114908786, + "grad_norm": 5.972632852655691, + "learning_rate": 1.854112400404112e-05, + "loss": 2.2266, + "step": 2846 + }, + { + "epoch": 0.5969804990564059, + "grad_norm": 5.2835076907116845, + "learning_rate": 1.85399463649217e-05, + "loss": 1.6403, + "step": 2847 + }, + { + "epoch": 0.5971901866219333, + "grad_norm": 6.784861006432264, + "learning_rate": 1.8538768288115213e-05, + "loss": 1.8356, + "step": 2848 + }, + { + "epoch": 0.5973998741874607, + "grad_norm": 6.129010218309208, + "learning_rate": 1.8537589773682046e-05, + "loss": 1.6441, + "step": 2849 + }, + { + "epoch": 0.5976095617529881, + "grad_norm": 6.100929593146235, + "learning_rate": 1.8536410821682594e-05, + "loss": 1.9326, + "step": 2850 + }, + { + "epoch": 0.5978192493185154, + "grad_norm": 5.885765651449141, + "learning_rate": 1.853523143217728e-05, + "loss": 2.0494, + "step": 2851 + }, + { + "epoch": 0.5980289368840428, + "grad_norm": 5.801941437890225, + "learning_rate": 1.8534051605226554e-05, + "loss": 1.9244, + "step": 2852 + }, + { + "epoch": 0.5982386244495701, + "grad_norm": 6.280821335752855, + "learning_rate": 1.853287134089088e-05, + "loss": 2.4128, + "step": 2853 + }, + { + "epoch": 0.5984483120150975, + "grad_norm": 6.000581315497626, + "learning_rate": 1.8531690639230754e-05, + "loss": 1.9193, + "step": 2854 + }, + { + "epoch": 0.5986579995806248, + "grad_norm": 6.420621630717793, + "learning_rate": 1.853050950030668e-05, + "loss": 2.0262, + "step": 2855 + }, + { + "epoch": 0.5988676871461522, + "grad_norm": 7.9537214163745915, + "learning_rate": 1.8529327924179203e-05, + "loss": 2.1248, + "step": 2856 + }, + { + "epoch": 0.5990773747116795, + "grad_norm": 7.0640750717408975, + "learning_rate": 1.8528145910908874e-05, + "loss": 1.679, + "step": 2857 + }, + { + "epoch": 0.599287062277207, + "grad_norm": 5.370730147105455, + "learning_rate": 1.852696346055628e-05, + "loss": 1.8568, + "step": 2858 + }, + { + "epoch": 0.5994967498427344, + "grad_norm": 5.685459794465621, + "learning_rate": 1.852578057318201e-05, + "loss": 2.1262, + "step": 2859 + }, + { + "epoch": 0.5997064374082617, + "grad_norm": 6.3106112522630236, + "learning_rate": 1.8524597248846707e-05, + "loss": 2.0098, + "step": 2860 + }, + { + "epoch": 0.5999161249737891, + "grad_norm": 6.074111782677629, + "learning_rate": 1.8523413487611005e-05, + "loss": 2.224, + "step": 2861 + }, + { + "epoch": 0.6001258125393164, + "grad_norm": 6.143776159094985, + "learning_rate": 1.8522229289535583e-05, + "loss": 1.6457, + "step": 2862 + }, + { + "epoch": 0.6003355001048438, + "grad_norm": 7.949057583030364, + "learning_rate": 1.8521044654681124e-05, + "loss": 1.8428, + "step": 2863 + }, + { + "epoch": 0.6005451876703711, + "grad_norm": 7.86701714819926, + "learning_rate": 1.8519859583108353e-05, + "loss": 2.2839, + "step": 2864 + }, + { + "epoch": 0.6007548752358985, + "grad_norm": 6.391031902563666, + "learning_rate": 1.8518674074877995e-05, + "loss": 1.9156, + "step": 2865 + }, + { + "epoch": 0.6009645628014259, + "grad_norm": 6.683752131355188, + "learning_rate": 1.8517488130050824e-05, + "loss": 1.7305, + "step": 2866 + }, + { + "epoch": 0.6011742503669533, + "grad_norm": 6.644589719481444, + "learning_rate": 1.8516301748687608e-05, + "loss": 2.104, + "step": 2867 + }, + { + "epoch": 0.6013839379324806, + "grad_norm": 5.623284751753787, + "learning_rate": 1.8515114930849158e-05, + "loss": 1.7125, + "step": 2868 + }, + { + "epoch": 0.601593625498008, + "grad_norm": 6.229321658809429, + "learning_rate": 1.8513927676596294e-05, + "loss": 2.1369, + "step": 2869 + }, + { + "epoch": 0.6018033130635353, + "grad_norm": 6.357073562502818, + "learning_rate": 1.8512739985989874e-05, + "loss": 1.9917, + "step": 2870 + }, + { + "epoch": 0.6020130006290627, + "grad_norm": 6.526375644843321, + "learning_rate": 1.8511551859090766e-05, + "loss": 2.0576, + "step": 2871 + }, + { + "epoch": 0.60222268819459, + "grad_norm": 6.229562658983859, + "learning_rate": 1.8510363295959865e-05, + "loss": 2.0787, + "step": 2872 + }, + { + "epoch": 0.6024323757601174, + "grad_norm": 6.152929085028943, + "learning_rate": 1.8509174296658078e-05, + "loss": 1.9562, + "step": 2873 + }, + { + "epoch": 0.6026420633256447, + "grad_norm": 5.188991159961233, + "learning_rate": 1.8507984861246356e-05, + "loss": 2.0261, + "step": 2874 + }, + { + "epoch": 0.6028517508911722, + "grad_norm": 6.406112501076073, + "learning_rate": 1.850679498978565e-05, + "loss": 2.1274, + "step": 2875 + }, + { + "epoch": 0.6030614384566996, + "grad_norm": 7.39611509009168, + "learning_rate": 1.8505604682336947e-05, + "loss": 1.88, + "step": 2876 + }, + { + "epoch": 0.6032711260222269, + "grad_norm": 5.798044942484015, + "learning_rate": 1.850441393896125e-05, + "loss": 1.9593, + "step": 2877 + }, + { + "epoch": 0.6034808135877543, + "grad_norm": 6.173909061279093, + "learning_rate": 1.8503222759719595e-05, + "loss": 1.9724, + "step": 2878 + }, + { + "epoch": 0.6036905011532816, + "grad_norm": 6.292733087099212, + "learning_rate": 1.850203114467302e-05, + "loss": 2.2413, + "step": 2879 + }, + { + "epoch": 0.603900188718809, + "grad_norm": 6.825306134314194, + "learning_rate": 1.8500839093882605e-05, + "loss": 2.3622, + "step": 2880 + }, + { + "epoch": 0.6041098762843363, + "grad_norm": 6.064936351106037, + "learning_rate": 1.849964660740944e-05, + "loss": 2.0736, + "step": 2881 + }, + { + "epoch": 0.6043195638498637, + "grad_norm": 5.821697498833978, + "learning_rate": 1.849845368531465e-05, + "loss": 1.7887, + "step": 2882 + }, + { + "epoch": 0.604529251415391, + "grad_norm": 5.6173388166632225, + "learning_rate": 1.8497260327659365e-05, + "loss": 2.0661, + "step": 2883 + }, + { + "epoch": 0.6047389389809185, + "grad_norm": 5.878711660941405, + "learning_rate": 1.849606653450475e-05, + "loss": 1.6939, + "step": 2884 + }, + { + "epoch": 0.6049486265464458, + "grad_norm": 5.9113748114240945, + "learning_rate": 1.8494872305911985e-05, + "loss": 1.9963, + "step": 2885 + }, + { + "epoch": 0.6051583141119732, + "grad_norm": 6.994643935741634, + "learning_rate": 1.8493677641942285e-05, + "loss": 1.9344, + "step": 2886 + }, + { + "epoch": 0.6053680016775005, + "grad_norm": 6.691061982530917, + "learning_rate": 1.8492482542656876e-05, + "loss": 1.8457, + "step": 2887 + }, + { + "epoch": 0.6055776892430279, + "grad_norm": 7.030763620402425, + "learning_rate": 1.8491287008117006e-05, + "loss": 2.0861, + "step": 2888 + }, + { + "epoch": 0.6057873768085552, + "grad_norm": 7.005274911508568, + "learning_rate": 1.849009103838395e-05, + "loss": 2.2697, + "step": 2889 + }, + { + "epoch": 0.6059970643740826, + "grad_norm": 6.241309158168311, + "learning_rate": 1.8488894633519003e-05, + "loss": 2.1214, + "step": 2890 + }, + { + "epoch": 0.6062067519396099, + "grad_norm": 5.830015435991998, + "learning_rate": 1.8487697793583482e-05, + "loss": 1.8558, + "step": 2891 + }, + { + "epoch": 0.6064164395051374, + "grad_norm": 7.224704941738584, + "learning_rate": 1.8486500518638728e-05, + "loss": 1.9305, + "step": 2892 + }, + { + "epoch": 0.6066261270706647, + "grad_norm": 6.260624604496244, + "learning_rate": 1.8485302808746105e-05, + "loss": 2.2322, + "step": 2893 + }, + { + "epoch": 0.6068358146361921, + "grad_norm": 6.8111114191392, + "learning_rate": 1.8484104663966995e-05, + "loss": 2.18, + "step": 2894 + }, + { + "epoch": 0.6070455022017195, + "grad_norm": 6.657392986031021, + "learning_rate": 1.8482906084362803e-05, + "loss": 1.7316, + "step": 2895 + }, + { + "epoch": 0.6072551897672468, + "grad_norm": 6.201499580975748, + "learning_rate": 1.8481707069994966e-05, + "loss": 1.9609, + "step": 2896 + }, + { + "epoch": 0.6074648773327742, + "grad_norm": 5.470043526155994, + "learning_rate": 1.8480507620924928e-05, + "loss": 2.0646, + "step": 2897 + }, + { + "epoch": 0.6076745648983015, + "grad_norm": 6.796724121156567, + "learning_rate": 1.8479307737214167e-05, + "loss": 2.095, + "step": 2898 + }, + { + "epoch": 0.6078842524638289, + "grad_norm": 6.560112781251169, + "learning_rate": 1.8478107418924177e-05, + "loss": 2.0357, + "step": 2899 + }, + { + "epoch": 0.6080939400293562, + "grad_norm": 5.4305111553655925, + "learning_rate": 1.8476906666116477e-05, + "loss": 1.8864, + "step": 2900 + }, + { + "epoch": 0.6083036275948837, + "grad_norm": 5.465648062777567, + "learning_rate": 1.847570547885261e-05, + "loss": 1.869, + "step": 2901 + }, + { + "epoch": 0.608513315160411, + "grad_norm": 5.57374208627059, + "learning_rate": 1.8474503857194133e-05, + "loss": 1.8677, + "step": 2902 + }, + { + "epoch": 0.6087230027259384, + "grad_norm": 5.6505789539138025, + "learning_rate": 1.8473301801202638e-05, + "loss": 1.9986, + "step": 2903 + }, + { + "epoch": 0.6089326902914657, + "grad_norm": 6.01513487906172, + "learning_rate": 1.847209931093973e-05, + "loss": 2.021, + "step": 2904 + }, + { + "epoch": 0.6091423778569931, + "grad_norm": 5.466153132829219, + "learning_rate": 1.8470896386467038e-05, + "loss": 1.7392, + "step": 2905 + }, + { + "epoch": 0.6093520654225204, + "grad_norm": 5.883750662777273, + "learning_rate": 1.8469693027846213e-05, + "loss": 2.1267, + "step": 2906 + }, + { + "epoch": 0.6095617529880478, + "grad_norm": 5.2252626889403935, + "learning_rate": 1.846848923513893e-05, + "loss": 2.1191, + "step": 2907 + }, + { + "epoch": 0.6097714405535751, + "grad_norm": 6.119880745658791, + "learning_rate": 1.8467285008406884e-05, + "loss": 2.03, + "step": 2908 + }, + { + "epoch": 0.6099811281191025, + "grad_norm": 5.6369860322041, + "learning_rate": 1.84660803477118e-05, + "loss": 1.9044, + "step": 2909 + }, + { + "epoch": 0.6101908156846299, + "grad_norm": 4.9913954117609975, + "learning_rate": 1.846487525311541e-05, + "loss": 1.7837, + "step": 2910 + }, + { + "epoch": 0.6104005032501573, + "grad_norm": 6.466961569951358, + "learning_rate": 1.8463669724679488e-05, + "loss": 2.0469, + "step": 2911 + }, + { + "epoch": 0.6106101908156847, + "grad_norm": 5.69077776061508, + "learning_rate": 1.8462463762465808e-05, + "loss": 1.6832, + "step": 2912 + }, + { + "epoch": 0.610819878381212, + "grad_norm": 5.577767689681743, + "learning_rate": 1.8461257366536186e-05, + "loss": 1.7935, + "step": 2913 + }, + { + "epoch": 0.6110295659467394, + "grad_norm": 5.793788948469559, + "learning_rate": 1.8460050536952443e-05, + "loss": 1.9819, + "step": 2914 + }, + { + "epoch": 0.6112392535122667, + "grad_norm": 6.475328337335335, + "learning_rate": 1.8458843273776442e-05, + "loss": 2.1086, + "step": 2915 + }, + { + "epoch": 0.6114489410777941, + "grad_norm": 6.185528942463991, + "learning_rate": 1.845763557707005e-05, + "loss": 2.3214, + "step": 2916 + }, + { + "epoch": 0.6116586286433214, + "grad_norm": 6.6081351299207345, + "learning_rate": 1.8456427446895168e-05, + "loss": 2.0832, + "step": 2917 + }, + { + "epoch": 0.6118683162088489, + "grad_norm": 5.908546059484214, + "learning_rate": 1.845521888331371e-05, + "loss": 2.1168, + "step": 2918 + }, + { + "epoch": 0.6120780037743762, + "grad_norm": 6.0208292537308825, + "learning_rate": 1.8454009886387617e-05, + "loss": 2.0028, + "step": 2919 + }, + { + "epoch": 0.6122876913399036, + "grad_norm": 6.273060215600239, + "learning_rate": 1.845280045617886e-05, + "loss": 2.4412, + "step": 2920 + }, + { + "epoch": 0.6124973789054309, + "grad_norm": 5.8018969512882945, + "learning_rate": 1.845159059274942e-05, + "loss": 1.9969, + "step": 2921 + }, + { + "epoch": 0.6127070664709583, + "grad_norm": 6.174908830481501, + "learning_rate": 1.84503802961613e-05, + "loss": 2.0131, + "step": 2922 + }, + { + "epoch": 0.6129167540364856, + "grad_norm": 7.121011420175259, + "learning_rate": 1.844916956647653e-05, + "loss": 2.2599, + "step": 2923 + }, + { + "epoch": 0.613126441602013, + "grad_norm": 5.676948897530075, + "learning_rate": 1.8447958403757172e-05, + "loss": 1.6529, + "step": 2924 + }, + { + "epoch": 0.6133361291675403, + "grad_norm": 6.242963608964129, + "learning_rate": 1.844674680806529e-05, + "loss": 1.6443, + "step": 2925 + }, + { + "epoch": 0.6135458167330677, + "grad_norm": 6.116517228764057, + "learning_rate": 1.8445534779462982e-05, + "loss": 2.2042, + "step": 2926 + }, + { + "epoch": 0.613755504298595, + "grad_norm": 6.036515936567746, + "learning_rate": 1.8444322318012373e-05, + "loss": 2.0309, + "step": 2927 + }, + { + "epoch": 0.6139651918641225, + "grad_norm": 6.007903017768495, + "learning_rate": 1.84431094237756e-05, + "loss": 1.963, + "step": 2928 + }, + { + "epoch": 0.6141748794296498, + "grad_norm": 5.136990563195864, + "learning_rate": 1.844189609681482e-05, + "loss": 1.8762, + "step": 2929 + }, + { + "epoch": 0.6143845669951772, + "grad_norm": 6.140688288583342, + "learning_rate": 1.8440682337192224e-05, + "loss": 2.1942, + "step": 2930 + }, + { + "epoch": 0.6145942545607046, + "grad_norm": 5.538283354234424, + "learning_rate": 1.8439468144970022e-05, + "loss": 1.882, + "step": 2931 + }, + { + "epoch": 0.6148039421262319, + "grad_norm": 5.2441375597442255, + "learning_rate": 1.843825352021044e-05, + "loss": 2.1779, + "step": 2932 + }, + { + "epoch": 0.6150136296917593, + "grad_norm": 5.956895783713266, + "learning_rate": 1.8437038462975728e-05, + "loss": 1.7479, + "step": 2933 + }, + { + "epoch": 0.6152233172572866, + "grad_norm": 6.392289428765457, + "learning_rate": 1.843582297332816e-05, + "loss": 1.8556, + "step": 2934 + }, + { + "epoch": 0.615433004822814, + "grad_norm": 6.78811861493254, + "learning_rate": 1.8434607051330034e-05, + "loss": 1.9108, + "step": 2935 + }, + { + "epoch": 0.6156426923883414, + "grad_norm": 6.404002829575354, + "learning_rate": 1.8433390697043668e-05, + "loss": 1.9437, + "step": 2936 + }, + { + "epoch": 0.6158523799538688, + "grad_norm": 6.311719832625062, + "learning_rate": 1.8432173910531405e-05, + "loss": 2.08, + "step": 2937 + }, + { + "epoch": 0.6160620675193961, + "grad_norm": 5.801134861337478, + "learning_rate": 1.84309566918556e-05, + "loss": 1.8332, + "step": 2938 + }, + { + "epoch": 0.6162717550849235, + "grad_norm": 5.843811089819831, + "learning_rate": 1.8429739041078646e-05, + "loss": 2.1143, + "step": 2939 + }, + { + "epoch": 0.6164814426504508, + "grad_norm": 5.484120365084043, + "learning_rate": 1.8428520958262942e-05, + "loss": 1.8145, + "step": 2940 + }, + { + "epoch": 0.6166911302159782, + "grad_norm": 6.407809558323341, + "learning_rate": 1.842730244347092e-05, + "loss": 1.8818, + "step": 2941 + }, + { + "epoch": 0.6169008177815055, + "grad_norm": 6.272648428434445, + "learning_rate": 1.8426083496765033e-05, + "loss": 2.0181, + "step": 2942 + }, + { + "epoch": 0.6171105053470329, + "grad_norm": 6.0755994469483285, + "learning_rate": 1.842486411820775e-05, + "loss": 2.0487, + "step": 2943 + }, + { + "epoch": 0.6173201929125602, + "grad_norm": 5.792865816084123, + "learning_rate": 1.8423644307861573e-05, + "loss": 2.2358, + "step": 2944 + }, + { + "epoch": 0.6175298804780877, + "grad_norm": 5.925759022180038, + "learning_rate": 1.8422424065789012e-05, + "loss": 1.8529, + "step": 2945 + }, + { + "epoch": 0.617739568043615, + "grad_norm": 6.70922830796199, + "learning_rate": 1.8421203392052612e-05, + "loss": 2.1349, + "step": 2946 + }, + { + "epoch": 0.6179492556091424, + "grad_norm": 4.754658103579967, + "learning_rate": 1.841998228671493e-05, + "loss": 1.9735, + "step": 2947 + }, + { + "epoch": 0.6181589431746698, + "grad_norm": 5.870961551358106, + "learning_rate": 1.8418760749838554e-05, + "loss": 2.1093, + "step": 2948 + }, + { + "epoch": 0.6183686307401971, + "grad_norm": 6.554197846675178, + "learning_rate": 1.8417538781486087e-05, + "loss": 2.0117, + "step": 2949 + }, + { + "epoch": 0.6185783183057245, + "grad_norm": 5.816861469325661, + "learning_rate": 1.8416316381720157e-05, + "loss": 1.8815, + "step": 2950 + }, + { + "epoch": 0.6187880058712518, + "grad_norm": 5.599796922541357, + "learning_rate": 1.8415093550603417e-05, + "loss": 1.8963, + "step": 2951 + }, + { + "epoch": 0.6189976934367792, + "grad_norm": 7.049869665461245, + "learning_rate": 1.8413870288198537e-05, + "loss": 2.0832, + "step": 2952 + }, + { + "epoch": 0.6192073810023065, + "grad_norm": 6.1240343175288645, + "learning_rate": 1.841264659456821e-05, + "loss": 1.8794, + "step": 2953 + }, + { + "epoch": 0.619417068567834, + "grad_norm": 6.781324922899264, + "learning_rate": 1.8411422469775157e-05, + "loss": 1.9303, + "step": 2954 + }, + { + "epoch": 0.6196267561333613, + "grad_norm": 6.103287801229008, + "learning_rate": 1.841019791388211e-05, + "loss": 1.6427, + "step": 2955 + }, + { + "epoch": 0.6198364436988887, + "grad_norm": 7.014494920410526, + "learning_rate": 1.8408972926951833e-05, + "loss": 2.0703, + "step": 2956 + }, + { + "epoch": 0.620046131264416, + "grad_norm": 6.659788083223147, + "learning_rate": 1.840774750904711e-05, + "loss": 2.0225, + "step": 2957 + }, + { + "epoch": 0.6202558188299434, + "grad_norm": 6.841402058639845, + "learning_rate": 1.840652166023075e-05, + "loss": 2.2322, + "step": 2958 + }, + { + "epoch": 0.6204655063954707, + "grad_norm": 5.490412568410373, + "learning_rate": 1.840529538056557e-05, + "loss": 1.8545, + "step": 2959 + }, + { + "epoch": 0.6206751939609981, + "grad_norm": 5.641754294915338, + "learning_rate": 1.840406867011442e-05, + "loss": 1.9678, + "step": 2960 + }, + { + "epoch": 0.6208848815265254, + "grad_norm": 6.423084300056922, + "learning_rate": 1.8402841528940178e-05, + "loss": 1.9471, + "step": 2961 + }, + { + "epoch": 0.6210945690920529, + "grad_norm": 6.243026150690286, + "learning_rate": 1.8401613957105735e-05, + "loss": 1.8033, + "step": 2962 + }, + { + "epoch": 0.6213042566575802, + "grad_norm": 5.355698002207967, + "learning_rate": 1.8400385954674002e-05, + "loss": 1.8766, + "step": 2963 + }, + { + "epoch": 0.6215139442231076, + "grad_norm": 7.491708880422434, + "learning_rate": 1.8399157521707918e-05, + "loss": 1.963, + "step": 2964 + }, + { + "epoch": 0.621723631788635, + "grad_norm": 6.001839891011696, + "learning_rate": 1.839792865827045e-05, + "loss": 1.9395, + "step": 2965 + }, + { + "epoch": 0.6219333193541623, + "grad_norm": 7.076759551407105, + "learning_rate": 1.839669936442457e-05, + "loss": 1.9208, + "step": 2966 + }, + { + "epoch": 0.6221430069196897, + "grad_norm": 6.499574473097217, + "learning_rate": 1.8395469640233277e-05, + "loss": 2.0457, + "step": 2967 + }, + { + "epoch": 0.622352694485217, + "grad_norm": 6.242789360030616, + "learning_rate": 1.8394239485759613e-05, + "loss": 1.8231, + "step": 2968 + }, + { + "epoch": 0.6225623820507444, + "grad_norm": 6.082853195416612, + "learning_rate": 1.8393008901066612e-05, + "loss": 1.9834, + "step": 2969 + }, + { + "epoch": 0.6227720696162717, + "grad_norm": 6.362345890634872, + "learning_rate": 1.839177788621735e-05, + "loss": 2.1581, + "step": 2970 + }, + { + "epoch": 0.6229817571817992, + "grad_norm": 5.440557014113412, + "learning_rate": 1.8390546441274914e-05, + "loss": 1.9935, + "step": 2971 + }, + { + "epoch": 0.6231914447473265, + "grad_norm": 5.413578860244006, + "learning_rate": 1.8389314566302422e-05, + "loss": 2.0807, + "step": 2972 + }, + { + "epoch": 0.6234011323128539, + "grad_norm": 5.253535948966556, + "learning_rate": 1.838808226136301e-05, + "loss": 1.8844, + "step": 2973 + }, + { + "epoch": 0.6236108198783812, + "grad_norm": 6.026476148171401, + "learning_rate": 1.838684952651983e-05, + "loss": 1.7923, + "step": 2974 + }, + { + "epoch": 0.6238205074439086, + "grad_norm": 6.031666843340524, + "learning_rate": 1.8385616361836066e-05, + "loss": 2.072, + "step": 2975 + }, + { + "epoch": 0.6240301950094359, + "grad_norm": 5.831357455822318, + "learning_rate": 1.8384382767374923e-05, + "loss": 1.6445, + "step": 2976 + }, + { + "epoch": 0.6242398825749633, + "grad_norm": 5.873151157900796, + "learning_rate": 1.8383148743199615e-05, + "loss": 2.2271, + "step": 2977 + }, + { + "epoch": 0.6244495701404906, + "grad_norm": 6.399140507553066, + "learning_rate": 1.8381914289373396e-05, + "loss": 1.931, + "step": 2978 + }, + { + "epoch": 0.624659257706018, + "grad_norm": 5.939862297556723, + "learning_rate": 1.8380679405959535e-05, + "loss": 1.4384, + "step": 2979 + }, + { + "epoch": 0.6248689452715454, + "grad_norm": 7.985760661391244, + "learning_rate": 1.8379444093021317e-05, + "loss": 2.1693, + "step": 2980 + }, + { + "epoch": 0.6250786328370728, + "grad_norm": 6.134582472468553, + "learning_rate": 1.8378208350622056e-05, + "loss": 2.2313, + "step": 2981 + }, + { + "epoch": 0.6252883204026001, + "grad_norm": 7.330600879974096, + "learning_rate": 1.8376972178825085e-05, + "loss": 2.4532, + "step": 2982 + }, + { + "epoch": 0.6254980079681275, + "grad_norm": 7.01349886967936, + "learning_rate": 1.8375735577693763e-05, + "loss": 2.2038, + "step": 2983 + }, + { + "epoch": 0.6257076955336549, + "grad_norm": 6.566217796768941, + "learning_rate": 1.8374498547291467e-05, + "loss": 1.8165, + "step": 2984 + }, + { + "epoch": 0.6259173830991822, + "grad_norm": 6.2582402601021885, + "learning_rate": 1.837326108768159e-05, + "loss": 1.7524, + "step": 2985 + }, + { + "epoch": 0.6261270706647096, + "grad_norm": 6.265951401991545, + "learning_rate": 1.837202319892756e-05, + "loss": 1.9066, + "step": 2986 + }, + { + "epoch": 0.6263367582302369, + "grad_norm": 6.368107366229069, + "learning_rate": 1.8370784881092823e-05, + "loss": 2.1225, + "step": 2987 + }, + { + "epoch": 0.6265464457957644, + "grad_norm": 5.564079796347212, + "learning_rate": 1.8369546134240844e-05, + "loss": 2.1955, + "step": 2988 + }, + { + "epoch": 0.6267561333612917, + "grad_norm": 6.5435462682551595, + "learning_rate": 1.8368306958435108e-05, + "loss": 1.8207, + "step": 2989 + }, + { + "epoch": 0.6269658209268191, + "grad_norm": 6.430563097358877, + "learning_rate": 1.8367067353739125e-05, + "loss": 2.0898, + "step": 2990 + }, + { + "epoch": 0.6271755084923464, + "grad_norm": 6.154208578228744, + "learning_rate": 1.836582732021643e-05, + "loss": 1.6919, + "step": 2991 + }, + { + "epoch": 0.6273851960578738, + "grad_norm": 5.764625633442794, + "learning_rate": 1.8364586857930574e-05, + "loss": 1.7333, + "step": 2992 + }, + { + "epoch": 0.6275948836234011, + "grad_norm": 5.602025141125753, + "learning_rate": 1.8363345966945136e-05, + "loss": 1.7396, + "step": 2993 + }, + { + "epoch": 0.6278045711889285, + "grad_norm": 6.205866619625006, + "learning_rate": 1.836210464732371e-05, + "loss": 2.002, + "step": 2994 + }, + { + "epoch": 0.6280142587544558, + "grad_norm": 5.489230757671925, + "learning_rate": 1.8360862899129917e-05, + "loss": 1.9006, + "step": 2995 + }, + { + "epoch": 0.6282239463199832, + "grad_norm": 6.075765108712726, + "learning_rate": 1.83596207224274e-05, + "loss": 2.0385, + "step": 2996 + }, + { + "epoch": 0.6284336338855105, + "grad_norm": 5.7565748011513165, + "learning_rate": 1.835837811727982e-05, + "loss": 2.3448, + "step": 2997 + }, + { + "epoch": 0.628643321451038, + "grad_norm": 5.5024067343373435, + "learning_rate": 1.8357135083750867e-05, + "loss": 1.973, + "step": 2998 + }, + { + "epoch": 0.6288530090165653, + "grad_norm": 5.86692485272245, + "learning_rate": 1.8355891621904245e-05, + "loss": 1.9835, + "step": 2999 + }, + { + "epoch": 0.6290626965820927, + "grad_norm": 5.805337766616843, + "learning_rate": 1.8354647731803687e-05, + "loss": 2.1475, + "step": 3000 + }, + { + "epoch": 0.6292723841476201, + "grad_norm": 6.881706936242785, + "learning_rate": 1.835340341351294e-05, + "loss": 2.1763, + "step": 3001 + }, + { + "epoch": 0.6294820717131474, + "grad_norm": 6.445833120415932, + "learning_rate": 1.835215866709578e-05, + "loss": 1.9523, + "step": 3002 + }, + { + "epoch": 0.6296917592786748, + "grad_norm": 7.641125533660694, + "learning_rate": 1.8350913492616e-05, + "loss": 2.1747, + "step": 3003 + }, + { + "epoch": 0.6299014468442021, + "grad_norm": 6.585319856334905, + "learning_rate": 1.8349667890137425e-05, + "loss": 2.3402, + "step": 3004 + }, + { + "epoch": 0.6301111344097295, + "grad_norm": 5.86932865879389, + "learning_rate": 1.8348421859723886e-05, + "loss": 1.8128, + "step": 3005 + }, + { + "epoch": 0.6303208219752569, + "grad_norm": 6.8128799455015585, + "learning_rate": 1.8347175401439245e-05, + "loss": 2.2549, + "step": 3006 + }, + { + "epoch": 0.6305305095407843, + "grad_norm": 6.235016165116718, + "learning_rate": 1.834592851534739e-05, + "loss": 2.0222, + "step": 3007 + }, + { + "epoch": 0.6307401971063116, + "grad_norm": 5.670669593581825, + "learning_rate": 1.8344681201512223e-05, + "loss": 1.9152, + "step": 3008 + }, + { + "epoch": 0.630949884671839, + "grad_norm": 6.0532093687989414, + "learning_rate": 1.834343345999767e-05, + "loss": 2.2976, + "step": 3009 + }, + { + "epoch": 0.6311595722373663, + "grad_norm": 5.627434316039015, + "learning_rate": 1.8342185290867683e-05, + "loss": 1.6983, + "step": 3010 + }, + { + "epoch": 0.6313692598028937, + "grad_norm": 5.4178916979648415, + "learning_rate": 1.8340936694186232e-05, + "loss": 2.02, + "step": 3011 + }, + { + "epoch": 0.631578947368421, + "grad_norm": 5.8437081006531875, + "learning_rate": 1.8339687670017306e-05, + "loss": 2.0354, + "step": 3012 + }, + { + "epoch": 0.6317886349339484, + "grad_norm": 5.661663000980645, + "learning_rate": 1.8338438218424923e-05, + "loss": 2.3399, + "step": 3013 + }, + { + "epoch": 0.6319983224994757, + "grad_norm": 6.353765604564318, + "learning_rate": 1.833718833947312e-05, + "loss": 2.0716, + "step": 3014 + }, + { + "epoch": 0.6322080100650032, + "grad_norm": 5.75083910948988, + "learning_rate": 1.8335938033225954e-05, + "loss": 1.99, + "step": 3015 + }, + { + "epoch": 0.6324176976305305, + "grad_norm": 5.747361601818475, + "learning_rate": 1.8334687299747505e-05, + "loss": 1.9127, + "step": 3016 + }, + { + "epoch": 0.6326273851960579, + "grad_norm": 6.457989111053694, + "learning_rate": 1.8333436139101876e-05, + "loss": 1.9246, + "step": 3017 + }, + { + "epoch": 0.6328370727615852, + "grad_norm": 6.231392402406366, + "learning_rate": 1.8332184551353193e-05, + "loss": 2.2198, + "step": 3018 + }, + { + "epoch": 0.6330467603271126, + "grad_norm": 6.310874894429705, + "learning_rate": 1.83309325365656e-05, + "loss": 1.7845, + "step": 3019 + }, + { + "epoch": 0.63325644789264, + "grad_norm": 5.795178709519663, + "learning_rate": 1.8329680094803265e-05, + "loss": 1.7572, + "step": 3020 + }, + { + "epoch": 0.6334661354581673, + "grad_norm": 6.35618531789427, + "learning_rate": 1.832842722613038e-05, + "loss": 2.0481, + "step": 3021 + }, + { + "epoch": 0.6336758230236947, + "grad_norm": 6.0666129248367655, + "learning_rate": 1.832717393061115e-05, + "loss": 2.2085, + "step": 3022 + }, + { + "epoch": 0.633885510589222, + "grad_norm": 6.085193114885912, + "learning_rate": 1.8325920208309815e-05, + "loss": 1.9534, + "step": 3023 + }, + { + "epoch": 0.6340951981547495, + "grad_norm": 6.6428480631854425, + "learning_rate": 1.832466605929063e-05, + "loss": 2.0879, + "step": 3024 + }, + { + "epoch": 0.6343048857202768, + "grad_norm": 5.973788170683085, + "learning_rate": 1.832341148361787e-05, + "loss": 2.2544, + "step": 3025 + }, + { + "epoch": 0.6345145732858042, + "grad_norm": 5.8470876802096745, + "learning_rate": 1.8322156481355836e-05, + "loss": 1.5665, + "step": 3026 + }, + { + "epoch": 0.6347242608513315, + "grad_norm": 5.952277485838748, + "learning_rate": 1.832090105256885e-05, + "loss": 2.2062, + "step": 3027 + }, + { + "epoch": 0.6349339484168589, + "grad_norm": 6.991806731698505, + "learning_rate": 1.831964519732125e-05, + "loss": 1.8965, + "step": 3028 + }, + { + "epoch": 0.6351436359823862, + "grad_norm": 5.56202844364899, + "learning_rate": 1.8318388915677406e-05, + "loss": 1.7468, + "step": 3029 + }, + { + "epoch": 0.6353533235479136, + "grad_norm": 6.136952608218808, + "learning_rate": 1.8317132207701703e-05, + "loss": 2.3373, + "step": 3030 + }, + { + "epoch": 0.6355630111134409, + "grad_norm": 6.120249658991906, + "learning_rate": 1.8315875073458546e-05, + "loss": 1.7408, + "step": 3031 + }, + { + "epoch": 0.6357726986789684, + "grad_norm": 5.20310201475147, + "learning_rate": 1.831461751301237e-05, + "loss": 1.8269, + "step": 3032 + }, + { + "epoch": 0.6359823862444957, + "grad_norm": 5.952074442396482, + "learning_rate": 1.8313359526427627e-05, + "loss": 1.5961, + "step": 3033 + }, + { + "epoch": 0.6361920738100231, + "grad_norm": 5.316538412482053, + "learning_rate": 1.8312101113768786e-05, + "loss": 1.9048, + "step": 3034 + }, + { + "epoch": 0.6364017613755504, + "grad_norm": 6.7088514617458594, + "learning_rate": 1.8310842275100352e-05, + "loss": 2.0127, + "step": 3035 + }, + { + "epoch": 0.6366114489410778, + "grad_norm": 6.0326708649622764, + "learning_rate": 1.8309583010486832e-05, + "loss": 2.2911, + "step": 3036 + }, + { + "epoch": 0.6368211365066052, + "grad_norm": 6.820922445631784, + "learning_rate": 1.8308323319992773e-05, + "loss": 2.1033, + "step": 3037 + }, + { + "epoch": 0.6370308240721325, + "grad_norm": 6.996388408391846, + "learning_rate": 1.830706320368273e-05, + "loss": 1.8596, + "step": 3038 + }, + { + "epoch": 0.6372405116376599, + "grad_norm": 5.5482173129981245, + "learning_rate": 1.8305802661621293e-05, + "loss": 1.947, + "step": 3039 + }, + { + "epoch": 0.6374501992031872, + "grad_norm": 5.6801980060072745, + "learning_rate": 1.8304541693873066e-05, + "loss": 2.1226, + "step": 3040 + }, + { + "epoch": 0.6376598867687147, + "grad_norm": 5.497440890326025, + "learning_rate": 1.830328030050267e-05, + "loss": 1.744, + "step": 3041 + }, + { + "epoch": 0.637869574334242, + "grad_norm": 6.902921978402993, + "learning_rate": 1.8302018481574756e-05, + "loss": 2.0559, + "step": 3042 + }, + { + "epoch": 0.6380792618997694, + "grad_norm": 6.088228193402139, + "learning_rate": 1.8300756237154e-05, + "loss": 1.9644, + "step": 3043 + }, + { + "epoch": 0.6382889494652967, + "grad_norm": 6.510881861310974, + "learning_rate": 1.8299493567305088e-05, + "loss": 2.0613, + "step": 3044 + }, + { + "epoch": 0.6384986370308241, + "grad_norm": 5.9988705028592815, + "learning_rate": 1.8298230472092738e-05, + "loss": 1.8969, + "step": 3045 + }, + { + "epoch": 0.6387083245963514, + "grad_norm": 6.041217302938949, + "learning_rate": 1.829696695158168e-05, + "loss": 1.7567, + "step": 3046 + }, + { + "epoch": 0.6389180121618788, + "grad_norm": 7.05934025622722, + "learning_rate": 1.8295703005836677e-05, + "loss": 1.9822, + "step": 3047 + }, + { + "epoch": 0.6391276997274061, + "grad_norm": 6.24086378561901, + "learning_rate": 1.8294438634922507e-05, + "loss": 1.8842, + "step": 3048 + }, + { + "epoch": 0.6393373872929335, + "grad_norm": 4.860105322229299, + "learning_rate": 1.829317383890397e-05, + "loss": 1.7116, + "step": 3049 + }, + { + "epoch": 0.6395470748584609, + "grad_norm": 5.947512241161165, + "learning_rate": 1.829190861784589e-05, + "loss": 1.863, + "step": 3050 + }, + { + "epoch": 0.6397567624239883, + "grad_norm": 6.027982894227071, + "learning_rate": 1.8290642971813113e-05, + "loss": 2.1154, + "step": 3051 + }, + { + "epoch": 0.6399664499895156, + "grad_norm": 6.550770136585487, + "learning_rate": 1.8289376900870506e-05, + "loss": 1.801, + "step": 3052 + }, + { + "epoch": 0.640176137555043, + "grad_norm": 6.877263008135769, + "learning_rate": 1.8288110405082956e-05, + "loss": 1.7381, + "step": 3053 + }, + { + "epoch": 0.6403858251205703, + "grad_norm": 5.87457880262923, + "learning_rate": 1.828684348451537e-05, + "loss": 2.0457, + "step": 3054 + }, + { + "epoch": 0.6405955126860977, + "grad_norm": 6.61273343295273, + "learning_rate": 1.828557613923268e-05, + "loss": 2.2301, + "step": 3055 + }, + { + "epoch": 0.6408052002516251, + "grad_norm": 6.761473456770218, + "learning_rate": 1.8284308369299845e-05, + "loss": 1.7939, + "step": 3056 + }, + { + "epoch": 0.6410148878171524, + "grad_norm": 5.568846549803721, + "learning_rate": 1.828304017478184e-05, + "loss": 1.9008, + "step": 3057 + }, + { + "epoch": 0.6412245753826799, + "grad_norm": 7.1327175395719555, + "learning_rate": 1.8281771555743655e-05, + "loss": 1.7472, + "step": 3058 + }, + { + "epoch": 0.6414342629482072, + "grad_norm": 6.028220563181006, + "learning_rate": 1.828050251225032e-05, + "loss": 2.1141, + "step": 3059 + }, + { + "epoch": 0.6416439505137346, + "grad_norm": 7.268321161030788, + "learning_rate": 1.8279233044366864e-05, + "loss": 2.0093, + "step": 3060 + }, + { + "epoch": 0.6418536380792619, + "grad_norm": 6.604429447192501, + "learning_rate": 1.827796315215836e-05, + "loss": 2.2255, + "step": 3061 + }, + { + "epoch": 0.6420633256447893, + "grad_norm": 6.111561761445589, + "learning_rate": 1.827669283568988e-05, + "loss": 1.9586, + "step": 3062 + }, + { + "epoch": 0.6422730132103166, + "grad_norm": 6.961044945941901, + "learning_rate": 1.8275422095026544e-05, + "loss": 2.1318, + "step": 3063 + }, + { + "epoch": 0.642482700775844, + "grad_norm": 6.753699687823845, + "learning_rate": 1.827415093023347e-05, + "loss": 2.2353, + "step": 3064 + }, + { + "epoch": 0.6426923883413713, + "grad_norm": 6.806147524542361, + "learning_rate": 1.8272879341375808e-05, + "loss": 2.1543, + "step": 3065 + }, + { + "epoch": 0.6429020759068987, + "grad_norm": 5.336034839169556, + "learning_rate": 1.8271607328518732e-05, + "loss": 1.8104, + "step": 3066 + }, + { + "epoch": 0.643111763472426, + "grad_norm": 6.742453529827101, + "learning_rate": 1.8270334891727433e-05, + "loss": 1.9633, + "step": 3067 + }, + { + "epoch": 0.6433214510379535, + "grad_norm": 6.12031899943188, + "learning_rate": 1.826906203106713e-05, + "loss": 1.8327, + "step": 3068 + }, + { + "epoch": 0.6435311386034808, + "grad_norm": 6.136631624615883, + "learning_rate": 1.8267788746603055e-05, + "loss": 1.734, + "step": 3069 + }, + { + "epoch": 0.6437408261690082, + "grad_norm": 7.398613128319719, + "learning_rate": 1.826651503840047e-05, + "loss": 2.2307, + "step": 3070 + }, + { + "epoch": 0.6439505137345355, + "grad_norm": 6.067959128754426, + "learning_rate": 1.8265240906524644e-05, + "loss": 1.9187, + "step": 3071 + }, + { + "epoch": 0.6441602013000629, + "grad_norm": 5.438652440044395, + "learning_rate": 1.826396635104089e-05, + "loss": 2.0244, + "step": 3072 + }, + { + "epoch": 0.6443698888655903, + "grad_norm": 5.793018233781782, + "learning_rate": 1.826269137201453e-05, + "loss": 1.9017, + "step": 3073 + }, + { + "epoch": 0.6445795764311176, + "grad_norm": 6.288640526570729, + "learning_rate": 1.82614159695109e-05, + "loss": 2.2585, + "step": 3074 + }, + { + "epoch": 0.644789263996645, + "grad_norm": 5.862238751583479, + "learning_rate": 1.826014014359538e-05, + "loss": 1.8447, + "step": 3075 + }, + { + "epoch": 0.6449989515621724, + "grad_norm": 5.642957180478201, + "learning_rate": 1.825886389433335e-05, + "loss": 1.861, + "step": 3076 + }, + { + "epoch": 0.6452086391276998, + "grad_norm": 6.10056544438305, + "learning_rate": 1.8257587221790218e-05, + "loss": 1.928, + "step": 3077 + }, + { + "epoch": 0.6454183266932271, + "grad_norm": 5.919663961991319, + "learning_rate": 1.825631012603142e-05, + "loss": 1.6966, + "step": 3078 + }, + { + "epoch": 0.6456280142587545, + "grad_norm": 5.9090245248108, + "learning_rate": 1.8255032607122406e-05, + "loss": 1.8555, + "step": 3079 + }, + { + "epoch": 0.6458377018242818, + "grad_norm": 6.233306103273052, + "learning_rate": 1.8253754665128656e-05, + "loss": 1.6421, + "step": 3080 + }, + { + "epoch": 0.6460473893898092, + "grad_norm": 5.937586057974155, + "learning_rate": 1.825247630011566e-05, + "loss": 2.4088, + "step": 3081 + }, + { + "epoch": 0.6462570769553365, + "grad_norm": 5.363858077334569, + "learning_rate": 1.825119751214895e-05, + "loss": 2.2506, + "step": 3082 + }, + { + "epoch": 0.6464667645208639, + "grad_norm": 5.544056082084755, + "learning_rate": 1.8249918301294046e-05, + "loss": 1.6543, + "step": 3083 + }, + { + "epoch": 0.6466764520863912, + "grad_norm": 5.871512424746385, + "learning_rate": 1.8248638667616526e-05, + "loss": 2.0845, + "step": 3084 + }, + { + "epoch": 0.6468861396519187, + "grad_norm": 6.684895913363462, + "learning_rate": 1.8247358611181965e-05, + "loss": 2.2184, + "step": 3085 + }, + { + "epoch": 0.647095827217446, + "grad_norm": 5.45504528562288, + "learning_rate": 1.8246078132055975e-05, + "loss": 2.1095, + "step": 3086 + }, + { + "epoch": 0.6473055147829734, + "grad_norm": 8.272426907997717, + "learning_rate": 1.8244797230304176e-05, + "loss": 1.7937, + "step": 3087 + }, + { + "epoch": 0.6475152023485007, + "grad_norm": 5.8539053275208035, + "learning_rate": 1.8243515905992217e-05, + "loss": 2.13, + "step": 3088 + }, + { + "epoch": 0.6477248899140281, + "grad_norm": 7.058377458417762, + "learning_rate": 1.8242234159185774e-05, + "loss": 1.6912, + "step": 3089 + }, + { + "epoch": 0.6479345774795554, + "grad_norm": 6.326104161111226, + "learning_rate": 1.8240951989950532e-05, + "loss": 1.988, + "step": 3090 + }, + { + "epoch": 0.6481442650450828, + "grad_norm": 5.299640883892419, + "learning_rate": 1.823966939835221e-05, + "loss": 1.6287, + "step": 3091 + }, + { + "epoch": 0.6483539526106102, + "grad_norm": 6.780734969667526, + "learning_rate": 1.823838638445654e-05, + "loss": 1.8447, + "step": 3092 + }, + { + "epoch": 0.6485636401761375, + "grad_norm": 6.071306870888296, + "learning_rate": 1.8237102948329276e-05, + "loss": 2.0756, + "step": 3093 + }, + { + "epoch": 0.648773327741665, + "grad_norm": 6.553205057201114, + "learning_rate": 1.82358190900362e-05, + "loss": 1.8641, + "step": 3094 + }, + { + "epoch": 0.6489830153071923, + "grad_norm": 6.238687972865546, + "learning_rate": 1.8234534809643113e-05, + "loss": 1.7182, + "step": 3095 + }, + { + "epoch": 0.6491927028727197, + "grad_norm": 5.996039594722565, + "learning_rate": 1.8233250107215835e-05, + "loss": 1.8194, + "step": 3096 + }, + { + "epoch": 0.649402390438247, + "grad_norm": 5.756966442994074, + "learning_rate": 1.823196498282021e-05, + "loss": 1.6929, + "step": 3097 + }, + { + "epoch": 0.6496120780037744, + "grad_norm": 7.032618136284973, + "learning_rate": 1.82306794365221e-05, + "loss": 2.0718, + "step": 3098 + }, + { + "epoch": 0.6498217655693017, + "grad_norm": 6.879859607243169, + "learning_rate": 1.822939346838739e-05, + "loss": 1.7797, + "step": 3099 + }, + { + "epoch": 0.6500314531348291, + "grad_norm": 7.483046624586785, + "learning_rate": 1.8228107078482e-05, + "loss": 2.0442, + "step": 3100 + }, + { + "epoch": 0.6502411407003564, + "grad_norm": 6.199752119143285, + "learning_rate": 1.8226820266871846e-05, + "loss": 1.9179, + "step": 3101 + }, + { + "epoch": 0.6504508282658839, + "grad_norm": 6.330111137169876, + "learning_rate": 1.8225533033622885e-05, + "loss": 2.1492, + "step": 3102 + }, + { + "epoch": 0.6506605158314112, + "grad_norm": 6.355188291172677, + "learning_rate": 1.822424537880109e-05, + "loss": 2.0496, + "step": 3103 + }, + { + "epoch": 0.6508702033969386, + "grad_norm": 6.083462761673524, + "learning_rate": 1.8222957302472454e-05, + "loss": 1.7815, + "step": 3104 + }, + { + "epoch": 0.6510798909624659, + "grad_norm": 6.823633269025465, + "learning_rate": 1.8221668804702995e-05, + "loss": 1.8783, + "step": 3105 + }, + { + "epoch": 0.6512895785279933, + "grad_norm": 7.418728386520889, + "learning_rate": 1.822037988555875e-05, + "loss": 1.7743, + "step": 3106 + }, + { + "epoch": 0.6514992660935206, + "grad_norm": 6.2550496740962105, + "learning_rate": 1.821909054510578e-05, + "loss": 2.0167, + "step": 3107 + }, + { + "epoch": 0.651708953659048, + "grad_norm": 6.050550669136302, + "learning_rate": 1.821780078341016e-05, + "loss": 1.8372, + "step": 3108 + }, + { + "epoch": 0.6519186412245754, + "grad_norm": 5.559842794261759, + "learning_rate": 1.8216510600538006e-05, + "loss": 1.8968, + "step": 3109 + }, + { + "epoch": 0.6521283287901027, + "grad_norm": 6.758661495195963, + "learning_rate": 1.821521999655542e-05, + "loss": 1.9693, + "step": 3110 + }, + { + "epoch": 0.6523380163556302, + "grad_norm": 6.1762410322381225, + "learning_rate": 1.821392897152857e-05, + "loss": 1.9382, + "step": 3111 + }, + { + "epoch": 0.6525477039211575, + "grad_norm": 5.5700317873183645, + "learning_rate": 1.821263752552361e-05, + "loss": 1.9557, + "step": 3112 + }, + { + "epoch": 0.6527573914866849, + "grad_norm": 6.422690278154433, + "learning_rate": 1.8211345658606737e-05, + "loss": 2.1507, + "step": 3113 + }, + { + "epoch": 0.6529670790522122, + "grad_norm": 5.913869206930769, + "learning_rate": 1.8210053370844152e-05, + "loss": 2.03, + "step": 3114 + }, + { + "epoch": 0.6531767666177396, + "grad_norm": 8.213758666560123, + "learning_rate": 1.8208760662302097e-05, + "loss": 1.5969, + "step": 3115 + }, + { + "epoch": 0.6533864541832669, + "grad_norm": 5.581285935631441, + "learning_rate": 1.8207467533046817e-05, + "loss": 2.1011, + "step": 3116 + }, + { + "epoch": 0.6535961417487943, + "grad_norm": 6.218754340953845, + "learning_rate": 1.820617398314459e-05, + "loss": 2.1834, + "step": 3117 + }, + { + "epoch": 0.6538058293143216, + "grad_norm": 5.9414474657047425, + "learning_rate": 1.8204880012661717e-05, + "loss": 1.9557, + "step": 3118 + }, + { + "epoch": 0.654015516879849, + "grad_norm": 5.312276878650203, + "learning_rate": 1.820358562166451e-05, + "loss": 2.1391, + "step": 3119 + }, + { + "epoch": 0.6542252044453764, + "grad_norm": 5.73033774277521, + "learning_rate": 1.8202290810219313e-05, + "loss": 1.9547, + "step": 3120 + }, + { + "epoch": 0.6544348920109038, + "grad_norm": 5.908047660799942, + "learning_rate": 1.8200995578392484e-05, + "loss": 2.2818, + "step": 3121 + }, + { + "epoch": 0.6546445795764311, + "grad_norm": 5.761058241598259, + "learning_rate": 1.8199699926250408e-05, + "loss": 1.9812, + "step": 3122 + }, + { + "epoch": 0.6548542671419585, + "grad_norm": 5.0810986758352845, + "learning_rate": 1.819840385385949e-05, + "loss": 1.9485, + "step": 3123 + }, + { + "epoch": 0.6550639547074858, + "grad_norm": 6.67215763714508, + "learning_rate": 1.8197107361286157e-05, + "loss": 1.9369, + "step": 3124 + }, + { + "epoch": 0.6552736422730132, + "grad_norm": 6.30665138181098, + "learning_rate": 1.8195810448596852e-05, + "loss": 2.0806, + "step": 3125 + }, + { + "epoch": 0.6554833298385406, + "grad_norm": 6.313203246063494, + "learning_rate": 1.8194513115858046e-05, + "loss": 2.2184, + "step": 3126 + }, + { + "epoch": 0.6556930174040679, + "grad_norm": 5.971525330938641, + "learning_rate": 1.8193215363136233e-05, + "loss": 2.0386, + "step": 3127 + }, + { + "epoch": 0.6559027049695954, + "grad_norm": 6.704912562677806, + "learning_rate": 1.8191917190497917e-05, + "loss": 1.9626, + "step": 3128 + }, + { + "epoch": 0.6561123925351227, + "grad_norm": 7.055745721364275, + "learning_rate": 1.8190618598009642e-05, + "loss": 2.0895, + "step": 3129 + }, + { + "epoch": 0.6563220801006501, + "grad_norm": 5.705712211192097, + "learning_rate": 1.8189319585737958e-05, + "loss": 1.6538, + "step": 3130 + }, + { + "epoch": 0.6565317676661774, + "grad_norm": 6.206017737800097, + "learning_rate": 1.818802015374944e-05, + "loss": 2.0474, + "step": 3131 + }, + { + "epoch": 0.6567414552317048, + "grad_norm": 5.416482164873523, + "learning_rate": 1.8186720302110687e-05, + "loss": 1.6146, + "step": 3132 + }, + { + "epoch": 0.6569511427972321, + "grad_norm": 5.663122953888279, + "learning_rate": 1.8185420030888322e-05, + "loss": 1.9269, + "step": 3133 + }, + { + "epoch": 0.6571608303627595, + "grad_norm": 6.132219005526236, + "learning_rate": 1.8184119340148983e-05, + "loss": 1.8909, + "step": 3134 + }, + { + "epoch": 0.6573705179282868, + "grad_norm": 5.323081498737843, + "learning_rate": 1.8182818229959332e-05, + "loss": 1.6997, + "step": 3135 + }, + { + "epoch": 0.6575802054938142, + "grad_norm": 6.3772133343287765, + "learning_rate": 1.818151670038605e-05, + "loss": 2.0824, + "step": 3136 + }, + { + "epoch": 0.6577898930593415, + "grad_norm": 6.099044204103873, + "learning_rate": 1.818021475149585e-05, + "loss": 1.9255, + "step": 3137 + }, + { + "epoch": 0.657999580624869, + "grad_norm": 6.425465340660839, + "learning_rate": 1.817891238335546e-05, + "loss": 1.8174, + "step": 3138 + }, + { + "epoch": 0.6582092681903963, + "grad_norm": 6.248321496759999, + "learning_rate": 1.817760959603162e-05, + "loss": 2.1289, + "step": 3139 + }, + { + "epoch": 0.6584189557559237, + "grad_norm": 6.09749499931928, + "learning_rate": 1.817630638959111e-05, + "loss": 1.7929, + "step": 3140 + }, + { + "epoch": 0.658628643321451, + "grad_norm": 6.431517971119232, + "learning_rate": 1.8175002764100714e-05, + "loss": 2.0492, + "step": 3141 + }, + { + "epoch": 0.6588383308869784, + "grad_norm": 5.302483425033261, + "learning_rate": 1.8173698719627245e-05, + "loss": 1.7843, + "step": 3142 + }, + { + "epoch": 0.6590480184525057, + "grad_norm": 6.6870306880329515, + "learning_rate": 1.8172394256237542e-05, + "loss": 1.8519, + "step": 3143 + }, + { + "epoch": 0.6592577060180331, + "grad_norm": 5.73647608556456, + "learning_rate": 1.817108937399846e-05, + "loss": 1.5772, + "step": 3144 + }, + { + "epoch": 0.6594673935835605, + "grad_norm": 6.183413960499327, + "learning_rate": 1.8169784072976875e-05, + "loss": 1.9048, + "step": 3145 + }, + { + "epoch": 0.6596770811490879, + "grad_norm": 6.016127980556942, + "learning_rate": 1.8168478353239685e-05, + "loss": 1.699, + "step": 3146 + }, + { + "epoch": 0.6598867687146153, + "grad_norm": 5.651412614976321, + "learning_rate": 1.816717221485381e-05, + "loss": 1.9678, + "step": 3147 + }, + { + "epoch": 0.6600964562801426, + "grad_norm": 5.941819015814283, + "learning_rate": 1.81658656578862e-05, + "loss": 1.769, + "step": 3148 + }, + { + "epoch": 0.66030614384567, + "grad_norm": 7.040237613396364, + "learning_rate": 1.8164558682403807e-05, + "loss": 1.7248, + "step": 3149 + }, + { + "epoch": 0.6605158314111973, + "grad_norm": 7.07162331659186, + "learning_rate": 1.8163251288473623e-05, + "loss": 1.9441, + "step": 3150 + }, + { + "epoch": 0.6607255189767247, + "grad_norm": 6.6492587134559065, + "learning_rate": 1.8161943476162652e-05, + "loss": 2.1139, + "step": 3151 + }, + { + "epoch": 0.660935206542252, + "grad_norm": 6.119506534146399, + "learning_rate": 1.816063524553792e-05, + "loss": 1.8542, + "step": 3152 + }, + { + "epoch": 0.6611448941077794, + "grad_norm": 5.76246901298754, + "learning_rate": 1.815932659666648e-05, + "loss": 1.6826, + "step": 3153 + }, + { + "epoch": 0.6613545816733067, + "grad_norm": 5.955827511190346, + "learning_rate": 1.8158017529615398e-05, + "loss": 2.3246, + "step": 3154 + }, + { + "epoch": 0.6615642692388342, + "grad_norm": 6.276874995340356, + "learning_rate": 1.815670804445177e-05, + "loss": 1.9315, + "step": 3155 + }, + { + "epoch": 0.6617739568043615, + "grad_norm": 6.094621171179124, + "learning_rate": 1.8155398141242707e-05, + "loss": 2.1074, + "step": 3156 + }, + { + "epoch": 0.6619836443698889, + "grad_norm": 5.979021126257763, + "learning_rate": 1.8154087820055345e-05, + "loss": 1.9156, + "step": 3157 + }, + { + "epoch": 0.6621933319354162, + "grad_norm": 6.205656242935968, + "learning_rate": 1.815277708095684e-05, + "loss": 1.8972, + "step": 3158 + }, + { + "epoch": 0.6624030195009436, + "grad_norm": 6.610765105239977, + "learning_rate": 1.815146592401437e-05, + "loss": 1.9385, + "step": 3159 + }, + { + "epoch": 0.6626127070664709, + "grad_norm": 6.963971902854117, + "learning_rate": 1.815015434929513e-05, + "loss": 2.191, + "step": 3160 + }, + { + "epoch": 0.6628223946319983, + "grad_norm": 6.085067625033411, + "learning_rate": 1.8148842356866344e-05, + "loss": 2.1619, + "step": 3161 + }, + { + "epoch": 0.6630320821975257, + "grad_norm": 6.375680670326424, + "learning_rate": 1.8147529946795256e-05, + "loss": 2.0022, + "step": 3162 + }, + { + "epoch": 0.663241769763053, + "grad_norm": 5.889061394899102, + "learning_rate": 1.814621711914913e-05, + "loss": 1.8135, + "step": 3163 + }, + { + "epoch": 0.6634514573285805, + "grad_norm": 6.977891226784837, + "learning_rate": 1.8144903873995244e-05, + "loss": 1.9745, + "step": 3164 + }, + { + "epoch": 0.6636611448941078, + "grad_norm": 5.614293695853713, + "learning_rate": 1.8143590211400907e-05, + "loss": 2.1332, + "step": 3165 + }, + { + "epoch": 0.6638708324596352, + "grad_norm": 6.494376532493229, + "learning_rate": 1.814227613143345e-05, + "loss": 1.9431, + "step": 3166 + }, + { + "epoch": 0.6640805200251625, + "grad_norm": 5.978828161545406, + "learning_rate": 1.8140961634160216e-05, + "loss": 1.7568, + "step": 3167 + }, + { + "epoch": 0.6642902075906899, + "grad_norm": 6.363451461464999, + "learning_rate": 1.813964671964858e-05, + "loss": 1.9357, + "step": 3168 + }, + { + "epoch": 0.6644998951562172, + "grad_norm": 6.428008802473638, + "learning_rate": 1.8138331387965932e-05, + "loss": 1.7691, + "step": 3169 + }, + { + "epoch": 0.6647095827217446, + "grad_norm": 7.025479013111277, + "learning_rate": 1.813701563917969e-05, + "loss": 2.3041, + "step": 3170 + }, + { + "epoch": 0.6649192702872719, + "grad_norm": 5.905271174198737, + "learning_rate": 1.8135699473357275e-05, + "loss": 1.9377, + "step": 3171 + }, + { + "epoch": 0.6651289578527994, + "grad_norm": 5.912755324416049, + "learning_rate": 1.8134382890566156e-05, + "loss": 1.9143, + "step": 3172 + }, + { + "epoch": 0.6653386454183267, + "grad_norm": 6.108354556888889, + "learning_rate": 1.8133065890873806e-05, + "loss": 1.7986, + "step": 3173 + }, + { + "epoch": 0.6655483329838541, + "grad_norm": 6.954863788148386, + "learning_rate": 1.813174847434772e-05, + "loss": 1.999, + "step": 3174 + }, + { + "epoch": 0.6657580205493814, + "grad_norm": 7.120107211779747, + "learning_rate": 1.8130430641055423e-05, + "loss": 2.0092, + "step": 3175 + }, + { + "epoch": 0.6659677081149088, + "grad_norm": 6.256929362352591, + "learning_rate": 1.8129112391064454e-05, + "loss": 1.9962, + "step": 3176 + }, + { + "epoch": 0.6661773956804361, + "grad_norm": 5.832168941411809, + "learning_rate": 1.8127793724442374e-05, + "loss": 1.7261, + "step": 3177 + }, + { + "epoch": 0.6663870832459635, + "grad_norm": 6.043609628309598, + "learning_rate": 1.812647464125677e-05, + "loss": 1.8731, + "step": 3178 + }, + { + "epoch": 0.6665967708114908, + "grad_norm": 6.106789644038843, + "learning_rate": 1.8125155141575246e-05, + "loss": 2.1362, + "step": 3179 + }, + { + "epoch": 0.6668064583770182, + "grad_norm": 5.632182617457892, + "learning_rate": 1.8123835225465432e-05, + "loss": 1.7333, + "step": 3180 + }, + { + "epoch": 0.6670161459425457, + "grad_norm": 5.444439523677967, + "learning_rate": 1.812251489299497e-05, + "loss": 1.6666, + "step": 3181 + }, + { + "epoch": 0.667225833508073, + "grad_norm": 6.804852434386522, + "learning_rate": 1.812119414423153e-05, + "loss": 1.627, + "step": 3182 + }, + { + "epoch": 0.6674355210736004, + "grad_norm": 5.893428175041616, + "learning_rate": 1.811987297924281e-05, + "loss": 2.4804, + "step": 3183 + }, + { + "epoch": 0.6676452086391277, + "grad_norm": 5.722632087161769, + "learning_rate": 1.811855139809651e-05, + "loss": 1.7029, + "step": 3184 + }, + { + "epoch": 0.6678548962046551, + "grad_norm": 5.863186777028515, + "learning_rate": 1.811722940086037e-05, + "loss": 2.3825, + "step": 3185 + }, + { + "epoch": 0.6680645837701824, + "grad_norm": 6.259101407820408, + "learning_rate": 1.811590698760215e-05, + "loss": 2.1024, + "step": 3186 + }, + { + "epoch": 0.6682742713357098, + "grad_norm": 5.515913406165411, + "learning_rate": 1.8114584158389615e-05, + "loss": 1.8011, + "step": 3187 + }, + { + "epoch": 0.6684839589012371, + "grad_norm": 5.745546846859845, + "learning_rate": 1.811326091329057e-05, + "loss": 1.6167, + "step": 3188 + }, + { + "epoch": 0.6686936464667645, + "grad_norm": 6.530695489850643, + "learning_rate": 1.8111937252372833e-05, + "loss": 1.674, + "step": 3189 + }, + { + "epoch": 0.6689033340322919, + "grad_norm": 6.235472766979741, + "learning_rate": 1.8110613175704243e-05, + "loss": 1.7064, + "step": 3190 + }, + { + "epoch": 0.6691130215978193, + "grad_norm": 5.609018975139266, + "learning_rate": 1.8109288683352656e-05, + "loss": 1.9706, + "step": 3191 + }, + { + "epoch": 0.6693227091633466, + "grad_norm": 6.415608593325858, + "learning_rate": 1.8107963775385963e-05, + "loss": 1.869, + "step": 3192 + }, + { + "epoch": 0.669532396728874, + "grad_norm": 5.981825121751691, + "learning_rate": 1.8106638451872065e-05, + "loss": 1.8882, + "step": 3193 + }, + { + "epoch": 0.6697420842944013, + "grad_norm": 5.759014511741097, + "learning_rate": 1.8105312712878884e-05, + "loss": 1.7248, + "step": 3194 + }, + { + "epoch": 0.6699517718599287, + "grad_norm": 6.165726035914848, + "learning_rate": 1.810398655847437e-05, + "loss": 1.8585, + "step": 3195 + }, + { + "epoch": 0.670161459425456, + "grad_norm": 5.623744311347049, + "learning_rate": 1.8102659988726485e-05, + "loss": 1.7314, + "step": 3196 + }, + { + "epoch": 0.6703711469909834, + "grad_norm": 6.7093902158353655, + "learning_rate": 1.8101333003703224e-05, + "loss": 1.8175, + "step": 3197 + }, + { + "epoch": 0.6705808345565109, + "grad_norm": 6.684429856044499, + "learning_rate": 1.8100005603472597e-05, + "loss": 1.9869, + "step": 3198 + }, + { + "epoch": 0.6707905221220382, + "grad_norm": 6.364278707254858, + "learning_rate": 1.8098677788102636e-05, + "loss": 1.9899, + "step": 3199 + }, + { + "epoch": 0.6710002096875656, + "grad_norm": 6.140847072729251, + "learning_rate": 1.809734955766139e-05, + "loss": 1.7042, + "step": 3200 + }, + { + "epoch": 0.6712098972530929, + "grad_norm": 6.21746854063012, + "learning_rate": 1.8096020912216932e-05, + "loss": 1.8085, + "step": 3201 + }, + { + "epoch": 0.6714195848186203, + "grad_norm": 6.033990021401269, + "learning_rate": 1.8094691851837366e-05, + "loss": 2.097, + "step": 3202 + }, + { + "epoch": 0.6716292723841476, + "grad_norm": 5.946716949760104, + "learning_rate": 1.80933623765908e-05, + "loss": 1.9133, + "step": 3203 + }, + { + "epoch": 0.671838959949675, + "grad_norm": 6.340688326596011, + "learning_rate": 1.8092032486545374e-05, + "loss": 1.8363, + "step": 3204 + }, + { + "epoch": 0.6720486475152023, + "grad_norm": 7.012621078767644, + "learning_rate": 1.8090702181769247e-05, + "loss": 2.3349, + "step": 3205 + }, + { + "epoch": 0.6722583350807297, + "grad_norm": 8.638722903648386, + "learning_rate": 1.8089371462330607e-05, + "loss": 1.5909, + "step": 3206 + }, + { + "epoch": 0.672468022646257, + "grad_norm": 6.362569720148795, + "learning_rate": 1.8088040328297644e-05, + "loss": 2.1324, + "step": 3207 + }, + { + "epoch": 0.6726777102117845, + "grad_norm": 5.389985791607408, + "learning_rate": 1.8086708779738588e-05, + "loss": 1.7115, + "step": 3208 + }, + { + "epoch": 0.6728873977773118, + "grad_norm": 7.0497891618394135, + "learning_rate": 1.808537681672168e-05, + "loss": 1.7978, + "step": 3209 + }, + { + "epoch": 0.6730970853428392, + "grad_norm": 5.81800661749597, + "learning_rate": 1.808404443931519e-05, + "loss": 1.9682, + "step": 3210 + }, + { + "epoch": 0.6733067729083665, + "grad_norm": 7.809124529242358, + "learning_rate": 1.8082711647587397e-05, + "loss": 1.9835, + "step": 3211 + }, + { + "epoch": 0.6735164604738939, + "grad_norm": 6.127695232328836, + "learning_rate": 1.8081378441606613e-05, + "loss": 2.1506, + "step": 3212 + }, + { + "epoch": 0.6737261480394212, + "grad_norm": 6.012062530311532, + "learning_rate": 1.808004482144117e-05, + "loss": 1.7999, + "step": 3213 + }, + { + "epoch": 0.6739358356049486, + "grad_norm": 5.805935547418881, + "learning_rate": 1.8078710787159415e-05, + "loss": 1.8612, + "step": 3214 + }, + { + "epoch": 0.6741455231704759, + "grad_norm": 6.755489030972783, + "learning_rate": 1.807737633882972e-05, + "loss": 1.9856, + "step": 3215 + }, + { + "epoch": 0.6743552107360034, + "grad_norm": 7.412176658113902, + "learning_rate": 1.8076041476520482e-05, + "loss": 2.2692, + "step": 3216 + }, + { + "epoch": 0.6745648983015308, + "grad_norm": 6.040010209966778, + "learning_rate": 1.807470620030011e-05, + "loss": 1.8721, + "step": 3217 + }, + { + "epoch": 0.6747745858670581, + "grad_norm": 5.886847429444838, + "learning_rate": 1.8073370510237035e-05, + "loss": 2.2316, + "step": 3218 + }, + { + "epoch": 0.6749842734325855, + "grad_norm": 5.880141646145295, + "learning_rate": 1.8072034406399722e-05, + "loss": 1.9596, + "step": 3219 + }, + { + "epoch": 0.6751939609981128, + "grad_norm": 6.2079691587909664, + "learning_rate": 1.8070697888856647e-05, + "loss": 1.8824, + "step": 3220 + }, + { + "epoch": 0.6754036485636402, + "grad_norm": 6.046700516667364, + "learning_rate": 1.8069360957676305e-05, + "loss": 2.1636, + "step": 3221 + }, + { + "epoch": 0.6756133361291675, + "grad_norm": 5.892219682228525, + "learning_rate": 1.806802361292722e-05, + "loss": 1.9598, + "step": 3222 + }, + { + "epoch": 0.6758230236946949, + "grad_norm": 6.7077331847729385, + "learning_rate": 1.806668585467793e-05, + "loss": 1.8905, + "step": 3223 + }, + { + "epoch": 0.6760327112602222, + "grad_norm": 5.604560933553084, + "learning_rate": 1.8065347682997e-05, + "loss": 1.7328, + "step": 3224 + }, + { + "epoch": 0.6762423988257497, + "grad_norm": 5.858242686694441, + "learning_rate": 1.8064009097953014e-05, + "loss": 2.0333, + "step": 3225 + }, + { + "epoch": 0.676452086391277, + "grad_norm": 5.285037726619053, + "learning_rate": 1.806267009961457e-05, + "loss": 1.8122, + "step": 3226 + }, + { + "epoch": 0.6766617739568044, + "grad_norm": 6.393781963263367, + "learning_rate": 1.8061330688050306e-05, + "loss": 1.9621, + "step": 3227 + }, + { + "epoch": 0.6768714615223317, + "grad_norm": 6.007219534272445, + "learning_rate": 1.8059990863328857e-05, + "loss": 2.2242, + "step": 3228 + }, + { + "epoch": 0.6770811490878591, + "grad_norm": 5.996645683955281, + "learning_rate": 1.8058650625518904e-05, + "loss": 1.7635, + "step": 3229 + }, + { + "epoch": 0.6772908366533864, + "grad_norm": 6.45365201126418, + "learning_rate": 1.8057309974689122e-05, + "loss": 2.1099, + "step": 3230 + }, + { + "epoch": 0.6775005242189138, + "grad_norm": 7.635443419482907, + "learning_rate": 1.8055968910908234e-05, + "loss": 1.7415, + "step": 3231 + }, + { + "epoch": 0.6777102117844411, + "grad_norm": 5.724378847655993, + "learning_rate": 1.805462743424496e-05, + "loss": 1.9009, + "step": 3232 + }, + { + "epoch": 0.6779198993499685, + "grad_norm": 6.324330869650399, + "learning_rate": 1.8053285544768065e-05, + "loss": 1.9163, + "step": 3233 + }, + { + "epoch": 0.678129586915496, + "grad_norm": 5.699776196850578, + "learning_rate": 1.8051943242546323e-05, + "loss": 1.7713, + "step": 3234 + }, + { + "epoch": 0.6783392744810233, + "grad_norm": 5.710073861730657, + "learning_rate": 1.805060052764852e-05, + "loss": 2.0643, + "step": 3235 + }, + { + "epoch": 0.6785489620465507, + "grad_norm": 5.462464488146465, + "learning_rate": 1.8049257400143476e-05, + "loss": 2.0281, + "step": 3236 + }, + { + "epoch": 0.678758649612078, + "grad_norm": 6.461800779497199, + "learning_rate": 1.804791386010003e-05, + "loss": 1.9406, + "step": 3237 + }, + { + "epoch": 0.6789683371776054, + "grad_norm": 5.9276722915766795, + "learning_rate": 1.804656990758704e-05, + "loss": 2.043, + "step": 3238 + }, + { + "epoch": 0.6791780247431327, + "grad_norm": 5.386339130503794, + "learning_rate": 1.8045225542673385e-05, + "loss": 1.8506, + "step": 3239 + }, + { + "epoch": 0.6793877123086601, + "grad_norm": 5.333098945509606, + "learning_rate": 1.804388076542797e-05, + "loss": 1.7762, + "step": 3240 + }, + { + "epoch": 0.6795973998741874, + "grad_norm": 5.939505899239394, + "learning_rate": 1.804253557591971e-05, + "loss": 2.145, + "step": 3241 + }, + { + "epoch": 0.6798070874397149, + "grad_norm": 5.453731557953743, + "learning_rate": 1.804118997421756e-05, + "loss": 1.7836, + "step": 3242 + }, + { + "epoch": 0.6800167750052422, + "grad_norm": 6.029157950361201, + "learning_rate": 1.803984396039047e-05, + "loss": 2.0945, + "step": 3243 + }, + { + "epoch": 0.6802264625707696, + "grad_norm": 6.497347529231123, + "learning_rate": 1.8038497534507438e-05, + "loss": 1.9885, + "step": 3244 + }, + { + "epoch": 0.6804361501362969, + "grad_norm": 6.014855966395795, + "learning_rate": 1.803715069663746e-05, + "loss": 1.9769, + "step": 3245 + }, + { + "epoch": 0.6806458377018243, + "grad_norm": 5.2524134894148045, + "learning_rate": 1.8035803446849574e-05, + "loss": 1.5727, + "step": 3246 + }, + { + "epoch": 0.6808555252673516, + "grad_norm": 6.957197559763476, + "learning_rate": 1.8034455785212822e-05, + "loss": 1.8986, + "step": 3247 + }, + { + "epoch": 0.681065212832879, + "grad_norm": 5.697668530796028, + "learning_rate": 1.8033107711796274e-05, + "loss": 1.8886, + "step": 3248 + }, + { + "epoch": 0.6812749003984063, + "grad_norm": 5.716232562746799, + "learning_rate": 1.8031759226669026e-05, + "loss": 1.8254, + "step": 3249 + }, + { + "epoch": 0.6814845879639337, + "grad_norm": 6.935459690598786, + "learning_rate": 1.8030410329900183e-05, + "loss": 2.2387, + "step": 3250 + }, + { + "epoch": 0.6816942755294612, + "grad_norm": 6.100784979379584, + "learning_rate": 1.802906102155889e-05, + "loss": 1.7785, + "step": 3251 + }, + { + "epoch": 0.6819039630949885, + "grad_norm": 6.244975202197563, + "learning_rate": 1.8027711301714288e-05, + "loss": 1.9904, + "step": 3252 + }, + { + "epoch": 0.6821136506605159, + "grad_norm": 6.529715717151556, + "learning_rate": 1.802636117043556e-05, + "loss": 2.036, + "step": 3253 + }, + { + "epoch": 0.6823233382260432, + "grad_norm": 6.07735667683381, + "learning_rate": 1.8025010627791898e-05, + "loss": 1.9323, + "step": 3254 + }, + { + "epoch": 0.6825330257915706, + "grad_norm": 5.38011765413242, + "learning_rate": 1.8023659673852527e-05, + "loss": 2.1259, + "step": 3255 + }, + { + "epoch": 0.6827427133570979, + "grad_norm": 7.150730400588157, + "learning_rate": 1.802230830868668e-05, + "loss": 2.0635, + "step": 3256 + }, + { + "epoch": 0.6829524009226253, + "grad_norm": 6.081813390930069, + "learning_rate": 1.8020956532363618e-05, + "loss": 1.9236, + "step": 3257 + }, + { + "epoch": 0.6831620884881526, + "grad_norm": 6.672804195158333, + "learning_rate": 1.8019604344952623e-05, + "loss": 1.914, + "step": 3258 + }, + { + "epoch": 0.68337177605368, + "grad_norm": 6.065829739684739, + "learning_rate": 1.8018251746522994e-05, + "loss": 1.9232, + "step": 3259 + }, + { + "epoch": 0.6835814636192074, + "grad_norm": 6.6451259402415985, + "learning_rate": 1.801689873714406e-05, + "loss": 2.0083, + "step": 3260 + }, + { + "epoch": 0.6837911511847348, + "grad_norm": 5.730980029079904, + "learning_rate": 1.8015545316885156e-05, + "loss": 1.7769, + "step": 3261 + }, + { + "epoch": 0.6840008387502621, + "grad_norm": 7.169657659611324, + "learning_rate": 1.8014191485815655e-05, + "loss": 2.0141, + "step": 3262 + }, + { + "epoch": 0.6842105263157895, + "grad_norm": 6.441308945278863, + "learning_rate": 1.8012837244004943e-05, + "loss": 1.9117, + "step": 3263 + }, + { + "epoch": 0.6844202138813168, + "grad_norm": 6.518712670344159, + "learning_rate": 1.801148259152242e-05, + "loss": 1.7167, + "step": 3264 + }, + { + "epoch": 0.6846299014468442, + "grad_norm": 6.976697898546187, + "learning_rate": 1.801012752843752e-05, + "loss": 2.0057, + "step": 3265 + }, + { + "epoch": 0.6848395890123715, + "grad_norm": 7.201288513312941, + "learning_rate": 1.800877205481969e-05, + "loss": 1.597, + "step": 3266 + }, + { + "epoch": 0.6850492765778989, + "grad_norm": 7.798083684687535, + "learning_rate": 1.8007416170738407e-05, + "loss": 2.0361, + "step": 3267 + }, + { + "epoch": 0.6852589641434262, + "grad_norm": 8.54899695963752, + "learning_rate": 1.800605987626315e-05, + "loss": 2.1742, + "step": 3268 + }, + { + "epoch": 0.6854686517089537, + "grad_norm": 7.230219947881708, + "learning_rate": 1.8004703171463445e-05, + "loss": 2.1699, + "step": 3269 + }, + { + "epoch": 0.6856783392744811, + "grad_norm": 7.296853109874198, + "learning_rate": 1.8003346056408813e-05, + "loss": 2.0253, + "step": 3270 + }, + { + "epoch": 0.6858880268400084, + "grad_norm": 6.671660838051233, + "learning_rate": 1.800198853116882e-05, + "loss": 2.1672, + "step": 3271 + }, + { + "epoch": 0.6860977144055358, + "grad_norm": 7.096620651572791, + "learning_rate": 1.8000630595813035e-05, + "loss": 2.0051, + "step": 3272 + }, + { + "epoch": 0.6863074019710631, + "grad_norm": 6.669286440996504, + "learning_rate": 1.7999272250411053e-05, + "loss": 1.9375, + "step": 3273 + }, + { + "epoch": 0.6865170895365905, + "grad_norm": 6.084563485225219, + "learning_rate": 1.7997913495032497e-05, + "loss": 2.0118, + "step": 3274 + }, + { + "epoch": 0.6867267771021178, + "grad_norm": 6.532932599629294, + "learning_rate": 1.7996554329747003e-05, + "loss": 1.777, + "step": 3275 + }, + { + "epoch": 0.6869364646676452, + "grad_norm": 5.56489706745577, + "learning_rate": 1.7995194754624225e-05, + "loss": 2.3794, + "step": 3276 + }, + { + "epoch": 0.6871461522331725, + "grad_norm": 5.824595590753818, + "learning_rate": 1.7993834769733856e-05, + "loss": 2.0992, + "step": 3277 + }, + { + "epoch": 0.6873558397987, + "grad_norm": 6.207846223883358, + "learning_rate": 1.7992474375145586e-05, + "loss": 1.8199, + "step": 3278 + }, + { + "epoch": 0.6875655273642273, + "grad_norm": 5.706951818363814, + "learning_rate": 1.7991113570929145e-05, + "loss": 1.8976, + "step": 3279 + }, + { + "epoch": 0.6877752149297547, + "grad_norm": 7.082112765803049, + "learning_rate": 1.798975235715427e-05, + "loss": 2.2045, + "step": 3280 + }, + { + "epoch": 0.687984902495282, + "grad_norm": 6.59657977612307, + "learning_rate": 1.798839073389073e-05, + "loss": 2.2119, + "step": 3281 + }, + { + "epoch": 0.6881945900608094, + "grad_norm": 6.075742386494613, + "learning_rate": 1.798702870120831e-05, + "loss": 2.0579, + "step": 3282 + }, + { + "epoch": 0.6884042776263367, + "grad_norm": 5.895622096392346, + "learning_rate": 1.7985666259176815e-05, + "loss": 1.8532, + "step": 3283 + }, + { + "epoch": 0.6886139651918641, + "grad_norm": 6.498667180294317, + "learning_rate": 1.7984303407866078e-05, + "loss": 1.6078, + "step": 3284 + }, + { + "epoch": 0.6888236527573914, + "grad_norm": 7.0110623323380565, + "learning_rate": 1.798294014734594e-05, + "loss": 1.9654, + "step": 3285 + }, + { + "epoch": 0.6890333403229189, + "grad_norm": 7.400397006988793, + "learning_rate": 1.7981576477686272e-05, + "loss": 1.725, + "step": 3286 + }, + { + "epoch": 0.6892430278884463, + "grad_norm": 6.1842453610868775, + "learning_rate": 1.798021239895697e-05, + "loss": 1.9258, + "step": 3287 + }, + { + "epoch": 0.6894527154539736, + "grad_norm": 5.813718125604629, + "learning_rate": 1.7978847911227938e-05, + "loss": 2.0203, + "step": 3288 + }, + { + "epoch": 0.689662403019501, + "grad_norm": 5.997215289742811, + "learning_rate": 1.7977483014569118e-05, + "loss": 2.0934, + "step": 3289 + }, + { + "epoch": 0.6898720905850283, + "grad_norm": 6.216867165912658, + "learning_rate": 1.7976117709050452e-05, + "loss": 1.9093, + "step": 3290 + }, + { + "epoch": 0.6900817781505557, + "grad_norm": 5.771666782904703, + "learning_rate": 1.7974751994741922e-05, + "loss": 1.7338, + "step": 3291 + }, + { + "epoch": 0.690291465716083, + "grad_norm": 6.089641578852998, + "learning_rate": 1.797338587171352e-05, + "loss": 1.7365, + "step": 3292 + }, + { + "epoch": 0.6905011532816104, + "grad_norm": 6.292746586876868, + "learning_rate": 1.797201934003526e-05, + "loss": 1.7147, + "step": 3293 + }, + { + "epoch": 0.6907108408471377, + "grad_norm": 6.270441726279288, + "learning_rate": 1.7970652399777185e-05, + "loss": 1.375, + "step": 3294 + }, + { + "epoch": 0.6909205284126652, + "grad_norm": 7.530413820118031, + "learning_rate": 1.7969285051009354e-05, + "loss": 1.7615, + "step": 3295 + }, + { + "epoch": 0.6911302159781925, + "grad_norm": 6.462696144853078, + "learning_rate": 1.7967917293801836e-05, + "loss": 1.9219, + "step": 3296 + }, + { + "epoch": 0.6913399035437199, + "grad_norm": 6.042842423204209, + "learning_rate": 1.796654912822474e-05, + "loss": 2.1346, + "step": 3297 + }, + { + "epoch": 0.6915495911092472, + "grad_norm": 5.807817323679866, + "learning_rate": 1.796518055434818e-05, + "loss": 1.6588, + "step": 3298 + }, + { + "epoch": 0.6917592786747746, + "grad_norm": 6.759429940389732, + "learning_rate": 1.7963811572242306e-05, + "loss": 1.8292, + "step": 3299 + }, + { + "epoch": 0.6919689662403019, + "grad_norm": 6.467017050341589, + "learning_rate": 1.7962442181977274e-05, + "loss": 1.9037, + "step": 3300 + }, + { + "epoch": 0.6921786538058293, + "grad_norm": 6.969881479071096, + "learning_rate": 1.796107238362327e-05, + "loss": 2.156, + "step": 3301 + }, + { + "epoch": 0.6923883413713566, + "grad_norm": 6.724932934258681, + "learning_rate": 1.7959702177250503e-05, + "loss": 1.8348, + "step": 3302 + }, + { + "epoch": 0.692598028936884, + "grad_norm": 6.577170506067217, + "learning_rate": 1.7958331562929188e-05, + "loss": 1.926, + "step": 3303 + }, + { + "epoch": 0.6928077165024114, + "grad_norm": 5.9009958327439005, + "learning_rate": 1.7956960540729582e-05, + "loss": 1.6556, + "step": 3304 + }, + { + "epoch": 0.6930174040679388, + "grad_norm": 7.116648502987047, + "learning_rate": 1.7955589110721946e-05, + "loss": 2.2525, + "step": 3305 + }, + { + "epoch": 0.6932270916334662, + "grad_norm": 6.606844015841355, + "learning_rate": 1.795421727297657e-05, + "loss": 2.0384, + "step": 3306 + }, + { + "epoch": 0.6934367791989935, + "grad_norm": 5.968572739395609, + "learning_rate": 1.7952845027563758e-05, + "loss": 1.8678, + "step": 3307 + }, + { + "epoch": 0.6936464667645209, + "grad_norm": 7.133876930285767, + "learning_rate": 1.7951472374553852e-05, + "loss": 1.7548, + "step": 3308 + }, + { + "epoch": 0.6938561543300482, + "grad_norm": 6.348168747093362, + "learning_rate": 1.7950099314017194e-05, + "loss": 1.812, + "step": 3309 + }, + { + "epoch": 0.6940658418955756, + "grad_norm": 5.773455863040737, + "learning_rate": 1.7948725846024153e-05, + "loss": 1.8128, + "step": 3310 + }, + { + "epoch": 0.6942755294611029, + "grad_norm": 5.904478604261118, + "learning_rate": 1.794735197064513e-05, + "loss": 1.7768, + "step": 3311 + }, + { + "epoch": 0.6944852170266304, + "grad_norm": 6.033866059869585, + "learning_rate": 1.7945977687950535e-05, + "loss": 1.7166, + "step": 3312 + }, + { + "epoch": 0.6946949045921577, + "grad_norm": 6.425278205080608, + "learning_rate": 1.79446029980108e-05, + "loss": 1.5628, + "step": 3313 + }, + { + "epoch": 0.6949045921576851, + "grad_norm": 6.628248806756656, + "learning_rate": 1.794322790089638e-05, + "loss": 2.0108, + "step": 3314 + }, + { + "epoch": 0.6951142797232124, + "grad_norm": 6.9408348481054825, + "learning_rate": 1.7941852396677757e-05, + "loss": 1.9478, + "step": 3315 + }, + { + "epoch": 0.6953239672887398, + "grad_norm": 6.896784300131047, + "learning_rate": 1.7940476485425424e-05, + "loss": 1.9928, + "step": 3316 + }, + { + "epoch": 0.6955336548542671, + "grad_norm": 6.506923866164512, + "learning_rate": 1.79391001672099e-05, + "loss": 2.1301, + "step": 3317 + }, + { + "epoch": 0.6957433424197945, + "grad_norm": 6.314360375140578, + "learning_rate": 1.793772344210172e-05, + "loss": 1.9989, + "step": 3318 + }, + { + "epoch": 0.6959530299853218, + "grad_norm": 5.6281101347474385, + "learning_rate": 1.793634631017145e-05, + "loss": 2.0314, + "step": 3319 + }, + { + "epoch": 0.6961627175508492, + "grad_norm": 5.279294921431496, + "learning_rate": 1.7934968771489663e-05, + "loss": 1.6397, + "step": 3320 + }, + { + "epoch": 0.6963724051163765, + "grad_norm": 5.926645901085874, + "learning_rate": 1.7933590826126964e-05, + "loss": 1.5286, + "step": 3321 + }, + { + "epoch": 0.696582092681904, + "grad_norm": 6.326367018370093, + "learning_rate": 1.7932212474153977e-05, + "loss": 2.4343, + "step": 3322 + }, + { + "epoch": 0.6967917802474314, + "grad_norm": 6.550728000977204, + "learning_rate": 1.793083371564134e-05, + "loss": 1.9441, + "step": 3323 + }, + { + "epoch": 0.6970014678129587, + "grad_norm": 5.863870606832622, + "learning_rate": 1.7929454550659723e-05, + "loss": 1.9915, + "step": 3324 + }, + { + "epoch": 0.6972111553784861, + "grad_norm": 5.236499753861973, + "learning_rate": 1.7928074979279806e-05, + "loss": 1.5367, + "step": 3325 + }, + { + "epoch": 0.6974208429440134, + "grad_norm": 6.350525551722535, + "learning_rate": 1.7926695001572295e-05, + "loss": 2.2273, + "step": 3326 + }, + { + "epoch": 0.6976305305095408, + "grad_norm": 7.140496641874133, + "learning_rate": 1.7925314617607918e-05, + "loss": 2.6077, + "step": 3327 + }, + { + "epoch": 0.6978402180750681, + "grad_norm": 6.743028823358207, + "learning_rate": 1.792393382745742e-05, + "loss": 2.1445, + "step": 3328 + }, + { + "epoch": 0.6980499056405955, + "grad_norm": 6.235060124169527, + "learning_rate": 1.7922552631191573e-05, + "loss": 1.9662, + "step": 3329 + }, + { + "epoch": 0.6982595932061229, + "grad_norm": 6.530651720778435, + "learning_rate": 1.7921171028881158e-05, + "loss": 1.7024, + "step": 3330 + }, + { + "epoch": 0.6984692807716503, + "grad_norm": 5.902371859499319, + "learning_rate": 1.7919789020596993e-05, + "loss": 1.891, + "step": 3331 + }, + { + "epoch": 0.6986789683371776, + "grad_norm": 6.127509526290477, + "learning_rate": 1.7918406606409905e-05, + "loss": 1.7999, + "step": 3332 + }, + { + "epoch": 0.698888655902705, + "grad_norm": 6.2433513930696405, + "learning_rate": 1.7917023786390742e-05, + "loss": 2.0957, + "step": 3333 + }, + { + "epoch": 0.6990983434682323, + "grad_norm": 5.741783854511772, + "learning_rate": 1.791564056061038e-05, + "loss": 2.03, + "step": 3334 + }, + { + "epoch": 0.6993080310337597, + "grad_norm": 7.270415587957947, + "learning_rate": 1.7914256929139712e-05, + "loss": 2.281, + "step": 3335 + }, + { + "epoch": 0.699517718599287, + "grad_norm": 5.771215892814484, + "learning_rate": 1.7912872892049648e-05, + "loss": 2.257, + "step": 3336 + }, + { + "epoch": 0.6997274061648144, + "grad_norm": 6.500445579308896, + "learning_rate": 1.791148844941112e-05, + "loss": 1.8889, + "step": 3337 + }, + { + "epoch": 0.6999370937303417, + "grad_norm": 5.245848412344272, + "learning_rate": 1.791010360129509e-05, + "loss": 1.5329, + "step": 3338 + }, + { + "epoch": 0.7001467812958692, + "grad_norm": 5.709369565352258, + "learning_rate": 1.7908718347772534e-05, + "loss": 2.0267, + "step": 3339 + }, + { + "epoch": 0.7003564688613965, + "grad_norm": 5.615565906771907, + "learning_rate": 1.7907332688914442e-05, + "loss": 1.8282, + "step": 3340 + }, + { + "epoch": 0.7005661564269239, + "grad_norm": 6.3906253558790675, + "learning_rate": 1.7905946624791837e-05, + "loss": 1.9056, + "step": 3341 + }, + { + "epoch": 0.7007758439924513, + "grad_norm": 5.33055691632164, + "learning_rate": 1.7904560155475756e-05, + "loss": 1.7913, + "step": 3342 + }, + { + "epoch": 0.7009855315579786, + "grad_norm": 6.921507024779086, + "learning_rate": 1.7903173281037256e-05, + "loss": 2.0845, + "step": 3343 + }, + { + "epoch": 0.701195219123506, + "grad_norm": 6.781939569117684, + "learning_rate": 1.7901786001547418e-05, + "loss": 2.1192, + "step": 3344 + }, + { + "epoch": 0.7014049066890333, + "grad_norm": 5.337855876491265, + "learning_rate": 1.7900398317077345e-05, + "loss": 1.3577, + "step": 3345 + }, + { + "epoch": 0.7016145942545607, + "grad_norm": 5.3120907022779305, + "learning_rate": 1.789901022769815e-05, + "loss": 1.6773, + "step": 3346 + }, + { + "epoch": 0.701824281820088, + "grad_norm": 6.891698336317076, + "learning_rate": 1.7897621733480988e-05, + "loss": 2.3032, + "step": 3347 + }, + { + "epoch": 0.7020339693856155, + "grad_norm": 6.200992405614746, + "learning_rate": 1.789623283449701e-05, + "loss": 2.0126, + "step": 3348 + }, + { + "epoch": 0.7022436569511428, + "grad_norm": 6.494642524254688, + "learning_rate": 1.7894843530817402e-05, + "loss": 2.3382, + "step": 3349 + }, + { + "epoch": 0.7024533445166702, + "grad_norm": 6.633325064273848, + "learning_rate": 1.7893453822513375e-05, + "loss": 1.7037, + "step": 3350 + }, + { + "epoch": 0.7026630320821975, + "grad_norm": 6.718264609507503, + "learning_rate": 1.7892063709656146e-05, + "loss": 2.0064, + "step": 3351 + }, + { + "epoch": 0.7028727196477249, + "grad_norm": 6.074534570205052, + "learning_rate": 1.7890673192316965e-05, + "loss": 2.0068, + "step": 3352 + }, + { + "epoch": 0.7030824072132522, + "grad_norm": 6.501024863482129, + "learning_rate": 1.7889282270567096e-05, + "loss": 1.8064, + "step": 3353 + }, + { + "epoch": 0.7032920947787796, + "grad_norm": 6.045726403512422, + "learning_rate": 1.788789094447783e-05, + "loss": 1.7423, + "step": 3354 + }, + { + "epoch": 0.7035017823443069, + "grad_norm": 6.051485253260123, + "learning_rate": 1.788649921412047e-05, + "loss": 2.0434, + "step": 3355 + }, + { + "epoch": 0.7037114699098344, + "grad_norm": 5.729579933642841, + "learning_rate": 1.788510707956635e-05, + "loss": 1.7226, + "step": 3356 + }, + { + "epoch": 0.7039211574753617, + "grad_norm": 6.192300906660107, + "learning_rate": 1.7883714540886808e-05, + "loss": 1.6466, + "step": 3357 + }, + { + "epoch": 0.7041308450408891, + "grad_norm": 7.558846996806186, + "learning_rate": 1.7882321598153228e-05, + "loss": 2.1446, + "step": 3358 + }, + { + "epoch": 0.7043405326064165, + "grad_norm": 7.608434490734879, + "learning_rate": 1.7880928251436994e-05, + "loss": 2.1729, + "step": 3359 + }, + { + "epoch": 0.7045502201719438, + "grad_norm": 6.796783426780534, + "learning_rate": 1.787953450080952e-05, + "loss": 1.9216, + "step": 3360 + }, + { + "epoch": 0.7047599077374712, + "grad_norm": 5.877450019657079, + "learning_rate": 1.7878140346342238e-05, + "loss": 1.8575, + "step": 3361 + }, + { + "epoch": 0.7049695953029985, + "grad_norm": 6.210527339023781, + "learning_rate": 1.7876745788106594e-05, + "loss": 1.912, + "step": 3362 + }, + { + "epoch": 0.7051792828685259, + "grad_norm": 6.347573630162063, + "learning_rate": 1.7875350826174068e-05, + "loss": 1.8913, + "step": 3363 + }, + { + "epoch": 0.7053889704340532, + "grad_norm": 4.937858577764903, + "learning_rate": 1.7873955460616156e-05, + "loss": 1.4774, + "step": 3364 + }, + { + "epoch": 0.7055986579995807, + "grad_norm": 6.480753068912573, + "learning_rate": 1.787255969150437e-05, + "loss": 2.1709, + "step": 3365 + }, + { + "epoch": 0.705808345565108, + "grad_norm": 6.426998139006233, + "learning_rate": 1.787116351891024e-05, + "loss": 2.1211, + "step": 3366 + }, + { + "epoch": 0.7060180331306354, + "grad_norm": 5.990185495027281, + "learning_rate": 1.7869766942905334e-05, + "loss": 1.705, + "step": 3367 + }, + { + "epoch": 0.7062277206961627, + "grad_norm": 6.031116383127443, + "learning_rate": 1.786836996356122e-05, + "loss": 1.9427, + "step": 3368 + }, + { + "epoch": 0.7064374082616901, + "grad_norm": 5.174960391881702, + "learning_rate": 1.78669725809495e-05, + "loss": 1.8994, + "step": 3369 + }, + { + "epoch": 0.7066470958272174, + "grad_norm": 5.470374618937298, + "learning_rate": 1.786557479514179e-05, + "loss": 1.9317, + "step": 3370 + }, + { + "epoch": 0.7068567833927448, + "grad_norm": 6.4224013499481245, + "learning_rate": 1.786417660620973e-05, + "loss": 1.8964, + "step": 3371 + }, + { + "epoch": 0.7070664709582721, + "grad_norm": 5.973755420207963, + "learning_rate": 1.7862778014224977e-05, + "loss": 2.202, + "step": 3372 + }, + { + "epoch": 0.7072761585237995, + "grad_norm": 6.481451465039927, + "learning_rate": 1.7861379019259216e-05, + "loss": 1.8383, + "step": 3373 + }, + { + "epoch": 0.7074858460893269, + "grad_norm": 5.86573343697531, + "learning_rate": 1.7859979621384144e-05, + "loss": 2.0065, + "step": 3374 + }, + { + "epoch": 0.7076955336548543, + "grad_norm": 6.145310765648565, + "learning_rate": 1.785857982067148e-05, + "loss": 2.0286, + "step": 3375 + }, + { + "epoch": 0.7079052212203817, + "grad_norm": 5.788095836267299, + "learning_rate": 1.7857179617192978e-05, + "loss": 2.0236, + "step": 3376 + }, + { + "epoch": 0.708114908785909, + "grad_norm": 5.39349603457023, + "learning_rate": 1.785577901102039e-05, + "loss": 2.1388, + "step": 3377 + }, + { + "epoch": 0.7083245963514364, + "grad_norm": 5.881033799477456, + "learning_rate": 1.78543780022255e-05, + "loss": 2.013, + "step": 3378 + }, + { + "epoch": 0.7085342839169637, + "grad_norm": 5.5549538177314455, + "learning_rate": 1.785297659088012e-05, + "loss": 1.8628, + "step": 3379 + }, + { + "epoch": 0.7087439714824911, + "grad_norm": 5.727954416263593, + "learning_rate": 1.7851574777056065e-05, + "loss": 1.9335, + "step": 3380 + }, + { + "epoch": 0.7089536590480184, + "grad_norm": 5.699244342926899, + "learning_rate": 1.7850172560825184e-05, + "loss": 2.0594, + "step": 3381 + }, + { + "epoch": 0.7091633466135459, + "grad_norm": 5.4522084708225105, + "learning_rate": 1.784876994225935e-05, + "loss": 1.7892, + "step": 3382 + }, + { + "epoch": 0.7093730341790732, + "grad_norm": 6.225143918071901, + "learning_rate": 1.7847366921430434e-05, + "loss": 1.6638, + "step": 3383 + }, + { + "epoch": 0.7095827217446006, + "grad_norm": 5.682056341800336, + "learning_rate": 1.7845963498410358e-05, + "loss": 1.8477, + "step": 3384 + }, + { + "epoch": 0.7097924093101279, + "grad_norm": 6.382984546014219, + "learning_rate": 1.7844559673271046e-05, + "loss": 1.8111, + "step": 3385 + }, + { + "epoch": 0.7100020968756553, + "grad_norm": 6.405746068556613, + "learning_rate": 1.7843155446084442e-05, + "loss": 1.9071, + "step": 3386 + }, + { + "epoch": 0.7102117844411826, + "grad_norm": 6.666519173822794, + "learning_rate": 1.784175081692252e-05, + "loss": 2.1207, + "step": 3387 + }, + { + "epoch": 0.71042147200671, + "grad_norm": 7.325758610660215, + "learning_rate": 1.7840345785857266e-05, + "loss": 2.1251, + "step": 3388 + }, + { + "epoch": 0.7106311595722373, + "grad_norm": 5.717875962677902, + "learning_rate": 1.7838940352960695e-05, + "loss": 2.2822, + "step": 3389 + }, + { + "epoch": 0.7108408471377647, + "grad_norm": 6.736777192336412, + "learning_rate": 1.7837534518304834e-05, + "loss": 1.768, + "step": 3390 + }, + { + "epoch": 0.711050534703292, + "grad_norm": 5.278442364109976, + "learning_rate": 1.7836128281961737e-05, + "loss": 1.8642, + "step": 3391 + }, + { + "epoch": 0.7112602222688195, + "grad_norm": 8.095060226811032, + "learning_rate": 1.783472164400347e-05, + "loss": 1.8471, + "step": 3392 + }, + { + "epoch": 0.7114699098343468, + "grad_norm": 7.429160335813844, + "learning_rate": 1.783331460450213e-05, + "loss": 1.7619, + "step": 3393 + }, + { + "epoch": 0.7116795973998742, + "grad_norm": 6.196982050832858, + "learning_rate": 1.7831907163529835e-05, + "loss": 1.6366, + "step": 3394 + }, + { + "epoch": 0.7118892849654016, + "grad_norm": 6.708373573212274, + "learning_rate": 1.7830499321158714e-05, + "loss": 1.9427, + "step": 3395 + }, + { + "epoch": 0.7120989725309289, + "grad_norm": 5.854080512468067, + "learning_rate": 1.782909107746092e-05, + "loss": 2.0476, + "step": 3396 + }, + { + "epoch": 0.7123086600964563, + "grad_norm": 5.8416101560721065, + "learning_rate": 1.7827682432508626e-05, + "loss": 2.0414, + "step": 3397 + }, + { + "epoch": 0.7125183476619836, + "grad_norm": 6.213846082530136, + "learning_rate": 1.7826273386374033e-05, + "loss": 1.8577, + "step": 3398 + }, + { + "epoch": 0.712728035227511, + "grad_norm": 6.5905807246393655, + "learning_rate": 1.7824863939129357e-05, + "loss": 1.8998, + "step": 3399 + }, + { + "epoch": 0.7129377227930384, + "grad_norm": 5.497870526272098, + "learning_rate": 1.782345409084683e-05, + "loss": 1.9242, + "step": 3400 + }, + { + "epoch": 0.7131474103585658, + "grad_norm": 6.723813363271084, + "learning_rate": 1.7822043841598713e-05, + "loss": 2.1381, + "step": 3401 + }, + { + "epoch": 0.7133570979240931, + "grad_norm": 6.9534229885764836, + "learning_rate": 1.7820633191457283e-05, + "loss": 2.1141, + "step": 3402 + }, + { + "epoch": 0.7135667854896205, + "grad_norm": 5.605395298320408, + "learning_rate": 1.7819222140494838e-05, + "loss": 1.9769, + "step": 3403 + }, + { + "epoch": 0.7137764730551478, + "grad_norm": 6.677481636063306, + "learning_rate": 1.7817810688783694e-05, + "loss": 2.0158, + "step": 3404 + }, + { + "epoch": 0.7139861606206752, + "grad_norm": 6.128042331587318, + "learning_rate": 1.7816398836396194e-05, + "loss": 1.9847, + "step": 3405 + }, + { + "epoch": 0.7141958481862025, + "grad_norm": 6.991839667544735, + "learning_rate": 1.78149865834047e-05, + "loss": 1.6439, + "step": 3406 + }, + { + "epoch": 0.7144055357517299, + "grad_norm": 5.997705132795919, + "learning_rate": 1.7813573929881583e-05, + "loss": 1.5963, + "step": 3407 + }, + { + "epoch": 0.7146152233172572, + "grad_norm": 6.755305063871415, + "learning_rate": 1.7812160875899257e-05, + "loss": 1.9644, + "step": 3408 + }, + { + "epoch": 0.7148249108827847, + "grad_norm": 6.491557307878263, + "learning_rate": 1.781074742153013e-05, + "loss": 1.7461, + "step": 3409 + }, + { + "epoch": 0.715034598448312, + "grad_norm": 5.939749072308881, + "learning_rate": 1.7809333566846655e-05, + "loss": 1.9942, + "step": 3410 + }, + { + "epoch": 0.7152442860138394, + "grad_norm": 6.709428110954093, + "learning_rate": 1.780791931192129e-05, + "loss": 1.9898, + "step": 3411 + }, + { + "epoch": 0.7154539735793668, + "grad_norm": 6.380867611272872, + "learning_rate": 1.780650465682652e-05, + "loss": 1.7385, + "step": 3412 + }, + { + "epoch": 0.7156636611448941, + "grad_norm": 8.219747417994995, + "learning_rate": 1.780508960163484e-05, + "loss": 1.8948, + "step": 3413 + }, + { + "epoch": 0.7158733487104215, + "grad_norm": 7.883292899507356, + "learning_rate": 1.780367414641879e-05, + "loss": 2.2643, + "step": 3414 + }, + { + "epoch": 0.7160830362759488, + "grad_norm": 6.66438142305715, + "learning_rate": 1.78022582912509e-05, + "loss": 1.5956, + "step": 3415 + }, + { + "epoch": 0.7162927238414762, + "grad_norm": 7.205451132554181, + "learning_rate": 1.7800842036203735e-05, + "loss": 2.2804, + "step": 3416 + }, + { + "epoch": 0.7165024114070035, + "grad_norm": 6.1311294261339455, + "learning_rate": 1.7799425381349892e-05, + "loss": 1.8552, + "step": 3417 + }, + { + "epoch": 0.716712098972531, + "grad_norm": 6.580376636123548, + "learning_rate": 1.779800832676197e-05, + "loss": 1.929, + "step": 3418 + }, + { + "epoch": 0.7169217865380583, + "grad_norm": 5.8636877000384375, + "learning_rate": 1.7796590872512597e-05, + "loss": 1.982, + "step": 3419 + }, + { + "epoch": 0.7171314741035857, + "grad_norm": 6.151850748745358, + "learning_rate": 1.7795173018674415e-05, + "loss": 1.9775, + "step": 3420 + }, + { + "epoch": 0.717341161669113, + "grad_norm": 6.301660568891588, + "learning_rate": 1.77937547653201e-05, + "loss": 2.0825, + "step": 3421 + }, + { + "epoch": 0.7175508492346404, + "grad_norm": 6.690238399078991, + "learning_rate": 1.7792336112522334e-05, + "loss": 2.2239, + "step": 3422 + }, + { + "epoch": 0.7177605368001677, + "grad_norm": 6.869769473789559, + "learning_rate": 1.7790917060353828e-05, + "loss": 1.9454, + "step": 3423 + }, + { + "epoch": 0.7179702243656951, + "grad_norm": 6.579389203467918, + "learning_rate": 1.7789497608887308e-05, + "loss": 1.7717, + "step": 3424 + }, + { + "epoch": 0.7181799119312224, + "grad_norm": 6.36883170452858, + "learning_rate": 1.778807775819553e-05, + "loss": 2.1342, + "step": 3425 + }, + { + "epoch": 0.7183895994967499, + "grad_norm": 5.822765740596492, + "learning_rate": 1.7786657508351254e-05, + "loss": 1.6524, + "step": 3426 + }, + { + "epoch": 0.7185992870622772, + "grad_norm": 6.091814132077698, + "learning_rate": 1.778523685942728e-05, + "loss": 2.0153, + "step": 3427 + }, + { + "epoch": 0.7188089746278046, + "grad_norm": 5.9271552982944895, + "learning_rate": 1.778381581149641e-05, + "loss": 1.8, + "step": 3428 + }, + { + "epoch": 0.7190186621933319, + "grad_norm": 6.657554735609322, + "learning_rate": 1.778239436463148e-05, + "loss": 2.0461, + "step": 3429 + }, + { + "epoch": 0.7192283497588593, + "grad_norm": 5.640561712555884, + "learning_rate": 1.778097251890534e-05, + "loss": 1.6521, + "step": 3430 + }, + { + "epoch": 0.7194380373243867, + "grad_norm": 6.788700653261892, + "learning_rate": 1.7779550274390867e-05, + "loss": 2.0177, + "step": 3431 + }, + { + "epoch": 0.719647724889914, + "grad_norm": 6.7073661504524225, + "learning_rate": 1.7778127631160948e-05, + "loss": 2.2163, + "step": 3432 + }, + { + "epoch": 0.7198574124554414, + "grad_norm": 6.190375259371967, + "learning_rate": 1.77767045892885e-05, + "loss": 2.1506, + "step": 3433 + }, + { + "epoch": 0.7200671000209687, + "grad_norm": 6.040513721093316, + "learning_rate": 1.7775281148846447e-05, + "loss": 1.6988, + "step": 3434 + }, + { + "epoch": 0.7202767875864962, + "grad_norm": 6.030079023648161, + "learning_rate": 1.7773857309907754e-05, + "loss": 1.831, + "step": 3435 + }, + { + "epoch": 0.7204864751520235, + "grad_norm": 6.79727168323244, + "learning_rate": 1.777243307254539e-05, + "loss": 1.6697, + "step": 3436 + }, + { + "epoch": 0.7206961627175509, + "grad_norm": 4.873972454541965, + "learning_rate": 1.7771008436832348e-05, + "loss": 1.3224, + "step": 3437 + }, + { + "epoch": 0.7209058502830782, + "grad_norm": 6.427984954055331, + "learning_rate": 1.776958340284165e-05, + "loss": 1.7845, + "step": 3438 + }, + { + "epoch": 0.7211155378486056, + "grad_norm": 7.004834857927032, + "learning_rate": 1.7768157970646326e-05, + "loss": 1.9751, + "step": 3439 + }, + { + "epoch": 0.7213252254141329, + "grad_norm": 5.917936110327826, + "learning_rate": 1.7766732140319432e-05, + "loss": 1.8962, + "step": 3440 + }, + { + "epoch": 0.7215349129796603, + "grad_norm": 7.552700500372858, + "learning_rate": 1.7765305911934043e-05, + "loss": 1.8722, + "step": 3441 + }, + { + "epoch": 0.7217446005451876, + "grad_norm": 8.553586489881292, + "learning_rate": 1.776387928556326e-05, + "loss": 2.1712, + "step": 3442 + }, + { + "epoch": 0.721954288110715, + "grad_norm": 6.550936595278142, + "learning_rate": 1.7762452261280198e-05, + "loss": 1.793, + "step": 3443 + }, + { + "epoch": 0.7221639756762424, + "grad_norm": 5.722258004857852, + "learning_rate": 1.7761024839157994e-05, + "loss": 1.8859, + "step": 3444 + }, + { + "epoch": 0.7223736632417698, + "grad_norm": 6.050880026365092, + "learning_rate": 1.7759597019269807e-05, + "loss": 1.8545, + "step": 3445 + }, + { + "epoch": 0.7225833508072971, + "grad_norm": 5.970337553784131, + "learning_rate": 1.7758168801688812e-05, + "loss": 1.627, + "step": 3446 + }, + { + "epoch": 0.7227930383728245, + "grad_norm": 5.497133917244876, + "learning_rate": 1.775674018648821e-05, + "loss": 1.6159, + "step": 3447 + }, + { + "epoch": 0.7230027259383519, + "grad_norm": 5.879714716802198, + "learning_rate": 1.7755311173741222e-05, + "loss": 1.4422, + "step": 3448 + }, + { + "epoch": 0.7232124135038792, + "grad_norm": 6.272006030678814, + "learning_rate": 1.7753881763521083e-05, + "loss": 2.1879, + "step": 3449 + }, + { + "epoch": 0.7234221010694066, + "grad_norm": 5.315662421164023, + "learning_rate": 1.775245195590106e-05, + "loss": 1.6826, + "step": 3450 + }, + { + "epoch": 0.7236317886349339, + "grad_norm": 6.182095613654355, + "learning_rate": 1.775102175095442e-05, + "loss": 1.9166, + "step": 3451 + }, + { + "epoch": 0.7238414762004614, + "grad_norm": 6.204622240673169, + "learning_rate": 1.774959114875448e-05, + "loss": 1.9598, + "step": 3452 + }, + { + "epoch": 0.7240511637659887, + "grad_norm": 7.08086173344824, + "learning_rate": 1.7748160149374553e-05, + "loss": 2.2209, + "step": 3453 + }, + { + "epoch": 0.7242608513315161, + "grad_norm": 5.474799221709456, + "learning_rate": 1.7746728752887973e-05, + "loss": 1.7998, + "step": 3454 + }, + { + "epoch": 0.7244705388970434, + "grad_norm": 6.905784551197978, + "learning_rate": 1.7745296959368114e-05, + "loss": 1.6539, + "step": 3455 + }, + { + "epoch": 0.7246802264625708, + "grad_norm": 5.870819503536297, + "learning_rate": 1.774386476888835e-05, + "loss": 2.0138, + "step": 3456 + }, + { + "epoch": 0.7248899140280981, + "grad_norm": 7.692670084785689, + "learning_rate": 1.7742432181522087e-05, + "loss": 1.9974, + "step": 3457 + }, + { + "epoch": 0.7250996015936255, + "grad_norm": 5.801440544263104, + "learning_rate": 1.7740999197342745e-05, + "loss": 1.5593, + "step": 3458 + }, + { + "epoch": 0.7253092891591528, + "grad_norm": 6.798539837196515, + "learning_rate": 1.7739565816423768e-05, + "loss": 2.2256, + "step": 3459 + }, + { + "epoch": 0.7255189767246802, + "grad_norm": 6.745810902187541, + "learning_rate": 1.773813203883862e-05, + "loss": 1.8935, + "step": 3460 + }, + { + "epoch": 0.7257286642902075, + "grad_norm": 6.2690184499351895, + "learning_rate": 1.773669786466078e-05, + "loss": 1.8115, + "step": 3461 + }, + { + "epoch": 0.725938351855735, + "grad_norm": 5.979825686970501, + "learning_rate": 1.7735263293963762e-05, + "loss": 1.8996, + "step": 3462 + }, + { + "epoch": 0.7261480394212623, + "grad_norm": 6.069687177172668, + "learning_rate": 1.7733828326821082e-05, + "loss": 1.5786, + "step": 3463 + }, + { + "epoch": 0.7263577269867897, + "grad_norm": 6.616850687043835, + "learning_rate": 1.7732392963306283e-05, + "loss": 1.9543, + "step": 3464 + }, + { + "epoch": 0.726567414552317, + "grad_norm": 6.054120991260907, + "learning_rate": 1.773095720349294e-05, + "loss": 1.9633, + "step": 3465 + }, + { + "epoch": 0.7267771021178444, + "grad_norm": 6.646569966368481, + "learning_rate": 1.7729521047454626e-05, + "loss": 1.6973, + "step": 3466 + }, + { + "epoch": 0.7269867896833718, + "grad_norm": 6.299582360081332, + "learning_rate": 1.7728084495264952e-05, + "loss": 1.9104, + "step": 3467 + }, + { + "epoch": 0.7271964772488991, + "grad_norm": 6.17159329839842, + "learning_rate": 1.7726647546997546e-05, + "loss": 2.1055, + "step": 3468 + }, + { + "epoch": 0.7274061648144265, + "grad_norm": 6.994537329881502, + "learning_rate": 1.7725210202726052e-05, + "loss": 1.6156, + "step": 3469 + }, + { + "epoch": 0.7276158523799539, + "grad_norm": 6.857116781932151, + "learning_rate": 1.7723772462524138e-05, + "loss": 1.7521, + "step": 3470 + }, + { + "epoch": 0.7278255399454813, + "grad_norm": 6.637617934924966, + "learning_rate": 1.7722334326465486e-05, + "loss": 2.0023, + "step": 3471 + }, + { + "epoch": 0.7280352275110086, + "grad_norm": 8.33253284881359, + "learning_rate": 1.7720895794623808e-05, + "loss": 1.8228, + "step": 3472 + }, + { + "epoch": 0.728244915076536, + "grad_norm": 5.9730835066963035, + "learning_rate": 1.7719456867072827e-05, + "loss": 1.6865, + "step": 3473 + }, + { + "epoch": 0.7284546026420633, + "grad_norm": 6.621109317941283, + "learning_rate": 1.7718017543886298e-05, + "loss": 1.3069, + "step": 3474 + }, + { + "epoch": 0.7286642902075907, + "grad_norm": 6.48723851726757, + "learning_rate": 1.771657782513798e-05, + "loss": 2.0635, + "step": 3475 + }, + { + "epoch": 0.728873977773118, + "grad_norm": 6.755567726469187, + "learning_rate": 1.7715137710901665e-05, + "loss": 2.0183, + "step": 3476 + }, + { + "epoch": 0.7290836653386454, + "grad_norm": 6.522208902453115, + "learning_rate": 1.771369720125116e-05, + "loss": 2.0141, + "step": 3477 + }, + { + "epoch": 0.7292933529041727, + "grad_norm": 7.386968784195653, + "learning_rate": 1.7712256296260296e-05, + "loss": 1.9449, + "step": 3478 + }, + { + "epoch": 0.7295030404697002, + "grad_norm": 7.402302800520711, + "learning_rate": 1.771081499600292e-05, + "loss": 2.2637, + "step": 3479 + }, + { + "epoch": 0.7297127280352275, + "grad_norm": 6.237474209578294, + "learning_rate": 1.7709373300552904e-05, + "loss": 1.6653, + "step": 3480 + }, + { + "epoch": 0.7299224156007549, + "grad_norm": 5.4814038472193305, + "learning_rate": 1.770793120998413e-05, + "loss": 1.6939, + "step": 3481 + }, + { + "epoch": 0.7301321031662822, + "grad_norm": 6.025095160487646, + "learning_rate": 1.770648872437052e-05, + "loss": 1.4771, + "step": 3482 + }, + { + "epoch": 0.7303417907318096, + "grad_norm": 6.046824635867647, + "learning_rate": 1.770504584378599e-05, + "loss": 1.6821, + "step": 3483 + }, + { + "epoch": 0.730551478297337, + "grad_norm": 5.785323693565155, + "learning_rate": 1.77036025683045e-05, + "loss": 1.4389, + "step": 3484 + }, + { + "epoch": 0.7307611658628643, + "grad_norm": 6.790880000296394, + "learning_rate": 1.7702158898000017e-05, + "loss": 2.0217, + "step": 3485 + }, + { + "epoch": 0.7309708534283917, + "grad_norm": 8.108517432852935, + "learning_rate": 1.770071483294653e-05, + "loss": 1.9364, + "step": 3486 + }, + { + "epoch": 0.731180540993919, + "grad_norm": 5.326151728198533, + "learning_rate": 1.7699270373218054e-05, + "loss": 1.7126, + "step": 3487 + }, + { + "epoch": 0.7313902285594465, + "grad_norm": 7.089225512732965, + "learning_rate": 1.7697825518888616e-05, + "loss": 2.2343, + "step": 3488 + }, + { + "epoch": 0.7315999161249738, + "grad_norm": 6.322263670807007, + "learning_rate": 1.7696380270032268e-05, + "loss": 2.0287, + "step": 3489 + }, + { + "epoch": 0.7318096036905012, + "grad_norm": 6.158213524366527, + "learning_rate": 1.7694934626723088e-05, + "loss": 1.9255, + "step": 3490 + }, + { + "epoch": 0.7320192912560285, + "grad_norm": 6.843910307944522, + "learning_rate": 1.7693488589035158e-05, + "loss": 1.5493, + "step": 3491 + }, + { + "epoch": 0.7322289788215559, + "grad_norm": 7.633805551519663, + "learning_rate": 1.7692042157042593e-05, + "loss": 1.6487, + "step": 3492 + }, + { + "epoch": 0.7324386663870832, + "grad_norm": 7.45311166854082, + "learning_rate": 1.769059533081953e-05, + "loss": 2.0295, + "step": 3493 + }, + { + "epoch": 0.7326483539526106, + "grad_norm": 6.25105710000813, + "learning_rate": 1.7689148110440114e-05, + "loss": 1.5065, + "step": 3494 + }, + { + "epoch": 0.7328580415181379, + "grad_norm": 6.930273741112612, + "learning_rate": 1.7687700495978523e-05, + "loss": 1.5519, + "step": 3495 + }, + { + "epoch": 0.7330677290836654, + "grad_norm": 6.481925845384008, + "learning_rate": 1.7686252487508948e-05, + "loss": 1.7427, + "step": 3496 + }, + { + "epoch": 0.7332774166491927, + "grad_norm": 6.829471511972074, + "learning_rate": 1.76848040851056e-05, + "loss": 2.1041, + "step": 3497 + }, + { + "epoch": 0.7334871042147201, + "grad_norm": 6.590349377260757, + "learning_rate": 1.7683355288842715e-05, + "loss": 1.8367, + "step": 3498 + }, + { + "epoch": 0.7336967917802474, + "grad_norm": 6.882213541841275, + "learning_rate": 1.768190609879455e-05, + "loss": 1.955, + "step": 3499 + }, + { + "epoch": 0.7339064793457748, + "grad_norm": 5.139295179718113, + "learning_rate": 1.7680456515035366e-05, + "loss": 1.6745, + "step": 3500 + }, + { + "epoch": 0.7341161669113022, + "grad_norm": 7.716769189430109, + "learning_rate": 1.7679006537639467e-05, + "loss": 2.0494, + "step": 3501 + }, + { + "epoch": 0.7343258544768295, + "grad_norm": 6.271225379507168, + "learning_rate": 1.7677556166681164e-05, + "loss": 1.6788, + "step": 3502 + }, + { + "epoch": 0.7345355420423569, + "grad_norm": 6.164585256650097, + "learning_rate": 1.7676105402234793e-05, + "loss": 2.1201, + "step": 3503 + }, + { + "epoch": 0.7347452296078842, + "grad_norm": 7.786761850429331, + "learning_rate": 1.7674654244374707e-05, + "loss": 2.1877, + "step": 3504 + }, + { + "epoch": 0.7349549171734117, + "grad_norm": 7.573615646008374, + "learning_rate": 1.767320269317528e-05, + "loss": 1.8912, + "step": 3505 + }, + { + "epoch": 0.735164604738939, + "grad_norm": 6.349820106832358, + "learning_rate": 1.7671750748710907e-05, + "loss": 1.7058, + "step": 3506 + }, + { + "epoch": 0.7353742923044664, + "grad_norm": 7.643900051261221, + "learning_rate": 1.7670298411056003e-05, + "loss": 2.1291, + "step": 3507 + }, + { + "epoch": 0.7355839798699937, + "grad_norm": 6.287763047285248, + "learning_rate": 1.7668845680284997e-05, + "loss": 1.7857, + "step": 3508 + }, + { + "epoch": 0.7357936674355211, + "grad_norm": 7.508387035284511, + "learning_rate": 1.7667392556472353e-05, + "loss": 1.9974, + "step": 3509 + }, + { + "epoch": 0.7360033550010484, + "grad_norm": 6.425994815871797, + "learning_rate": 1.766593903969254e-05, + "loss": 2.0255, + "step": 3510 + }, + { + "epoch": 0.7362130425665758, + "grad_norm": 6.639558369019603, + "learning_rate": 1.7664485130020058e-05, + "loss": 1.7627, + "step": 3511 + }, + { + "epoch": 0.7364227301321031, + "grad_norm": 6.453931121376507, + "learning_rate": 1.7663030827529416e-05, + "loss": 1.8288, + "step": 3512 + }, + { + "epoch": 0.7366324176976305, + "grad_norm": 6.327640958216904, + "learning_rate": 1.7661576132295157e-05, + "loss": 2.0013, + "step": 3513 + }, + { + "epoch": 0.7368421052631579, + "grad_norm": 6.290983683195035, + "learning_rate": 1.7660121044391833e-05, + "loss": 1.8652, + "step": 3514 + }, + { + "epoch": 0.7370517928286853, + "grad_norm": 6.515871861467369, + "learning_rate": 1.7658665563894016e-05, + "loss": 1.7338, + "step": 3515 + }, + { + "epoch": 0.7372614803942126, + "grad_norm": 6.119074266501888, + "learning_rate": 1.7657209690876307e-05, + "loss": 1.6997, + "step": 3516 + }, + { + "epoch": 0.73747116795974, + "grad_norm": 7.466766158790882, + "learning_rate": 1.7655753425413322e-05, + "loss": 1.9961, + "step": 3517 + }, + { + "epoch": 0.7376808555252673, + "grad_norm": 7.272743425278089, + "learning_rate": 1.7654296767579695e-05, + "loss": 1.7095, + "step": 3518 + }, + { + "epoch": 0.7378905430907947, + "grad_norm": 6.297719906303153, + "learning_rate": 1.7652839717450084e-05, + "loss": 1.4778, + "step": 3519 + }, + { + "epoch": 0.7381002306563221, + "grad_norm": 6.545851691498975, + "learning_rate": 1.7651382275099163e-05, + "loss": 1.9275, + "step": 3520 + }, + { + "epoch": 0.7383099182218494, + "grad_norm": 5.975519901383928, + "learning_rate": 1.764992444060163e-05, + "loss": 1.5358, + "step": 3521 + }, + { + "epoch": 0.7385196057873769, + "grad_norm": 7.350344199701822, + "learning_rate": 1.76484662140322e-05, + "loss": 1.9711, + "step": 3522 + }, + { + "epoch": 0.7387292933529042, + "grad_norm": 5.780397576399049, + "learning_rate": 1.7647007595465613e-05, + "loss": 1.679, + "step": 3523 + }, + { + "epoch": 0.7389389809184316, + "grad_norm": 6.3756078183149505, + "learning_rate": 1.764554858497662e-05, + "loss": 1.8165, + "step": 3524 + }, + { + "epoch": 0.7391486684839589, + "grad_norm": 6.423580888657026, + "learning_rate": 1.7644089182640005e-05, + "loss": 2.0505, + "step": 3525 + }, + { + "epoch": 0.7393583560494863, + "grad_norm": 6.029032841688427, + "learning_rate": 1.7642629388530563e-05, + "loss": 1.6834, + "step": 3526 + }, + { + "epoch": 0.7395680436150136, + "grad_norm": 6.420524565976833, + "learning_rate": 1.7641169202723104e-05, + "loss": 2.0454, + "step": 3527 + }, + { + "epoch": 0.739777731180541, + "grad_norm": 7.882987302948785, + "learning_rate": 1.7639708625292473e-05, + "loss": 2.2519, + "step": 3528 + }, + { + "epoch": 0.7399874187460683, + "grad_norm": 6.210174798919756, + "learning_rate": 1.7638247656313527e-05, + "loss": 1.687, + "step": 3529 + }, + { + "epoch": 0.7401971063115957, + "grad_norm": 6.067006147804724, + "learning_rate": 1.7636786295861136e-05, + "loss": 2.0509, + "step": 3530 + }, + { + "epoch": 0.740406793877123, + "grad_norm": 5.873371250416261, + "learning_rate": 1.7635324544010206e-05, + "loss": 1.6983, + "step": 3531 + }, + { + "epoch": 0.7406164814426505, + "grad_norm": 5.680615814440841, + "learning_rate": 1.763386240083565e-05, + "loss": 1.9874, + "step": 3532 + }, + { + "epoch": 0.7408261690081778, + "grad_norm": 5.766275247099527, + "learning_rate": 1.7632399866412405e-05, + "loss": 1.8346, + "step": 3533 + }, + { + "epoch": 0.7410358565737052, + "grad_norm": 4.865901757488386, + "learning_rate": 1.7630936940815425e-05, + "loss": 1.8454, + "step": 3534 + }, + { + "epoch": 0.7412455441392325, + "grad_norm": 6.497715455120673, + "learning_rate": 1.7629473624119697e-05, + "loss": 1.7203, + "step": 3535 + }, + { + "epoch": 0.7414552317047599, + "grad_norm": 7.4412090555731, + "learning_rate": 1.762800991640021e-05, + "loss": 1.9602, + "step": 3536 + }, + { + "epoch": 0.7416649192702873, + "grad_norm": 6.283537151889722, + "learning_rate": 1.762654581773199e-05, + "loss": 1.8062, + "step": 3537 + }, + { + "epoch": 0.7418746068358146, + "grad_norm": 6.136979182549098, + "learning_rate": 1.7625081328190064e-05, + "loss": 1.8542, + "step": 3538 + }, + { + "epoch": 0.742084294401342, + "grad_norm": 7.438125356661845, + "learning_rate": 1.7623616447849496e-05, + "loss": 1.9401, + "step": 3539 + }, + { + "epoch": 0.7422939819668694, + "grad_norm": 6.139641260440515, + "learning_rate": 1.7622151176785363e-05, + "loss": 1.9455, + "step": 3540 + }, + { + "epoch": 0.7425036695323968, + "grad_norm": 5.922848312930531, + "learning_rate": 1.762068551507276e-05, + "loss": 1.564, + "step": 3541 + }, + { + "epoch": 0.7427133570979241, + "grad_norm": 6.214216444154346, + "learning_rate": 1.7619219462786813e-05, + "loss": 1.8344, + "step": 3542 + }, + { + "epoch": 0.7429230446634515, + "grad_norm": 6.262885624515905, + "learning_rate": 1.7617753020002647e-05, + "loss": 1.9541, + "step": 3543 + }, + { + "epoch": 0.7431327322289788, + "grad_norm": 7.424142984649746, + "learning_rate": 1.761628618679543e-05, + "loss": 1.9596, + "step": 3544 + }, + { + "epoch": 0.7433424197945062, + "grad_norm": 6.728248078324123, + "learning_rate": 1.7614818963240337e-05, + "loss": 1.53, + "step": 3545 + }, + { + "epoch": 0.7435521073600335, + "grad_norm": 6.549823976519318, + "learning_rate": 1.7613351349412562e-05, + "loss": 1.8263, + "step": 3546 + }, + { + "epoch": 0.7437617949255609, + "grad_norm": 6.87609312026156, + "learning_rate": 1.7611883345387327e-05, + "loss": 1.975, + "step": 3547 + }, + { + "epoch": 0.7439714824910882, + "grad_norm": 6.842420137979952, + "learning_rate": 1.761041495123987e-05, + "loss": 1.5892, + "step": 3548 + }, + { + "epoch": 0.7441811700566157, + "grad_norm": 7.148329183404065, + "learning_rate": 1.7608946167045444e-05, + "loss": 1.7235, + "step": 3549 + }, + { + "epoch": 0.744390857622143, + "grad_norm": 6.07406070172432, + "learning_rate": 1.760747699287933e-05, + "loss": 2.239, + "step": 3550 + }, + { + "epoch": 0.7446005451876704, + "grad_norm": 5.908149400892406, + "learning_rate": 1.7606007428816826e-05, + "loss": 2.0016, + "step": 3551 + }, + { + "epoch": 0.7448102327531977, + "grad_norm": 5.9254212630017555, + "learning_rate": 1.7604537474933252e-05, + "loss": 1.7518, + "step": 3552 + }, + { + "epoch": 0.7450199203187251, + "grad_norm": 6.14527800512368, + "learning_rate": 1.760306713130394e-05, + "loss": 2.2464, + "step": 3553 + }, + { + "epoch": 0.7452296078842524, + "grad_norm": 6.76904976285918, + "learning_rate": 1.7601596398004255e-05, + "loss": 1.6643, + "step": 3554 + }, + { + "epoch": 0.7454392954497798, + "grad_norm": 5.550949931182336, + "learning_rate": 1.7600125275109565e-05, + "loss": 1.6049, + "step": 3555 + }, + { + "epoch": 0.7456489830153072, + "grad_norm": 6.552466739811548, + "learning_rate": 1.7598653762695275e-05, + "loss": 1.8545, + "step": 3556 + }, + { + "epoch": 0.7458586705808345, + "grad_norm": 7.405786551213001, + "learning_rate": 1.7597181860836802e-05, + "loss": 1.9707, + "step": 3557 + }, + { + "epoch": 0.746068358146362, + "grad_norm": 7.280680392664758, + "learning_rate": 1.759570956960958e-05, + "loss": 1.9457, + "step": 3558 + }, + { + "epoch": 0.7462780457118893, + "grad_norm": 6.350353415331434, + "learning_rate": 1.759423688908907e-05, + "loss": 1.6103, + "step": 3559 + }, + { + "epoch": 0.7464877332774167, + "grad_norm": 7.251341324953078, + "learning_rate": 1.7592763819350746e-05, + "loss": 2.1117, + "step": 3560 + }, + { + "epoch": 0.746697420842944, + "grad_norm": 7.440507141373042, + "learning_rate": 1.759129036047011e-05, + "loss": 2.0942, + "step": 3561 + }, + { + "epoch": 0.7469071084084714, + "grad_norm": 6.746596763248542, + "learning_rate": 1.758981651252267e-05, + "loss": 1.8331, + "step": 3562 + }, + { + "epoch": 0.7471167959739987, + "grad_norm": 6.687846637146918, + "learning_rate": 1.758834227558398e-05, + "loss": 1.6062, + "step": 3563 + }, + { + "epoch": 0.7473264835395261, + "grad_norm": 7.250030562987938, + "learning_rate": 1.758686764972958e-05, + "loss": 1.9847, + "step": 3564 + }, + { + "epoch": 0.7475361711050534, + "grad_norm": 6.317939530622629, + "learning_rate": 1.7585392635035055e-05, + "loss": 1.7993, + "step": 3565 + }, + { + "epoch": 0.7477458586705809, + "grad_norm": 6.9416785543652155, + "learning_rate": 1.7583917231576002e-05, + "loss": 2.0835, + "step": 3566 + }, + { + "epoch": 0.7479555462361082, + "grad_norm": 6.030863805436057, + "learning_rate": 1.7582441439428037e-05, + "loss": 2.0406, + "step": 3567 + }, + { + "epoch": 0.7481652338016356, + "grad_norm": 7.284693143816317, + "learning_rate": 1.7580965258666797e-05, + "loss": 1.6491, + "step": 3568 + }, + { + "epoch": 0.7483749213671629, + "grad_norm": 6.137328105493726, + "learning_rate": 1.757948868936794e-05, + "loss": 1.6784, + "step": 3569 + }, + { + "epoch": 0.7485846089326903, + "grad_norm": 6.261534894003862, + "learning_rate": 1.757801173160714e-05, + "loss": 1.7973, + "step": 3570 + }, + { + "epoch": 0.7487942964982176, + "grad_norm": 5.872047246636773, + "learning_rate": 1.7576534385460096e-05, + "loss": 1.7136, + "step": 3571 + }, + { + "epoch": 0.749003984063745, + "grad_norm": 7.304200893656662, + "learning_rate": 1.757505665100253e-05, + "loss": 1.9465, + "step": 3572 + }, + { + "epoch": 0.7492136716292724, + "grad_norm": 7.148031510096267, + "learning_rate": 1.7573578528310164e-05, + "loss": 1.865, + "step": 3573 + }, + { + "epoch": 0.7494233591947997, + "grad_norm": 6.7531473109553275, + "learning_rate": 1.7572100017458766e-05, + "loss": 2.1642, + "step": 3574 + }, + { + "epoch": 0.7496330467603272, + "grad_norm": 5.944393437657226, + "learning_rate": 1.757062111852411e-05, + "loss": 1.5782, + "step": 3575 + }, + { + "epoch": 0.7498427343258545, + "grad_norm": 5.125368828084269, + "learning_rate": 1.756914183158199e-05, + "loss": 1.6213, + "step": 3576 + }, + { + "epoch": 0.7500524218913819, + "grad_norm": 5.98840513150999, + "learning_rate": 1.756766215670822e-05, + "loss": 1.8489, + "step": 3577 + }, + { + "epoch": 0.7502621094569092, + "grad_norm": 6.5295129281532995, + "learning_rate": 1.7566182093978643e-05, + "loss": 1.9036, + "step": 3578 + }, + { + "epoch": 0.7504717970224366, + "grad_norm": 6.830166823498713, + "learning_rate": 1.756470164346911e-05, + "loss": 1.8673, + "step": 3579 + }, + { + "epoch": 0.7506814845879639, + "grad_norm": 7.089647736742841, + "learning_rate": 1.7563220805255496e-05, + "loss": 1.7548, + "step": 3580 + }, + { + "epoch": 0.7508911721534913, + "grad_norm": 6.053219240037933, + "learning_rate": 1.75617395794137e-05, + "loss": 1.9925, + "step": 3581 + }, + { + "epoch": 0.7511008597190186, + "grad_norm": 7.012176842986784, + "learning_rate": 1.7560257966019634e-05, + "loss": 1.9095, + "step": 3582 + }, + { + "epoch": 0.751310547284546, + "grad_norm": 5.987066930418948, + "learning_rate": 1.7558775965149236e-05, + "loss": 1.4474, + "step": 3583 + }, + { + "epoch": 0.7515202348500734, + "grad_norm": 7.393547883923078, + "learning_rate": 1.755729357687846e-05, + "loss": 2.1521, + "step": 3584 + }, + { + "epoch": 0.7517299224156008, + "grad_norm": 6.740133743507663, + "learning_rate": 1.7555810801283278e-05, + "loss": 1.9401, + "step": 3585 + }, + { + "epoch": 0.7519396099811281, + "grad_norm": 6.389116286352003, + "learning_rate": 1.755432763843969e-05, + "loss": 1.7206, + "step": 3586 + }, + { + "epoch": 0.7521492975466555, + "grad_norm": 6.19748635459507, + "learning_rate": 1.7552844088423707e-05, + "loss": 1.8184, + "step": 3587 + }, + { + "epoch": 0.7523589851121828, + "grad_norm": 6.940194945997136, + "learning_rate": 1.7551360151311364e-05, + "loss": 1.7933, + "step": 3588 + }, + { + "epoch": 0.7525686726777102, + "grad_norm": 6.176398439168147, + "learning_rate": 1.7549875827178716e-05, + "loss": 1.8535, + "step": 3589 + }, + { + "epoch": 0.7527783602432375, + "grad_norm": 7.055827621359366, + "learning_rate": 1.754839111610184e-05, + "loss": 1.8103, + "step": 3590 + }, + { + "epoch": 0.7529880478087649, + "grad_norm": 4.980668637808491, + "learning_rate": 1.7546906018156825e-05, + "loss": 1.7236, + "step": 3591 + }, + { + "epoch": 0.7531977353742924, + "grad_norm": 6.5901203155426495, + "learning_rate": 1.754542053341979e-05, + "loss": 2.0351, + "step": 3592 + }, + { + "epoch": 0.7534074229398197, + "grad_norm": 7.355011806283256, + "learning_rate": 1.7543934661966863e-05, + "loss": 1.9339, + "step": 3593 + }, + { + "epoch": 0.7536171105053471, + "grad_norm": 7.146201136194959, + "learning_rate": 1.75424484038742e-05, + "loss": 2.0235, + "step": 3594 + }, + { + "epoch": 0.7538267980708744, + "grad_norm": 7.047761182349562, + "learning_rate": 1.7540961759217976e-05, + "loss": 1.8004, + "step": 3595 + }, + { + "epoch": 0.7540364856364018, + "grad_norm": 6.45243307882741, + "learning_rate": 1.7539474728074382e-05, + "loss": 1.6887, + "step": 3596 + }, + { + "epoch": 0.7542461732019291, + "grad_norm": 6.175569473365782, + "learning_rate": 1.7537987310519634e-05, + "loss": 1.9449, + "step": 3597 + }, + { + "epoch": 0.7544558607674565, + "grad_norm": 6.7896714808473515, + "learning_rate": 1.7536499506629956e-05, + "loss": 1.9703, + "step": 3598 + }, + { + "epoch": 0.7546655483329838, + "grad_norm": 6.844532950541191, + "learning_rate": 1.7535011316481613e-05, + "loss": 1.9322, + "step": 3599 + }, + { + "epoch": 0.7548752358985112, + "grad_norm": 7.528371728842008, + "learning_rate": 1.753352274015087e-05, + "loss": 1.904, + "step": 3600 + }, + { + "epoch": 0.7550849234640385, + "grad_norm": 6.098817790442983, + "learning_rate": 1.753203377771402e-05, + "loss": 1.7005, + "step": 3601 + }, + { + "epoch": 0.755294611029566, + "grad_norm": 6.266547019433409, + "learning_rate": 1.7530544429247375e-05, + "loss": 1.7496, + "step": 3602 + }, + { + "epoch": 0.7555042985950933, + "grad_norm": 6.297473338903792, + "learning_rate": 1.7529054694827265e-05, + "loss": 1.7812, + "step": 3603 + }, + { + "epoch": 0.7557139861606207, + "grad_norm": 6.227443771207519, + "learning_rate": 1.7527564574530047e-05, + "loss": 1.2683, + "step": 3604 + }, + { + "epoch": 0.755923673726148, + "grad_norm": 6.79213724523078, + "learning_rate": 1.7526074068432087e-05, + "loss": 1.6328, + "step": 3605 + }, + { + "epoch": 0.7561333612916754, + "grad_norm": 6.015713224600076, + "learning_rate": 1.7524583176609776e-05, + "loss": 1.9586, + "step": 3606 + }, + { + "epoch": 0.7563430488572027, + "grad_norm": 8.028247469594895, + "learning_rate": 1.7523091899139528e-05, + "loss": 1.9516, + "step": 3607 + }, + { + "epoch": 0.7565527364227301, + "grad_norm": 6.925459387659107, + "learning_rate": 1.7521600236097774e-05, + "loss": 1.8299, + "step": 3608 + }, + { + "epoch": 0.7567624239882575, + "grad_norm": 7.0091158756271685, + "learning_rate": 1.752010818756096e-05, + "loss": 2.0454, + "step": 3609 + }, + { + "epoch": 0.7569721115537849, + "grad_norm": 8.159946168228625, + "learning_rate": 1.751861575360556e-05, + "loss": 2.2354, + "step": 3610 + }, + { + "epoch": 0.7571817991193123, + "grad_norm": 6.359549518540783, + "learning_rate": 1.751712293430806e-05, + "loss": 1.5468, + "step": 3611 + }, + { + "epoch": 0.7573914866848396, + "grad_norm": 6.2245972049427785, + "learning_rate": 1.751562972974497e-05, + "loss": 1.8496, + "step": 3612 + }, + { + "epoch": 0.757601174250367, + "grad_norm": 5.1841139289842575, + "learning_rate": 1.7514136139992826e-05, + "loss": 1.8064, + "step": 3613 + }, + { + "epoch": 0.7578108618158943, + "grad_norm": 5.859895466259266, + "learning_rate": 1.7512642165128172e-05, + "loss": 2.0262, + "step": 3614 + }, + { + "epoch": 0.7580205493814217, + "grad_norm": 6.5730561736042175, + "learning_rate": 1.7511147805227573e-05, + "loss": 1.9988, + "step": 3615 + }, + { + "epoch": 0.758230236946949, + "grad_norm": 6.218980701638524, + "learning_rate": 1.7509653060367626e-05, + "loss": 1.8955, + "step": 3616 + }, + { + "epoch": 0.7584399245124764, + "grad_norm": 6.192123380880596, + "learning_rate": 1.750815793062493e-05, + "loss": 2.0516, + "step": 3617 + }, + { + "epoch": 0.7586496120780037, + "grad_norm": 7.56922979210302, + "learning_rate": 1.750666241607612e-05, + "loss": 1.911, + "step": 3618 + }, + { + "epoch": 0.7588592996435312, + "grad_norm": 5.477849151068813, + "learning_rate": 1.7505166516797844e-05, + "loss": 1.0648, + "step": 3619 + }, + { + "epoch": 0.7590689872090585, + "grad_norm": 6.693238166822524, + "learning_rate": 1.7503670232866767e-05, + "loss": 1.5961, + "step": 3620 + }, + { + "epoch": 0.7592786747745859, + "grad_norm": 6.3325741439253935, + "learning_rate": 1.750217356435957e-05, + "loss": 1.646, + "step": 3621 + }, + { + "epoch": 0.7594883623401132, + "grad_norm": 7.216636352147544, + "learning_rate": 1.750067651135297e-05, + "loss": 1.9933, + "step": 3622 + }, + { + "epoch": 0.7596980499056406, + "grad_norm": 7.530584128066125, + "learning_rate": 1.7499179073923692e-05, + "loss": 1.7248, + "step": 3623 + }, + { + "epoch": 0.7599077374711679, + "grad_norm": 6.77064345312104, + "learning_rate": 1.7497681252148476e-05, + "loss": 2.3199, + "step": 3624 + }, + { + "epoch": 0.7601174250366953, + "grad_norm": 6.554710793652707, + "learning_rate": 1.7496183046104095e-05, + "loss": 1.9308, + "step": 3625 + }, + { + "epoch": 0.7603271126022227, + "grad_norm": 5.676704730342347, + "learning_rate": 1.7494684455867328e-05, + "loss": 1.2719, + "step": 3626 + }, + { + "epoch": 0.76053680016775, + "grad_norm": 6.521686600516974, + "learning_rate": 1.7493185481514984e-05, + "loss": 1.5967, + "step": 3627 + }, + { + "epoch": 0.7607464877332775, + "grad_norm": 6.2899336584694625, + "learning_rate": 1.7491686123123888e-05, + "loss": 1.8855, + "step": 3628 + }, + { + "epoch": 0.7609561752988048, + "grad_norm": 6.364351430230438, + "learning_rate": 1.7490186380770887e-05, + "loss": 1.8863, + "step": 3629 + }, + { + "epoch": 0.7611658628643322, + "grad_norm": 6.720176726465069, + "learning_rate": 1.7488686254532844e-05, + "loss": 2.0107, + "step": 3630 + }, + { + "epoch": 0.7613755504298595, + "grad_norm": 6.056017334625059, + "learning_rate": 1.7487185744486636e-05, + "loss": 1.7627, + "step": 3631 + }, + { + "epoch": 0.7615852379953869, + "grad_norm": 6.045644452104308, + "learning_rate": 1.748568485070918e-05, + "loss": 1.512, + "step": 3632 + }, + { + "epoch": 0.7617949255609142, + "grad_norm": 5.78160406555623, + "learning_rate": 1.7484183573277385e-05, + "loss": 1.7835, + "step": 3633 + }, + { + "epoch": 0.7620046131264416, + "grad_norm": 5.908498260913524, + "learning_rate": 1.7482681912268205e-05, + "loss": 1.6271, + "step": 3634 + }, + { + "epoch": 0.7622143006919689, + "grad_norm": 6.836265825833793, + "learning_rate": 1.74811798677586e-05, + "loss": 2.1131, + "step": 3635 + }, + { + "epoch": 0.7624239882574964, + "grad_norm": 6.789212375288109, + "learning_rate": 1.747967743982555e-05, + "loss": 1.5101, + "step": 3636 + }, + { + "epoch": 0.7626336758230237, + "grad_norm": 6.35568114238226, + "learning_rate": 1.7478174628546057e-05, + "loss": 1.571, + "step": 3637 + }, + { + "epoch": 0.7628433633885511, + "grad_norm": 6.271275619464627, + "learning_rate": 1.7476671433997147e-05, + "loss": 1.3386, + "step": 3638 + }, + { + "epoch": 0.7630530509540784, + "grad_norm": 7.236527020633516, + "learning_rate": 1.7475167856255857e-05, + "loss": 2.1999, + "step": 3639 + }, + { + "epoch": 0.7632627385196058, + "grad_norm": 7.770764995641788, + "learning_rate": 1.7473663895399247e-05, + "loss": 2.0176, + "step": 3640 + }, + { + "epoch": 0.7634724260851331, + "grad_norm": 6.471162610445809, + "learning_rate": 1.7472159551504405e-05, + "loss": 1.8678, + "step": 3641 + }, + { + "epoch": 0.7636821136506605, + "grad_norm": 7.479839497561408, + "learning_rate": 1.747065482464842e-05, + "loss": 1.9362, + "step": 3642 + }, + { + "epoch": 0.7638918012161878, + "grad_norm": 10.262946999574746, + "learning_rate": 1.746914971490842e-05, + "loss": 2.2274, + "step": 3643 + }, + { + "epoch": 0.7641014887817152, + "grad_norm": 9.258574287443022, + "learning_rate": 1.7467644222361547e-05, + "loss": 2.5424, + "step": 3644 + }, + { + "epoch": 0.7643111763472427, + "grad_norm": 6.536040300734493, + "learning_rate": 1.7466138347084953e-05, + "loss": 1.9496, + "step": 3645 + }, + { + "epoch": 0.76452086391277, + "grad_norm": 6.365878404129815, + "learning_rate": 1.7464632089155818e-05, + "loss": 2.0183, + "step": 3646 + }, + { + "epoch": 0.7647305514782974, + "grad_norm": 5.868973886244549, + "learning_rate": 1.746312544865134e-05, + "loss": 2.0426, + "step": 3647 + }, + { + "epoch": 0.7649402390438247, + "grad_norm": 5.61115259340127, + "learning_rate": 1.7461618425648742e-05, + "loss": 1.7539, + "step": 3648 + }, + { + "epoch": 0.7651499266093521, + "grad_norm": 4.7846516726578665, + "learning_rate": 1.746011102022526e-05, + "loss": 1.7948, + "step": 3649 + }, + { + "epoch": 0.7653596141748794, + "grad_norm": 5.760766993002182, + "learning_rate": 1.7458603232458144e-05, + "loss": 2.0803, + "step": 3650 + }, + { + "epoch": 0.7655693017404068, + "grad_norm": 6.278621467586301, + "learning_rate": 1.7457095062424677e-05, + "loss": 1.7772, + "step": 3651 + }, + { + "epoch": 0.7657789893059341, + "grad_norm": 5.59161583500458, + "learning_rate": 1.745558651020216e-05, + "loss": 1.8771, + "step": 3652 + }, + { + "epoch": 0.7659886768714615, + "grad_norm": 5.981280062825638, + "learning_rate": 1.7454077575867895e-05, + "loss": 1.7596, + "step": 3653 + }, + { + "epoch": 0.7661983644369889, + "grad_norm": 6.09898091736078, + "learning_rate": 1.745256825949923e-05, + "loss": 2.0797, + "step": 3654 + }, + { + "epoch": 0.7664080520025163, + "grad_norm": 5.841052371563669, + "learning_rate": 1.745105856117352e-05, + "loss": 1.7735, + "step": 3655 + }, + { + "epoch": 0.7666177395680436, + "grad_norm": 5.19666873342045, + "learning_rate": 1.7449548480968127e-05, + "loss": 1.7499, + "step": 3656 + }, + { + "epoch": 0.766827427133571, + "grad_norm": 6.338726029347207, + "learning_rate": 1.744803801896046e-05, + "loss": 2.0768, + "step": 3657 + }, + { + "epoch": 0.7670371146990983, + "grad_norm": 5.7325334663187295, + "learning_rate": 1.7446527175227925e-05, + "loss": 2.0459, + "step": 3658 + }, + { + "epoch": 0.7672468022646257, + "grad_norm": 5.453141994128758, + "learning_rate": 1.7445015949847955e-05, + "loss": 1.7382, + "step": 3659 + }, + { + "epoch": 0.767456489830153, + "grad_norm": 5.527711132901874, + "learning_rate": 1.744350434289801e-05, + "loss": 1.7456, + "step": 3660 + }, + { + "epoch": 0.7676661773956804, + "grad_norm": 5.536202616847874, + "learning_rate": 1.744199235445555e-05, + "loss": 1.4765, + "step": 3661 + }, + { + "epoch": 0.7678758649612079, + "grad_norm": 6.9415768305610355, + "learning_rate": 1.744047998459808e-05, + "loss": 2.0375, + "step": 3662 + }, + { + "epoch": 0.7680855525267352, + "grad_norm": 6.401075433619252, + "learning_rate": 1.7438967233403103e-05, + "loss": 1.9118, + "step": 3663 + }, + { + "epoch": 0.7682952400922626, + "grad_norm": 6.857058952525928, + "learning_rate": 1.7437454100948158e-05, + "loss": 1.8915, + "step": 3664 + }, + { + "epoch": 0.7685049276577899, + "grad_norm": 7.411573346172245, + "learning_rate": 1.7435940587310785e-05, + "loss": 1.8935, + "step": 3665 + }, + { + "epoch": 0.7687146152233173, + "grad_norm": 7.339349900624808, + "learning_rate": 1.7434426692568563e-05, + "loss": 1.808, + "step": 3666 + }, + { + "epoch": 0.7689243027888446, + "grad_norm": 6.073993859773652, + "learning_rate": 1.7432912416799075e-05, + "loss": 1.6403, + "step": 3667 + }, + { + "epoch": 0.769133990354372, + "grad_norm": 6.538471392037432, + "learning_rate": 1.743139776007994e-05, + "loss": 1.8015, + "step": 3668 + }, + { + "epoch": 0.7693436779198993, + "grad_norm": 6.679344107350826, + "learning_rate": 1.7429882722488775e-05, + "loss": 2.023, + "step": 3669 + }, + { + "epoch": 0.7695533654854267, + "grad_norm": 7.6094687659735305, + "learning_rate": 1.7428367304103236e-05, + "loss": 2.0213, + "step": 3670 + }, + { + "epoch": 0.769763053050954, + "grad_norm": 6.779798778428214, + "learning_rate": 1.742685150500099e-05, + "loss": 1.7587, + "step": 3671 + }, + { + "epoch": 0.7699727406164815, + "grad_norm": 6.757766399887811, + "learning_rate": 1.742533532525972e-05, + "loss": 2.0425, + "step": 3672 + }, + { + "epoch": 0.7701824281820088, + "grad_norm": 6.743488083277184, + "learning_rate": 1.742381876495714e-05, + "loss": 1.788, + "step": 3673 + }, + { + "epoch": 0.7703921157475362, + "grad_norm": 5.709147954066134, + "learning_rate": 1.742230182417097e-05, + "loss": 1.4902, + "step": 3674 + }, + { + "epoch": 0.7706018033130635, + "grad_norm": 6.977914522852681, + "learning_rate": 1.742078450297896e-05, + "loss": 1.9757, + "step": 3675 + }, + { + "epoch": 0.7708114908785909, + "grad_norm": 6.111881360596249, + "learning_rate": 1.7419266801458866e-05, + "loss": 1.7127, + "step": 3676 + }, + { + "epoch": 0.7710211784441182, + "grad_norm": 5.510855756553036, + "learning_rate": 1.7417748719688488e-05, + "loss": 1.7443, + "step": 3677 + }, + { + "epoch": 0.7712308660096456, + "grad_norm": 6.536702921585937, + "learning_rate": 1.7416230257745624e-05, + "loss": 1.5566, + "step": 3678 + }, + { + "epoch": 0.7714405535751729, + "grad_norm": 5.716671615169279, + "learning_rate": 1.741471141570809e-05, + "loss": 1.7709, + "step": 3679 + }, + { + "epoch": 0.7716502411407004, + "grad_norm": 5.555081395778038, + "learning_rate": 1.741319219365374e-05, + "loss": 1.6331, + "step": 3680 + }, + { + "epoch": 0.7718599287062278, + "grad_norm": 6.750869221208491, + "learning_rate": 1.7411672591660433e-05, + "loss": 1.7384, + "step": 3681 + }, + { + "epoch": 0.7720696162717551, + "grad_norm": 6.695160836802208, + "learning_rate": 1.741015260980605e-05, + "loss": 1.7865, + "step": 3682 + }, + { + "epoch": 0.7722793038372825, + "grad_norm": 5.9770800413181036, + "learning_rate": 1.7408632248168492e-05, + "loss": 1.613, + "step": 3683 + }, + { + "epoch": 0.7724889914028098, + "grad_norm": 6.793188038358102, + "learning_rate": 1.7407111506825686e-05, + "loss": 1.9104, + "step": 3684 + }, + { + "epoch": 0.7726986789683372, + "grad_norm": 6.325997286887496, + "learning_rate": 1.7405590385855563e-05, + "loss": 1.698, + "step": 3685 + }, + { + "epoch": 0.7729083665338645, + "grad_norm": 6.374183867551127, + "learning_rate": 1.7404068885336093e-05, + "loss": 1.9877, + "step": 3686 + }, + { + "epoch": 0.7731180540993919, + "grad_norm": 6.552711821913816, + "learning_rate": 1.7402547005345246e-05, + "loss": 1.7765, + "step": 3687 + }, + { + "epoch": 0.7733277416649192, + "grad_norm": 6.14122226227425, + "learning_rate": 1.740102474596103e-05, + "loss": 1.7039, + "step": 3688 + }, + { + "epoch": 0.7735374292304467, + "grad_norm": 7.122926686964567, + "learning_rate": 1.7399502107261453e-05, + "loss": 1.6352, + "step": 3689 + }, + { + "epoch": 0.773747116795974, + "grad_norm": 6.355239932185493, + "learning_rate": 1.739797908932457e-05, + "loss": 1.8225, + "step": 3690 + }, + { + "epoch": 0.7739568043615014, + "grad_norm": 6.239632884365097, + "learning_rate": 1.739645569222842e-05, + "loss": 1.8271, + "step": 3691 + }, + { + "epoch": 0.7741664919270287, + "grad_norm": 7.116278558799536, + "learning_rate": 1.739493191605109e-05, + "loss": 2.053, + "step": 3692 + }, + { + "epoch": 0.7743761794925561, + "grad_norm": 6.131043029445476, + "learning_rate": 1.739340776087067e-05, + "loss": 1.7674, + "step": 3693 + }, + { + "epoch": 0.7745858670580834, + "grad_norm": 6.052566275181074, + "learning_rate": 1.7391883226765278e-05, + "loss": 1.5088, + "step": 3694 + }, + { + "epoch": 0.7747955546236108, + "grad_norm": 7.494807947346358, + "learning_rate": 1.7390358313813056e-05, + "loss": 1.6442, + "step": 3695 + }, + { + "epoch": 0.7750052421891381, + "grad_norm": 5.8679602513475855, + "learning_rate": 1.738883302209215e-05, + "loss": 1.9317, + "step": 3696 + }, + { + "epoch": 0.7752149297546655, + "grad_norm": 6.314531817242334, + "learning_rate": 1.7387307351680735e-05, + "loss": 1.676, + "step": 3697 + }, + { + "epoch": 0.775424617320193, + "grad_norm": 6.215495584880176, + "learning_rate": 1.7385781302657005e-05, + "loss": 2.1957, + "step": 3698 + }, + { + "epoch": 0.7756343048857203, + "grad_norm": 6.845501418141876, + "learning_rate": 1.7384254875099177e-05, + "loss": 2.0808, + "step": 3699 + }, + { + "epoch": 0.7758439924512477, + "grad_norm": 6.139472042533223, + "learning_rate": 1.7382728069085476e-05, + "loss": 1.8114, + "step": 3700 + }, + { + "epoch": 0.776053680016775, + "grad_norm": 5.8290804514132795, + "learning_rate": 1.738120088469416e-05, + "loss": 2.2186, + "step": 3701 + }, + { + "epoch": 0.7762633675823024, + "grad_norm": 5.486323133570247, + "learning_rate": 1.7379673322003493e-05, + "loss": 1.6502, + "step": 3702 + }, + { + "epoch": 0.7764730551478297, + "grad_norm": 5.420312803535227, + "learning_rate": 1.737814538109177e-05, + "loss": 1.5772, + "step": 3703 + }, + { + "epoch": 0.7766827427133571, + "grad_norm": 5.707300103941764, + "learning_rate": 1.7376617062037298e-05, + "loss": 2.0236, + "step": 3704 + }, + { + "epoch": 0.7768924302788844, + "grad_norm": 8.624140557111666, + "learning_rate": 1.7375088364918412e-05, + "loss": 1.8287, + "step": 3705 + }, + { + "epoch": 0.7771021178444119, + "grad_norm": 6.016758332069924, + "learning_rate": 1.737355928981345e-05, + "loss": 1.8598, + "step": 3706 + }, + { + "epoch": 0.7773118054099392, + "grad_norm": 5.197530421822921, + "learning_rate": 1.737202983680079e-05, + "loss": 1.7728, + "step": 3707 + }, + { + "epoch": 0.7775214929754666, + "grad_norm": 7.3863502485668295, + "learning_rate": 1.7370500005958818e-05, + "loss": 1.9952, + "step": 3708 + }, + { + "epoch": 0.7777311805409939, + "grad_norm": 5.797472213560914, + "learning_rate": 1.736896979736593e-05, + "loss": 1.9066, + "step": 3709 + }, + { + "epoch": 0.7779408681065213, + "grad_norm": 5.548546574467694, + "learning_rate": 1.7367439211100562e-05, + "loss": 1.787, + "step": 3710 + }, + { + "epoch": 0.7781505556720486, + "grad_norm": 6.678654679124065, + "learning_rate": 1.7365908247241156e-05, + "loss": 1.8541, + "step": 3711 + }, + { + "epoch": 0.778360243237576, + "grad_norm": 5.937774431010136, + "learning_rate": 1.736437690586618e-05, + "loss": 1.823, + "step": 3712 + }, + { + "epoch": 0.7785699308031033, + "grad_norm": 5.77062241835383, + "learning_rate": 1.736284518705411e-05, + "loss": 1.8848, + "step": 3713 + }, + { + "epoch": 0.7787796183686307, + "grad_norm": 6.066334595045192, + "learning_rate": 1.7361313090883455e-05, + "loss": 1.7502, + "step": 3714 + }, + { + "epoch": 0.778989305934158, + "grad_norm": 5.94859261906079, + "learning_rate": 1.7359780617432735e-05, + "loss": 1.7828, + "step": 3715 + }, + { + "epoch": 0.7791989934996855, + "grad_norm": 5.674002493380154, + "learning_rate": 1.7358247766780497e-05, + "loss": 1.9576, + "step": 3716 + }, + { + "epoch": 0.7794086810652129, + "grad_norm": 6.161203207592451, + "learning_rate": 1.7356714539005297e-05, + "loss": 1.8563, + "step": 3717 + }, + { + "epoch": 0.7796183686307402, + "grad_norm": 6.422434939327867, + "learning_rate": 1.7355180934185717e-05, + "loss": 1.89, + "step": 3718 + }, + { + "epoch": 0.7798280561962676, + "grad_norm": 6.316184690673304, + "learning_rate": 1.735364695240036e-05, + "loss": 1.9439, + "step": 3719 + }, + { + "epoch": 0.7800377437617949, + "grad_norm": 6.434339532347986, + "learning_rate": 1.7352112593727838e-05, + "loss": 1.7838, + "step": 3720 + }, + { + "epoch": 0.7802474313273223, + "grad_norm": 6.067686122163278, + "learning_rate": 1.7350577858246796e-05, + "loss": 1.7625, + "step": 3721 + }, + { + "epoch": 0.7804571188928496, + "grad_norm": 5.60892959814783, + "learning_rate": 1.734904274603589e-05, + "loss": 1.818, + "step": 3722 + }, + { + "epoch": 0.780666806458377, + "grad_norm": 5.462586681528775, + "learning_rate": 1.7347507257173797e-05, + "loss": 1.9281, + "step": 3723 + }, + { + "epoch": 0.7808764940239044, + "grad_norm": 5.959990016291894, + "learning_rate": 1.7345971391739213e-05, + "loss": 1.4341, + "step": 3724 + }, + { + "epoch": 0.7810861815894318, + "grad_norm": 5.857288788698317, + "learning_rate": 1.7344435149810857e-05, + "loss": 1.8696, + "step": 3725 + }, + { + "epoch": 0.7812958691549591, + "grad_norm": 5.7259854691807615, + "learning_rate": 1.734289853146746e-05, + "loss": 1.9614, + "step": 3726 + }, + { + "epoch": 0.7815055567204865, + "grad_norm": 6.366311799426714, + "learning_rate": 1.7341361536787774e-05, + "loss": 1.8454, + "step": 3727 + }, + { + "epoch": 0.7817152442860138, + "grad_norm": 6.5359142095354645, + "learning_rate": 1.733982416585058e-05, + "loss": 2.0712, + "step": 3728 + }, + { + "epoch": 0.7819249318515412, + "grad_norm": 7.403335527495739, + "learning_rate": 1.7338286418734666e-05, + "loss": 1.7362, + "step": 3729 + }, + { + "epoch": 0.7821346194170685, + "grad_norm": 6.759457254223462, + "learning_rate": 1.7336748295518846e-05, + "loss": 1.5543, + "step": 3730 + }, + { + "epoch": 0.7823443069825959, + "grad_norm": 5.809230554526833, + "learning_rate": 1.733520979628195e-05, + "loss": 1.8014, + "step": 3731 + }, + { + "epoch": 0.7825539945481232, + "grad_norm": 6.467669548226693, + "learning_rate": 1.733367092110283e-05, + "loss": 1.6486, + "step": 3732 + }, + { + "epoch": 0.7827636821136507, + "grad_norm": 6.74810538342474, + "learning_rate": 1.7332131670060356e-05, + "loss": 1.9013, + "step": 3733 + }, + { + "epoch": 0.7829733696791781, + "grad_norm": 5.6481619832997865, + "learning_rate": 1.733059204323342e-05, + "loss": 1.5805, + "step": 3734 + }, + { + "epoch": 0.7831830572447054, + "grad_norm": 6.93072857651362, + "learning_rate": 1.732905204070092e-05, + "loss": 2.0013, + "step": 3735 + }, + { + "epoch": 0.7833927448102328, + "grad_norm": 7.386761936291472, + "learning_rate": 1.73275116625418e-05, + "loss": 1.8713, + "step": 3736 + }, + { + "epoch": 0.7836024323757601, + "grad_norm": 5.721379355546761, + "learning_rate": 1.7325970908834993e-05, + "loss": 1.8693, + "step": 3737 + }, + { + "epoch": 0.7838121199412875, + "grad_norm": 6.504881031715918, + "learning_rate": 1.732442977965947e-05, + "loss": 1.7702, + "step": 3738 + }, + { + "epoch": 0.7840218075068148, + "grad_norm": 7.277498499545157, + "learning_rate": 1.732288827509422e-05, + "loss": 2.3072, + "step": 3739 + }, + { + "epoch": 0.7842314950723422, + "grad_norm": 7.121452781804273, + "learning_rate": 1.7321346395218243e-05, + "loss": 1.7482, + "step": 3740 + }, + { + "epoch": 0.7844411826378695, + "grad_norm": 6.204376078176178, + "learning_rate": 1.7319804140110567e-05, + "loss": 1.9128, + "step": 3741 + }, + { + "epoch": 0.784650870203397, + "grad_norm": 6.678226094692333, + "learning_rate": 1.7318261509850232e-05, + "loss": 1.8391, + "step": 3742 + }, + { + "epoch": 0.7848605577689243, + "grad_norm": 6.767069065605165, + "learning_rate": 1.7316718504516302e-05, + "loss": 1.8829, + "step": 3743 + }, + { + "epoch": 0.7850702453344517, + "grad_norm": 6.548352483353616, + "learning_rate": 1.731517512418786e-05, + "loss": 1.8328, + "step": 3744 + }, + { + "epoch": 0.785279932899979, + "grad_norm": 6.359480852574694, + "learning_rate": 1.7313631368944002e-05, + "loss": 1.7449, + "step": 3745 + }, + { + "epoch": 0.7854896204655064, + "grad_norm": 7.126567059130508, + "learning_rate": 1.731208723886385e-05, + "loss": 2.0378, + "step": 3746 + }, + { + "epoch": 0.7856993080310337, + "grad_norm": 5.497025126889722, + "learning_rate": 1.7310542734026553e-05, + "loss": 1.5895, + "step": 3747 + }, + { + "epoch": 0.7859089955965611, + "grad_norm": 5.915108898695909, + "learning_rate": 1.7308997854511257e-05, + "loss": 1.552, + "step": 3748 + }, + { + "epoch": 0.7861186831620884, + "grad_norm": 5.559981812177856, + "learning_rate": 1.730745260039714e-05, + "loss": 1.8118, + "step": 3749 + }, + { + "epoch": 0.7863283707276159, + "grad_norm": 7.626200982493386, + "learning_rate": 1.7305906971763407e-05, + "loss": 1.651, + "step": 3750 + }, + { + "epoch": 0.7865380582931433, + "grad_norm": 5.723157443469827, + "learning_rate": 1.730436096868927e-05, + "loss": 1.6888, + "step": 3751 + }, + { + "epoch": 0.7867477458586706, + "grad_norm": 5.5653901139750195, + "learning_rate": 1.7302814591253966e-05, + "loss": 2.141, + "step": 3752 + }, + { + "epoch": 0.786957433424198, + "grad_norm": 5.281258842465091, + "learning_rate": 1.7301267839536746e-05, + "loss": 1.5765, + "step": 3753 + }, + { + "epoch": 0.7871671209897253, + "grad_norm": 5.661900767812405, + "learning_rate": 1.7299720713616887e-05, + "loss": 1.7015, + "step": 3754 + }, + { + "epoch": 0.7873768085552527, + "grad_norm": 6.806577795223069, + "learning_rate": 1.7298173213573682e-05, + "loss": 1.479, + "step": 3755 + }, + { + "epoch": 0.78758649612078, + "grad_norm": 5.844847643625029, + "learning_rate": 1.729662533948644e-05, + "loss": 1.8283, + "step": 3756 + }, + { + "epoch": 0.7877961836863074, + "grad_norm": 7.5478040015298715, + "learning_rate": 1.729507709143449e-05, + "loss": 2.0444, + "step": 3757 + }, + { + "epoch": 0.7880058712518347, + "grad_norm": 5.697051218338185, + "learning_rate": 1.7293528469497193e-05, + "loss": 1.5551, + "step": 3758 + }, + { + "epoch": 0.7882155588173622, + "grad_norm": 7.272403623970555, + "learning_rate": 1.729197947375391e-05, + "loss": 1.9502, + "step": 3759 + }, + { + "epoch": 0.7884252463828895, + "grad_norm": 7.690605762464217, + "learning_rate": 1.7290430104284032e-05, + "loss": 1.729, + "step": 3760 + }, + { + "epoch": 0.7886349339484169, + "grad_norm": 6.282304951578839, + "learning_rate": 1.7288880361166963e-05, + "loss": 1.5239, + "step": 3761 + }, + { + "epoch": 0.7888446215139442, + "grad_norm": 5.372825395608513, + "learning_rate": 1.7287330244482137e-05, + "loss": 1.3797, + "step": 3762 + }, + { + "epoch": 0.7890543090794716, + "grad_norm": 6.700864206515308, + "learning_rate": 1.7285779754308995e-05, + "loss": 2.0115, + "step": 3763 + }, + { + "epoch": 0.7892639966449989, + "grad_norm": 6.475946005773212, + "learning_rate": 1.7284228890727006e-05, + "loss": 1.6781, + "step": 3764 + }, + { + "epoch": 0.7894736842105263, + "grad_norm": 6.967050518608623, + "learning_rate": 1.7282677653815652e-05, + "loss": 1.5084, + "step": 3765 + }, + { + "epoch": 0.7896833717760536, + "grad_norm": 5.3013809157615635, + "learning_rate": 1.7281126043654436e-05, + "loss": 1.7113, + "step": 3766 + }, + { + "epoch": 0.789893059341581, + "grad_norm": 6.502806611506876, + "learning_rate": 1.727957406032288e-05, + "loss": 1.8401, + "step": 3767 + }, + { + "epoch": 0.7901027469071084, + "grad_norm": 5.561432351562957, + "learning_rate": 1.727802170390053e-05, + "loss": 1.3767, + "step": 3768 + }, + { + "epoch": 0.7903124344726358, + "grad_norm": 6.774195578611065, + "learning_rate": 1.7276468974466946e-05, + "loss": 1.761, + "step": 3769 + }, + { + "epoch": 0.7905221220381632, + "grad_norm": 7.25839027326519, + "learning_rate": 1.7274915872101704e-05, + "loss": 1.871, + "step": 3770 + }, + { + "epoch": 0.7907318096036905, + "grad_norm": 7.185594100387391, + "learning_rate": 1.7273362396884405e-05, + "loss": 2.09, + "step": 3771 + }, + { + "epoch": 0.7909414971692179, + "grad_norm": 7.020451992151008, + "learning_rate": 1.727180854889467e-05, + "loss": 1.8139, + "step": 3772 + }, + { + "epoch": 0.7911511847347452, + "grad_norm": 7.6327971279649365, + "learning_rate": 1.727025432821214e-05, + "loss": 1.8714, + "step": 3773 + }, + { + "epoch": 0.7913608723002726, + "grad_norm": 5.8341109856249895, + "learning_rate": 1.7268699734916458e-05, + "loss": 1.833, + "step": 3774 + }, + { + "epoch": 0.7915705598657999, + "grad_norm": 7.177776879424799, + "learning_rate": 1.7267144769087308e-05, + "loss": 1.8145, + "step": 3775 + }, + { + "epoch": 0.7917802474313274, + "grad_norm": 6.312583535356928, + "learning_rate": 1.726558943080439e-05, + "loss": 1.8524, + "step": 3776 + }, + { + "epoch": 0.7919899349968547, + "grad_norm": 6.73636148457771, + "learning_rate": 1.7264033720147408e-05, + "loss": 1.8153, + "step": 3777 + }, + { + "epoch": 0.7921996225623821, + "grad_norm": 6.9189433647696825, + "learning_rate": 1.7262477637196096e-05, + "loss": 1.8173, + "step": 3778 + }, + { + "epoch": 0.7924093101279094, + "grad_norm": 5.790322560463186, + "learning_rate": 1.7260921182030216e-05, + "loss": 2.053, + "step": 3779 + }, + { + "epoch": 0.7926189976934368, + "grad_norm": 6.389957297059367, + "learning_rate": 1.725936435472953e-05, + "loss": 1.7552, + "step": 3780 + }, + { + "epoch": 0.7928286852589641, + "grad_norm": 6.314598199232086, + "learning_rate": 1.7257807155373827e-05, + "loss": 1.8706, + "step": 3781 + }, + { + "epoch": 0.7930383728244915, + "grad_norm": 5.787078024360537, + "learning_rate": 1.7256249584042922e-05, + "loss": 1.7596, + "step": 3782 + }, + { + "epoch": 0.7932480603900188, + "grad_norm": 6.425191335934589, + "learning_rate": 1.7254691640816642e-05, + "loss": 1.6373, + "step": 3783 + }, + { + "epoch": 0.7934577479555462, + "grad_norm": 6.39785336382828, + "learning_rate": 1.725313332577483e-05, + "loss": 1.8341, + "step": 3784 + }, + { + "epoch": 0.7936674355210735, + "grad_norm": 6.318111129730518, + "learning_rate": 1.7251574638997354e-05, + "loss": 2.1615, + "step": 3785 + }, + { + "epoch": 0.793877123086601, + "grad_norm": 6.1883330846969855, + "learning_rate": 1.7250015580564104e-05, + "loss": 1.8955, + "step": 3786 + }, + { + "epoch": 0.7940868106521284, + "grad_norm": 6.750690227888648, + "learning_rate": 1.724845615055498e-05, + "loss": 1.671, + "step": 3787 + }, + { + "epoch": 0.7942964982176557, + "grad_norm": 6.535233279981005, + "learning_rate": 1.7246896349049902e-05, + "loss": 1.8024, + "step": 3788 + }, + { + "epoch": 0.7945061857831831, + "grad_norm": 6.772774932749703, + "learning_rate": 1.7245336176128823e-05, + "loss": 2.1947, + "step": 3789 + }, + { + "epoch": 0.7947158733487104, + "grad_norm": 6.2559571036721415, + "learning_rate": 1.7243775631871695e-05, + "loss": 1.9779, + "step": 3790 + }, + { + "epoch": 0.7949255609142378, + "grad_norm": 5.986316117092616, + "learning_rate": 1.7242214716358505e-05, + "loss": 1.8951, + "step": 3791 + }, + { + "epoch": 0.7951352484797651, + "grad_norm": 6.613841955465014, + "learning_rate": 1.7240653429669247e-05, + "loss": 1.6175, + "step": 3792 + }, + { + "epoch": 0.7953449360452925, + "grad_norm": 5.713812453568993, + "learning_rate": 1.723909177188394e-05, + "loss": 1.877, + "step": 3793 + }, + { + "epoch": 0.7955546236108199, + "grad_norm": 6.0688181772144505, + "learning_rate": 1.7237529743082633e-05, + "loss": 1.5416, + "step": 3794 + }, + { + "epoch": 0.7957643111763473, + "grad_norm": 6.072981791258866, + "learning_rate": 1.7235967343345366e-05, + "loss": 1.5986, + "step": 3795 + }, + { + "epoch": 0.7959739987418746, + "grad_norm": 6.063548487148982, + "learning_rate": 1.7234404572752227e-05, + "loss": 1.942, + "step": 3796 + }, + { + "epoch": 0.796183686307402, + "grad_norm": 6.4782408324516165, + "learning_rate": 1.7232841431383303e-05, + "loss": 1.9494, + "step": 3797 + }, + { + "epoch": 0.7963933738729293, + "grad_norm": 6.003525252312476, + "learning_rate": 1.723127791931871e-05, + "loss": 1.8673, + "step": 3798 + }, + { + "epoch": 0.7966030614384567, + "grad_norm": 6.030061542015949, + "learning_rate": 1.7229714036638587e-05, + "loss": 1.6838, + "step": 3799 + }, + { + "epoch": 0.796812749003984, + "grad_norm": 6.674114090550902, + "learning_rate": 1.7228149783423076e-05, + "loss": 1.4886, + "step": 3800 + }, + { + "epoch": 0.7970224365695114, + "grad_norm": 6.668637716461723, + "learning_rate": 1.7226585159752353e-05, + "loss": 1.8793, + "step": 3801 + }, + { + "epoch": 0.7972321241350387, + "grad_norm": 5.416851967880576, + "learning_rate": 1.7225020165706604e-05, + "loss": 1.8698, + "step": 3802 + }, + { + "epoch": 0.7974418117005662, + "grad_norm": 6.978915757041792, + "learning_rate": 1.7223454801366046e-05, + "loss": 1.7246, + "step": 3803 + }, + { + "epoch": 0.7976514992660935, + "grad_norm": 6.571909494372986, + "learning_rate": 1.72218890668109e-05, + "loss": 2.016, + "step": 3804 + }, + { + "epoch": 0.7978611868316209, + "grad_norm": 6.892446555062564, + "learning_rate": 1.722032296212141e-05, + "loss": 1.7742, + "step": 3805 + }, + { + "epoch": 0.7980708743971483, + "grad_norm": 7.083615288691697, + "learning_rate": 1.7218756487377847e-05, + "loss": 1.7427, + "step": 3806 + }, + { + "epoch": 0.7982805619626756, + "grad_norm": 6.218556958226318, + "learning_rate": 1.7217189642660494e-05, + "loss": 1.6799, + "step": 3807 + }, + { + "epoch": 0.798490249528203, + "grad_norm": 5.886471881923572, + "learning_rate": 1.7215622428049654e-05, + "loss": 1.7002, + "step": 3808 + }, + { + "epoch": 0.7986999370937303, + "grad_norm": 6.745564840928233, + "learning_rate": 1.7214054843625648e-05, + "loss": 1.941, + "step": 3809 + }, + { + "epoch": 0.7989096246592577, + "grad_norm": 5.579330799862577, + "learning_rate": 1.721248688946882e-05, + "loss": 1.2416, + "step": 3810 + }, + { + "epoch": 0.799119312224785, + "grad_norm": 6.8048380758167655, + "learning_rate": 1.721091856565953e-05, + "loss": 1.8795, + "step": 3811 + }, + { + "epoch": 0.7993289997903125, + "grad_norm": 5.601752762941683, + "learning_rate": 1.720934987227816e-05, + "loss": 1.7188, + "step": 3812 + }, + { + "epoch": 0.7995386873558398, + "grad_norm": 6.9495853592189265, + "learning_rate": 1.72077808094051e-05, + "loss": 1.843, + "step": 3813 + }, + { + "epoch": 0.7997483749213672, + "grad_norm": 5.690448920199598, + "learning_rate": 1.7206211377120774e-05, + "loss": 1.7054, + "step": 3814 + }, + { + "epoch": 0.7999580624868945, + "grad_norm": 5.356620538465669, + "learning_rate": 1.720464157550562e-05, + "loss": 1.5814, + "step": 3815 + }, + { + "epoch": 0.8001677500524219, + "grad_norm": 7.624145202548761, + "learning_rate": 1.7203071404640082e-05, + "loss": 1.7814, + "step": 3816 + }, + { + "epoch": 0.8003774376179492, + "grad_norm": 6.451469813819988, + "learning_rate": 1.7201500864604647e-05, + "loss": 1.5567, + "step": 3817 + }, + { + "epoch": 0.8005871251834766, + "grad_norm": 6.022100448953644, + "learning_rate": 1.7199929955479802e-05, + "loss": 1.8386, + "step": 3818 + }, + { + "epoch": 0.8007968127490039, + "grad_norm": 7.464575530393103, + "learning_rate": 1.7198358677346057e-05, + "loss": 1.7831, + "step": 3819 + }, + { + "epoch": 0.8010065003145314, + "grad_norm": 6.10489966943075, + "learning_rate": 1.7196787030283947e-05, + "loss": 1.7321, + "step": 3820 + }, + { + "epoch": 0.8012161878800587, + "grad_norm": 6.857979364748703, + "learning_rate": 1.719521501437402e-05, + "loss": 2.0479, + "step": 3821 + }, + { + "epoch": 0.8014258754455861, + "grad_norm": 6.66497312338493, + "learning_rate": 1.719364262969684e-05, + "loss": 1.6613, + "step": 3822 + }, + { + "epoch": 0.8016355630111135, + "grad_norm": 6.7263985000781386, + "learning_rate": 1.7192069876333002e-05, + "loss": 1.7643, + "step": 3823 + }, + { + "epoch": 0.8018452505766408, + "grad_norm": 5.851714349019155, + "learning_rate": 1.7190496754363108e-05, + "loss": 1.7006, + "step": 3824 + }, + { + "epoch": 0.8020549381421682, + "grad_norm": 6.070106510216911, + "learning_rate": 1.7188923263867782e-05, + "loss": 1.7015, + "step": 3825 + }, + { + "epoch": 0.8022646257076955, + "grad_norm": 6.589466145770719, + "learning_rate": 1.7187349404927673e-05, + "loss": 1.602, + "step": 3826 + }, + { + "epoch": 0.8024743132732229, + "grad_norm": 6.712943345196664, + "learning_rate": 1.718577517762344e-05, + "loss": 1.7956, + "step": 3827 + }, + { + "epoch": 0.8026840008387502, + "grad_norm": 6.534884201755805, + "learning_rate": 1.7184200582035766e-05, + "loss": 1.8752, + "step": 3828 + }, + { + "epoch": 0.8028936884042777, + "grad_norm": 7.079045462705081, + "learning_rate": 1.7182625618245352e-05, + "loss": 1.7322, + "step": 3829 + }, + { + "epoch": 0.803103375969805, + "grad_norm": 6.346393789780253, + "learning_rate": 1.7181050286332915e-05, + "loss": 1.5294, + "step": 3830 + }, + { + "epoch": 0.8033130635353324, + "grad_norm": 6.867774565332408, + "learning_rate": 1.7179474586379198e-05, + "loss": 1.7358, + "step": 3831 + }, + { + "epoch": 0.8035227511008597, + "grad_norm": 6.2550411729918824, + "learning_rate": 1.717789851846496e-05, + "loss": 1.7556, + "step": 3832 + }, + { + "epoch": 0.8037324386663871, + "grad_norm": 7.030671005251954, + "learning_rate": 1.7176322082670968e-05, + "loss": 1.886, + "step": 3833 + }, + { + "epoch": 0.8039421262319144, + "grad_norm": 5.85199752738826, + "learning_rate": 1.7174745279078027e-05, + "loss": 1.8604, + "step": 3834 + }, + { + "epoch": 0.8041518137974418, + "grad_norm": 5.817270856429654, + "learning_rate": 1.717316810776694e-05, + "loss": 1.9208, + "step": 3835 + }, + { + "epoch": 0.8043615013629691, + "grad_norm": 6.008708085348807, + "learning_rate": 1.7171590568818555e-05, + "loss": 1.5347, + "step": 3836 + }, + { + "epoch": 0.8045711889284966, + "grad_norm": 5.462490751806413, + "learning_rate": 1.717001266231371e-05, + "loss": 1.4815, + "step": 3837 + }, + { + "epoch": 0.8047808764940239, + "grad_norm": 5.619333441743936, + "learning_rate": 1.716843438833328e-05, + "loss": 1.627, + "step": 3838 + }, + { + "epoch": 0.8049905640595513, + "grad_norm": 5.176482233766037, + "learning_rate": 1.7166855746958156e-05, + "loss": 1.5186, + "step": 3839 + }, + { + "epoch": 0.8052002516250786, + "grad_norm": 6.600445941310134, + "learning_rate": 1.7165276738269242e-05, + "loss": 1.9792, + "step": 3840 + }, + { + "epoch": 0.805409939190606, + "grad_norm": 6.689859704797481, + "learning_rate": 1.716369736234747e-05, + "loss": 2.1147, + "step": 3841 + }, + { + "epoch": 0.8056196267561334, + "grad_norm": 6.1384239941711405, + "learning_rate": 1.7162117619273783e-05, + "loss": 2.0201, + "step": 3842 + }, + { + "epoch": 0.8058293143216607, + "grad_norm": 6.22079975226795, + "learning_rate": 1.7160537509129147e-05, + "loss": 1.9786, + "step": 3843 + }, + { + "epoch": 0.8060390018871881, + "grad_norm": 5.354616251386052, + "learning_rate": 1.715895703199454e-05, + "loss": 1.4581, + "step": 3844 + }, + { + "epoch": 0.8062486894527154, + "grad_norm": 6.161687289078216, + "learning_rate": 1.7157376187950974e-05, + "loss": 1.7461, + "step": 3845 + }, + { + "epoch": 0.8064583770182429, + "grad_norm": 6.240097575561236, + "learning_rate": 1.715579497707946e-05, + "loss": 1.7106, + "step": 3846 + }, + { + "epoch": 0.8066680645837702, + "grad_norm": 6.304997890600722, + "learning_rate": 1.7154213399461048e-05, + "loss": 1.6415, + "step": 3847 + }, + { + "epoch": 0.8068777521492976, + "grad_norm": 7.168846557616687, + "learning_rate": 1.7152631455176787e-05, + "loss": 2.2949, + "step": 3848 + }, + { + "epoch": 0.8070874397148249, + "grad_norm": 6.610468118676679, + "learning_rate": 1.715104914430776e-05, + "loss": 1.7858, + "step": 3849 + }, + { + "epoch": 0.8072971272803523, + "grad_norm": 5.937503205474189, + "learning_rate": 1.714946646693506e-05, + "loss": 1.7341, + "step": 3850 + }, + { + "epoch": 0.8075068148458796, + "grad_norm": 5.387170622602409, + "learning_rate": 1.71478834231398e-05, + "loss": 1.924, + "step": 3851 + }, + { + "epoch": 0.807716502411407, + "grad_norm": 6.879081460648234, + "learning_rate": 1.714630001300312e-05, + "loss": 1.9383, + "step": 3852 + }, + { + "epoch": 0.8079261899769343, + "grad_norm": 7.208809971669895, + "learning_rate": 1.714471623660617e-05, + "loss": 2.1168, + "step": 3853 + }, + { + "epoch": 0.8081358775424617, + "grad_norm": 5.8625241374675925, + "learning_rate": 1.7143132094030123e-05, + "loss": 1.7497, + "step": 3854 + }, + { + "epoch": 0.808345565107989, + "grad_norm": 7.3765272204907495, + "learning_rate": 1.7141547585356166e-05, + "loss": 1.7477, + "step": 3855 + }, + { + "epoch": 0.8085552526735165, + "grad_norm": 5.0374632483112265, + "learning_rate": 1.713996271066551e-05, + "loss": 1.5288, + "step": 3856 + }, + { + "epoch": 0.8087649402390438, + "grad_norm": 7.362723808791006, + "learning_rate": 1.7138377470039378e-05, + "loss": 2.1388, + "step": 3857 + }, + { + "epoch": 0.8089746278045712, + "grad_norm": 6.797606879188635, + "learning_rate": 1.7136791863559025e-05, + "loss": 1.3778, + "step": 3858 + }, + { + "epoch": 0.8091843153700986, + "grad_norm": 5.72627689035103, + "learning_rate": 1.7135205891305708e-05, + "loss": 1.6881, + "step": 3859 + }, + { + "epoch": 0.8093940029356259, + "grad_norm": 5.933320309137687, + "learning_rate": 1.7133619553360712e-05, + "loss": 2.0796, + "step": 3860 + }, + { + "epoch": 0.8096036905011533, + "grad_norm": 6.0464941947261375, + "learning_rate": 1.713203284980534e-05, + "loss": 1.7245, + "step": 3861 + }, + { + "epoch": 0.8098133780666806, + "grad_norm": 6.020859638122137, + "learning_rate": 1.7130445780720917e-05, + "loss": 2.0015, + "step": 3862 + }, + { + "epoch": 0.810023065632208, + "grad_norm": 5.749231695495465, + "learning_rate": 1.7128858346188782e-05, + "loss": 1.7884, + "step": 3863 + }, + { + "epoch": 0.8102327531977354, + "grad_norm": 5.616404901822202, + "learning_rate": 1.712727054629029e-05, + "loss": 1.5936, + "step": 3864 + }, + { + "epoch": 0.8104424407632628, + "grad_norm": 5.783498764050352, + "learning_rate": 1.7125682381106823e-05, + "loss": 1.8613, + "step": 3865 + }, + { + "epoch": 0.8106521283287901, + "grad_norm": 5.839088678026966, + "learning_rate": 1.7124093850719772e-05, + "loss": 1.7901, + "step": 3866 + }, + { + "epoch": 0.8108618158943175, + "grad_norm": 7.5869478501642735, + "learning_rate": 1.712250495521056e-05, + "loss": 1.9588, + "step": 3867 + }, + { + "epoch": 0.8110715034598448, + "grad_norm": 6.871929242603936, + "learning_rate": 1.7120915694660612e-05, + "loss": 2.0037, + "step": 3868 + }, + { + "epoch": 0.8112811910253722, + "grad_norm": 7.309677272184801, + "learning_rate": 1.7119326069151382e-05, + "loss": 1.7913, + "step": 3869 + }, + { + "epoch": 0.8114908785908995, + "grad_norm": 6.666497868615008, + "learning_rate": 1.711773607876435e-05, + "loss": 1.8072, + "step": 3870 + }, + { + "epoch": 0.8117005661564269, + "grad_norm": 7.2474070086423295, + "learning_rate": 1.7116145723580992e-05, + "loss": 1.6869, + "step": 3871 + }, + { + "epoch": 0.8119102537219542, + "grad_norm": 7.089398637308205, + "learning_rate": 1.7114555003682825e-05, + "loss": 1.7287, + "step": 3872 + }, + { + "epoch": 0.8121199412874817, + "grad_norm": 6.206714235520265, + "learning_rate": 1.7112963919151378e-05, + "loss": 1.9023, + "step": 3873 + }, + { + "epoch": 0.812329628853009, + "grad_norm": 7.228809344924215, + "learning_rate": 1.7111372470068192e-05, + "loss": 1.7569, + "step": 3874 + }, + { + "epoch": 0.8125393164185364, + "grad_norm": 7.5493383469573345, + "learning_rate": 1.710978065651483e-05, + "loss": 1.9262, + "step": 3875 + }, + { + "epoch": 0.8127490039840638, + "grad_norm": 7.468431627598546, + "learning_rate": 1.710818847857288e-05, + "loss": 2.2593, + "step": 3876 + }, + { + "epoch": 0.8129586915495911, + "grad_norm": 7.746975675404207, + "learning_rate": 1.710659593632394e-05, + "loss": 1.8683, + "step": 3877 + }, + { + "epoch": 0.8131683791151185, + "grad_norm": 7.772874909079346, + "learning_rate": 1.7105003029849635e-05, + "loss": 1.9158, + "step": 3878 + }, + { + "epoch": 0.8133780666806458, + "grad_norm": 6.075576999572013, + "learning_rate": 1.7103409759231603e-05, + "loss": 1.8004, + "step": 3879 + }, + { + "epoch": 0.8135877542461732, + "grad_norm": 6.7395463574091385, + "learning_rate": 1.71018161245515e-05, + "loss": 2.1537, + "step": 3880 + }, + { + "epoch": 0.8137974418117006, + "grad_norm": 7.057123390469567, + "learning_rate": 1.7100222125891e-05, + "loss": 1.6438, + "step": 3881 + }, + { + "epoch": 0.814007129377228, + "grad_norm": 7.59638470230116, + "learning_rate": 1.7098627763331804e-05, + "loss": 1.7968, + "step": 3882 + }, + { + "epoch": 0.8142168169427553, + "grad_norm": 6.369594802852877, + "learning_rate": 1.709703303695562e-05, + "loss": 2.0111, + "step": 3883 + }, + { + "epoch": 0.8144265045082827, + "grad_norm": 5.617026892781355, + "learning_rate": 1.7095437946844186e-05, + "loss": 1.6553, + "step": 3884 + }, + { + "epoch": 0.81463619207381, + "grad_norm": 6.712605791095175, + "learning_rate": 1.709384249307925e-05, + "loss": 1.7758, + "step": 3885 + }, + { + "epoch": 0.8148458796393374, + "grad_norm": 5.636393609553181, + "learning_rate": 1.7092246675742585e-05, + "loss": 1.7874, + "step": 3886 + }, + { + "epoch": 0.8150555672048647, + "grad_norm": 6.201993124639991, + "learning_rate": 1.7090650494915976e-05, + "loss": 1.5989, + "step": 3887 + }, + { + "epoch": 0.8152652547703921, + "grad_norm": 9.040236563336762, + "learning_rate": 1.708905395068123e-05, + "loss": 1.4851, + "step": 3888 + }, + { + "epoch": 0.8154749423359194, + "grad_norm": 5.858930164935578, + "learning_rate": 1.708745704312017e-05, + "loss": 1.7632, + "step": 3889 + }, + { + "epoch": 0.8156846299014469, + "grad_norm": 6.393560494320786, + "learning_rate": 1.708585977231465e-05, + "loss": 2.247, + "step": 3890 + }, + { + "epoch": 0.8158943174669742, + "grad_norm": 6.340133599402949, + "learning_rate": 1.7084262138346526e-05, + "loss": 1.5463, + "step": 3891 + }, + { + "epoch": 0.8161040050325016, + "grad_norm": 6.937659462205507, + "learning_rate": 1.708266414129768e-05, + "loss": 1.8473, + "step": 3892 + }, + { + "epoch": 0.8163136925980289, + "grad_norm": 6.741601316501367, + "learning_rate": 1.7081065781250013e-05, + "loss": 1.8479, + "step": 3893 + }, + { + "epoch": 0.8165233801635563, + "grad_norm": 7.502901551758546, + "learning_rate": 1.707946705828544e-05, + "loss": 1.709, + "step": 3894 + }, + { + "epoch": 0.8167330677290837, + "grad_norm": 6.9527500790872425, + "learning_rate": 1.7077867972485904e-05, + "loss": 1.4605, + "step": 3895 + }, + { + "epoch": 0.816942755294611, + "grad_norm": 5.7770182074210075, + "learning_rate": 1.7076268523933358e-05, + "loss": 1.5681, + "step": 3896 + }, + { + "epoch": 0.8171524428601384, + "grad_norm": 6.378681629755651, + "learning_rate": 1.7074668712709778e-05, + "loss": 1.5815, + "step": 3897 + }, + { + "epoch": 0.8173621304256657, + "grad_norm": 6.238428664092997, + "learning_rate": 1.7073068538897154e-05, + "loss": 1.6234, + "step": 3898 + }, + { + "epoch": 0.8175718179911932, + "grad_norm": 6.794791184725831, + "learning_rate": 1.7071468002577504e-05, + "loss": 2.0315, + "step": 3899 + }, + { + "epoch": 0.8177815055567205, + "grad_norm": 5.975337225159249, + "learning_rate": 1.706986710383285e-05, + "loss": 1.8659, + "step": 3900 + }, + { + "epoch": 0.8179911931222479, + "grad_norm": 6.962117465458161, + "learning_rate": 1.7068265842745244e-05, + "loss": 1.9934, + "step": 3901 + }, + { + "epoch": 0.8182008806877752, + "grad_norm": 6.797723912795713, + "learning_rate": 1.7066664219396755e-05, + "loss": 2.1104, + "step": 3902 + }, + { + "epoch": 0.8184105682533026, + "grad_norm": 5.802479631795701, + "learning_rate": 1.706506223386947e-05, + "loss": 1.5743, + "step": 3903 + }, + { + "epoch": 0.8186202558188299, + "grad_norm": 6.666724039398372, + "learning_rate": 1.7063459886245492e-05, + "loss": 1.8826, + "step": 3904 + }, + { + "epoch": 0.8188299433843573, + "grad_norm": 6.72251864907007, + "learning_rate": 1.706185717660694e-05, + "loss": 2.0979, + "step": 3905 + }, + { + "epoch": 0.8190396309498846, + "grad_norm": 6.48987150541069, + "learning_rate": 1.7060254105035966e-05, + "loss": 2.0513, + "step": 3906 + }, + { + "epoch": 0.819249318515412, + "grad_norm": 6.942590678729585, + "learning_rate": 1.7058650671614717e-05, + "loss": 1.8838, + "step": 3907 + }, + { + "epoch": 0.8194590060809394, + "grad_norm": 7.025102360484922, + "learning_rate": 1.7057046876425386e-05, + "loss": 1.8276, + "step": 3908 + }, + { + "epoch": 0.8196686936464668, + "grad_norm": 5.287910968566388, + "learning_rate": 1.7055442719550156e-05, + "loss": 1.4727, + "step": 3909 + }, + { + "epoch": 0.8198783812119941, + "grad_norm": 7.321205499189208, + "learning_rate": 1.7053838201071252e-05, + "loss": 1.867, + "step": 3910 + }, + { + "epoch": 0.8200880687775215, + "grad_norm": 5.485169547078051, + "learning_rate": 1.7052233321070904e-05, + "loss": 1.6076, + "step": 3911 + }, + { + "epoch": 0.8202977563430489, + "grad_norm": 5.793018704988025, + "learning_rate": 1.705062807963137e-05, + "loss": 1.689, + "step": 3912 + }, + { + "epoch": 0.8205074439085762, + "grad_norm": 5.7599728572233575, + "learning_rate": 1.7049022476834917e-05, + "loss": 1.8892, + "step": 3913 + }, + { + "epoch": 0.8207171314741036, + "grad_norm": 5.787770412984968, + "learning_rate": 1.704741651276384e-05, + "loss": 2.2209, + "step": 3914 + }, + { + "epoch": 0.8209268190396309, + "grad_norm": 5.779157183874328, + "learning_rate": 1.7045810187500436e-05, + "loss": 1.4783, + "step": 3915 + }, + { + "epoch": 0.8211365066051584, + "grad_norm": 8.111830612672588, + "learning_rate": 1.7044203501127043e-05, + "loss": 2.0416, + "step": 3916 + }, + { + "epoch": 0.8213461941706857, + "grad_norm": 6.648864842883792, + "learning_rate": 1.7042596453726003e-05, + "loss": 1.5495, + "step": 3917 + }, + { + "epoch": 0.8215558817362131, + "grad_norm": 6.475013067424919, + "learning_rate": 1.704098904537968e-05, + "loss": 1.7308, + "step": 3918 + }, + { + "epoch": 0.8217655693017404, + "grad_norm": 6.441652061371102, + "learning_rate": 1.703938127617046e-05, + "loss": 1.9492, + "step": 3919 + }, + { + "epoch": 0.8219752568672678, + "grad_norm": 8.202785891608146, + "learning_rate": 1.7037773146180737e-05, + "loss": 1.8701, + "step": 3920 + }, + { + "epoch": 0.8221849444327951, + "grad_norm": 6.405510752463787, + "learning_rate": 1.7036164655492932e-05, + "loss": 1.7116, + "step": 3921 + }, + { + "epoch": 0.8223946319983225, + "grad_norm": 5.307028005512784, + "learning_rate": 1.7034555804189487e-05, + "loss": 1.6826, + "step": 3922 + }, + { + "epoch": 0.8226043195638498, + "grad_norm": 6.980848627048561, + "learning_rate": 1.7032946592352857e-05, + "loss": 2.0592, + "step": 3923 + }, + { + "epoch": 0.8228140071293772, + "grad_norm": 6.6322796335710805, + "learning_rate": 1.7031337020065515e-05, + "loss": 1.8762, + "step": 3924 + }, + { + "epoch": 0.8230236946949046, + "grad_norm": 6.664248156728514, + "learning_rate": 1.7029727087409952e-05, + "loss": 1.8921, + "step": 3925 + }, + { + "epoch": 0.823233382260432, + "grad_norm": 5.679375082721564, + "learning_rate": 1.702811679446869e-05, + "loss": 1.9246, + "step": 3926 + }, + { + "epoch": 0.8234430698259593, + "grad_norm": 6.261660824279909, + "learning_rate": 1.702650614132425e-05, + "loss": 1.569, + "step": 3927 + }, + { + "epoch": 0.8236527573914867, + "grad_norm": 6.147221898437719, + "learning_rate": 1.702489512805918e-05, + "loss": 1.6396, + "step": 3928 + }, + { + "epoch": 0.823862444957014, + "grad_norm": 5.3295980078599285, + "learning_rate": 1.7023283754756053e-05, + "loss": 1.7624, + "step": 3929 + }, + { + "epoch": 0.8240721325225414, + "grad_norm": 6.609587325806969, + "learning_rate": 1.7021672021497456e-05, + "loss": 1.9656, + "step": 3930 + }, + { + "epoch": 0.8242818200880688, + "grad_norm": 6.434584650598884, + "learning_rate": 1.7020059928365987e-05, + "loss": 1.7911, + "step": 3931 + }, + { + "epoch": 0.8244915076535961, + "grad_norm": 5.40550898630555, + "learning_rate": 1.7018447475444267e-05, + "loss": 1.3427, + "step": 3932 + }, + { + "epoch": 0.8247011952191236, + "grad_norm": 7.199681556426338, + "learning_rate": 1.7016834662814946e-05, + "loss": 1.6897, + "step": 3933 + }, + { + "epoch": 0.8249108827846509, + "grad_norm": 7.2762209158635445, + "learning_rate": 1.701522149056068e-05, + "loss": 1.4452, + "step": 3934 + }, + { + "epoch": 0.8251205703501783, + "grad_norm": 6.721100600599509, + "learning_rate": 1.701360795876414e-05, + "loss": 1.6193, + "step": 3935 + }, + { + "epoch": 0.8253302579157056, + "grad_norm": 7.226238926267577, + "learning_rate": 1.7011994067508027e-05, + "loss": 1.7999, + "step": 3936 + }, + { + "epoch": 0.825539945481233, + "grad_norm": 5.948006949237822, + "learning_rate": 1.701037981687506e-05, + "loss": 1.655, + "step": 3937 + }, + { + "epoch": 0.8257496330467603, + "grad_norm": 6.595746328253617, + "learning_rate": 1.7008765206947967e-05, + "loss": 1.785, + "step": 3938 + }, + { + "epoch": 0.8259593206122877, + "grad_norm": 7.013954677420128, + "learning_rate": 1.7007150237809503e-05, + "loss": 1.7213, + "step": 3939 + }, + { + "epoch": 0.826169008177815, + "grad_norm": 7.462444451069835, + "learning_rate": 1.7005534909542435e-05, + "loss": 2.3463, + "step": 3940 + }, + { + "epoch": 0.8263786957433424, + "grad_norm": 6.6141827406002935, + "learning_rate": 1.700391922222955e-05, + "loss": 1.3879, + "step": 3941 + }, + { + "epoch": 0.8265883833088697, + "grad_norm": 7.3848168808440375, + "learning_rate": 1.700230317595366e-05, + "loss": 1.7175, + "step": 3942 + }, + { + "epoch": 0.8267980708743972, + "grad_norm": 7.790187907916004, + "learning_rate": 1.7000686770797586e-05, + "loss": 1.6395, + "step": 3943 + }, + { + "epoch": 0.8270077584399245, + "grad_norm": 6.812089362203391, + "learning_rate": 1.6999070006844174e-05, + "loss": 1.8498, + "step": 3944 + }, + { + "epoch": 0.8272174460054519, + "grad_norm": 7.2699440434590175, + "learning_rate": 1.6997452884176287e-05, + "loss": 1.7232, + "step": 3945 + }, + { + "epoch": 0.8274271335709792, + "grad_norm": 6.639489202137102, + "learning_rate": 1.6995835402876798e-05, + "loss": 1.8514, + "step": 3946 + }, + { + "epoch": 0.8276368211365066, + "grad_norm": 6.435596860653872, + "learning_rate": 1.6994217563028615e-05, + "loss": 1.729, + "step": 3947 + }, + { + "epoch": 0.827846508702034, + "grad_norm": 6.4058472140387686, + "learning_rate": 1.699259936471465e-05, + "loss": 1.8428, + "step": 3948 + }, + { + "epoch": 0.8280561962675613, + "grad_norm": 6.3039675545852445, + "learning_rate": 1.699098080801784e-05, + "loss": 2.1255, + "step": 3949 + }, + { + "epoch": 0.8282658838330887, + "grad_norm": 7.53350247338998, + "learning_rate": 1.6989361893021134e-05, + "loss": 2.2175, + "step": 3950 + }, + { + "epoch": 0.828475571398616, + "grad_norm": 6.950589553255795, + "learning_rate": 1.698774261980751e-05, + "loss": 2.2171, + "step": 3951 + }, + { + "epoch": 0.8286852589641435, + "grad_norm": 6.656253907041343, + "learning_rate": 1.698612298845996e-05, + "loss": 1.7233, + "step": 3952 + }, + { + "epoch": 0.8288949465296708, + "grad_norm": 6.2427764791960945, + "learning_rate": 1.6984502999061492e-05, + "loss": 1.8896, + "step": 3953 + }, + { + "epoch": 0.8291046340951982, + "grad_norm": 5.606495194544615, + "learning_rate": 1.698288265169513e-05, + "loss": 1.7047, + "step": 3954 + }, + { + "epoch": 0.8293143216607255, + "grad_norm": 6.912351971044972, + "learning_rate": 1.6981261946443915e-05, + "loss": 1.3881, + "step": 3955 + }, + { + "epoch": 0.8295240092262529, + "grad_norm": 5.754447320623951, + "learning_rate": 1.6979640883390924e-05, + "loss": 1.8004, + "step": 3956 + }, + { + "epoch": 0.8297336967917802, + "grad_norm": 6.475571073481391, + "learning_rate": 1.697801946261923e-05, + "loss": 1.8933, + "step": 3957 + }, + { + "epoch": 0.8299433843573076, + "grad_norm": 6.393119396321377, + "learning_rate": 1.6976397684211935e-05, + "loss": 1.9184, + "step": 3958 + }, + { + "epoch": 0.8301530719228349, + "grad_norm": 6.1871064110886165, + "learning_rate": 1.697477554825216e-05, + "loss": 1.5414, + "step": 3959 + }, + { + "epoch": 0.8303627594883624, + "grad_norm": 6.078185500252048, + "learning_rate": 1.697315305482304e-05, + "loss": 1.6762, + "step": 3960 + }, + { + "epoch": 0.8305724470538897, + "grad_norm": 7.887267942186894, + "learning_rate": 1.697153020400773e-05, + "loss": 2.2682, + "step": 3961 + }, + { + "epoch": 0.8307821346194171, + "grad_norm": 6.757760424495994, + "learning_rate": 1.6969906995889407e-05, + "loss": 1.7142, + "step": 3962 + }, + { + "epoch": 0.8309918221849444, + "grad_norm": 6.817190135598574, + "learning_rate": 1.6968283430551263e-05, + "loss": 1.6131, + "step": 3963 + }, + { + "epoch": 0.8312015097504718, + "grad_norm": 5.500258609668544, + "learning_rate": 1.6966659508076504e-05, + "loss": 1.673, + "step": 3964 + }, + { + "epoch": 0.8314111973159991, + "grad_norm": 5.33586380362758, + "learning_rate": 1.6965035228548365e-05, + "loss": 1.4146, + "step": 3965 + }, + { + "epoch": 0.8316208848815265, + "grad_norm": 6.961738659664671, + "learning_rate": 1.696341059205009e-05, + "loss": 1.89, + "step": 3966 + }, + { + "epoch": 0.8318305724470539, + "grad_norm": 6.475552082764284, + "learning_rate": 1.696178559866494e-05, + "loss": 1.7037, + "step": 3967 + }, + { + "epoch": 0.8320402600125812, + "grad_norm": 7.576856132343177, + "learning_rate": 1.6960160248476212e-05, + "loss": 1.5367, + "step": 3968 + }, + { + "epoch": 0.8322499475781087, + "grad_norm": 6.96790883441301, + "learning_rate": 1.6958534541567195e-05, + "loss": 1.9419, + "step": 3969 + }, + { + "epoch": 0.832459635143636, + "grad_norm": 5.95952315193859, + "learning_rate": 1.6956908478021213e-05, + "loss": 1.9999, + "step": 3970 + }, + { + "epoch": 0.8326693227091634, + "grad_norm": 6.2708041834424355, + "learning_rate": 1.6955282057921608e-05, + "loss": 1.9492, + "step": 3971 + }, + { + "epoch": 0.8328790102746907, + "grad_norm": 7.082938415490325, + "learning_rate": 1.695365528135173e-05, + "loss": 1.8262, + "step": 3972 + }, + { + "epoch": 0.8330886978402181, + "grad_norm": 8.493882233002296, + "learning_rate": 1.695202814839496e-05, + "loss": 1.5482, + "step": 3973 + }, + { + "epoch": 0.8332983854057454, + "grad_norm": 5.696394128221247, + "learning_rate": 1.6950400659134693e-05, + "loss": 1.6053, + "step": 3974 + }, + { + "epoch": 0.8335080729712728, + "grad_norm": 7.213121116884791, + "learning_rate": 1.6948772813654336e-05, + "loss": 1.7486, + "step": 3975 + }, + { + "epoch": 0.8337177605368001, + "grad_norm": 7.0743770787021765, + "learning_rate": 1.6947144612037316e-05, + "loss": 1.4862, + "step": 3976 + }, + { + "epoch": 0.8339274481023276, + "grad_norm": 6.5501205717384385, + "learning_rate": 1.694551605436709e-05, + "loss": 1.8104, + "step": 3977 + }, + { + "epoch": 0.8341371356678549, + "grad_norm": 5.81805469112791, + "learning_rate": 1.6943887140727118e-05, + "loss": 1.9099, + "step": 3978 + }, + { + "epoch": 0.8343468232333823, + "grad_norm": 6.016732702821326, + "learning_rate": 1.6942257871200885e-05, + "loss": 1.4017, + "step": 3979 + }, + { + "epoch": 0.8345565107989096, + "grad_norm": 6.349253790757424, + "learning_rate": 1.69406282458719e-05, + "loss": 1.632, + "step": 3980 + }, + { + "epoch": 0.834766198364437, + "grad_norm": 6.592541998110839, + "learning_rate": 1.6938998264823675e-05, + "loss": 1.6978, + "step": 3981 + }, + { + "epoch": 0.8349758859299643, + "grad_norm": 5.551872739046872, + "learning_rate": 1.6937367928139758e-05, + "loss": 1.3985, + "step": 3982 + }, + { + "epoch": 0.8351855734954917, + "grad_norm": 5.8725837130670255, + "learning_rate": 1.69357372359037e-05, + "loss": 1.9824, + "step": 3983 + }, + { + "epoch": 0.8353952610610191, + "grad_norm": 6.690071727941447, + "learning_rate": 1.693410618819908e-05, + "loss": 2.1046, + "step": 3984 + }, + { + "epoch": 0.8356049486265464, + "grad_norm": 6.812457278647027, + "learning_rate": 1.693247478510949e-05, + "loss": 1.751, + "step": 3985 + }, + { + "epoch": 0.8358146361920739, + "grad_norm": 5.752509665320992, + "learning_rate": 1.693084302671854e-05, + "loss": 1.7108, + "step": 3986 + }, + { + "epoch": 0.8360243237576012, + "grad_norm": 7.0222242488131235, + "learning_rate": 1.692921091310987e-05, + "loss": 1.8449, + "step": 3987 + }, + { + "epoch": 0.8362340113231286, + "grad_norm": 7.896002349747242, + "learning_rate": 1.692757844436712e-05, + "loss": 1.7794, + "step": 3988 + }, + { + "epoch": 0.8364436988886559, + "grad_norm": 6.149762195566721, + "learning_rate": 1.692594562057396e-05, + "loss": 1.7636, + "step": 3989 + }, + { + "epoch": 0.8366533864541833, + "grad_norm": 6.336346190062133, + "learning_rate": 1.6924312441814075e-05, + "loss": 2.0715, + "step": 3990 + }, + { + "epoch": 0.8368630740197106, + "grad_norm": 7.62156232109994, + "learning_rate": 1.692267890817117e-05, + "loss": 1.5241, + "step": 3991 + }, + { + "epoch": 0.837072761585238, + "grad_norm": 6.80000315404917, + "learning_rate": 1.692104501972896e-05, + "loss": 1.8459, + "step": 3992 + }, + { + "epoch": 0.8372824491507653, + "grad_norm": 6.525447921706902, + "learning_rate": 1.691941077657119e-05, + "loss": 1.7569, + "step": 3993 + }, + { + "epoch": 0.8374921367162927, + "grad_norm": 6.305492333798289, + "learning_rate": 1.6917776178781618e-05, + "loss": 1.7838, + "step": 3994 + }, + { + "epoch": 0.83770182428182, + "grad_norm": 6.260333980163938, + "learning_rate": 1.6916141226444017e-05, + "loss": 1.7514, + "step": 3995 + }, + { + "epoch": 0.8379115118473475, + "grad_norm": 7.387058605029511, + "learning_rate": 1.6914505919642183e-05, + "loss": 2.1554, + "step": 3996 + }, + { + "epoch": 0.8381211994128748, + "grad_norm": 6.764726114907841, + "learning_rate": 1.6912870258459927e-05, + "loss": 1.6502, + "step": 3997 + }, + { + "epoch": 0.8383308869784022, + "grad_norm": 5.634243443455935, + "learning_rate": 1.6911234242981082e-05, + "loss": 2.0902, + "step": 3998 + }, + { + "epoch": 0.8385405745439295, + "grad_norm": 6.549890450228507, + "learning_rate": 1.6909597873289495e-05, + "loss": 1.6364, + "step": 3999 + }, + { + "epoch": 0.8387502621094569, + "grad_norm": 6.183494656841771, + "learning_rate": 1.6907961149469034e-05, + "loss": 1.602, + "step": 4000 + }, + { + "epoch": 0.8389599496749843, + "grad_norm": 7.217535730101063, + "learning_rate": 1.690632407160358e-05, + "loss": 1.9851, + "step": 4001 + }, + { + "epoch": 0.8391696372405116, + "grad_norm": 5.9332054416389015, + "learning_rate": 1.6904686639777042e-05, + "loss": 1.5317, + "step": 4002 + }, + { + "epoch": 0.839379324806039, + "grad_norm": 6.715253099310353, + "learning_rate": 1.6903048854073335e-05, + "loss": 1.8186, + "step": 4003 + }, + { + "epoch": 0.8395890123715664, + "grad_norm": 6.753639254937967, + "learning_rate": 1.6901410714576406e-05, + "loss": 1.7341, + "step": 4004 + }, + { + "epoch": 0.8397986999370938, + "grad_norm": 6.1121809819360955, + "learning_rate": 1.6899772221370205e-05, + "loss": 2.0645, + "step": 4005 + }, + { + "epoch": 0.8400083875026211, + "grad_norm": 6.906744394471587, + "learning_rate": 1.6898133374538708e-05, + "loss": 1.9306, + "step": 4006 + }, + { + "epoch": 0.8402180750681485, + "grad_norm": 6.477071918635669, + "learning_rate": 1.6896494174165915e-05, + "loss": 1.8979, + "step": 4007 + }, + { + "epoch": 0.8404277626336758, + "grad_norm": 7.576741766780498, + "learning_rate": 1.689485462033583e-05, + "loss": 1.7563, + "step": 4008 + }, + { + "epoch": 0.8406374501992032, + "grad_norm": 7.50509130129985, + "learning_rate": 1.6893214713132493e-05, + "loss": 1.9493, + "step": 4009 + }, + { + "epoch": 0.8408471377647305, + "grad_norm": 7.121248469291038, + "learning_rate": 1.6891574452639945e-05, + "loss": 1.8676, + "step": 4010 + }, + { + "epoch": 0.8410568253302579, + "grad_norm": 7.452978984434102, + "learning_rate": 1.688993383894225e-05, + "loss": 1.8357, + "step": 4011 + }, + { + "epoch": 0.8412665128957852, + "grad_norm": 6.115838218404241, + "learning_rate": 1.68882928721235e-05, + "loss": 1.6665, + "step": 4012 + }, + { + "epoch": 0.8414762004613127, + "grad_norm": 6.181234890005197, + "learning_rate": 1.688665155226779e-05, + "loss": 1.7065, + "step": 4013 + }, + { + "epoch": 0.84168588802684, + "grad_norm": 6.813608444389689, + "learning_rate": 1.6885009879459245e-05, + "loss": 1.867, + "step": 4014 + }, + { + "epoch": 0.8418955755923674, + "grad_norm": 6.522394137976414, + "learning_rate": 1.6883367853782e-05, + "loss": 1.7219, + "step": 4015 + }, + { + "epoch": 0.8421052631578947, + "grad_norm": 6.85252620738763, + "learning_rate": 1.688172547532022e-05, + "loss": 1.8185, + "step": 4016 + }, + { + "epoch": 0.8423149507234221, + "grad_norm": 7.209103110221067, + "learning_rate": 1.6880082744158067e-05, + "loss": 1.7135, + "step": 4017 + }, + { + "epoch": 0.8425246382889494, + "grad_norm": 6.931465671789631, + "learning_rate": 1.687843966037974e-05, + "loss": 1.789, + "step": 4018 + }, + { + "epoch": 0.8427343258544768, + "grad_norm": 6.916898807033722, + "learning_rate": 1.6876796224069457e-05, + "loss": 2.0119, + "step": 4019 + }, + { + "epoch": 0.8429440134200042, + "grad_norm": 7.305549369292113, + "learning_rate": 1.6875152435311437e-05, + "loss": 1.7291, + "step": 4020 + }, + { + "epoch": 0.8431537009855316, + "grad_norm": 7.192916222582827, + "learning_rate": 1.687350829418993e-05, + "loss": 1.6291, + "step": 4021 + }, + { + "epoch": 0.843363388551059, + "grad_norm": 7.396340310351018, + "learning_rate": 1.68718638007892e-05, + "loss": 1.8393, + "step": 4022 + }, + { + "epoch": 0.8435730761165863, + "grad_norm": 7.770199906827693, + "learning_rate": 1.687021895519353e-05, + "loss": 1.931, + "step": 4023 + }, + { + "epoch": 0.8437827636821137, + "grad_norm": 6.771345182781149, + "learning_rate": 1.686857375748723e-05, + "loss": 1.9987, + "step": 4024 + }, + { + "epoch": 0.843992451247641, + "grad_norm": 6.369968201595189, + "learning_rate": 1.6866928207754606e-05, + "loss": 1.8065, + "step": 4025 + }, + { + "epoch": 0.8442021388131684, + "grad_norm": 6.939187416800364, + "learning_rate": 1.6865282306080004e-05, + "loss": 1.5416, + "step": 4026 + }, + { + "epoch": 0.8444118263786957, + "grad_norm": 6.160319056442851, + "learning_rate": 1.6863636052547777e-05, + "loss": 1.6808, + "step": 4027 + }, + { + "epoch": 0.8446215139442231, + "grad_norm": 6.671515785380993, + "learning_rate": 1.6861989447242292e-05, + "loss": 1.9519, + "step": 4028 + }, + { + "epoch": 0.8448312015097504, + "grad_norm": 5.624884097408585, + "learning_rate": 1.6860342490247955e-05, + "loss": 1.3278, + "step": 4029 + }, + { + "epoch": 0.8450408890752779, + "grad_norm": 6.1733010812894324, + "learning_rate": 1.685869518164916e-05, + "loss": 1.9412, + "step": 4030 + }, + { + "epoch": 0.8452505766408052, + "grad_norm": 5.765314732838086, + "learning_rate": 1.6857047521530346e-05, + "loss": 1.6069, + "step": 4031 + }, + { + "epoch": 0.8454602642063326, + "grad_norm": 6.374809020935715, + "learning_rate": 1.685539950997595e-05, + "loss": 1.7501, + "step": 4032 + }, + { + "epoch": 0.8456699517718599, + "grad_norm": 5.9228495312441165, + "learning_rate": 1.685375114707044e-05, + "loss": 1.7501, + "step": 4033 + }, + { + "epoch": 0.8458796393373873, + "grad_norm": 7.218462249616903, + "learning_rate": 1.6852102432898298e-05, + "loss": 2.0852, + "step": 4034 + }, + { + "epoch": 0.8460893269029146, + "grad_norm": 7.838262731900389, + "learning_rate": 1.685045336754402e-05, + "loss": 2.3111, + "step": 4035 + }, + { + "epoch": 0.846299014468442, + "grad_norm": 6.747390285628549, + "learning_rate": 1.684880395109213e-05, + "loss": 1.8842, + "step": 4036 + }, + { + "epoch": 0.8465087020339694, + "grad_norm": 5.644998930329347, + "learning_rate": 1.684715418362715e-05, + "loss": 1.7946, + "step": 4037 + }, + { + "epoch": 0.8467183895994967, + "grad_norm": 6.49390528388707, + "learning_rate": 1.684550406523365e-05, + "loss": 1.8395, + "step": 4038 + }, + { + "epoch": 0.8469280771650242, + "grad_norm": 6.15223090537594, + "learning_rate": 1.6843853595996188e-05, + "loss": 1.7943, + "step": 4039 + }, + { + "epoch": 0.8471377647305515, + "grad_norm": 6.118079923338906, + "learning_rate": 1.6842202775999363e-05, + "loss": 1.6471, + "step": 4040 + }, + { + "epoch": 0.8473474522960789, + "grad_norm": 6.45871309394736, + "learning_rate": 1.6840551605327778e-05, + "loss": 1.7374, + "step": 4041 + }, + { + "epoch": 0.8475571398616062, + "grad_norm": 5.340383660209119, + "learning_rate": 1.6838900084066057e-05, + "loss": 1.5416, + "step": 4042 + }, + { + "epoch": 0.8477668274271336, + "grad_norm": 5.431163753877188, + "learning_rate": 1.683724821229885e-05, + "loss": 2.0014, + "step": 4043 + }, + { + "epoch": 0.8479765149926609, + "grad_norm": 5.786637531431312, + "learning_rate": 1.683559599011081e-05, + "loss": 1.7262, + "step": 4044 + }, + { + "epoch": 0.8481862025581883, + "grad_norm": 5.728800571441821, + "learning_rate": 1.683394341758662e-05, + "loss": 1.6645, + "step": 4045 + }, + { + "epoch": 0.8483958901237156, + "grad_norm": 5.84242924625343, + "learning_rate": 1.6832290494810973e-05, + "loss": 1.6721, + "step": 4046 + }, + { + "epoch": 0.848605577689243, + "grad_norm": 5.762757560058926, + "learning_rate": 1.6830637221868594e-05, + "loss": 1.5264, + "step": 4047 + }, + { + "epoch": 0.8488152652547704, + "grad_norm": 6.494943925464844, + "learning_rate": 1.682898359884421e-05, + "loss": 1.4834, + "step": 4048 + }, + { + "epoch": 0.8490249528202978, + "grad_norm": 6.178539836281641, + "learning_rate": 1.682732962582257e-05, + "loss": 2.0168, + "step": 4049 + }, + { + "epoch": 0.8492346403858251, + "grad_norm": 7.5972464468103595, + "learning_rate": 1.6825675302888447e-05, + "loss": 1.8575, + "step": 4050 + }, + { + "epoch": 0.8494443279513525, + "grad_norm": 5.69115397775228, + "learning_rate": 1.6824020630126625e-05, + "loss": 1.5521, + "step": 4051 + }, + { + "epoch": 0.8496540155168798, + "grad_norm": 7.801675329561413, + "learning_rate": 1.682236560762191e-05, + "loss": 1.9698, + "step": 4052 + }, + { + "epoch": 0.8498637030824072, + "grad_norm": 5.769053570789696, + "learning_rate": 1.6820710235459127e-05, + "loss": 1.5013, + "step": 4053 + }, + { + "epoch": 0.8500733906479345, + "grad_norm": 6.546824269564784, + "learning_rate": 1.681905451372311e-05, + "loss": 1.8828, + "step": 4054 + }, + { + "epoch": 0.8502830782134619, + "grad_norm": 6.04430494919629, + "learning_rate": 1.6817398442498723e-05, + "loss": 1.6741, + "step": 4055 + }, + { + "epoch": 0.8504927657789894, + "grad_norm": 6.193996276395604, + "learning_rate": 1.6815742021870845e-05, + "loss": 1.6034, + "step": 4056 + }, + { + "epoch": 0.8507024533445167, + "grad_norm": 6.5374210177670955, + "learning_rate": 1.6814085251924364e-05, + "loss": 1.5482, + "step": 4057 + }, + { + "epoch": 0.8509121409100441, + "grad_norm": 6.514336917218364, + "learning_rate": 1.68124281327442e-05, + "loss": 1.6008, + "step": 4058 + }, + { + "epoch": 0.8511218284755714, + "grad_norm": 6.7995146920074685, + "learning_rate": 1.6810770664415276e-05, + "loss": 1.8645, + "step": 4059 + }, + { + "epoch": 0.8513315160410988, + "grad_norm": 6.312212077918305, + "learning_rate": 1.680911284702254e-05, + "loss": 2.003, + "step": 4060 + }, + { + "epoch": 0.8515412036066261, + "grad_norm": 6.688409603756761, + "learning_rate": 1.6807454680650966e-05, + "loss": 1.8913, + "step": 4061 + }, + { + "epoch": 0.8517508911721535, + "grad_norm": 6.369184361389251, + "learning_rate": 1.6805796165385528e-05, + "loss": 1.8479, + "step": 4062 + }, + { + "epoch": 0.8519605787376808, + "grad_norm": 7.5700002444483285, + "learning_rate": 1.6804137301311234e-05, + "loss": 2.1044, + "step": 4063 + }, + { + "epoch": 0.8521702663032082, + "grad_norm": 6.817948679081982, + "learning_rate": 1.68024780885131e-05, + "loss": 1.5958, + "step": 4064 + }, + { + "epoch": 0.8523799538687356, + "grad_norm": 7.506265727778259, + "learning_rate": 1.680081852707617e-05, + "loss": 1.727, + "step": 4065 + }, + { + "epoch": 0.852589641434263, + "grad_norm": 7.280909309786979, + "learning_rate": 1.6799158617085493e-05, + "loss": 2.0798, + "step": 4066 + }, + { + "epoch": 0.8527993289997903, + "grad_norm": 6.67376062868688, + "learning_rate": 1.6797498358626144e-05, + "loss": 1.3141, + "step": 4067 + }, + { + "epoch": 0.8530090165653177, + "grad_norm": 6.738049115867896, + "learning_rate": 1.6795837751783215e-05, + "loss": 1.7618, + "step": 4068 + }, + { + "epoch": 0.853218704130845, + "grad_norm": 5.461867226881196, + "learning_rate": 1.6794176796641816e-05, + "loss": 1.3658, + "step": 4069 + }, + { + "epoch": 0.8534283916963724, + "grad_norm": 6.064585966998766, + "learning_rate": 1.679251549328707e-05, + "loss": 1.2837, + "step": 4070 + }, + { + "epoch": 0.8536380792618997, + "grad_norm": 5.288320535297466, + "learning_rate": 1.679085384180412e-05, + "loss": 1.7139, + "step": 4071 + }, + { + "epoch": 0.8538477668274271, + "grad_norm": 7.042558412663411, + "learning_rate": 1.678919184227814e-05, + "loss": 1.6952, + "step": 4072 + }, + { + "epoch": 0.8540574543929546, + "grad_norm": 5.768677494000512, + "learning_rate": 1.67875294947943e-05, + "loss": 1.7048, + "step": 4073 + }, + { + "epoch": 0.8542671419584819, + "grad_norm": 6.605477401303626, + "learning_rate": 1.6785866799437802e-05, + "loss": 1.8234, + "step": 4074 + }, + { + "epoch": 0.8544768295240093, + "grad_norm": 6.333116552791403, + "learning_rate": 1.6784203756293858e-05, + "loss": 2.0078, + "step": 4075 + }, + { + "epoch": 0.8546865170895366, + "grad_norm": 6.347038795224373, + "learning_rate": 1.6782540365447706e-05, + "loss": 1.9784, + "step": 4076 + }, + { + "epoch": 0.854896204655064, + "grad_norm": 5.442063158279843, + "learning_rate": 1.67808766269846e-05, + "loss": 1.5329, + "step": 4077 + }, + { + "epoch": 0.8551058922205913, + "grad_norm": 6.584376713015203, + "learning_rate": 1.6779212540989802e-05, + "loss": 1.8931, + "step": 4078 + }, + { + "epoch": 0.8553155797861187, + "grad_norm": 6.132185835620608, + "learning_rate": 1.6777548107548605e-05, + "loss": 1.3404, + "step": 4079 + }, + { + "epoch": 0.855525267351646, + "grad_norm": 5.875433445790861, + "learning_rate": 1.677588332674631e-05, + "loss": 1.702, + "step": 4080 + }, + { + "epoch": 0.8557349549171734, + "grad_norm": 6.516010102519589, + "learning_rate": 1.6774218198668246e-05, + "loss": 1.7293, + "step": 4081 + }, + { + "epoch": 0.8559446424827007, + "grad_norm": 7.416113122299674, + "learning_rate": 1.677255272339975e-05, + "loss": 1.6949, + "step": 4082 + }, + { + "epoch": 0.8561543300482282, + "grad_norm": 6.27239336850154, + "learning_rate": 1.6770886901026175e-05, + "loss": 1.7851, + "step": 4083 + }, + { + "epoch": 0.8563640176137555, + "grad_norm": 6.645256940827824, + "learning_rate": 1.676922073163291e-05, + "loss": 1.8578, + "step": 4084 + }, + { + "epoch": 0.8565737051792829, + "grad_norm": 8.519694567238316, + "learning_rate": 1.6767554215305336e-05, + "loss": 1.9412, + "step": 4085 + }, + { + "epoch": 0.8567833927448102, + "grad_norm": 8.558611013620533, + "learning_rate": 1.6765887352128875e-05, + "loss": 1.7994, + "step": 4086 + }, + { + "epoch": 0.8569930803103376, + "grad_norm": 7.269436595798622, + "learning_rate": 1.676422014218895e-05, + "loss": 1.6119, + "step": 4087 + }, + { + "epoch": 0.8572027678758649, + "grad_norm": 8.358283549429638, + "learning_rate": 1.6762552585571013e-05, + "loss": 1.8244, + "step": 4088 + }, + { + "epoch": 0.8574124554413923, + "grad_norm": 7.274918469975, + "learning_rate": 1.6760884682360524e-05, + "loss": 1.6623, + "step": 4089 + }, + { + "epoch": 0.8576221430069196, + "grad_norm": 6.267100562341706, + "learning_rate": 1.675921643264297e-05, + "loss": 1.5391, + "step": 4090 + }, + { + "epoch": 0.857831830572447, + "grad_norm": 7.294849565446345, + "learning_rate": 1.6757547836503852e-05, + "loss": 1.6781, + "step": 4091 + }, + { + "epoch": 0.8580415181379745, + "grad_norm": 6.300140674071617, + "learning_rate": 1.6755878894028686e-05, + "loss": 1.758, + "step": 4092 + }, + { + "epoch": 0.8582512057035018, + "grad_norm": 6.416570693730129, + "learning_rate": 1.675420960530301e-05, + "loss": 1.7945, + "step": 4093 + }, + { + "epoch": 0.8584608932690292, + "grad_norm": 7.767231207369511, + "learning_rate": 1.6752539970412372e-05, + "loss": 1.684, + "step": 4094 + }, + { + "epoch": 0.8586705808345565, + "grad_norm": 8.243118646417733, + "learning_rate": 1.6750869989442355e-05, + "loss": 2.0537, + "step": 4095 + }, + { + "epoch": 0.8588802684000839, + "grad_norm": 6.190850573492643, + "learning_rate": 1.674919966247854e-05, + "loss": 1.8319, + "step": 4096 + }, + { + "epoch": 0.8590899559656112, + "grad_norm": 6.042247186646617, + "learning_rate": 1.6747528989606538e-05, + "loss": 1.7628, + "step": 4097 + }, + { + "epoch": 0.8592996435311386, + "grad_norm": 6.424365228216792, + "learning_rate": 1.674585797091197e-05, + "loss": 1.6779, + "step": 4098 + }, + { + "epoch": 0.8595093310966659, + "grad_norm": 6.080976025170756, + "learning_rate": 1.6744186606480482e-05, + "loss": 2.0233, + "step": 4099 + }, + { + "epoch": 0.8597190186621934, + "grad_norm": 5.227385829929483, + "learning_rate": 1.6742514896397732e-05, + "loss": 1.3561, + "step": 4100 + }, + { + "epoch": 0.8599287062277207, + "grad_norm": 6.460993572808233, + "learning_rate": 1.67408428407494e-05, + "loss": 2.0457, + "step": 4101 + }, + { + "epoch": 0.8601383937932481, + "grad_norm": 7.001027589011534, + "learning_rate": 1.6739170439621184e-05, + "loss": 1.5379, + "step": 4102 + }, + { + "epoch": 0.8603480813587754, + "grad_norm": 5.66578736761006, + "learning_rate": 1.6737497693098792e-05, + "loss": 1.764, + "step": 4103 + }, + { + "epoch": 0.8605577689243028, + "grad_norm": 6.535439473448954, + "learning_rate": 1.6735824601267956e-05, + "loss": 2.085, + "step": 4104 + }, + { + "epoch": 0.8607674564898301, + "grad_norm": 5.602150166482043, + "learning_rate": 1.6734151164214426e-05, + "loss": 1.5367, + "step": 4105 + }, + { + "epoch": 0.8609771440553575, + "grad_norm": 6.345825360859426, + "learning_rate": 1.6732477382023975e-05, + "loss": 1.5763, + "step": 4106 + }, + { + "epoch": 0.8611868316208848, + "grad_norm": 6.197337714905158, + "learning_rate": 1.6730803254782375e-05, + "loss": 1.9616, + "step": 4107 + }, + { + "epoch": 0.8613965191864122, + "grad_norm": 6.704681234172346, + "learning_rate": 1.672912878257544e-05, + "loss": 1.7714, + "step": 4108 + }, + { + "epoch": 0.8616062067519397, + "grad_norm": 6.84404509215181, + "learning_rate": 1.6727453965488977e-05, + "loss": 1.8884, + "step": 4109 + }, + { + "epoch": 0.861815894317467, + "grad_norm": 5.759213560494223, + "learning_rate": 1.6725778803608836e-05, + "loss": 1.6507, + "step": 4110 + }, + { + "epoch": 0.8620255818829944, + "grad_norm": 6.43857197204098, + "learning_rate": 1.672410329702086e-05, + "loss": 1.8624, + "step": 4111 + }, + { + "epoch": 0.8622352694485217, + "grad_norm": 6.617445157629611, + "learning_rate": 1.6722427445810935e-05, + "loss": 1.9023, + "step": 4112 + }, + { + "epoch": 0.8624449570140491, + "grad_norm": 6.347460036252912, + "learning_rate": 1.6720751250064936e-05, + "loss": 2.1338, + "step": 4113 + }, + { + "epoch": 0.8626546445795764, + "grad_norm": 5.932569865442243, + "learning_rate": 1.6719074709868785e-05, + "loss": 1.6801, + "step": 4114 + }, + { + "epoch": 0.8628643321451038, + "grad_norm": 6.403342864378055, + "learning_rate": 1.67173978253084e-05, + "loss": 1.4328, + "step": 4115 + }, + { + "epoch": 0.8630740197106311, + "grad_norm": 5.670422689912964, + "learning_rate": 1.6715720596469723e-05, + "loss": 1.6441, + "step": 4116 + }, + { + "epoch": 0.8632837072761586, + "grad_norm": 6.065047645298659, + "learning_rate": 1.6714043023438724e-05, + "loss": 2.0748, + "step": 4117 + }, + { + "epoch": 0.8634933948416859, + "grad_norm": 5.580019512609728, + "learning_rate": 1.6712365106301368e-05, + "loss": 1.831, + "step": 4118 + }, + { + "epoch": 0.8637030824072133, + "grad_norm": 6.417617044226933, + "learning_rate": 1.671068684514366e-05, + "loss": 1.8771, + "step": 4119 + }, + { + "epoch": 0.8639127699727406, + "grad_norm": 5.883432892905316, + "learning_rate": 1.6709008240051615e-05, + "loss": 1.73, + "step": 4120 + }, + { + "epoch": 0.864122457538268, + "grad_norm": 5.939082319613713, + "learning_rate": 1.670732929111126e-05, + "loss": 1.8674, + "step": 4121 + }, + { + "epoch": 0.8643321451037953, + "grad_norm": 5.769408061985854, + "learning_rate": 1.6705649998408644e-05, + "loss": 1.8751, + "step": 4122 + }, + { + "epoch": 0.8645418326693227, + "grad_norm": 5.939437567638129, + "learning_rate": 1.6703970362029836e-05, + "loss": 1.7763, + "step": 4123 + }, + { + "epoch": 0.86475152023485, + "grad_norm": 6.864290094889894, + "learning_rate": 1.6702290382060923e-05, + "loss": 1.5722, + "step": 4124 + }, + { + "epoch": 0.8649612078003774, + "grad_norm": 5.900771778414244, + "learning_rate": 1.6700610058588e-05, + "loss": 1.6329, + "step": 4125 + }, + { + "epoch": 0.8651708953659049, + "grad_norm": 6.397590505603625, + "learning_rate": 1.669892939169719e-05, + "loss": 1.5469, + "step": 4126 + }, + { + "epoch": 0.8653805829314322, + "grad_norm": 6.4875939528809115, + "learning_rate": 1.6697248381474628e-05, + "loss": 1.8049, + "step": 4127 + }, + { + "epoch": 0.8655902704969596, + "grad_norm": 6.224547445409898, + "learning_rate": 1.6695567028006473e-05, + "loss": 1.7051, + "step": 4128 + }, + { + "epoch": 0.8657999580624869, + "grad_norm": 7.289401450254317, + "learning_rate": 1.6693885331378895e-05, + "loss": 1.9078, + "step": 4129 + }, + { + "epoch": 0.8660096456280143, + "grad_norm": 10.925646727562006, + "learning_rate": 1.6692203291678083e-05, + "loss": 1.874, + "step": 4130 + }, + { + "epoch": 0.8662193331935416, + "grad_norm": 5.957501322225471, + "learning_rate": 1.6690520908990246e-05, + "loss": 1.775, + "step": 4131 + }, + { + "epoch": 0.866429020759069, + "grad_norm": 5.8350814328062155, + "learning_rate": 1.668883818340161e-05, + "loss": 1.4979, + "step": 4132 + }, + { + "epoch": 0.8666387083245963, + "grad_norm": 5.83128349361276, + "learning_rate": 1.6687155114998416e-05, + "loss": 1.8158, + "step": 4133 + }, + { + "epoch": 0.8668483958901237, + "grad_norm": 6.214088500148104, + "learning_rate": 1.668547170386692e-05, + "loss": 1.6123, + "step": 4134 + }, + { + "epoch": 0.867058083455651, + "grad_norm": 5.345833789676633, + "learning_rate": 1.6683787950093407e-05, + "loss": 1.6547, + "step": 4135 + }, + { + "epoch": 0.8672677710211785, + "grad_norm": 6.295755869637431, + "learning_rate": 1.6682103853764168e-05, + "loss": 1.5095, + "step": 4136 + }, + { + "epoch": 0.8674774585867058, + "grad_norm": 6.917409491913672, + "learning_rate": 1.6680419414965516e-05, + "loss": 1.9964, + "step": 4137 + }, + { + "epoch": 0.8676871461522332, + "grad_norm": 7.6354402584142544, + "learning_rate": 1.6678734633783783e-05, + "loss": 1.7531, + "step": 4138 + }, + { + "epoch": 0.8678968337177605, + "grad_norm": 6.232529202633637, + "learning_rate": 1.667704951030532e-05, + "loss": 1.7983, + "step": 4139 + }, + { + "epoch": 0.8681065212832879, + "grad_norm": 6.085741507743412, + "learning_rate": 1.6675364044616486e-05, + "loss": 1.7778, + "step": 4140 + }, + { + "epoch": 0.8683162088488152, + "grad_norm": 7.259996606017139, + "learning_rate": 1.6673678236803662e-05, + "loss": 1.9203, + "step": 4141 + }, + { + "epoch": 0.8685258964143426, + "grad_norm": 6.085718675555314, + "learning_rate": 1.667199208695326e-05, + "loss": 1.6597, + "step": 4142 + }, + { + "epoch": 0.8687355839798699, + "grad_norm": 6.70760805124773, + "learning_rate": 1.667030559515169e-05, + "loss": 1.7846, + "step": 4143 + }, + { + "epoch": 0.8689452715453974, + "grad_norm": 7.066188953890235, + "learning_rate": 1.6668618761485385e-05, + "loss": 1.6989, + "step": 4144 + }, + { + "epoch": 0.8691549591109248, + "grad_norm": 6.444540020861225, + "learning_rate": 1.6666931586040808e-05, + "loss": 2.0588, + "step": 4145 + }, + { + "epoch": 0.8693646466764521, + "grad_norm": 6.910348421867361, + "learning_rate": 1.6665244068904424e-05, + "loss": 1.5115, + "step": 4146 + }, + { + "epoch": 0.8695743342419795, + "grad_norm": 6.370747231989148, + "learning_rate": 1.6663556210162715e-05, + "loss": 1.5102, + "step": 4147 + }, + { + "epoch": 0.8697840218075068, + "grad_norm": 7.13844079649498, + "learning_rate": 1.66618680099022e-05, + "loss": 2.0964, + "step": 4148 + }, + { + "epoch": 0.8699937093730342, + "grad_norm": 7.20659292614585, + "learning_rate": 1.666017946820939e-05, + "loss": 1.5944, + "step": 4149 + }, + { + "epoch": 0.8702033969385615, + "grad_norm": 6.446851760976773, + "learning_rate": 1.6658490585170834e-05, + "loss": 1.7971, + "step": 4150 + }, + { + "epoch": 0.8704130845040889, + "grad_norm": 6.933176390484422, + "learning_rate": 1.6656801360873087e-05, + "loss": 1.985, + "step": 4151 + }, + { + "epoch": 0.8706227720696162, + "grad_norm": 7.044346910190448, + "learning_rate": 1.6655111795402724e-05, + "loss": 1.9448, + "step": 4152 + }, + { + "epoch": 0.8708324596351437, + "grad_norm": 6.666922888668738, + "learning_rate": 1.665342188884634e-05, + "loss": 1.8658, + "step": 4153 + }, + { + "epoch": 0.871042147200671, + "grad_norm": 5.815692865868926, + "learning_rate": 1.665173164129054e-05, + "loss": 1.6452, + "step": 4154 + }, + { + "epoch": 0.8712518347661984, + "grad_norm": 6.616894657237418, + "learning_rate": 1.665004105282196e-05, + "loss": 1.7552, + "step": 4155 + }, + { + "epoch": 0.8714615223317257, + "grad_norm": 5.943510227564201, + "learning_rate": 1.664835012352724e-05, + "loss": 1.658, + "step": 4156 + }, + { + "epoch": 0.8716712098972531, + "grad_norm": 6.728526701869622, + "learning_rate": 1.664665885349305e-05, + "loss": 1.6258, + "step": 4157 + }, + { + "epoch": 0.8718808974627804, + "grad_norm": 6.927190531965406, + "learning_rate": 1.664496724280606e-05, + "loss": 2.2536, + "step": 4158 + }, + { + "epoch": 0.8720905850283078, + "grad_norm": 6.256615811841955, + "learning_rate": 1.6643275291552978e-05, + "loss": 1.7736, + "step": 4159 + }, + { + "epoch": 0.8723002725938351, + "grad_norm": 7.404477606781132, + "learning_rate": 1.6641582999820514e-05, + "loss": 1.8992, + "step": 4160 + }, + { + "epoch": 0.8725099601593626, + "grad_norm": 6.600040278912407, + "learning_rate": 1.66398903676954e-05, + "loss": 1.4958, + "step": 4161 + }, + { + "epoch": 0.87271964772489, + "grad_norm": 6.533917000318811, + "learning_rate": 1.663819739526439e-05, + "loss": 1.6696, + "step": 4162 + }, + { + "epoch": 0.8729293352904173, + "grad_norm": 6.33142490413681, + "learning_rate": 1.6636504082614247e-05, + "loss": 1.9349, + "step": 4163 + }, + { + "epoch": 0.8731390228559447, + "grad_norm": 5.912891503656744, + "learning_rate": 1.6634810429831764e-05, + "loss": 1.5889, + "step": 4164 + }, + { + "epoch": 0.873348710421472, + "grad_norm": 5.8095362101669865, + "learning_rate": 1.6633116437003737e-05, + "loss": 1.9217, + "step": 4165 + }, + { + "epoch": 0.8735583979869994, + "grad_norm": 6.691066716627281, + "learning_rate": 1.663142210421699e-05, + "loss": 1.7573, + "step": 4166 + }, + { + "epoch": 0.8737680855525267, + "grad_norm": 6.795445764657608, + "learning_rate": 1.6629727431558355e-05, + "loss": 1.9464, + "step": 4167 + }, + { + "epoch": 0.8739777731180541, + "grad_norm": 7.396434034826449, + "learning_rate": 1.662803241911469e-05, + "loss": 1.4333, + "step": 4168 + }, + { + "epoch": 0.8741874606835814, + "grad_norm": 6.144683077864973, + "learning_rate": 1.662633706697287e-05, + "loss": 1.8239, + "step": 4169 + }, + { + "epoch": 0.8743971482491089, + "grad_norm": 5.607598988224902, + "learning_rate": 1.6624641375219782e-05, + "loss": 1.5463, + "step": 4170 + }, + { + "epoch": 0.8746068358146362, + "grad_norm": 6.436820754770758, + "learning_rate": 1.6622945343942336e-05, + "loss": 1.7603, + "step": 4171 + }, + { + "epoch": 0.8748165233801636, + "grad_norm": 6.193795991470847, + "learning_rate": 1.6621248973227454e-05, + "loss": 1.5974, + "step": 4172 + }, + { + "epoch": 0.8750262109456909, + "grad_norm": 6.503986743050873, + "learning_rate": 1.6619552263162076e-05, + "loss": 1.9723, + "step": 4173 + }, + { + "epoch": 0.8752358985112183, + "grad_norm": 8.326753326907651, + "learning_rate": 1.6617855213833162e-05, + "loss": 1.9234, + "step": 4174 + }, + { + "epoch": 0.8754455860767456, + "grad_norm": 6.146816444851391, + "learning_rate": 1.661615782532769e-05, + "loss": 1.7026, + "step": 4175 + }, + { + "epoch": 0.875655273642273, + "grad_norm": 6.895539325954905, + "learning_rate": 1.661446009773266e-05, + "loss": 1.6319, + "step": 4176 + }, + { + "epoch": 0.8758649612078003, + "grad_norm": 6.481939533611673, + "learning_rate": 1.6612762031135075e-05, + "loss": 1.953, + "step": 4177 + }, + { + "epoch": 0.8760746487733277, + "grad_norm": 6.156340351809294, + "learning_rate": 1.6611063625621963e-05, + "loss": 1.3428, + "step": 4178 + }, + { + "epoch": 0.876284336338855, + "grad_norm": 6.244744926077012, + "learning_rate": 1.6609364881280375e-05, + "loss": 1.4629, + "step": 4179 + }, + { + "epoch": 0.8764940239043825, + "grad_norm": 6.177259750176226, + "learning_rate": 1.6607665798197376e-05, + "loss": 1.4052, + "step": 4180 + }, + { + "epoch": 0.8767037114699099, + "grad_norm": 6.388589274637202, + "learning_rate": 1.6605966376460046e-05, + "loss": 1.8215, + "step": 4181 + }, + { + "epoch": 0.8769133990354372, + "grad_norm": 6.286666901060453, + "learning_rate": 1.6604266616155476e-05, + "loss": 1.607, + "step": 4182 + }, + { + "epoch": 0.8771230866009646, + "grad_norm": 6.310579091458862, + "learning_rate": 1.6602566517370793e-05, + "loss": 1.5558, + "step": 4183 + }, + { + "epoch": 0.8773327741664919, + "grad_norm": 6.313980374218684, + "learning_rate": 1.660086608019312e-05, + "loss": 1.8328, + "step": 4184 + }, + { + "epoch": 0.8775424617320193, + "grad_norm": 6.595329085417071, + "learning_rate": 1.6599165304709616e-05, + "loss": 1.6355, + "step": 4185 + }, + { + "epoch": 0.8777521492975466, + "grad_norm": 7.286760683543323, + "learning_rate": 1.659746419100744e-05, + "loss": 1.6846, + "step": 4186 + }, + { + "epoch": 0.877961836863074, + "grad_norm": 6.035817162734676, + "learning_rate": 1.6595762739173785e-05, + "loss": 1.7408, + "step": 4187 + }, + { + "epoch": 0.8781715244286014, + "grad_norm": 5.5316037705403724, + "learning_rate": 1.659406094929585e-05, + "loss": 1.4016, + "step": 4188 + }, + { + "epoch": 0.8783812119941288, + "grad_norm": 5.674474812781397, + "learning_rate": 1.6592358821460856e-05, + "loss": 1.4591, + "step": 4189 + }, + { + "epoch": 0.8785908995596561, + "grad_norm": 8.007116948368038, + "learning_rate": 1.6590656355756033e-05, + "loss": 2.2284, + "step": 4190 + }, + { + "epoch": 0.8788005871251835, + "grad_norm": 6.299430611650503, + "learning_rate": 1.6588953552268644e-05, + "loss": 1.792, + "step": 4191 + }, + { + "epoch": 0.8790102746907108, + "grad_norm": 6.975303432966199, + "learning_rate": 1.6587250411085956e-05, + "loss": 1.7971, + "step": 4192 + }, + { + "epoch": 0.8792199622562382, + "grad_norm": 5.836448985337462, + "learning_rate": 1.658554693229526e-05, + "loss": 1.4385, + "step": 4193 + }, + { + "epoch": 0.8794296498217655, + "grad_norm": 5.998244069465465, + "learning_rate": 1.6583843115983862e-05, + "loss": 1.7741, + "step": 4194 + }, + { + "epoch": 0.8796393373872929, + "grad_norm": 5.695771881860834, + "learning_rate": 1.6582138962239087e-05, + "loss": 1.6615, + "step": 4195 + }, + { + "epoch": 0.8798490249528202, + "grad_norm": 6.657903649702837, + "learning_rate": 1.658043447114827e-05, + "loss": 1.731, + "step": 4196 + }, + { + "epoch": 0.8800587125183477, + "grad_norm": 6.231902850585657, + "learning_rate": 1.6578729642798776e-05, + "loss": 1.7567, + "step": 4197 + }, + { + "epoch": 0.8802684000838751, + "grad_norm": 6.241851435191944, + "learning_rate": 1.657702447727798e-05, + "loss": 2.1111, + "step": 4198 + }, + { + "epoch": 0.8804780876494024, + "grad_norm": 6.194869427874186, + "learning_rate": 1.657531897467327e-05, + "loss": 1.8413, + "step": 4199 + }, + { + "epoch": 0.8806877752149298, + "grad_norm": 8.090924345094702, + "learning_rate": 1.6573613135072054e-05, + "loss": 2.0217, + "step": 4200 + }, + { + "epoch": 0.8808974627804571, + "grad_norm": 7.145242538444357, + "learning_rate": 1.657190695856177e-05, + "loss": 1.6433, + "step": 4201 + }, + { + "epoch": 0.8811071503459845, + "grad_norm": 6.783958492129735, + "learning_rate": 1.657020044522985e-05, + "loss": 1.8909, + "step": 4202 + }, + { + "epoch": 0.8813168379115118, + "grad_norm": 7.084025728111844, + "learning_rate": 1.6568493595163767e-05, + "loss": 1.8949, + "step": 4203 + }, + { + "epoch": 0.8815265254770392, + "grad_norm": 5.511609751078453, + "learning_rate": 1.6566786408450994e-05, + "loss": 1.6822, + "step": 4204 + }, + { + "epoch": 0.8817362130425666, + "grad_norm": 5.861108809171588, + "learning_rate": 1.6565078885179023e-05, + "loss": 1.6471, + "step": 4205 + }, + { + "epoch": 0.881945900608094, + "grad_norm": 5.9138373410186, + "learning_rate": 1.6563371025435378e-05, + "loss": 1.8203, + "step": 4206 + }, + { + "epoch": 0.8821555881736213, + "grad_norm": 7.5940866238742375, + "learning_rate": 1.6561662829307584e-05, + "loss": 1.8631, + "step": 4207 + }, + { + "epoch": 0.8823652757391487, + "grad_norm": 5.678065033979423, + "learning_rate": 1.6559954296883186e-05, + "loss": 1.259, + "step": 4208 + }, + { + "epoch": 0.882574963304676, + "grad_norm": 6.132008966117958, + "learning_rate": 1.6558245428249755e-05, + "loss": 1.7613, + "step": 4209 + }, + { + "epoch": 0.8827846508702034, + "grad_norm": 6.026407606866815, + "learning_rate": 1.6556536223494873e-05, + "loss": 1.8227, + "step": 4210 + }, + { + "epoch": 0.8829943384357307, + "grad_norm": 6.479886305033372, + "learning_rate": 1.6554826682706134e-05, + "loss": 1.7815, + "step": 4211 + }, + { + "epoch": 0.8832040260012581, + "grad_norm": 6.272687322304673, + "learning_rate": 1.6553116805971158e-05, + "loss": 1.1637, + "step": 4212 + }, + { + "epoch": 0.8834137135667854, + "grad_norm": 7.010320419088454, + "learning_rate": 1.6551406593377586e-05, + "loss": 2.0757, + "step": 4213 + }, + { + "epoch": 0.8836234011323129, + "grad_norm": 5.607070268571226, + "learning_rate": 1.6549696045013056e-05, + "loss": 1.4201, + "step": 4214 + }, + { + "epoch": 0.8838330886978402, + "grad_norm": 6.465862970307962, + "learning_rate": 1.6547985160965246e-05, + "loss": 1.7698, + "step": 4215 + }, + { + "epoch": 0.8840427762633676, + "grad_norm": 8.046636789694077, + "learning_rate": 1.6546273941321843e-05, + "loss": 1.7464, + "step": 4216 + }, + { + "epoch": 0.884252463828895, + "grad_norm": 6.871983865045377, + "learning_rate": 1.6544562386170544e-05, + "loss": 1.8556, + "step": 4217 + }, + { + "epoch": 0.8844621513944223, + "grad_norm": 7.02663216876567, + "learning_rate": 1.6542850495599073e-05, + "loss": 1.7666, + "step": 4218 + }, + { + "epoch": 0.8846718389599497, + "grad_norm": 6.364714001485779, + "learning_rate": 1.6541138269695166e-05, + "loss": 1.7579, + "step": 4219 + }, + { + "epoch": 0.884881526525477, + "grad_norm": 5.744510311875562, + "learning_rate": 1.653942570854658e-05, + "loss": 1.5383, + "step": 4220 + }, + { + "epoch": 0.8850912140910044, + "grad_norm": 6.202910582922797, + "learning_rate": 1.6537712812241078e-05, + "loss": 1.7455, + "step": 4221 + }, + { + "epoch": 0.8853009016565317, + "grad_norm": 6.274524909498673, + "learning_rate": 1.653599958086646e-05, + "loss": 1.503, + "step": 4222 + }, + { + "epoch": 0.8855105892220592, + "grad_norm": 8.240448855122168, + "learning_rate": 1.6534286014510525e-05, + "loss": 1.6262, + "step": 4223 + }, + { + "epoch": 0.8857202767875865, + "grad_norm": 5.3901477754515605, + "learning_rate": 1.65325721132611e-05, + "loss": 1.6222, + "step": 4224 + }, + { + "epoch": 0.8859299643531139, + "grad_norm": 5.496054191013193, + "learning_rate": 1.6530857877206023e-05, + "loss": 1.8457, + "step": 4225 + }, + { + "epoch": 0.8861396519186412, + "grad_norm": 6.721973920138175, + "learning_rate": 1.6529143306433154e-05, + "loss": 1.8193, + "step": 4226 + }, + { + "epoch": 0.8863493394841686, + "grad_norm": 6.789185424653191, + "learning_rate": 1.6527428401030365e-05, + "loss": 1.9454, + "step": 4227 + }, + { + "epoch": 0.8865590270496959, + "grad_norm": 7.414278783949691, + "learning_rate": 1.652571316108555e-05, + "loss": 2.1373, + "step": 4228 + }, + { + "epoch": 0.8867687146152233, + "grad_norm": 6.785899924502074, + "learning_rate": 1.6523997586686615e-05, + "loss": 1.7761, + "step": 4229 + }, + { + "epoch": 0.8869784021807506, + "grad_norm": 7.069587241695219, + "learning_rate": 1.6522281677921492e-05, + "loss": 1.9637, + "step": 4230 + }, + { + "epoch": 0.887188089746278, + "grad_norm": 6.102843097255973, + "learning_rate": 1.6520565434878116e-05, + "loss": 1.6843, + "step": 4231 + }, + { + "epoch": 0.8873977773118054, + "grad_norm": 6.775947982101723, + "learning_rate": 1.6518848857644457e-05, + "loss": 1.9318, + "step": 4232 + }, + { + "epoch": 0.8876074648773328, + "grad_norm": 6.376389146725458, + "learning_rate": 1.6517131946308485e-05, + "loss": 1.6781, + "step": 4233 + }, + { + "epoch": 0.8878171524428602, + "grad_norm": 6.307086498055612, + "learning_rate": 1.65154147009582e-05, + "loss": 1.5653, + "step": 4234 + }, + { + "epoch": 0.8880268400083875, + "grad_norm": 6.903887930849312, + "learning_rate": 1.6513697121681608e-05, + "loss": 1.8538, + "step": 4235 + }, + { + "epoch": 0.8882365275739149, + "grad_norm": 6.391011103727617, + "learning_rate": 1.6511979208566745e-05, + "loss": 1.8071, + "step": 4236 + }, + { + "epoch": 0.8884462151394422, + "grad_norm": 5.746264749818297, + "learning_rate": 1.6510260961701646e-05, + "loss": 1.9037, + "step": 4237 + }, + { + "epoch": 0.8886559027049696, + "grad_norm": 6.784776582153941, + "learning_rate": 1.6508542381174385e-05, + "loss": 2.0323, + "step": 4238 + }, + { + "epoch": 0.8888655902704969, + "grad_norm": 7.3612835810126604, + "learning_rate": 1.650682346707304e-05, + "loss": 1.9431, + "step": 4239 + }, + { + "epoch": 0.8890752778360244, + "grad_norm": 6.504269313988687, + "learning_rate": 1.6505104219485705e-05, + "loss": 1.918, + "step": 4240 + }, + { + "epoch": 0.8892849654015517, + "grad_norm": 5.167449973790571, + "learning_rate": 1.65033846385005e-05, + "loss": 1.4456, + "step": 4241 + }, + { + "epoch": 0.8894946529670791, + "grad_norm": 6.397935277898472, + "learning_rate": 1.650166472420555e-05, + "loss": 1.6406, + "step": 4242 + }, + { + "epoch": 0.8897043405326064, + "grad_norm": 6.162533108459571, + "learning_rate": 1.6499944476689006e-05, + "loss": 1.7019, + "step": 4243 + }, + { + "epoch": 0.8899140280981338, + "grad_norm": 7.077215973426679, + "learning_rate": 1.6498223896039035e-05, + "loss": 1.4343, + "step": 4244 + }, + { + "epoch": 0.8901237156636611, + "grad_norm": 6.474706000905946, + "learning_rate": 1.6496502982343817e-05, + "loss": 1.5677, + "step": 4245 + }, + { + "epoch": 0.8903334032291885, + "grad_norm": 6.030284493007642, + "learning_rate": 1.6494781735691555e-05, + "loss": 1.6542, + "step": 4246 + }, + { + "epoch": 0.8905430907947158, + "grad_norm": 7.082140868043991, + "learning_rate": 1.6493060156170463e-05, + "loss": 1.8232, + "step": 4247 + }, + { + "epoch": 0.8907527783602432, + "grad_norm": 6.145336002177126, + "learning_rate": 1.649133824386878e-05, + "loss": 1.5695, + "step": 4248 + }, + { + "epoch": 0.8909624659257706, + "grad_norm": 6.770046406997616, + "learning_rate": 1.648961599887475e-05, + "loss": 1.518, + "step": 4249 + }, + { + "epoch": 0.891172153491298, + "grad_norm": 6.376799723545786, + "learning_rate": 1.6487893421276643e-05, + "loss": 1.6271, + "step": 4250 + }, + { + "epoch": 0.8913818410568253, + "grad_norm": 7.292859050577482, + "learning_rate": 1.648617051116275e-05, + "loss": 1.7005, + "step": 4251 + }, + { + "epoch": 0.8915915286223527, + "grad_norm": 6.668519049889091, + "learning_rate": 1.6484447268621364e-05, + "loss": 1.5505, + "step": 4252 + }, + { + "epoch": 0.8918012161878801, + "grad_norm": 7.278232045203354, + "learning_rate": 1.648272369374081e-05, + "loss": 1.816, + "step": 4253 + }, + { + "epoch": 0.8920109037534074, + "grad_norm": 7.177546690738252, + "learning_rate": 1.6480999786609423e-05, + "loss": 1.8321, + "step": 4254 + }, + { + "epoch": 0.8922205913189348, + "grad_norm": 7.95248473626415, + "learning_rate": 1.6479275547315558e-05, + "loss": 1.5383, + "step": 4255 + }, + { + "epoch": 0.8924302788844621, + "grad_norm": 6.952452889383223, + "learning_rate": 1.6477550975947584e-05, + "loss": 1.4688, + "step": 4256 + }, + { + "epoch": 0.8926399664499896, + "grad_norm": 6.254976002686117, + "learning_rate": 1.6475826072593885e-05, + "loss": 1.5773, + "step": 4257 + }, + { + "epoch": 0.8928496540155169, + "grad_norm": 6.057883195296067, + "learning_rate": 1.647410083734287e-05, + "loss": 1.5676, + "step": 4258 + }, + { + "epoch": 0.8930593415810443, + "grad_norm": 6.095931765122487, + "learning_rate": 1.647237527028296e-05, + "loss": 1.5402, + "step": 4259 + }, + { + "epoch": 0.8932690291465716, + "grad_norm": 8.033310954860912, + "learning_rate": 1.6470649371502588e-05, + "loss": 1.8677, + "step": 4260 + }, + { + "epoch": 0.893478716712099, + "grad_norm": 6.165255631743311, + "learning_rate": 1.6468923141090214e-05, + "loss": 1.4886, + "step": 4261 + }, + { + "epoch": 0.8936884042776263, + "grad_norm": 7.286344461469252, + "learning_rate": 1.646719657913431e-05, + "loss": 1.9981, + "step": 4262 + }, + { + "epoch": 0.8938980918431537, + "grad_norm": 7.196530703907088, + "learning_rate": 1.6465469685723365e-05, + "loss": 1.9035, + "step": 4263 + }, + { + "epoch": 0.894107779408681, + "grad_norm": 7.4650745346189895, + "learning_rate": 1.646374246094588e-05, + "loss": 1.9181, + "step": 4264 + }, + { + "epoch": 0.8943174669742084, + "grad_norm": 7.725337489712534, + "learning_rate": 1.6462014904890387e-05, + "loss": 1.8381, + "step": 4265 + }, + { + "epoch": 0.8945271545397357, + "grad_norm": 5.39072190055967, + "learning_rate": 1.6460287017645424e-05, + "loss": 1.4073, + "step": 4266 + }, + { + "epoch": 0.8947368421052632, + "grad_norm": 6.030636488298617, + "learning_rate": 1.6458558799299544e-05, + "loss": 1.7902, + "step": 4267 + }, + { + "epoch": 0.8949465296707905, + "grad_norm": 5.771952500052789, + "learning_rate": 1.6456830249941327e-05, + "loss": 1.5278, + "step": 4268 + }, + { + "epoch": 0.8951562172363179, + "grad_norm": 8.109291201178733, + "learning_rate": 1.6455101369659355e-05, + "loss": 2.0938, + "step": 4269 + }, + { + "epoch": 0.8953659048018453, + "grad_norm": 6.619845440592631, + "learning_rate": 1.6453372158542248e-05, + "loss": 1.7231, + "step": 4270 + }, + { + "epoch": 0.8955755923673726, + "grad_norm": 7.555751792139053, + "learning_rate": 1.645164261667862e-05, + "loss": 1.6357, + "step": 4271 + }, + { + "epoch": 0.8957852799329, + "grad_norm": 6.473899216348755, + "learning_rate": 1.6449912744157122e-05, + "loss": 1.6218, + "step": 4272 + }, + { + "epoch": 0.8959949674984273, + "grad_norm": 8.22273344946025, + "learning_rate": 1.6448182541066405e-05, + "loss": 1.8389, + "step": 4273 + }, + { + "epoch": 0.8962046550639547, + "grad_norm": 6.256457764211807, + "learning_rate": 1.6446452007495155e-05, + "loss": 1.6211, + "step": 4274 + }, + { + "epoch": 0.896414342629482, + "grad_norm": 6.48400698732034, + "learning_rate": 1.6444721143532053e-05, + "loss": 1.8175, + "step": 4275 + }, + { + "epoch": 0.8966240301950095, + "grad_norm": 6.987088345497975, + "learning_rate": 1.6442989949265817e-05, + "loss": 1.919, + "step": 4276 + }, + { + "epoch": 0.8968337177605368, + "grad_norm": 6.462687716003618, + "learning_rate": 1.6441258424785173e-05, + "loss": 1.5565, + "step": 4277 + }, + { + "epoch": 0.8970434053260642, + "grad_norm": 5.79865585686706, + "learning_rate": 1.643952657017886e-05, + "loss": 1.7655, + "step": 4278 + }, + { + "epoch": 0.8972530928915915, + "grad_norm": 6.421857247942963, + "learning_rate": 1.6437794385535642e-05, + "loss": 1.7764, + "step": 4279 + }, + { + "epoch": 0.8974627804571189, + "grad_norm": 7.030308663351112, + "learning_rate": 1.64360618709443e-05, + "loss": 1.8292, + "step": 4280 + }, + { + "epoch": 0.8976724680226462, + "grad_norm": 6.993477496168257, + "learning_rate": 1.6434329026493617e-05, + "loss": 1.9691, + "step": 4281 + }, + { + "epoch": 0.8978821555881736, + "grad_norm": 6.635039965067503, + "learning_rate": 1.643259585227242e-05, + "loss": 1.8039, + "step": 4282 + }, + { + "epoch": 0.8980918431537009, + "grad_norm": 7.292677118914848, + "learning_rate": 1.643086234836952e-05, + "loss": 1.9571, + "step": 4283 + }, + { + "epoch": 0.8983015307192284, + "grad_norm": 7.020340822127908, + "learning_rate": 1.6429128514873776e-05, + "loss": 1.7692, + "step": 4284 + }, + { + "epoch": 0.8985112182847557, + "grad_norm": 6.617333002416124, + "learning_rate": 1.6427394351874046e-05, + "loss": 1.7467, + "step": 4285 + }, + { + "epoch": 0.8987209058502831, + "grad_norm": 6.045944329224551, + "learning_rate": 1.6425659859459204e-05, + "loss": 1.5847, + "step": 4286 + }, + { + "epoch": 0.8989305934158105, + "grad_norm": 6.015326204713119, + "learning_rate": 1.6423925037718152e-05, + "loss": 1.5621, + "step": 4287 + }, + { + "epoch": 0.8991402809813378, + "grad_norm": 6.608819163596873, + "learning_rate": 1.64221898867398e-05, + "loss": 1.5137, + "step": 4288 + }, + { + "epoch": 0.8993499685468652, + "grad_norm": 7.098223131107678, + "learning_rate": 1.6420454406613075e-05, + "loss": 1.7714, + "step": 4289 + }, + { + "epoch": 0.8995596561123925, + "grad_norm": 8.486568921824515, + "learning_rate": 1.641871859742693e-05, + "loss": 2.1459, + "step": 4290 + }, + { + "epoch": 0.8997693436779199, + "grad_norm": 6.34059069652641, + "learning_rate": 1.641698245927032e-05, + "loss": 1.608, + "step": 4291 + }, + { + "epoch": 0.8999790312434472, + "grad_norm": 6.233184821251443, + "learning_rate": 1.6415245992232234e-05, + "loss": 1.4338, + "step": 4292 + }, + { + "epoch": 0.9001887188089747, + "grad_norm": 7.810188675632155, + "learning_rate": 1.641350919640166e-05, + "loss": 1.7228, + "step": 4293 + }, + { + "epoch": 0.900398406374502, + "grad_norm": 7.736446474571807, + "learning_rate": 1.6411772071867616e-05, + "loss": 1.8537, + "step": 4294 + }, + { + "epoch": 0.9006080939400294, + "grad_norm": 5.750950421497219, + "learning_rate": 1.6410034618719136e-05, + "loss": 1.5984, + "step": 4295 + }, + { + "epoch": 0.9008177815055567, + "grad_norm": 9.687445975352974, + "learning_rate": 1.6408296837045262e-05, + "loss": 1.9094, + "step": 4296 + }, + { + "epoch": 0.9010274690710841, + "grad_norm": 7.057526666639502, + "learning_rate": 1.640655872693506e-05, + "loss": 1.6042, + "step": 4297 + }, + { + "epoch": 0.9012371566366114, + "grad_norm": 6.9696296304593, + "learning_rate": 1.6404820288477613e-05, + "loss": 1.4502, + "step": 4298 + }, + { + "epoch": 0.9014468442021388, + "grad_norm": 5.529859940652647, + "learning_rate": 1.6403081521762016e-05, + "loss": 1.9686, + "step": 4299 + }, + { + "epoch": 0.9016565317676661, + "grad_norm": 5.7892938212460825, + "learning_rate": 1.640134242687739e-05, + "loss": 1.5735, + "step": 4300 + }, + { + "epoch": 0.9018662193331936, + "grad_norm": 7.5895051966563525, + "learning_rate": 1.6399603003912855e-05, + "loss": 1.8772, + "step": 4301 + }, + { + "epoch": 0.9020759068987209, + "grad_norm": 7.324913500832093, + "learning_rate": 1.639786325295757e-05, + "loss": 1.8145, + "step": 4302 + }, + { + "epoch": 0.9022855944642483, + "grad_norm": 7.505297131292149, + "learning_rate": 1.63961231741007e-05, + "loss": 1.9549, + "step": 4303 + }, + { + "epoch": 0.9024952820297756, + "grad_norm": 6.298140946984438, + "learning_rate": 1.639438276743142e-05, + "loss": 1.7429, + "step": 4304 + }, + { + "epoch": 0.902704969595303, + "grad_norm": 7.627588090340259, + "learning_rate": 1.639264203303893e-05, + "loss": 2.0589, + "step": 4305 + }, + { + "epoch": 0.9029146571608304, + "grad_norm": 7.115784869860814, + "learning_rate": 1.6390900971012454e-05, + "loss": 1.8725, + "step": 4306 + }, + { + "epoch": 0.9031243447263577, + "grad_norm": 6.528479020808057, + "learning_rate": 1.638915958144122e-05, + "loss": 1.7042, + "step": 4307 + }, + { + "epoch": 0.9033340322918851, + "grad_norm": 6.041012901647171, + "learning_rate": 1.638741786441447e-05, + "loss": 1.6954, + "step": 4308 + }, + { + "epoch": 0.9035437198574124, + "grad_norm": 7.694660109251331, + "learning_rate": 1.638567582002148e-05, + "loss": 1.9204, + "step": 4309 + }, + { + "epoch": 0.9037534074229399, + "grad_norm": 6.694175874116762, + "learning_rate": 1.6383933448351525e-05, + "loss": 2.03, + "step": 4310 + }, + { + "epoch": 0.9039630949884672, + "grad_norm": 6.5073501046553845, + "learning_rate": 1.638219074949391e-05, + "loss": 1.5927, + "step": 4311 + }, + { + "epoch": 0.9041727825539946, + "grad_norm": 6.503760192136076, + "learning_rate": 1.638044772353795e-05, + "loss": 1.9448, + "step": 4312 + }, + { + "epoch": 0.9043824701195219, + "grad_norm": 5.492668909432616, + "learning_rate": 1.637870437057298e-05, + "loss": 1.3937, + "step": 4313 + }, + { + "epoch": 0.9045921576850493, + "grad_norm": 6.99995971419565, + "learning_rate": 1.6376960690688348e-05, + "loss": 1.5515, + "step": 4314 + }, + { + "epoch": 0.9048018452505766, + "grad_norm": 7.000553223129831, + "learning_rate": 1.6375216683973415e-05, + "loss": 1.552, + "step": 4315 + }, + { + "epoch": 0.905011532816104, + "grad_norm": 6.1831041876948145, + "learning_rate": 1.6373472350517573e-05, + "loss": 1.4621, + "step": 4316 + }, + { + "epoch": 0.9052212203816313, + "grad_norm": 7.191624789877629, + "learning_rate": 1.637172769041022e-05, + "loss": 1.596, + "step": 4317 + }, + { + "epoch": 0.9054309079471587, + "grad_norm": 7.218806387837956, + "learning_rate": 1.636998270374077e-05, + "loss": 2.148, + "step": 4318 + }, + { + "epoch": 0.905640595512686, + "grad_norm": 7.706265859111886, + "learning_rate": 1.636823739059866e-05, + "loss": 1.9338, + "step": 4319 + }, + { + "epoch": 0.9058502830782135, + "grad_norm": 5.536449607180294, + "learning_rate": 1.6366491751073335e-05, + "loss": 1.6005, + "step": 4320 + }, + { + "epoch": 0.9060599706437408, + "grad_norm": 7.182639283017236, + "learning_rate": 1.636474578525427e-05, + "loss": 2.0927, + "step": 4321 + }, + { + "epoch": 0.9062696582092682, + "grad_norm": 8.184584904694006, + "learning_rate": 1.636299949323094e-05, + "loss": 1.7981, + "step": 4322 + }, + { + "epoch": 0.9064793457747956, + "grad_norm": 7.799748304093303, + "learning_rate": 1.6361252875092856e-05, + "loss": 1.541, + "step": 4323 + }, + { + "epoch": 0.9066890333403229, + "grad_norm": 5.80287761151934, + "learning_rate": 1.6359505930929523e-05, + "loss": 1.7787, + "step": 4324 + }, + { + "epoch": 0.9068987209058503, + "grad_norm": 7.2185749565819854, + "learning_rate": 1.6357758660830483e-05, + "loss": 2.1301, + "step": 4325 + }, + { + "epoch": 0.9071084084713776, + "grad_norm": 6.546540913985563, + "learning_rate": 1.6356011064885286e-05, + "loss": 1.5384, + "step": 4326 + }, + { + "epoch": 0.907318096036905, + "grad_norm": 7.373222779960865, + "learning_rate": 1.63542631431835e-05, + "loss": 1.6398, + "step": 4327 + }, + { + "epoch": 0.9075277836024324, + "grad_norm": 6.678594007554756, + "learning_rate": 1.63525148958147e-05, + "loss": 1.4834, + "step": 4328 + }, + { + "epoch": 0.9077374711679598, + "grad_norm": 7.168803090071044, + "learning_rate": 1.63507663228685e-05, + "loss": 1.7961, + "step": 4329 + }, + { + "epoch": 0.9079471587334871, + "grad_norm": 7.303216501263846, + "learning_rate": 1.634901742443451e-05, + "loss": 1.7225, + "step": 4330 + }, + { + "epoch": 0.9081568462990145, + "grad_norm": 5.579091969326058, + "learning_rate": 1.6347268200602363e-05, + "loss": 1.7082, + "step": 4331 + }, + { + "epoch": 0.9083665338645418, + "grad_norm": 6.297473043422979, + "learning_rate": 1.6345518651461714e-05, + "loss": 1.7463, + "step": 4332 + }, + { + "epoch": 0.9085762214300692, + "grad_norm": 8.020946092659324, + "learning_rate": 1.6343768777102226e-05, + "loss": 1.6613, + "step": 4333 + }, + { + "epoch": 0.9087859089955965, + "grad_norm": 6.108860659930576, + "learning_rate": 1.6342018577613586e-05, + "loss": 1.4376, + "step": 4334 + }, + { + "epoch": 0.9089955965611239, + "grad_norm": 6.772799753351255, + "learning_rate": 1.63402680530855e-05, + "loss": 1.8513, + "step": 4335 + }, + { + "epoch": 0.9092052841266512, + "grad_norm": 6.692228437107451, + "learning_rate": 1.6338517203607672e-05, + "loss": 1.8041, + "step": 4336 + }, + { + "epoch": 0.9094149716921787, + "grad_norm": 6.304840139369475, + "learning_rate": 1.6336766029269846e-05, + "loss": 1.6344, + "step": 4337 + }, + { + "epoch": 0.909624659257706, + "grad_norm": 6.417303609540669, + "learning_rate": 1.633501453016177e-05, + "loss": 1.7031, + "step": 4338 + }, + { + "epoch": 0.9098343468232334, + "grad_norm": 7.516928062667972, + "learning_rate": 1.6333262706373214e-05, + "loss": 1.8159, + "step": 4339 + }, + { + "epoch": 0.9100440343887607, + "grad_norm": 5.853021315398659, + "learning_rate": 1.6331510557993957e-05, + "loss": 1.7995, + "step": 4340 + }, + { + "epoch": 0.9102537219542881, + "grad_norm": 8.19628553891524, + "learning_rate": 1.6329758085113803e-05, + "loss": 2.2299, + "step": 4341 + }, + { + "epoch": 0.9104634095198155, + "grad_norm": 6.239882178807351, + "learning_rate": 1.6328005287822572e-05, + "loss": 1.3961, + "step": 4342 + }, + { + "epoch": 0.9106730970853428, + "grad_norm": 7.0516206182275765, + "learning_rate": 1.6326252166210092e-05, + "loss": 1.6739, + "step": 4343 + }, + { + "epoch": 0.9108827846508702, + "grad_norm": 8.01990140970183, + "learning_rate": 1.6324498720366218e-05, + "loss": 2.0441, + "step": 4344 + }, + { + "epoch": 0.9110924722163976, + "grad_norm": 6.519041413412794, + "learning_rate": 1.6322744950380812e-05, + "loss": 1.85, + "step": 4345 + }, + { + "epoch": 0.911302159781925, + "grad_norm": 6.459089967252704, + "learning_rate": 1.6320990856343765e-05, + "loss": 1.6333, + "step": 4346 + }, + { + "epoch": 0.9115118473474523, + "grad_norm": 6.314516366446648, + "learning_rate": 1.631923643834497e-05, + "loss": 1.3681, + "step": 4347 + }, + { + "epoch": 0.9117215349129797, + "grad_norm": 7.099301769671213, + "learning_rate": 1.6317481696474347e-05, + "loss": 1.6458, + "step": 4348 + }, + { + "epoch": 0.911931222478507, + "grad_norm": 5.859678060164586, + "learning_rate": 1.6315726630821833e-05, + "loss": 1.532, + "step": 4349 + }, + { + "epoch": 0.9121409100440344, + "grad_norm": 7.486706669146943, + "learning_rate": 1.6313971241477374e-05, + "loss": 1.5815, + "step": 4350 + }, + { + "epoch": 0.9123505976095617, + "grad_norm": 5.507327539317999, + "learning_rate": 1.6312215528530938e-05, + "loss": 1.6873, + "step": 4351 + }, + { + "epoch": 0.9125602851750891, + "grad_norm": 6.212626975976258, + "learning_rate": 1.6310459492072505e-05, + "loss": 1.7321, + "step": 4352 + }, + { + "epoch": 0.9127699727406164, + "grad_norm": 8.281062523950995, + "learning_rate": 1.6308703132192082e-05, + "loss": 2.0378, + "step": 4353 + }, + { + "epoch": 0.9129796603061439, + "grad_norm": 6.381123499078793, + "learning_rate": 1.630694644897968e-05, + "loss": 1.9276, + "step": 4354 + }, + { + "epoch": 0.9131893478716712, + "grad_norm": 6.443945950221432, + "learning_rate": 1.6305189442525334e-05, + "loss": 1.8758, + "step": 4355 + }, + { + "epoch": 0.9133990354371986, + "grad_norm": 6.613472721991775, + "learning_rate": 1.6303432112919095e-05, + "loss": 1.5967, + "step": 4356 + }, + { + "epoch": 0.9136087230027259, + "grad_norm": 5.974271645654126, + "learning_rate": 1.6301674460251024e-05, + "loss": 1.9305, + "step": 4357 + }, + { + "epoch": 0.9138184105682533, + "grad_norm": 6.545512474139671, + "learning_rate": 1.629991648461121e-05, + "loss": 1.5447, + "step": 4358 + }, + { + "epoch": 0.9140280981337807, + "grad_norm": 5.865910773887542, + "learning_rate": 1.6298158186089748e-05, + "loss": 1.7877, + "step": 4359 + }, + { + "epoch": 0.914237785699308, + "grad_norm": 6.768863143133781, + "learning_rate": 1.6296399564776754e-05, + "loss": 1.6611, + "step": 4360 + }, + { + "epoch": 0.9144474732648354, + "grad_norm": 6.216003383808288, + "learning_rate": 1.6294640620762367e-05, + "loss": 1.6584, + "step": 4361 + }, + { + "epoch": 0.9146571608303627, + "grad_norm": 7.547124297060335, + "learning_rate": 1.6292881354136726e-05, + "loss": 2.0497, + "step": 4362 + }, + { + "epoch": 0.9148668483958902, + "grad_norm": 7.147278435862349, + "learning_rate": 1.629112176499e-05, + "loss": 1.5562, + "step": 4363 + }, + { + "epoch": 0.9150765359614175, + "grad_norm": 7.001188116167043, + "learning_rate": 1.6289361853412376e-05, + "loss": 1.7919, + "step": 4364 + }, + { + "epoch": 0.9152862235269449, + "grad_norm": 6.520064265585425, + "learning_rate": 1.628760161949405e-05, + "loss": 1.7465, + "step": 4365 + }, + { + "epoch": 0.9154959110924722, + "grad_norm": 7.615162229771016, + "learning_rate": 1.628584106332523e-05, + "loss": 1.673, + "step": 4366 + }, + { + "epoch": 0.9157055986579996, + "grad_norm": 6.512046843743319, + "learning_rate": 1.6284080184996157e-05, + "loss": 1.576, + "step": 4367 + }, + { + "epoch": 0.9159152862235269, + "grad_norm": 6.495102613086033, + "learning_rate": 1.6282318984597075e-05, + "loss": 1.7918, + "step": 4368 + }, + { + "epoch": 0.9161249737890543, + "grad_norm": 6.357872338896982, + "learning_rate": 1.6280557462218246e-05, + "loss": 1.7552, + "step": 4369 + }, + { + "epoch": 0.9163346613545816, + "grad_norm": 5.835192482795335, + "learning_rate": 1.627879561794996e-05, + "loss": 1.4341, + "step": 4370 + }, + { + "epoch": 0.916544348920109, + "grad_norm": 7.0697480943292215, + "learning_rate": 1.6277033451882503e-05, + "loss": 1.8518, + "step": 4371 + }, + { + "epoch": 0.9167540364856364, + "grad_norm": 5.288844709476438, + "learning_rate": 1.6275270964106197e-05, + "loss": 1.5693, + "step": 4372 + }, + { + "epoch": 0.9169637240511638, + "grad_norm": 6.362519194294375, + "learning_rate": 1.6273508154711372e-05, + "loss": 1.4956, + "step": 4373 + }, + { + "epoch": 0.9171734116166911, + "grad_norm": 5.7498749336798705, + "learning_rate": 1.627174502378837e-05, + "loss": 1.5004, + "step": 4374 + }, + { + "epoch": 0.9173830991822185, + "grad_norm": 5.991968662252214, + "learning_rate": 1.626998157142756e-05, + "loss": 1.6713, + "step": 4375 + }, + { + "epoch": 0.9175927867477458, + "grad_norm": 6.221388357799579, + "learning_rate": 1.6268217797719317e-05, + "loss": 1.7192, + "step": 4376 + }, + { + "epoch": 0.9178024743132732, + "grad_norm": 7.0164166629880915, + "learning_rate": 1.6266453702754043e-05, + "loss": 1.6554, + "step": 4377 + }, + { + "epoch": 0.9180121618788006, + "grad_norm": 5.866641158548073, + "learning_rate": 1.6264689286622143e-05, + "loss": 1.591, + "step": 4378 + }, + { + "epoch": 0.9182218494443279, + "grad_norm": 7.413725333290152, + "learning_rate": 1.626292454941406e-05, + "loss": 1.8048, + "step": 4379 + }, + { + "epoch": 0.9184315370098554, + "grad_norm": 8.298297650330117, + "learning_rate": 1.6261159491220224e-05, + "loss": 1.8587, + "step": 4380 + }, + { + "epoch": 0.9186412245753827, + "grad_norm": 8.295866874468631, + "learning_rate": 1.625939411213111e-05, + "loss": 1.808, + "step": 4381 + }, + { + "epoch": 0.9188509121409101, + "grad_norm": 6.552629265318845, + "learning_rate": 1.6257628412237188e-05, + "loss": 1.326, + "step": 4382 + }, + { + "epoch": 0.9190605997064374, + "grad_norm": 6.2615053461182875, + "learning_rate": 1.6255862391628957e-05, + "loss": 1.6671, + "step": 4383 + }, + { + "epoch": 0.9192702872719648, + "grad_norm": 6.614092219307581, + "learning_rate": 1.625409605039693e-05, + "loss": 1.7288, + "step": 4384 + }, + { + "epoch": 0.9194799748374921, + "grad_norm": 6.105639247716342, + "learning_rate": 1.6252329388631628e-05, + "loss": 1.6215, + "step": 4385 + }, + { + "epoch": 0.9196896624030195, + "grad_norm": 6.2749386543622645, + "learning_rate": 1.6250562406423606e-05, + "loss": 2.1094, + "step": 4386 + }, + { + "epoch": 0.9198993499685468, + "grad_norm": 6.847189163776937, + "learning_rate": 1.6248795103863416e-05, + "loss": 1.601, + "step": 4387 + }, + { + "epoch": 0.9201090375340742, + "grad_norm": 7.33543487838931, + "learning_rate": 1.624702748104164e-05, + "loss": 1.8603, + "step": 4388 + }, + { + "epoch": 0.9203187250996016, + "grad_norm": 6.213328629086073, + "learning_rate": 1.6245259538048873e-05, + "loss": 1.7125, + "step": 4389 + }, + { + "epoch": 0.920528412665129, + "grad_norm": 6.768527269479866, + "learning_rate": 1.624349127497572e-05, + "loss": 1.441, + "step": 4390 + }, + { + "epoch": 0.9207381002306563, + "grad_norm": 6.661431142157733, + "learning_rate": 1.6241722691912816e-05, + "loss": 2.0778, + "step": 4391 + }, + { + "epoch": 0.9209477877961837, + "grad_norm": 7.067486426076362, + "learning_rate": 1.6239953788950792e-05, + "loss": 1.5105, + "step": 4392 + }, + { + "epoch": 0.921157475361711, + "grad_norm": 7.059220657734244, + "learning_rate": 1.6238184566180315e-05, + "loss": 1.9681, + "step": 4393 + }, + { + "epoch": 0.9213671629272384, + "grad_norm": 5.990077654128932, + "learning_rate": 1.623641502369206e-05, + "loss": 1.4067, + "step": 4394 + }, + { + "epoch": 0.9215768504927658, + "grad_norm": 6.802717561780429, + "learning_rate": 1.623464516157672e-05, + "loss": 1.5818, + "step": 4395 + }, + { + "epoch": 0.9217865380582931, + "grad_norm": 5.818947875778048, + "learning_rate": 1.6232874979925e-05, + "loss": 1.3551, + "step": 4396 + }, + { + "epoch": 0.9219962256238206, + "grad_norm": 6.7915483422950835, + "learning_rate": 1.623110447882763e-05, + "loss": 1.6253, + "step": 4397 + }, + { + "epoch": 0.9222059131893479, + "grad_norm": 7.065093576468098, + "learning_rate": 1.6229333658375343e-05, + "loss": 1.8848, + "step": 4398 + }, + { + "epoch": 0.9224156007548753, + "grad_norm": 6.662458923393933, + "learning_rate": 1.6227562518658907e-05, + "loss": 1.7856, + "step": 4399 + }, + { + "epoch": 0.9226252883204026, + "grad_norm": 6.055326744451304, + "learning_rate": 1.6225791059769086e-05, + "loss": 1.3741, + "step": 4400 + }, + { + "epoch": 0.92283497588593, + "grad_norm": 6.428053918225503, + "learning_rate": 1.622401928179668e-05, + "loss": 1.6846, + "step": 4401 + }, + { + "epoch": 0.9230446634514573, + "grad_norm": 7.345024217483868, + "learning_rate": 1.6222247184832486e-05, + "loss": 1.7402, + "step": 4402 + }, + { + "epoch": 0.9232543510169847, + "grad_norm": 5.623346553951054, + "learning_rate": 1.6220474768967335e-05, + "loss": 1.4672, + "step": 4403 + }, + { + "epoch": 0.923464038582512, + "grad_norm": 7.850432426108634, + "learning_rate": 1.6218702034292064e-05, + "loss": 1.5921, + "step": 4404 + }, + { + "epoch": 0.9236737261480394, + "grad_norm": 7.114620783336331, + "learning_rate": 1.6216928980897526e-05, + "loss": 1.5157, + "step": 4405 + }, + { + "epoch": 0.9238834137135667, + "grad_norm": 7.478132435368127, + "learning_rate": 1.6215155608874595e-05, + "loss": 1.9406, + "step": 4406 + }, + { + "epoch": 0.9240931012790942, + "grad_norm": 6.836786602419382, + "learning_rate": 1.621338191831416e-05, + "loss": 2.0092, + "step": 4407 + }, + { + "epoch": 0.9243027888446215, + "grad_norm": 8.042484983343705, + "learning_rate": 1.6211607909307126e-05, + "loss": 2.0921, + "step": 4408 + }, + { + "epoch": 0.9245124764101489, + "grad_norm": 7.372346267507749, + "learning_rate": 1.620983358194441e-05, + "loss": 2.108, + "step": 4409 + }, + { + "epoch": 0.9247221639756762, + "grad_norm": 7.019074967021153, + "learning_rate": 1.6208058936316955e-05, + "loss": 1.8297, + "step": 4410 + }, + { + "epoch": 0.9249318515412036, + "grad_norm": 7.2236405393593435, + "learning_rate": 1.6206283972515713e-05, + "loss": 1.4667, + "step": 4411 + }, + { + "epoch": 0.925141539106731, + "grad_norm": 7.46208428836598, + "learning_rate": 1.620450869063165e-05, + "loss": 1.9439, + "step": 4412 + }, + { + "epoch": 0.9253512266722583, + "grad_norm": 6.024431963693183, + "learning_rate": 1.620273309075576e-05, + "loss": 1.5034, + "step": 4413 + }, + { + "epoch": 0.9255609142377857, + "grad_norm": 6.259155450799531, + "learning_rate": 1.6200957172979038e-05, + "loss": 1.4678, + "step": 4414 + }, + { + "epoch": 0.925770601803313, + "grad_norm": 6.548697144643111, + "learning_rate": 1.6199180937392505e-05, + "loss": 1.6529, + "step": 4415 + }, + { + "epoch": 0.9259802893688405, + "grad_norm": 6.117135598673797, + "learning_rate": 1.6197404384087198e-05, + "loss": 1.4944, + "step": 4416 + }, + { + "epoch": 0.9261899769343678, + "grad_norm": 6.261652688816058, + "learning_rate": 1.6195627513154167e-05, + "loss": 1.6021, + "step": 4417 + }, + { + "epoch": 0.9263996644998952, + "grad_norm": 7.107856936870377, + "learning_rate": 1.619385032468448e-05, + "loss": 1.6406, + "step": 4418 + }, + { + "epoch": 0.9266093520654225, + "grad_norm": 6.8613218708699355, + "learning_rate": 1.6192072818769223e-05, + "loss": 1.7249, + "step": 4419 + }, + { + "epoch": 0.9268190396309499, + "grad_norm": 5.616859411034356, + "learning_rate": 1.6190294995499493e-05, + "loss": 1.4266, + "step": 4420 + }, + { + "epoch": 0.9270287271964772, + "grad_norm": 5.610035598230709, + "learning_rate": 1.6188516854966407e-05, + "loss": 1.6625, + "step": 4421 + }, + { + "epoch": 0.9272384147620046, + "grad_norm": 7.508050911568481, + "learning_rate": 1.61867383972611e-05, + "loss": 2.2655, + "step": 4422 + }, + { + "epoch": 0.9274481023275319, + "grad_norm": 6.174966617556254, + "learning_rate": 1.6184959622474724e-05, + "loss": 1.5804, + "step": 4423 + }, + { + "epoch": 0.9276577898930594, + "grad_norm": 5.832840890122981, + "learning_rate": 1.6183180530698434e-05, + "loss": 1.7264, + "step": 4424 + }, + { + "epoch": 0.9278674774585867, + "grad_norm": 6.691305729856465, + "learning_rate": 1.618140112202342e-05, + "loss": 1.8952, + "step": 4425 + }, + { + "epoch": 0.9280771650241141, + "grad_norm": 6.378138513348659, + "learning_rate": 1.6179621396540874e-05, + "loss": 1.7918, + "step": 4426 + }, + { + "epoch": 0.9282868525896414, + "grad_norm": 6.83764016309168, + "learning_rate": 1.617784135434202e-05, + "loss": 1.8642, + "step": 4427 + }, + { + "epoch": 0.9284965401551688, + "grad_norm": 5.996763995698975, + "learning_rate": 1.6176060995518077e-05, + "loss": 1.5763, + "step": 4428 + }, + { + "epoch": 0.9287062277206961, + "grad_norm": 7.824905011652463, + "learning_rate": 1.6174280320160298e-05, + "loss": 1.6793, + "step": 4429 + }, + { + "epoch": 0.9289159152862235, + "grad_norm": 6.846672910165432, + "learning_rate": 1.617249932835994e-05, + "loss": 1.7084, + "step": 4430 + }, + { + "epoch": 0.9291256028517509, + "grad_norm": 5.353500356099213, + "learning_rate": 1.6170718020208292e-05, + "loss": 1.4591, + "step": 4431 + }, + { + "epoch": 0.9293352904172782, + "grad_norm": 6.606189606109909, + "learning_rate": 1.6168936395796638e-05, + "loss": 1.9792, + "step": 4432 + }, + { + "epoch": 0.9295449779828057, + "grad_norm": 7.740272371235913, + "learning_rate": 1.6167154455216296e-05, + "loss": 2.2183, + "step": 4433 + }, + { + "epoch": 0.929754665548333, + "grad_norm": 6.251280791506547, + "learning_rate": 1.6165372198558593e-05, + "loss": 1.5812, + "step": 4434 + }, + { + "epoch": 0.9299643531138604, + "grad_norm": 6.637952676385502, + "learning_rate": 1.616358962591487e-05, + "loss": 1.7054, + "step": 4435 + }, + { + "epoch": 0.9301740406793877, + "grad_norm": 7.23101764833605, + "learning_rate": 1.616180673737649e-05, + "loss": 2.3731, + "step": 4436 + }, + { + "epoch": 0.9303837282449151, + "grad_norm": 7.718208960777992, + "learning_rate": 1.6160023533034826e-05, + "loss": 1.9321, + "step": 4437 + }, + { + "epoch": 0.9305934158104424, + "grad_norm": 5.9818174332065865, + "learning_rate": 1.615824001298127e-05, + "loss": 1.4085, + "step": 4438 + }, + { + "epoch": 0.9308031033759698, + "grad_norm": 7.112599732605032, + "learning_rate": 1.6156456177307235e-05, + "loss": 1.8498, + "step": 4439 + }, + { + "epoch": 0.9310127909414971, + "grad_norm": 5.769584308947202, + "learning_rate": 1.6154672026104146e-05, + "loss": 1.6647, + "step": 4440 + }, + { + "epoch": 0.9312224785070246, + "grad_norm": 6.726857785313026, + "learning_rate": 1.6152887559463436e-05, + "loss": 1.8672, + "step": 4441 + }, + { + "epoch": 0.9314321660725519, + "grad_norm": 7.507906691618299, + "learning_rate": 1.6151102777476567e-05, + "loss": 2.1798, + "step": 4442 + }, + { + "epoch": 0.9316418536380793, + "grad_norm": 5.2920271644766865, + "learning_rate": 1.6149317680235014e-05, + "loss": 1.5182, + "step": 4443 + }, + { + "epoch": 0.9318515412036066, + "grad_norm": 6.5719985556222875, + "learning_rate": 1.6147532267830264e-05, + "loss": 2.1591, + "step": 4444 + }, + { + "epoch": 0.932061228769134, + "grad_norm": 6.428791949906032, + "learning_rate": 1.6145746540353826e-05, + "loss": 1.572, + "step": 4445 + }, + { + "epoch": 0.9322709163346613, + "grad_norm": 6.1152972462119966, + "learning_rate": 1.6143960497897216e-05, + "loss": 1.697, + "step": 4446 + }, + { + "epoch": 0.9324806039001887, + "grad_norm": 6.8090274655414955, + "learning_rate": 1.6142174140551973e-05, + "loss": 1.7058, + "step": 4447 + }, + { + "epoch": 0.9326902914657161, + "grad_norm": 5.831374745137305, + "learning_rate": 1.614038746840965e-05, + "loss": 1.4237, + "step": 4448 + }, + { + "epoch": 0.9328999790312434, + "grad_norm": 6.09164193779769, + "learning_rate": 1.6138600481561827e-05, + "loss": 1.8915, + "step": 4449 + }, + { + "epoch": 0.9331096665967709, + "grad_norm": 6.228854432734358, + "learning_rate": 1.6136813180100075e-05, + "loss": 1.7031, + "step": 4450 + }, + { + "epoch": 0.9333193541622982, + "grad_norm": 7.497199914688314, + "learning_rate": 1.6135025564116007e-05, + "loss": 1.6823, + "step": 4451 + }, + { + "epoch": 0.9335290417278256, + "grad_norm": 6.314755304800982, + "learning_rate": 1.6133237633701236e-05, + "loss": 2.1506, + "step": 4452 + }, + { + "epoch": 0.9337387292933529, + "grad_norm": 5.962640269927995, + "learning_rate": 1.6131449388947402e-05, + "loss": 1.5752, + "step": 4453 + }, + { + "epoch": 0.9339484168588803, + "grad_norm": 5.797161920662345, + "learning_rate": 1.612966082994615e-05, + "loss": 1.909, + "step": 4454 + }, + { + "epoch": 0.9341581044244076, + "grad_norm": 6.146511821637059, + "learning_rate": 1.6127871956789148e-05, + "loss": 1.7289, + "step": 4455 + }, + { + "epoch": 0.934367791989935, + "grad_norm": 6.877633850633511, + "learning_rate": 1.6126082769568086e-05, + "loss": 1.7565, + "step": 4456 + }, + { + "epoch": 0.9345774795554623, + "grad_norm": 6.9776631666589495, + "learning_rate": 1.6124293268374652e-05, + "loss": 1.789, + "step": 4457 + }, + { + "epoch": 0.9347871671209897, + "grad_norm": 6.131643937481767, + "learning_rate": 1.6122503453300566e-05, + "loss": 1.4719, + "step": 4458 + }, + { + "epoch": 0.934996854686517, + "grad_norm": 7.592613666484798, + "learning_rate": 1.612071332443756e-05, + "loss": 1.5795, + "step": 4459 + }, + { + "epoch": 0.9352065422520445, + "grad_norm": 5.855304175246936, + "learning_rate": 1.6118922881877385e-05, + "loss": 1.5437, + "step": 4460 + }, + { + "epoch": 0.9354162298175718, + "grad_norm": 6.998569760176806, + "learning_rate": 1.611713212571179e-05, + "loss": 1.3634, + "step": 4461 + }, + { + "epoch": 0.9356259173830992, + "grad_norm": 5.763690518963403, + "learning_rate": 1.6115341056032568e-05, + "loss": 1.7428, + "step": 4462 + }, + { + "epoch": 0.9358356049486265, + "grad_norm": 6.618312838524231, + "learning_rate": 1.6113549672931513e-05, + "loss": 1.92, + "step": 4463 + }, + { + "epoch": 0.9360452925141539, + "grad_norm": 8.841933727785282, + "learning_rate": 1.6111757976500432e-05, + "loss": 1.6921, + "step": 4464 + }, + { + "epoch": 0.9362549800796812, + "grad_norm": 6.20136682744596, + "learning_rate": 1.6109965966831156e-05, + "loss": 1.3835, + "step": 4465 + }, + { + "epoch": 0.9364646676452086, + "grad_norm": 6.6192848257958525, + "learning_rate": 1.6108173644015527e-05, + "loss": 1.638, + "step": 4466 + }, + { + "epoch": 0.936674355210736, + "grad_norm": 8.95348991362724, + "learning_rate": 1.61063810081454e-05, + "loss": 1.8111, + "step": 4467 + }, + { + "epoch": 0.9368840427762634, + "grad_norm": 6.547302454340777, + "learning_rate": 1.6104588059312658e-05, + "loss": 1.6214, + "step": 4468 + }, + { + "epoch": 0.9370937303417908, + "grad_norm": 7.014334918725269, + "learning_rate": 1.6102794797609193e-05, + "loss": 1.7615, + "step": 4469 + }, + { + "epoch": 0.9373034179073181, + "grad_norm": 5.708616849453963, + "learning_rate": 1.6101001223126903e-05, + "loss": 1.3088, + "step": 4470 + }, + { + "epoch": 0.9375131054728455, + "grad_norm": 7.246624871091289, + "learning_rate": 1.6099207335957725e-05, + "loss": 1.6269, + "step": 4471 + }, + { + "epoch": 0.9377227930383728, + "grad_norm": 7.429109919948727, + "learning_rate": 1.6097413136193592e-05, + "loss": 1.7847, + "step": 4472 + }, + { + "epoch": 0.9379324806039002, + "grad_norm": 6.939310604116794, + "learning_rate": 1.6095618623926456e-05, + "loss": 1.7187, + "step": 4473 + }, + { + "epoch": 0.9381421681694275, + "grad_norm": 7.465801384192674, + "learning_rate": 1.6093823799248296e-05, + "loss": 1.9004, + "step": 4474 + }, + { + "epoch": 0.9383518557349549, + "grad_norm": 6.514105338582315, + "learning_rate": 1.6092028662251098e-05, + "loss": 1.5694, + "step": 4475 + }, + { + "epoch": 0.9385615433004822, + "grad_norm": 5.677454220762757, + "learning_rate": 1.6090233213026866e-05, + "loss": 1.2352, + "step": 4476 + }, + { + "epoch": 0.9387712308660097, + "grad_norm": 7.951403162372309, + "learning_rate": 1.6088437451667615e-05, + "loss": 1.7742, + "step": 4477 + }, + { + "epoch": 0.938980918431537, + "grad_norm": 6.6573018463867255, + "learning_rate": 1.608664137826539e-05, + "loss": 1.4948, + "step": 4478 + }, + { + "epoch": 0.9391906059970644, + "grad_norm": 6.213036392454229, + "learning_rate": 1.608484499291223e-05, + "loss": 1.6946, + "step": 4479 + }, + { + "epoch": 0.9394002935625917, + "grad_norm": 6.038991206255365, + "learning_rate": 1.6083048295700218e-05, + "loss": 1.6243, + "step": 4480 + }, + { + "epoch": 0.9396099811281191, + "grad_norm": 6.066801593153094, + "learning_rate": 1.6081251286721426e-05, + "loss": 1.4496, + "step": 4481 + }, + { + "epoch": 0.9398196686936464, + "grad_norm": 7.372406252383504, + "learning_rate": 1.6079453966067957e-05, + "loss": 1.6421, + "step": 4482 + }, + { + "epoch": 0.9400293562591738, + "grad_norm": 7.151892012867839, + "learning_rate": 1.6077656333831934e-05, + "loss": 1.6256, + "step": 4483 + }, + { + "epoch": 0.9402390438247012, + "grad_norm": 5.8873624800887265, + "learning_rate": 1.6075858390105477e-05, + "loss": 1.768, + "step": 4484 + }, + { + "epoch": 0.9404487313902286, + "grad_norm": 6.76356366776248, + "learning_rate": 1.607406013498074e-05, + "loss": 1.9942, + "step": 4485 + }, + { + "epoch": 0.940658418955756, + "grad_norm": 6.402948221953007, + "learning_rate": 1.607226156854989e-05, + "loss": 1.5599, + "step": 4486 + }, + { + "epoch": 0.9408681065212833, + "grad_norm": 6.594757726520265, + "learning_rate": 1.6070462690905098e-05, + "loss": 1.4347, + "step": 4487 + }, + { + "epoch": 0.9410777940868107, + "grad_norm": 5.879376147903356, + "learning_rate": 1.606866350213857e-05, + "loss": 1.5229, + "step": 4488 + }, + { + "epoch": 0.941287481652338, + "grad_norm": 6.642760870292059, + "learning_rate": 1.6066864002342505e-05, + "loss": 1.7565, + "step": 4489 + }, + { + "epoch": 0.9414971692178654, + "grad_norm": 6.187776560876245, + "learning_rate": 1.606506419160914e-05, + "loss": 1.7074, + "step": 4490 + }, + { + "epoch": 0.9417068567833927, + "grad_norm": 6.410270377822095, + "learning_rate": 1.606326407003072e-05, + "loss": 1.7316, + "step": 4491 + }, + { + "epoch": 0.9419165443489201, + "grad_norm": 7.3406793803613954, + "learning_rate": 1.6061463637699495e-05, + "loss": 1.6427, + "step": 4492 + }, + { + "epoch": 0.9421262319144474, + "grad_norm": 5.63502249057864, + "learning_rate": 1.605966289470775e-05, + "loss": 1.5482, + "step": 4493 + }, + { + "epoch": 0.9423359194799749, + "grad_norm": 5.895473895118337, + "learning_rate": 1.605786184114777e-05, + "loss": 1.4428, + "step": 4494 + }, + { + "epoch": 0.9425456070455022, + "grad_norm": 6.343393969410663, + "learning_rate": 1.6056060477111864e-05, + "loss": 2.04, + "step": 4495 + }, + { + "epoch": 0.9427552946110296, + "grad_norm": 6.703423611785393, + "learning_rate": 1.6054258802692353e-05, + "loss": 1.6578, + "step": 4496 + }, + { + "epoch": 0.9429649821765569, + "grad_norm": 6.479350483314323, + "learning_rate": 1.605245681798158e-05, + "loss": 1.7604, + "step": 4497 + }, + { + "epoch": 0.9431746697420843, + "grad_norm": 6.485721806327503, + "learning_rate": 1.6050654523071894e-05, + "loss": 1.633, + "step": 4498 + }, + { + "epoch": 0.9433843573076116, + "grad_norm": 6.169951459210537, + "learning_rate": 1.604885191805567e-05, + "loss": 1.7649, + "step": 4499 + }, + { + "epoch": 0.943594044873139, + "grad_norm": 6.464780003616582, + "learning_rate": 1.6047049003025298e-05, + "loss": 1.6471, + "step": 4500 + }, + { + "epoch": 0.9438037324386663, + "grad_norm": 6.193191443638445, + "learning_rate": 1.6045245778073176e-05, + "loss": 1.5919, + "step": 4501 + }, + { + "epoch": 0.9440134200041937, + "grad_norm": 6.986993507371011, + "learning_rate": 1.604344224329172e-05, + "loss": 1.8541, + "step": 4502 + }, + { + "epoch": 0.9442231075697212, + "grad_norm": 6.040887580301437, + "learning_rate": 1.604163839877337e-05, + "loss": 1.8934, + "step": 4503 + }, + { + "epoch": 0.9444327951352485, + "grad_norm": 6.2020766958797395, + "learning_rate": 1.603983424461057e-05, + "loss": 1.5522, + "step": 4504 + }, + { + "epoch": 0.9446424827007759, + "grad_norm": 6.1808520910949065, + "learning_rate": 1.6038029780895792e-05, + "loss": 1.6089, + "step": 4505 + }, + { + "epoch": 0.9448521702663032, + "grad_norm": 7.3752134766282635, + "learning_rate": 1.6036225007721514e-05, + "loss": 2.0526, + "step": 4506 + }, + { + "epoch": 0.9450618578318306, + "grad_norm": 6.958233505604151, + "learning_rate": 1.6034419925180236e-05, + "loss": 1.571, + "step": 4507 + }, + { + "epoch": 0.9452715453973579, + "grad_norm": 6.980913167830229, + "learning_rate": 1.603261453336447e-05, + "loss": 1.6079, + "step": 4508 + }, + { + "epoch": 0.9454812329628853, + "grad_norm": 6.076386193352435, + "learning_rate": 1.6030808832366746e-05, + "loss": 1.7133, + "step": 4509 + }, + { + "epoch": 0.9456909205284126, + "grad_norm": 5.34987463263882, + "learning_rate": 1.6029002822279613e-05, + "loss": 1.6037, + "step": 4510 + }, + { + "epoch": 0.94590060809394, + "grad_norm": 8.577410465804133, + "learning_rate": 1.6027196503195623e-05, + "loss": 2.3296, + "step": 4511 + }, + { + "epoch": 0.9461102956594674, + "grad_norm": 6.872323653709511, + "learning_rate": 1.602538987520736e-05, + "loss": 1.7701, + "step": 4512 + }, + { + "epoch": 0.9463199832249948, + "grad_norm": 6.325983267266496, + "learning_rate": 1.6023582938407417e-05, + "loss": 1.6765, + "step": 4513 + }, + { + "epoch": 0.9465296707905221, + "grad_norm": 6.1655850707221935, + "learning_rate": 1.60217756928884e-05, + "loss": 1.9077, + "step": 4514 + }, + { + "epoch": 0.9467393583560495, + "grad_norm": 6.096162362692154, + "learning_rate": 1.6019968138742938e-05, + "loss": 1.697, + "step": 4515 + }, + { + "epoch": 0.9469490459215768, + "grad_norm": 7.415107139185473, + "learning_rate": 1.6018160276063663e-05, + "loss": 2.1718, + "step": 4516 + }, + { + "epoch": 0.9471587334871042, + "grad_norm": 6.1149474435779485, + "learning_rate": 1.601635210494324e-05, + "loss": 1.384, + "step": 4517 + }, + { + "epoch": 0.9473684210526315, + "grad_norm": 6.396663552752187, + "learning_rate": 1.6014543625474334e-05, + "loss": 1.695, + "step": 4518 + }, + { + "epoch": 0.9475781086181589, + "grad_norm": 5.784633772417223, + "learning_rate": 1.6012734837749632e-05, + "loss": 1.8603, + "step": 4519 + }, + { + "epoch": 0.9477877961836864, + "grad_norm": 6.499001302760238, + "learning_rate": 1.6010925741861848e-05, + "loss": 1.8, + "step": 4520 + }, + { + "epoch": 0.9479974837492137, + "grad_norm": 5.5183814267454885, + "learning_rate": 1.6009116337903692e-05, + "loss": 1.8045, + "step": 4521 + }, + { + "epoch": 0.9482071713147411, + "grad_norm": 5.994249173378188, + "learning_rate": 1.6007306625967897e-05, + "loss": 1.6435, + "step": 4522 + }, + { + "epoch": 0.9484168588802684, + "grad_norm": 5.995555086356783, + "learning_rate": 1.6005496606147223e-05, + "loss": 1.8597, + "step": 4523 + }, + { + "epoch": 0.9486265464457958, + "grad_norm": 6.152916742314905, + "learning_rate": 1.600368627853443e-05, + "loss": 1.8165, + "step": 4524 + }, + { + "epoch": 0.9488362340113231, + "grad_norm": 5.958726669185426, + "learning_rate": 1.6001875643222304e-05, + "loss": 1.9461, + "step": 4525 + }, + { + "epoch": 0.9490459215768505, + "grad_norm": 6.791255862748508, + "learning_rate": 1.600006470030364e-05, + "loss": 1.8562, + "step": 4526 + }, + { + "epoch": 0.9492556091423778, + "grad_norm": 5.988010677214452, + "learning_rate": 1.5998253449871254e-05, + "loss": 1.7902, + "step": 4527 + }, + { + "epoch": 0.9494652967079052, + "grad_norm": 6.099693587861169, + "learning_rate": 1.5996441892017976e-05, + "loss": 1.595, + "step": 4528 + }, + { + "epoch": 0.9496749842734326, + "grad_norm": 5.631905644757514, + "learning_rate": 1.599463002683665e-05, + "loss": 1.8557, + "step": 4529 + }, + { + "epoch": 0.94988467183896, + "grad_norm": 6.205132243304965, + "learning_rate": 1.599281785442014e-05, + "loss": 1.7436, + "step": 4530 + }, + { + "epoch": 0.9500943594044873, + "grad_norm": 6.708527451715467, + "learning_rate": 1.5991005374861316e-05, + "loss": 1.8037, + "step": 4531 + }, + { + "epoch": 0.9503040469700147, + "grad_norm": 6.623699121242019, + "learning_rate": 1.598919258825308e-05, + "loss": 2.0272, + "step": 4532 + }, + { + "epoch": 0.950513734535542, + "grad_norm": 5.851493749418373, + "learning_rate": 1.5987379494688337e-05, + "loss": 1.5981, + "step": 4533 + }, + { + "epoch": 0.9507234221010694, + "grad_norm": 6.873572559800961, + "learning_rate": 1.5985566094260007e-05, + "loss": 1.85, + "step": 4534 + }, + { + "epoch": 0.9509331096665967, + "grad_norm": 5.6998256818711965, + "learning_rate": 1.5983752387061032e-05, + "loss": 1.225, + "step": 4535 + }, + { + "epoch": 0.9511427972321241, + "grad_norm": 6.969952717142911, + "learning_rate": 1.5981938373184377e-05, + "loss": 1.7099, + "step": 4536 + }, + { + "epoch": 0.9513524847976516, + "grad_norm": 5.852717917542091, + "learning_rate": 1.5980124052722995e-05, + "loss": 1.6431, + "step": 4537 + }, + { + "epoch": 0.9515621723631789, + "grad_norm": 6.461349536437548, + "learning_rate": 1.597830942576989e-05, + "loss": 1.7041, + "step": 4538 + }, + { + "epoch": 0.9517718599287063, + "grad_norm": 6.691571438771863, + "learning_rate": 1.597649449241806e-05, + "loss": 1.9284, + "step": 4539 + }, + { + "epoch": 0.9519815474942336, + "grad_norm": 6.308227383453972, + "learning_rate": 1.597467925276051e-05, + "loss": 1.7432, + "step": 4540 + }, + { + "epoch": 0.952191235059761, + "grad_norm": 5.88915827088278, + "learning_rate": 1.5972863706890295e-05, + "loss": 1.2667, + "step": 4541 + }, + { + "epoch": 0.9524009226252883, + "grad_norm": 6.982953354675416, + "learning_rate": 1.5971047854900456e-05, + "loss": 1.6122, + "step": 4542 + }, + { + "epoch": 0.9526106101908157, + "grad_norm": 5.752859914088138, + "learning_rate": 1.5969231696884056e-05, + "loss": 1.4876, + "step": 4543 + }, + { + "epoch": 0.952820297756343, + "grad_norm": 6.107082342256429, + "learning_rate": 1.596741523293418e-05, + "loss": 1.4963, + "step": 4544 + }, + { + "epoch": 0.9530299853218704, + "grad_norm": 7.116168528640246, + "learning_rate": 1.596559846314392e-05, + "loss": 1.9147, + "step": 4545 + }, + { + "epoch": 0.9532396728873977, + "grad_norm": 7.594009396455394, + "learning_rate": 1.5963781387606392e-05, + "loss": 1.8329, + "step": 4546 + }, + { + "epoch": 0.9534493604529252, + "grad_norm": 6.3269892407735036, + "learning_rate": 1.596196400641473e-05, + "loss": 1.8286, + "step": 4547 + }, + { + "epoch": 0.9536590480184525, + "grad_norm": 6.394814109983584, + "learning_rate": 1.5960146319662065e-05, + "loss": 1.679, + "step": 4548 + }, + { + "epoch": 0.9538687355839799, + "grad_norm": 6.129156696230484, + "learning_rate": 1.5958328327441568e-05, + "loss": 1.818, + "step": 4549 + }, + { + "epoch": 0.9540784231495072, + "grad_norm": 5.626376146621979, + "learning_rate": 1.5956510029846408e-05, + "loss": 1.7069, + "step": 4550 + }, + { + "epoch": 0.9542881107150346, + "grad_norm": 6.8157582578920355, + "learning_rate": 1.595469142696978e-05, + "loss": 1.8133, + "step": 4551 + }, + { + "epoch": 0.9544977982805619, + "grad_norm": 6.412778511828664, + "learning_rate": 1.5952872518904887e-05, + "loss": 1.7625, + "step": 4552 + }, + { + "epoch": 0.9547074858460893, + "grad_norm": 6.260219544721869, + "learning_rate": 1.5951053305744954e-05, + "loss": 1.7489, + "step": 4553 + }, + { + "epoch": 0.9549171734116166, + "grad_norm": 6.6505690868592495, + "learning_rate": 1.5949233787583215e-05, + "loss": 2.095, + "step": 4554 + }, + { + "epoch": 0.955126860977144, + "grad_norm": 5.273690642091402, + "learning_rate": 1.594741396451293e-05, + "loss": 1.3721, + "step": 4555 + }, + { + "epoch": 0.9553365485426715, + "grad_norm": 5.504043299719686, + "learning_rate": 1.594559383662736e-05, + "loss": 1.2659, + "step": 4556 + }, + { + "epoch": 0.9555462361081988, + "grad_norm": 6.947730056911571, + "learning_rate": 1.5943773404019792e-05, + "loss": 1.7903, + "step": 4557 + }, + { + "epoch": 0.9557559236737262, + "grad_norm": 6.203974909715322, + "learning_rate": 1.5941952666783533e-05, + "loss": 1.472, + "step": 4558 + }, + { + "epoch": 0.9559656112392535, + "grad_norm": 7.61949736418357, + "learning_rate": 1.5940131625011895e-05, + "loss": 1.9565, + "step": 4559 + }, + { + "epoch": 0.9561752988047809, + "grad_norm": 5.814309019725462, + "learning_rate": 1.5938310278798205e-05, + "loss": 1.6565, + "step": 4560 + }, + { + "epoch": 0.9563849863703082, + "grad_norm": 5.757177845230832, + "learning_rate": 1.5936488628235815e-05, + "loss": 1.7122, + "step": 4561 + }, + { + "epoch": 0.9565946739358356, + "grad_norm": 5.422158293647324, + "learning_rate": 1.5934666673418087e-05, + "loss": 1.521, + "step": 4562 + }, + { + "epoch": 0.9568043615013629, + "grad_norm": 6.968081015071573, + "learning_rate": 1.5932844414438396e-05, + "loss": 1.536, + "step": 4563 + }, + { + "epoch": 0.9570140490668904, + "grad_norm": 7.639328884700881, + "learning_rate": 1.593102185139014e-05, + "loss": 1.829, + "step": 4564 + }, + { + "epoch": 0.9572237366324177, + "grad_norm": 7.6757736312789575, + "learning_rate": 1.592919898436673e-05, + "loss": 1.7969, + "step": 4565 + }, + { + "epoch": 0.9574334241979451, + "grad_norm": 7.060474843598261, + "learning_rate": 1.5927375813461587e-05, + "loss": 1.832, + "step": 4566 + }, + { + "epoch": 0.9576431117634724, + "grad_norm": 7.537485009525664, + "learning_rate": 1.5925552338768155e-05, + "loss": 1.7992, + "step": 4567 + }, + { + "epoch": 0.9578527993289998, + "grad_norm": 6.252338307285478, + "learning_rate": 1.5923728560379882e-05, + "loss": 1.4291, + "step": 4568 + }, + { + "epoch": 0.9580624868945271, + "grad_norm": 6.704310944156857, + "learning_rate": 1.592190447839025e-05, + "loss": 1.788, + "step": 4569 + }, + { + "epoch": 0.9582721744600545, + "grad_norm": 5.620503607854182, + "learning_rate": 1.5920080092892744e-05, + "loss": 1.127, + "step": 4570 + }, + { + "epoch": 0.9584818620255818, + "grad_norm": 6.257402605301782, + "learning_rate": 1.5918255403980865e-05, + "loss": 1.6515, + "step": 4571 + }, + { + "epoch": 0.9586915495911092, + "grad_norm": 6.3976192662948055, + "learning_rate": 1.591643041174813e-05, + "loss": 1.7114, + "step": 4572 + }, + { + "epoch": 0.9589012371566367, + "grad_norm": 6.032598021799135, + "learning_rate": 1.5914605116288075e-05, + "loss": 1.4725, + "step": 4573 + }, + { + "epoch": 0.959110924722164, + "grad_norm": 7.855252083920772, + "learning_rate": 1.591277951769425e-05, + "loss": 1.9874, + "step": 4574 + }, + { + "epoch": 0.9593206122876914, + "grad_norm": 5.998395434232328, + "learning_rate": 1.5910953616060217e-05, + "loss": 1.7291, + "step": 4575 + }, + { + "epoch": 0.9595302998532187, + "grad_norm": 5.9393630955492105, + "learning_rate": 1.5909127411479563e-05, + "loss": 1.7275, + "step": 4576 + }, + { + "epoch": 0.9597399874187461, + "grad_norm": 6.987690709005374, + "learning_rate": 1.5907300904045877e-05, + "loss": 1.4758, + "step": 4577 + }, + { + "epoch": 0.9599496749842734, + "grad_norm": 8.41346772379831, + "learning_rate": 1.5905474093852776e-05, + "loss": 1.7811, + "step": 4578 + }, + { + "epoch": 0.9601593625498008, + "grad_norm": 7.363456930144115, + "learning_rate": 1.590364698099389e-05, + "loss": 1.5923, + "step": 4579 + }, + { + "epoch": 0.9603690501153281, + "grad_norm": 7.055045643140238, + "learning_rate": 1.5901819565562848e-05, + "loss": 2.1664, + "step": 4580 + }, + { + "epoch": 0.9605787376808556, + "grad_norm": 6.356748596114354, + "learning_rate": 1.5899991847653323e-05, + "loss": 1.4083, + "step": 4581 + }, + { + "epoch": 0.9607884252463829, + "grad_norm": 6.319358262812723, + "learning_rate": 1.5898163827358983e-05, + "loss": 1.589, + "step": 4582 + }, + { + "epoch": 0.9609981128119103, + "grad_norm": 6.275077172947977, + "learning_rate": 1.589633550477351e-05, + "loss": 1.629, + "step": 4583 + }, + { + "epoch": 0.9612078003774376, + "grad_norm": 5.551624146851175, + "learning_rate": 1.589450687999062e-05, + "loss": 1.512, + "step": 4584 + }, + { + "epoch": 0.961417487942965, + "grad_norm": 5.818591173106327, + "learning_rate": 1.5892677953104032e-05, + "loss": 1.6917, + "step": 4585 + }, + { + "epoch": 0.9616271755084923, + "grad_norm": 6.530800859708341, + "learning_rate": 1.5890848724207476e-05, + "loss": 1.3499, + "step": 4586 + }, + { + "epoch": 0.9618368630740197, + "grad_norm": 6.317686780656677, + "learning_rate": 1.5889019193394703e-05, + "loss": 1.5654, + "step": 4587 + }, + { + "epoch": 0.962046550639547, + "grad_norm": 7.2755995189196625, + "learning_rate": 1.5887189360759487e-05, + "loss": 1.7316, + "step": 4588 + }, + { + "epoch": 0.9622562382050744, + "grad_norm": 6.768192470348803, + "learning_rate": 1.5885359226395603e-05, + "loss": 1.7223, + "step": 4589 + }, + { + "epoch": 0.9624659257706017, + "grad_norm": 7.503512442048614, + "learning_rate": 1.5883528790396853e-05, + "loss": 1.7287, + "step": 4590 + }, + { + "epoch": 0.9626756133361292, + "grad_norm": 6.511368591079411, + "learning_rate": 1.5881698052857047e-05, + "loss": 1.3884, + "step": 4591 + }, + { + "epoch": 0.9628853009016566, + "grad_norm": 7.347753474197154, + "learning_rate": 1.5879867013870012e-05, + "loss": 1.7772, + "step": 4592 + }, + { + "epoch": 0.9630949884671839, + "grad_norm": 6.191386099009408, + "learning_rate": 1.5878035673529597e-05, + "loss": 1.6401, + "step": 4593 + }, + { + "epoch": 0.9633046760327113, + "grad_norm": 7.46404297369746, + "learning_rate": 1.587620403192966e-05, + "loss": 1.937, + "step": 4594 + }, + { + "epoch": 0.9635143635982386, + "grad_norm": 5.94210095297413, + "learning_rate": 1.587437208916407e-05, + "loss": 1.3525, + "step": 4595 + }, + { + "epoch": 0.963724051163766, + "grad_norm": 6.098101893787626, + "learning_rate": 1.5872539845326726e-05, + "loss": 1.8992, + "step": 4596 + }, + { + "epoch": 0.9639337387292933, + "grad_norm": 5.886779164198539, + "learning_rate": 1.5870707300511528e-05, + "loss": 1.5636, + "step": 4597 + }, + { + "epoch": 0.9641434262948207, + "grad_norm": 7.063862744816543, + "learning_rate": 1.5868874454812398e-05, + "loss": 1.6299, + "step": 4598 + }, + { + "epoch": 0.964353113860348, + "grad_norm": 6.598440461335364, + "learning_rate": 1.5867041308323274e-05, + "loss": 1.718, + "step": 4599 + }, + { + "epoch": 0.9645628014258755, + "grad_norm": 7.584531756935614, + "learning_rate": 1.5865207861138107e-05, + "loss": 1.8372, + "step": 4600 + }, + { + "epoch": 0.9647724889914028, + "grad_norm": 6.928126624095986, + "learning_rate": 1.5863374113350864e-05, + "loss": 1.4052, + "step": 4601 + }, + { + "epoch": 0.9649821765569302, + "grad_norm": 6.863964059057822, + "learning_rate": 1.5861540065055532e-05, + "loss": 1.827, + "step": 4602 + }, + { + "epoch": 0.9651918641224575, + "grad_norm": 7.002270321597739, + "learning_rate": 1.58597057163461e-05, + "loss": 1.6869, + "step": 4603 + }, + { + "epoch": 0.9654015516879849, + "grad_norm": 6.43765929042971, + "learning_rate": 1.5857871067316592e-05, + "loss": 1.2841, + "step": 4604 + }, + { + "epoch": 0.9656112392535122, + "grad_norm": 7.623892005949478, + "learning_rate": 1.5856036118061028e-05, + "loss": 1.8407, + "step": 4605 + }, + { + "epoch": 0.9658209268190396, + "grad_norm": 6.474918336242843, + "learning_rate": 1.5854200868673455e-05, + "loss": 1.8253, + "step": 4606 + }, + { + "epoch": 0.9660306143845669, + "grad_norm": 6.463211750060084, + "learning_rate": 1.5852365319247936e-05, + "loss": 1.8542, + "step": 4607 + }, + { + "epoch": 0.9662403019500944, + "grad_norm": 6.25127139064424, + "learning_rate": 1.5850529469878544e-05, + "loss": 1.7682, + "step": 4608 + }, + { + "epoch": 0.9664499895156218, + "grad_norm": 8.370090315226767, + "learning_rate": 1.584869332065937e-05, + "loss": 1.5027, + "step": 4609 + }, + { + "epoch": 0.9666596770811491, + "grad_norm": 6.883920768953059, + "learning_rate": 1.584685687168452e-05, + "loss": 1.8097, + "step": 4610 + }, + { + "epoch": 0.9668693646466765, + "grad_norm": 7.066718086976704, + "learning_rate": 1.584502012304811e-05, + "loss": 1.2618, + "step": 4611 + }, + { + "epoch": 0.9670790522122038, + "grad_norm": 7.704975294289999, + "learning_rate": 1.5843183074844283e-05, + "loss": 1.624, + "step": 4612 + }, + { + "epoch": 0.9672887397777312, + "grad_norm": 6.712239176117779, + "learning_rate": 1.584134572716719e-05, + "loss": 1.8167, + "step": 4613 + }, + { + "epoch": 0.9674984273432585, + "grad_norm": 7.734517705930198, + "learning_rate": 1.5839508080110995e-05, + "loss": 1.4983, + "step": 4614 + }, + { + "epoch": 0.9677081149087859, + "grad_norm": 7.033353227657182, + "learning_rate": 1.5837670133769888e-05, + "loss": 1.8377, + "step": 4615 + }, + { + "epoch": 0.9679178024743132, + "grad_norm": 7.833608984245218, + "learning_rate": 1.5835831888238057e-05, + "loss": 1.7102, + "step": 4616 + }, + { + "epoch": 0.9681274900398407, + "grad_norm": 7.079612115190486, + "learning_rate": 1.5833993343609725e-05, + "loss": 1.8071, + "step": 4617 + }, + { + "epoch": 0.968337177605368, + "grad_norm": 8.97197578130918, + "learning_rate": 1.583215449997911e-05, + "loss": 1.9485, + "step": 4618 + }, + { + "epoch": 0.9685468651708954, + "grad_norm": 6.643944754498032, + "learning_rate": 1.5830315357440464e-05, + "loss": 1.165, + "step": 4619 + }, + { + "epoch": 0.9687565527364227, + "grad_norm": 7.5221030348114395, + "learning_rate": 1.5828475916088044e-05, + "loss": 1.7425, + "step": 4620 + }, + { + "epoch": 0.9689662403019501, + "grad_norm": 6.110224733321276, + "learning_rate": 1.5826636176016123e-05, + "loss": 1.5333, + "step": 4621 + }, + { + "epoch": 0.9691759278674774, + "grad_norm": 7.336909225708022, + "learning_rate": 1.5824796137318993e-05, + "loss": 1.8197, + "step": 4622 + }, + { + "epoch": 0.9693856154330048, + "grad_norm": 6.596170542005773, + "learning_rate": 1.582295580009096e-05, + "loss": 1.5548, + "step": 4623 + }, + { + "epoch": 0.9695953029985321, + "grad_norm": 6.548505595951554, + "learning_rate": 1.582111516442634e-05, + "loss": 1.5336, + "step": 4624 + }, + { + "epoch": 0.9698049905640596, + "grad_norm": 5.972803844713978, + "learning_rate": 1.5819274230419472e-05, + "loss": 1.2644, + "step": 4625 + }, + { + "epoch": 0.9700146781295869, + "grad_norm": 5.857470228695838, + "learning_rate": 1.581743299816471e-05, + "loss": 1.2006, + "step": 4626 + }, + { + "epoch": 0.9702243656951143, + "grad_norm": 6.334808805287599, + "learning_rate": 1.5815591467756412e-05, + "loss": 1.5085, + "step": 4627 + }, + { + "epoch": 0.9704340532606417, + "grad_norm": 6.813430107317001, + "learning_rate": 1.5813749639288967e-05, + "loss": 1.4082, + "step": 4628 + }, + { + "epoch": 0.970643740826169, + "grad_norm": 7.846657462999528, + "learning_rate": 1.581190751285677e-05, + "loss": 1.9396, + "step": 4629 + }, + { + "epoch": 0.9708534283916964, + "grad_norm": 6.147825416776932, + "learning_rate": 1.5810065088554232e-05, + "loss": 1.464, + "step": 4630 + }, + { + "epoch": 0.9710631159572237, + "grad_norm": 5.531721340604371, + "learning_rate": 1.5808222366475783e-05, + "loss": 1.5946, + "step": 4631 + }, + { + "epoch": 0.9712728035227511, + "grad_norm": 8.293966409640001, + "learning_rate": 1.580637934671586e-05, + "loss": 1.8766, + "step": 4632 + }, + { + "epoch": 0.9714824910882784, + "grad_norm": 7.463534787353004, + "learning_rate": 1.580453602936893e-05, + "loss": 1.3587, + "step": 4633 + }, + { + "epoch": 0.9716921786538059, + "grad_norm": 7.020277913397013, + "learning_rate": 1.5802692414529456e-05, + "loss": 2.1281, + "step": 4634 + }, + { + "epoch": 0.9719018662193332, + "grad_norm": 5.7496402752023466, + "learning_rate": 1.5800848502291935e-05, + "loss": 1.281, + "step": 4635 + }, + { + "epoch": 0.9721115537848606, + "grad_norm": 6.473100760349967, + "learning_rate": 1.5799004292750866e-05, + "loss": 1.6311, + "step": 4636 + }, + { + "epoch": 0.9723212413503879, + "grad_norm": 8.12287168595937, + "learning_rate": 1.5797159786000767e-05, + "loss": 1.7443, + "step": 4637 + }, + { + "epoch": 0.9725309289159153, + "grad_norm": 7.530576394355859, + "learning_rate": 1.579531498213618e-05, + "loss": 1.851, + "step": 4638 + }, + { + "epoch": 0.9727406164814426, + "grad_norm": 7.158474886576736, + "learning_rate": 1.579346988125164e-05, + "loss": 1.657, + "step": 4639 + }, + { + "epoch": 0.97295030404697, + "grad_norm": 6.788429752325326, + "learning_rate": 1.579162448344173e-05, + "loss": 1.5973, + "step": 4640 + }, + { + "epoch": 0.9731599916124973, + "grad_norm": 8.254831359054881, + "learning_rate": 1.5789778788801013e-05, + "loss": 1.7952, + "step": 4641 + }, + { + "epoch": 0.9733696791780247, + "grad_norm": 5.829114624100832, + "learning_rate": 1.5787932797424094e-05, + "loss": 1.5307, + "step": 4642 + }, + { + "epoch": 0.973579366743552, + "grad_norm": 8.938174126033726, + "learning_rate": 1.5786086509405584e-05, + "loss": 1.4481, + "step": 4643 + }, + { + "epoch": 0.9737890543090795, + "grad_norm": 6.163980375886287, + "learning_rate": 1.5784239924840097e-05, + "loss": 1.347, + "step": 4644 + }, + { + "epoch": 0.9739987418746069, + "grad_norm": 7.584073575320734, + "learning_rate": 1.5782393043822287e-05, + "loss": 1.5172, + "step": 4645 + }, + { + "epoch": 0.9742084294401342, + "grad_norm": 5.582625614393856, + "learning_rate": 1.5780545866446802e-05, + "loss": 1.6059, + "step": 4646 + }, + { + "epoch": 0.9744181170056616, + "grad_norm": 7.523203143265562, + "learning_rate": 1.5778698392808317e-05, + "loss": 1.8865, + "step": 4647 + }, + { + "epoch": 0.9746278045711889, + "grad_norm": 6.36718586983104, + "learning_rate": 1.577685062300152e-05, + "loss": 1.6981, + "step": 4648 + }, + { + "epoch": 0.9748374921367163, + "grad_norm": 5.694735695367508, + "learning_rate": 1.5775002557121104e-05, + "loss": 1.4363, + "step": 4649 + }, + { + "epoch": 0.9750471797022436, + "grad_norm": 6.021705808041867, + "learning_rate": 1.5773154195261793e-05, + "loss": 1.5108, + "step": 4650 + }, + { + "epoch": 0.975256867267771, + "grad_norm": 7.272458829094439, + "learning_rate": 1.5771305537518316e-05, + "loss": 1.7476, + "step": 4651 + }, + { + "epoch": 0.9754665548332984, + "grad_norm": 7.365379028599278, + "learning_rate": 1.576945658398542e-05, + "loss": 2.0312, + "step": 4652 + }, + { + "epoch": 0.9756762423988258, + "grad_norm": 5.925481968503258, + "learning_rate": 1.5767607334757867e-05, + "loss": 1.2852, + "step": 4653 + }, + { + "epoch": 0.9758859299643531, + "grad_norm": 6.671681388741526, + "learning_rate": 1.5765757789930434e-05, + "loss": 1.7179, + "step": 4654 + }, + { + "epoch": 0.9760956175298805, + "grad_norm": 6.304346432997182, + "learning_rate": 1.576390794959792e-05, + "loss": 1.9099, + "step": 4655 + }, + { + "epoch": 0.9763053050954078, + "grad_norm": 6.321280161926331, + "learning_rate": 1.576205781385512e-05, + "loss": 1.6482, + "step": 4656 + }, + { + "epoch": 0.9765149926609352, + "grad_norm": 6.978277251728305, + "learning_rate": 1.5760207382796867e-05, + "loss": 1.3028, + "step": 4657 + }, + { + "epoch": 0.9767246802264625, + "grad_norm": 7.211551033650958, + "learning_rate": 1.5758356656517993e-05, + "loss": 1.456, + "step": 4658 + }, + { + "epoch": 0.9769343677919899, + "grad_norm": 7.4160608843310705, + "learning_rate": 1.575650563511335e-05, + "loss": 1.7712, + "step": 4659 + }, + { + "epoch": 0.9771440553575172, + "grad_norm": 6.518884928478189, + "learning_rate": 1.5754654318677814e-05, + "loss": 1.8428, + "step": 4660 + }, + { + "epoch": 0.9773537429230447, + "grad_norm": 7.707212910732268, + "learning_rate": 1.575280270730626e-05, + "loss": 1.7144, + "step": 4661 + }, + { + "epoch": 0.9775634304885721, + "grad_norm": 6.501589529580172, + "learning_rate": 1.5750950801093592e-05, + "loss": 1.6494, + "step": 4662 + }, + { + "epoch": 0.9777731180540994, + "grad_norm": 7.6504116768994095, + "learning_rate": 1.574909860013472e-05, + "loss": 1.9984, + "step": 4663 + }, + { + "epoch": 0.9779828056196268, + "grad_norm": 6.1782548361359035, + "learning_rate": 1.5747246104524574e-05, + "loss": 1.957, + "step": 4664 + }, + { + "epoch": 0.9781924931851541, + "grad_norm": 6.233855322639774, + "learning_rate": 1.57453933143581e-05, + "loss": 1.3755, + "step": 4665 + }, + { + "epoch": 0.9784021807506815, + "grad_norm": 7.4175380207094355, + "learning_rate": 1.5743540229730245e-05, + "loss": 2.101, + "step": 4666 + }, + { + "epoch": 0.9786118683162088, + "grad_norm": 7.017750218503154, + "learning_rate": 1.5741686850735996e-05, + "loss": 1.6182, + "step": 4667 + }, + { + "epoch": 0.9788215558817362, + "grad_norm": 6.109148405615685, + "learning_rate": 1.5739833177470338e-05, + "loss": 1.7079, + "step": 4668 + }, + { + "epoch": 0.9790312434472636, + "grad_norm": 6.773429746029386, + "learning_rate": 1.5737979210028277e-05, + "loss": 1.8969, + "step": 4669 + }, + { + "epoch": 0.979240931012791, + "grad_norm": 6.06593542688359, + "learning_rate": 1.5736124948504827e-05, + "loss": 1.6872, + "step": 4670 + }, + { + "epoch": 0.9794506185783183, + "grad_norm": 5.456591584844584, + "learning_rate": 1.5734270392995024e-05, + "loss": 1.2133, + "step": 4671 + }, + { + "epoch": 0.9796603061438457, + "grad_norm": 6.542448073494205, + "learning_rate": 1.573241554359392e-05, + "loss": 1.748, + "step": 4672 + }, + { + "epoch": 0.979869993709373, + "grad_norm": 5.458869267173071, + "learning_rate": 1.5730560400396572e-05, + "loss": 1.628, + "step": 4673 + }, + { + "epoch": 0.9800796812749004, + "grad_norm": 5.624290066860508, + "learning_rate": 1.572870496349807e-05, + "loss": 1.6756, + "step": 4674 + }, + { + "epoch": 0.9802893688404277, + "grad_norm": 6.646962163372272, + "learning_rate": 1.5726849232993503e-05, + "loss": 2.0415, + "step": 4675 + }, + { + "epoch": 0.9804990564059551, + "grad_norm": 6.278787587664155, + "learning_rate": 1.5724993208977973e-05, + "loss": 1.8056, + "step": 4676 + }, + { + "epoch": 0.9807087439714824, + "grad_norm": 5.599934607581942, + "learning_rate": 1.5723136891546616e-05, + "loss": 1.311, + "step": 4677 + }, + { + "epoch": 0.9809184315370099, + "grad_norm": 6.713989285555248, + "learning_rate": 1.572128028079457e-05, + "loss": 1.9383, + "step": 4678 + }, + { + "epoch": 0.9811281191025372, + "grad_norm": 6.585273285797768, + "learning_rate": 1.5719423376816983e-05, + "loss": 1.7311, + "step": 4679 + }, + { + "epoch": 0.9813378066680646, + "grad_norm": 6.981889839176656, + "learning_rate": 1.5717566179709028e-05, + "loss": 1.3141, + "step": 4680 + }, + { + "epoch": 0.981547494233592, + "grad_norm": 7.482025653531907, + "learning_rate": 1.5715708689565895e-05, + "loss": 1.8212, + "step": 4681 + }, + { + "epoch": 0.9817571817991193, + "grad_norm": 7.479972131518916, + "learning_rate": 1.571385090648277e-05, + "loss": 1.6145, + "step": 4682 + }, + { + "epoch": 0.9819668693646467, + "grad_norm": 6.511970499706646, + "learning_rate": 1.571199283055488e-05, + "loss": 1.683, + "step": 4683 + }, + { + "epoch": 0.982176556930174, + "grad_norm": 8.46534380142125, + "learning_rate": 1.5710134461877454e-05, + "loss": 1.8409, + "step": 4684 + }, + { + "epoch": 0.9823862444957014, + "grad_norm": 5.746535991179911, + "learning_rate": 1.5708275800545724e-05, + "loss": 1.5957, + "step": 4685 + }, + { + "epoch": 0.9825959320612287, + "grad_norm": 6.090738886487232, + "learning_rate": 1.5706416846654967e-05, + "loss": 1.7316, + "step": 4686 + }, + { + "epoch": 0.9828056196267562, + "grad_norm": 6.283248584151236, + "learning_rate": 1.5704557600300447e-05, + "loss": 1.4384, + "step": 4687 + }, + { + "epoch": 0.9830153071922835, + "grad_norm": 5.536807851806472, + "learning_rate": 1.570269806157745e-05, + "loss": 1.4214, + "step": 4688 + }, + { + "epoch": 0.9832249947578109, + "grad_norm": 6.575722458961724, + "learning_rate": 1.5700838230581295e-05, + "loss": 1.4482, + "step": 4689 + }, + { + "epoch": 0.9834346823233382, + "grad_norm": 6.924901059445336, + "learning_rate": 1.569897810740729e-05, + "loss": 1.7714, + "step": 4690 + }, + { + "epoch": 0.9836443698888656, + "grad_norm": 6.706251417751598, + "learning_rate": 1.5697117692150773e-05, + "loss": 1.7371, + "step": 4691 + }, + { + "epoch": 0.9838540574543929, + "grad_norm": 6.7450677192777215, + "learning_rate": 1.5695256984907094e-05, + "loss": 1.5486, + "step": 4692 + }, + { + "epoch": 0.9840637450199203, + "grad_norm": 6.895724100984174, + "learning_rate": 1.5693395985771615e-05, + "loss": 1.8585, + "step": 4693 + }, + { + "epoch": 0.9842734325854476, + "grad_norm": 7.150128314685859, + "learning_rate": 1.5691534694839714e-05, + "loss": 1.7073, + "step": 4694 + }, + { + "epoch": 0.984483120150975, + "grad_norm": 5.175895080754177, + "learning_rate": 1.5689673112206793e-05, + "loss": 1.4491, + "step": 4695 + }, + { + "epoch": 0.9846928077165024, + "grad_norm": 6.223516134523837, + "learning_rate": 1.5687811237968254e-05, + "loss": 1.5606, + "step": 4696 + }, + { + "epoch": 0.9849024952820298, + "grad_norm": 6.561371979098763, + "learning_rate": 1.568594907221953e-05, + "loss": 1.6077, + "step": 4697 + }, + { + "epoch": 0.9851121828475572, + "grad_norm": 7.842256970049425, + "learning_rate": 1.5684086615056047e-05, + "loss": 1.3531, + "step": 4698 + }, + { + "epoch": 0.9853218704130845, + "grad_norm": 7.06878644138483, + "learning_rate": 1.568222386657327e-05, + "loss": 1.53, + "step": 4699 + }, + { + "epoch": 0.9855315579786119, + "grad_norm": 7.100400552780994, + "learning_rate": 1.5680360826866662e-05, + "loss": 1.8881, + "step": 4700 + }, + { + "epoch": 0.9857412455441392, + "grad_norm": 7.003120317100073, + "learning_rate": 1.5678497496031714e-05, + "loss": 1.8249, + "step": 4701 + }, + { + "epoch": 0.9859509331096666, + "grad_norm": 7.579406378747845, + "learning_rate": 1.5676633874163916e-05, + "loss": 1.731, + "step": 4702 + }, + { + "epoch": 0.9861606206751939, + "grad_norm": 5.992297313752019, + "learning_rate": 1.567476996135879e-05, + "loss": 1.5333, + "step": 4703 + }, + { + "epoch": 0.9863703082407214, + "grad_norm": 7.128370011162646, + "learning_rate": 1.5672905757711857e-05, + "loss": 1.3557, + "step": 4704 + }, + { + "epoch": 0.9865799958062487, + "grad_norm": 8.58932128755457, + "learning_rate": 1.5671041263318665e-05, + "loss": 1.5978, + "step": 4705 + }, + { + "epoch": 0.9867896833717761, + "grad_norm": 7.108877637254589, + "learning_rate": 1.5669176478274775e-05, + "loss": 1.688, + "step": 4706 + }, + { + "epoch": 0.9869993709373034, + "grad_norm": 6.23929333269962, + "learning_rate": 1.5667311402675758e-05, + "loss": 1.6544, + "step": 4707 + }, + { + "epoch": 0.9872090585028308, + "grad_norm": 6.506240507750133, + "learning_rate": 1.56654460366172e-05, + "loss": 1.7887, + "step": 4708 + }, + { + "epoch": 0.9874187460683581, + "grad_norm": 5.937561033901428, + "learning_rate": 1.5663580380194708e-05, + "loss": 1.5289, + "step": 4709 + }, + { + "epoch": 0.9876284336338855, + "grad_norm": 6.28247731839628, + "learning_rate": 1.56617144335039e-05, + "loss": 1.2513, + "step": 4710 + }, + { + "epoch": 0.9878381211994128, + "grad_norm": 7.625473899361198, + "learning_rate": 1.5659848196640403e-05, + "loss": 1.6971, + "step": 4711 + }, + { + "epoch": 0.9880478087649402, + "grad_norm": 6.204804902850047, + "learning_rate": 1.5657981669699874e-05, + "loss": 1.3682, + "step": 4712 + }, + { + "epoch": 0.9882574963304676, + "grad_norm": 7.00849610229715, + "learning_rate": 1.5656114852777968e-05, + "loss": 1.8106, + "step": 4713 + }, + { + "epoch": 0.988467183895995, + "grad_norm": 7.232532497026672, + "learning_rate": 1.565424774597037e-05, + "loss": 2.1506, + "step": 4714 + }, + { + "epoch": 0.9886768714615223, + "grad_norm": 6.126261473716029, + "learning_rate": 1.565238034937276e-05, + "loss": 1.5816, + "step": 4715 + }, + { + "epoch": 0.9888865590270497, + "grad_norm": 7.29537048186078, + "learning_rate": 1.5650512663080862e-05, + "loss": 1.7248, + "step": 4716 + }, + { + "epoch": 0.9890962465925771, + "grad_norm": 6.54284687184934, + "learning_rate": 1.5648644687190388e-05, + "loss": 1.5076, + "step": 4717 + }, + { + "epoch": 0.9893059341581044, + "grad_norm": 7.560713721326173, + "learning_rate": 1.5646776421797073e-05, + "loss": 1.8843, + "step": 4718 + }, + { + "epoch": 0.9895156217236318, + "grad_norm": 6.141870452737586, + "learning_rate": 1.5644907866996676e-05, + "loss": 1.7725, + "step": 4719 + }, + { + "epoch": 0.9897253092891591, + "grad_norm": 7.456775166813868, + "learning_rate": 1.564303902288496e-05, + "loss": 1.7688, + "step": 4720 + }, + { + "epoch": 0.9899349968546866, + "grad_norm": 5.829284499255656, + "learning_rate": 1.5641169889557705e-05, + "loss": 1.5737, + "step": 4721 + }, + { + "epoch": 0.9901446844202139, + "grad_norm": 6.595997551041795, + "learning_rate": 1.5639300467110706e-05, + "loss": 1.6998, + "step": 4722 + }, + { + "epoch": 0.9903543719857413, + "grad_norm": 6.998664899999526, + "learning_rate": 1.5637430755639782e-05, + "loss": 1.698, + "step": 4723 + }, + { + "epoch": 0.9905640595512686, + "grad_norm": 6.110093213570619, + "learning_rate": 1.5635560755240753e-05, + "loss": 1.4847, + "step": 4724 + }, + { + "epoch": 0.990773747116796, + "grad_norm": 7.483475634816272, + "learning_rate": 1.5633690466009462e-05, + "loss": 1.8429, + "step": 4725 + }, + { + "epoch": 0.9909834346823233, + "grad_norm": 7.636147999959991, + "learning_rate": 1.5631819888041763e-05, + "loss": 1.482, + "step": 4726 + }, + { + "epoch": 0.9911931222478507, + "grad_norm": 7.459828034141028, + "learning_rate": 1.5629949021433527e-05, + "loss": 1.7581, + "step": 4727 + }, + { + "epoch": 0.991402809813378, + "grad_norm": 7.035740233990961, + "learning_rate": 1.5628077866280637e-05, + "loss": 1.5578, + "step": 4728 + }, + { + "epoch": 0.9916124973789054, + "grad_norm": 6.84929234695881, + "learning_rate": 1.5626206422678995e-05, + "loss": 1.5546, + "step": 4729 + }, + { + "epoch": 0.9918221849444327, + "grad_norm": 6.529377440298712, + "learning_rate": 1.5624334690724518e-05, + "loss": 1.5729, + "step": 4730 + }, + { + "epoch": 0.9920318725099602, + "grad_norm": 6.4583840153164225, + "learning_rate": 1.5622462670513132e-05, + "loss": 1.7093, + "step": 4731 + }, + { + "epoch": 0.9922415600754875, + "grad_norm": 7.8356550577953685, + "learning_rate": 1.562059036214078e-05, + "loss": 1.73, + "step": 4732 + }, + { + "epoch": 0.9924512476410149, + "grad_norm": 6.777204223262935, + "learning_rate": 1.561871776570343e-05, + "loss": 1.9414, + "step": 4733 + }, + { + "epoch": 0.9926609352065423, + "grad_norm": 6.165735943166361, + "learning_rate": 1.5616844881297046e-05, + "loss": 1.5659, + "step": 4734 + }, + { + "epoch": 0.9928706227720696, + "grad_norm": 5.974788351642358, + "learning_rate": 1.5614971709017615e-05, + "loss": 1.77, + "step": 4735 + }, + { + "epoch": 0.993080310337597, + "grad_norm": 7.163448647259221, + "learning_rate": 1.5613098248961153e-05, + "loss": 1.8803, + "step": 4736 + }, + { + "epoch": 0.9932899979031243, + "grad_norm": 5.74652630865415, + "learning_rate": 1.5611224501223662e-05, + "loss": 1.624, + "step": 4737 + }, + { + "epoch": 0.9934996854686517, + "grad_norm": 6.302638991479064, + "learning_rate": 1.5609350465901192e-05, + "loss": 1.5766, + "step": 4738 + }, + { + "epoch": 0.993709373034179, + "grad_norm": 6.0251103955830425, + "learning_rate": 1.5607476143089777e-05, + "loss": 1.7713, + "step": 4739 + }, + { + "epoch": 0.9939190605997065, + "grad_norm": 5.991077590894774, + "learning_rate": 1.5605601532885484e-05, + "loss": 1.707, + "step": 4740 + }, + { + "epoch": 0.9941287481652338, + "grad_norm": 5.574971282067392, + "learning_rate": 1.560372663538439e-05, + "loss": 1.4232, + "step": 4741 + }, + { + "epoch": 0.9943384357307612, + "grad_norm": 6.935163289653459, + "learning_rate": 1.560185145068259e-05, + "loss": 1.6434, + "step": 4742 + }, + { + "epoch": 0.9945481232962885, + "grad_norm": 6.841965116816619, + "learning_rate": 1.559997597887618e-05, + "loss": 1.6611, + "step": 4743 + }, + { + "epoch": 0.9947578108618159, + "grad_norm": 6.521858204046354, + "learning_rate": 1.5598100220061297e-05, + "loss": 1.5034, + "step": 4744 + }, + { + "epoch": 0.9949674984273432, + "grad_norm": 5.902121992098352, + "learning_rate": 1.5596224174334062e-05, + "loss": 1.7083, + "step": 4745 + }, + { + "epoch": 0.9951771859928706, + "grad_norm": 6.624911567062908, + "learning_rate": 1.5594347841790636e-05, + "loss": 1.6365, + "step": 4746 + }, + { + "epoch": 0.9953868735583979, + "grad_norm": 6.787391084546669, + "learning_rate": 1.5592471222527177e-05, + "loss": 1.8257, + "step": 4747 + }, + { + "epoch": 0.9955965611239254, + "grad_norm": 7.896686393093549, + "learning_rate": 1.5590594316639872e-05, + "loss": 1.872, + "step": 4748 + }, + { + "epoch": 0.9958062486894527, + "grad_norm": 7.2389605839494315, + "learning_rate": 1.558871712422491e-05, + "loss": 1.6712, + "step": 4749 + }, + { + "epoch": 0.9960159362549801, + "grad_norm": 6.077597398922661, + "learning_rate": 1.55868396453785e-05, + "loss": 1.82, + "step": 4750 + }, + { + "epoch": 0.9962256238205074, + "grad_norm": 7.557189163290166, + "learning_rate": 1.5584961880196873e-05, + "loss": 1.9075, + "step": 4751 + }, + { + "epoch": 0.9964353113860348, + "grad_norm": 6.98272421469055, + "learning_rate": 1.5583083828776257e-05, + "loss": 1.6065, + "step": 4752 + }, + { + "epoch": 0.9966449989515622, + "grad_norm": 8.54891787340911, + "learning_rate": 1.5581205491212915e-05, + "loss": 1.5479, + "step": 4753 + }, + { + "epoch": 0.9968546865170895, + "grad_norm": 5.647946622961197, + "learning_rate": 1.557932686760311e-05, + "loss": 1.8178, + "step": 4754 + }, + { + "epoch": 0.9970643740826169, + "grad_norm": 6.763389138666307, + "learning_rate": 1.5577447958043127e-05, + "loss": 1.8128, + "step": 4755 + }, + { + "epoch": 0.9972740616481442, + "grad_norm": 5.655368189708778, + "learning_rate": 1.5575568762629265e-05, + "loss": 1.4167, + "step": 4756 + }, + { + "epoch": 0.9974837492136717, + "grad_norm": 6.129613806477013, + "learning_rate": 1.557368928145783e-05, + "loss": 1.8118, + "step": 4757 + }, + { + "epoch": 0.997693436779199, + "grad_norm": 7.183523443011787, + "learning_rate": 1.5571809514625154e-05, + "loss": 1.6398, + "step": 4758 + }, + { + "epoch": 0.9979031243447264, + "grad_norm": 6.0028636404971225, + "learning_rate": 1.5569929462227578e-05, + "loss": 1.4096, + "step": 4759 + }, + { + "epoch": 0.9981128119102537, + "grad_norm": 7.583668733915585, + "learning_rate": 1.556804912436145e-05, + "loss": 2.1504, + "step": 4760 + }, + { + "epoch": 0.9983224994757811, + "grad_norm": 6.538615120859601, + "learning_rate": 1.5566168501123153e-05, + "loss": 1.444, + "step": 4761 + }, + { + "epoch": 0.9985321870413084, + "grad_norm": 7.002487362281974, + "learning_rate": 1.5564287592609065e-05, + "loss": 1.7205, + "step": 4762 + }, + { + "epoch": 0.9987418746068358, + "grad_norm": 6.740027524260676, + "learning_rate": 1.556240639891559e-05, + "loss": 1.4106, + "step": 4763 + }, + { + "epoch": 0.9989515621723631, + "grad_norm": 7.469900794982689, + "learning_rate": 1.5560524920139135e-05, + "loss": 1.4109, + "step": 4764 + }, + { + "epoch": 0.9991612497378906, + "grad_norm": 7.930028683270726, + "learning_rate": 1.5558643156376138e-05, + "loss": 1.6216, + "step": 4765 + }, + { + "epoch": 0.9993709373034179, + "grad_norm": 7.080676572829301, + "learning_rate": 1.5556761107723034e-05, + "loss": 1.7556, + "step": 4766 + }, + { + "epoch": 0.9995806248689453, + "grad_norm": 6.68080725151864, + "learning_rate": 1.555487877427629e-05, + "loss": 1.576, + "step": 4767 + }, + { + "epoch": 0.9997903124344726, + "grad_norm": 6.608495204505117, + "learning_rate": 1.5552996156132368e-05, + "loss": 1.7584, + "step": 4768 + }, + { + "epoch": 1.0, + "grad_norm": 6.347222962776922, + "learning_rate": 1.5551113253387765e-05, + "loss": 1.6321, + "step": 4769 + }, + { + "epoch": 1.0002096875655273, + "grad_norm": 6.160992356998679, + "learning_rate": 1.554923006613898e-05, + "loss": 0.7599, + "step": 4770 + }, + { + "epoch": 1.0004193751310548, + "grad_norm": 7.003062246576398, + "learning_rate": 1.5547346594482527e-05, + "loss": 0.955, + "step": 4771 + }, + { + "epoch": 1.0006290626965821, + "grad_norm": 6.141910853326404, + "learning_rate": 1.5545462838514945e-05, + "loss": 0.9296, + "step": 4772 + }, + { + "epoch": 1.0008387502621094, + "grad_norm": 5.597492055554543, + "learning_rate": 1.554357879833277e-05, + "loss": 0.9148, + "step": 4773 + }, + { + "epoch": 1.0010484378276368, + "grad_norm": 5.528091974130121, + "learning_rate": 1.554169447403257e-05, + "loss": 0.8728, + "step": 4774 + }, + { + "epoch": 1.0012581253931643, + "grad_norm": 6.762864367165756, + "learning_rate": 1.5539809865710913e-05, + "loss": 1.016, + "step": 4775 + }, + { + "epoch": 1.0014678129586916, + "grad_norm": 6.127952313172415, + "learning_rate": 1.5537924973464397e-05, + "loss": 1.0382, + "step": 4776 + }, + { + "epoch": 1.0016775005242189, + "grad_norm": 6.0177848211495695, + "learning_rate": 1.553603979738962e-05, + "loss": 0.9296, + "step": 4777 + }, + { + "epoch": 1.0018871880897462, + "grad_norm": 6.430524103152137, + "learning_rate": 1.5534154337583198e-05, + "loss": 0.9276, + "step": 4778 + }, + { + "epoch": 1.0020968756552737, + "grad_norm": 6.902758896448298, + "learning_rate": 1.5532268594141767e-05, + "loss": 0.8018, + "step": 4779 + }, + { + "epoch": 1.002306563220801, + "grad_norm": 7.0240480556193985, + "learning_rate": 1.553038256716198e-05, + "loss": 0.8344, + "step": 4780 + }, + { + "epoch": 1.0025162507863283, + "grad_norm": 6.741271948755893, + "learning_rate": 1.5528496256740492e-05, + "loss": 0.8489, + "step": 4781 + }, + { + "epoch": 1.0027259383518556, + "grad_norm": 7.318285709842115, + "learning_rate": 1.552660966297398e-05, + "loss": 0.8857, + "step": 4782 + }, + { + "epoch": 1.0029356259173832, + "grad_norm": 6.400814764801574, + "learning_rate": 1.5524722785959142e-05, + "loss": 0.7388, + "step": 4783 + }, + { + "epoch": 1.0031453134829105, + "grad_norm": 8.782317484212461, + "learning_rate": 1.5522835625792674e-05, + "loss": 0.8743, + "step": 4784 + }, + { + "epoch": 1.0033550010484378, + "grad_norm": 8.458458066706296, + "learning_rate": 1.5520948182571305e-05, + "loss": 1.0624, + "step": 4785 + }, + { + "epoch": 1.003564688613965, + "grad_norm": 9.598469956844733, + "learning_rate": 1.5519060456391763e-05, + "loss": 0.9172, + "step": 4786 + }, + { + "epoch": 1.0037743761794926, + "grad_norm": 9.449742846720936, + "learning_rate": 1.5517172447350806e-05, + "loss": 1.0205, + "step": 4787 + }, + { + "epoch": 1.00398406374502, + "grad_norm": 7.628697144396201, + "learning_rate": 1.5515284155545187e-05, + "loss": 0.9323, + "step": 4788 + }, + { + "epoch": 1.0041937513105472, + "grad_norm": 10.023282079242447, + "learning_rate": 1.5513395581071692e-05, + "loss": 1.0517, + "step": 4789 + }, + { + "epoch": 1.0044034388760748, + "grad_norm": 11.597535446234218, + "learning_rate": 1.551150672402711e-05, + "loss": 1.3654, + "step": 4790 + }, + { + "epoch": 1.004613126441602, + "grad_norm": 7.89602725031278, + "learning_rate": 1.550961758450825e-05, + "loss": 0.8884, + "step": 4791 + }, + { + "epoch": 1.0048228140071294, + "grad_norm": 7.472540163553277, + "learning_rate": 1.550772816261193e-05, + "loss": 0.8363, + "step": 4792 + }, + { + "epoch": 1.0050325015726567, + "grad_norm": 7.979102850120129, + "learning_rate": 1.550583845843499e-05, + "loss": 1.0372, + "step": 4793 + }, + { + "epoch": 1.0052421891381842, + "grad_norm": 8.134542192639802, + "learning_rate": 1.550394847207428e-05, + "loss": 0.927, + "step": 4794 + }, + { + "epoch": 1.0054518767037115, + "grad_norm": 6.66423236170934, + "learning_rate": 1.5502058203626663e-05, + "loss": 0.858, + "step": 4795 + }, + { + "epoch": 1.0056615642692388, + "grad_norm": 6.1002395239030855, + "learning_rate": 1.5500167653189024e-05, + "loss": 0.7533, + "step": 4796 + }, + { + "epoch": 1.0058712518347661, + "grad_norm": 6.1888208626094965, + "learning_rate": 1.5498276820858255e-05, + "loss": 0.7957, + "step": 4797 + }, + { + "epoch": 1.0060809394002936, + "grad_norm": 8.430796112383831, + "learning_rate": 1.5496385706731255e-05, + "loss": 1.2532, + "step": 4798 + }, + { + "epoch": 1.006290626965821, + "grad_norm": 7.171822195544356, + "learning_rate": 1.549449431090496e-05, + "loss": 1.1588, + "step": 4799 + }, + { + "epoch": 1.0065003145313483, + "grad_norm": 7.101904096560574, + "learning_rate": 1.54926026334763e-05, + "loss": 0.7991, + "step": 4800 + }, + { + "epoch": 1.0067100020968756, + "grad_norm": 7.172039592377217, + "learning_rate": 1.5490710674542227e-05, + "loss": 0.9349, + "step": 4801 + }, + { + "epoch": 1.006919689662403, + "grad_norm": 7.570033783058164, + "learning_rate": 1.5488818434199713e-05, + "loss": 1.0008, + "step": 4802 + }, + { + "epoch": 1.0071293772279304, + "grad_norm": 6.4288680238348705, + "learning_rate": 1.5486925912545732e-05, + "loss": 0.6979, + "step": 4803 + }, + { + "epoch": 1.0073390647934577, + "grad_norm": 6.733630736384756, + "learning_rate": 1.548503310967728e-05, + "loss": 1.0022, + "step": 4804 + }, + { + "epoch": 1.0075487523589852, + "grad_norm": 6.710651848201421, + "learning_rate": 1.548314002569137e-05, + "loss": 1.038, + "step": 4805 + }, + { + "epoch": 1.0077584399245125, + "grad_norm": 7.017493004074286, + "learning_rate": 1.5481246660685024e-05, + "loss": 0.9196, + "step": 4806 + }, + { + "epoch": 1.0079681274900398, + "grad_norm": 7.356488083523028, + "learning_rate": 1.5479353014755278e-05, + "loss": 0.9708, + "step": 4807 + }, + { + "epoch": 1.0081778150555671, + "grad_norm": 8.866333547929878, + "learning_rate": 1.5477459087999188e-05, + "loss": 0.9884, + "step": 4808 + }, + { + "epoch": 1.0083875026210947, + "grad_norm": 6.163757802439368, + "learning_rate": 1.547556488051382e-05, + "loss": 0.8365, + "step": 4809 + }, + { + "epoch": 1.008597190186622, + "grad_norm": 8.021070729056651, + "learning_rate": 1.5473670392396255e-05, + "loss": 1.0373, + "step": 4810 + }, + { + "epoch": 1.0088068777521493, + "grad_norm": 7.068433734458395, + "learning_rate": 1.5471775623743588e-05, + "loss": 0.8432, + "step": 4811 + }, + { + "epoch": 1.0090165653176766, + "grad_norm": 9.830581462857863, + "learning_rate": 1.546988057465293e-05, + "loss": 1.0924, + "step": 4812 + }, + { + "epoch": 1.0092262528832041, + "grad_norm": 7.394931313614123, + "learning_rate": 1.5467985245221406e-05, + "loss": 0.8874, + "step": 4813 + }, + { + "epoch": 1.0094359404487314, + "grad_norm": 7.538013906143377, + "learning_rate": 1.5466089635546155e-05, + "loss": 1.0055, + "step": 4814 + }, + { + "epoch": 1.0096456280142587, + "grad_norm": 6.232427846893156, + "learning_rate": 1.5464193745724325e-05, + "loss": 0.6984, + "step": 4815 + }, + { + "epoch": 1.009855315579786, + "grad_norm": 5.877638188346615, + "learning_rate": 1.546229757585309e-05, + "loss": 0.7757, + "step": 4816 + }, + { + "epoch": 1.0100650031453136, + "grad_norm": 7.489816565219068, + "learning_rate": 1.5460401126029637e-05, + "loss": 0.9767, + "step": 4817 + }, + { + "epoch": 1.0102746907108409, + "grad_norm": 7.792365557288659, + "learning_rate": 1.545850439635115e-05, + "loss": 1.0158, + "step": 4818 + }, + { + "epoch": 1.0104843782763682, + "grad_norm": 8.341728234282202, + "learning_rate": 1.5456607386914845e-05, + "loss": 0.8665, + "step": 4819 + }, + { + "epoch": 1.0106940658418955, + "grad_norm": 8.47319788320541, + "learning_rate": 1.545471009781795e-05, + "loss": 0.984, + "step": 4820 + }, + { + "epoch": 1.010903753407423, + "grad_norm": 8.941107616003826, + "learning_rate": 1.54528125291577e-05, + "loss": 0.7165, + "step": 4821 + }, + { + "epoch": 1.0111134409729503, + "grad_norm": 11.124118945174159, + "learning_rate": 1.5450914681031352e-05, + "loss": 1.2522, + "step": 4822 + }, + { + "epoch": 1.0113231285384776, + "grad_norm": 6.265448431014745, + "learning_rate": 1.5449016553536168e-05, + "loss": 0.7148, + "step": 4823 + }, + { + "epoch": 1.0115328161040051, + "grad_norm": 7.978689641467229, + "learning_rate": 1.544711814676944e-05, + "loss": 0.9815, + "step": 4824 + }, + { + "epoch": 1.0117425036695324, + "grad_norm": 7.397987623192461, + "learning_rate": 1.544521946082846e-05, + "loss": 0.8813, + "step": 4825 + }, + { + "epoch": 1.0119521912350598, + "grad_norm": 7.171100212561379, + "learning_rate": 1.5443320495810535e-05, + "loss": 0.8061, + "step": 4826 + }, + { + "epoch": 1.012161878800587, + "grad_norm": 8.045166625457925, + "learning_rate": 1.5441421251812993e-05, + "loss": 0.9499, + "step": 4827 + }, + { + "epoch": 1.0123715663661146, + "grad_norm": 7.651080839085174, + "learning_rate": 1.5439521728933177e-05, + "loss": 1.0635, + "step": 4828 + }, + { + "epoch": 1.012581253931642, + "grad_norm": 9.857127619093347, + "learning_rate": 1.5437621927268435e-05, + "loss": 0.9477, + "step": 4829 + }, + { + "epoch": 1.0127909414971692, + "grad_norm": 7.729377406431391, + "learning_rate": 1.5435721846916137e-05, + "loss": 0.9344, + "step": 4830 + }, + { + "epoch": 1.0130006290626965, + "grad_norm": 6.78144221566299, + "learning_rate": 1.5433821487973674e-05, + "loss": 0.8084, + "step": 4831 + }, + { + "epoch": 1.013210316628224, + "grad_norm": 6.752617376832929, + "learning_rate": 1.5431920850538432e-05, + "loss": 0.9478, + "step": 4832 + }, + { + "epoch": 1.0134200041937513, + "grad_norm": 5.9980789959754555, + "learning_rate": 1.543001993470782e-05, + "loss": 0.6817, + "step": 4833 + }, + { + "epoch": 1.0136296917592786, + "grad_norm": 6.577932642380443, + "learning_rate": 1.5428118740579274e-05, + "loss": 0.7125, + "step": 4834 + }, + { + "epoch": 1.013839379324806, + "grad_norm": 5.711699780954579, + "learning_rate": 1.5426217268250226e-05, + "loss": 0.6333, + "step": 4835 + }, + { + "epoch": 1.0140490668903335, + "grad_norm": 8.122745473699483, + "learning_rate": 1.5424315517818134e-05, + "loss": 1.0306, + "step": 4836 + }, + { + "epoch": 1.0142587544558608, + "grad_norm": 5.818445579124551, + "learning_rate": 1.542241348938046e-05, + "loss": 0.8214, + "step": 4837 + }, + { + "epoch": 1.014468442021388, + "grad_norm": 7.497099912494813, + "learning_rate": 1.5420511183034695e-05, + "loss": 0.899, + "step": 4838 + }, + { + "epoch": 1.0146781295869154, + "grad_norm": 7.833021591244792, + "learning_rate": 1.5418608598878325e-05, + "loss": 0.8644, + "step": 4839 + }, + { + "epoch": 1.014887817152443, + "grad_norm": 7.221951250666633, + "learning_rate": 1.5416705737008868e-05, + "loss": 0.9763, + "step": 4840 + }, + { + "epoch": 1.0150975047179702, + "grad_norm": 7.060329361357559, + "learning_rate": 1.541480259752385e-05, + "loss": 0.8822, + "step": 4841 + }, + { + "epoch": 1.0153071922834975, + "grad_norm": 6.607377770868969, + "learning_rate": 1.5412899180520804e-05, + "loss": 0.7601, + "step": 4842 + }, + { + "epoch": 1.015516879849025, + "grad_norm": 7.128108755993761, + "learning_rate": 1.541099548609729e-05, + "loss": 0.8568, + "step": 4843 + }, + { + "epoch": 1.0157265674145524, + "grad_norm": 7.7933385718225106, + "learning_rate": 1.540909151435087e-05, + "loss": 0.9918, + "step": 4844 + }, + { + "epoch": 1.0159362549800797, + "grad_norm": 7.842364843020797, + "learning_rate": 1.5407187265379127e-05, + "loss": 0.9293, + "step": 4845 + }, + { + "epoch": 1.016145942545607, + "grad_norm": 9.015812971669483, + "learning_rate": 1.540528273927966e-05, + "loss": 0.9142, + "step": 4846 + }, + { + "epoch": 1.0163556301111345, + "grad_norm": 6.600221295929714, + "learning_rate": 1.5403377936150076e-05, + "loss": 0.602, + "step": 4847 + }, + { + "epoch": 1.0165653176766618, + "grad_norm": 9.039132498764376, + "learning_rate": 1.5401472856088006e-05, + "loss": 1.0537, + "step": 4848 + }, + { + "epoch": 1.0167750052421891, + "grad_norm": 7.79865091391446, + "learning_rate": 1.539956749919108e-05, + "loss": 0.9747, + "step": 4849 + }, + { + "epoch": 1.0169846928077164, + "grad_norm": 7.603094653958017, + "learning_rate": 1.539766186555695e-05, + "loss": 0.9491, + "step": 4850 + }, + { + "epoch": 1.017194380373244, + "grad_norm": 7.652528116781066, + "learning_rate": 1.5395755955283297e-05, + "loss": 0.8649, + "step": 4851 + }, + { + "epoch": 1.0174040679387713, + "grad_norm": 6.456569485212163, + "learning_rate": 1.5393849768467788e-05, + "loss": 0.8898, + "step": 4852 + }, + { + "epoch": 1.0176137555042986, + "grad_norm": 6.809272595758615, + "learning_rate": 1.5391943305208124e-05, + "loss": 0.7431, + "step": 4853 + }, + { + "epoch": 1.0178234430698259, + "grad_norm": 8.01741550480431, + "learning_rate": 1.539003656560201e-05, + "loss": 0.993, + "step": 4854 + }, + { + "epoch": 1.0180331306353534, + "grad_norm": 7.77373667146071, + "learning_rate": 1.538812954974718e-05, + "loss": 0.9156, + "step": 4855 + }, + { + "epoch": 1.0182428182008807, + "grad_norm": 6.572133287398302, + "learning_rate": 1.538622225774136e-05, + "loss": 0.7831, + "step": 4856 + }, + { + "epoch": 1.018452505766408, + "grad_norm": 7.216682451758032, + "learning_rate": 1.5384314689682313e-05, + "loss": 0.8784, + "step": 4857 + }, + { + "epoch": 1.0186621933319353, + "grad_norm": 7.243893114131416, + "learning_rate": 1.5382406845667794e-05, + "loss": 1.0223, + "step": 4858 + }, + { + "epoch": 1.0188718808974628, + "grad_norm": 7.749479472701532, + "learning_rate": 1.538049872579559e-05, + "loss": 0.9865, + "step": 4859 + }, + { + "epoch": 1.0190815684629901, + "grad_norm": 7.213357088359016, + "learning_rate": 1.53785903301635e-05, + "loss": 0.9423, + "step": 4860 + }, + { + "epoch": 1.0192912560285174, + "grad_norm": 7.7311774524613135, + "learning_rate": 1.537668165886932e-05, + "loss": 1.037, + "step": 4861 + }, + { + "epoch": 1.019500943594045, + "grad_norm": 6.939071631903374, + "learning_rate": 1.5374772712010878e-05, + "loss": 0.7832, + "step": 4862 + }, + { + "epoch": 1.0197106311595723, + "grad_norm": 7.304979051799655, + "learning_rate": 1.5372863489686022e-05, + "loss": 1.0671, + "step": 4863 + }, + { + "epoch": 1.0199203187250996, + "grad_norm": 7.393542107705038, + "learning_rate": 1.5370953991992584e-05, + "loss": 1.0416, + "step": 4864 + }, + { + "epoch": 1.020130006290627, + "grad_norm": 9.137503390076736, + "learning_rate": 1.5369044219028444e-05, + "loss": 1.1535, + "step": 4865 + }, + { + "epoch": 1.0203396938561544, + "grad_norm": 6.654196987292526, + "learning_rate": 1.5367134170891472e-05, + "loss": 0.8538, + "step": 4866 + }, + { + "epoch": 1.0205493814216817, + "grad_norm": 8.293336701776512, + "learning_rate": 1.5365223847679568e-05, + "loss": 0.901, + "step": 4867 + }, + { + "epoch": 1.020759068987209, + "grad_norm": 6.231442627086459, + "learning_rate": 1.5363313249490635e-05, + "loss": 0.7209, + "step": 4868 + }, + { + "epoch": 1.0209687565527363, + "grad_norm": 6.380867152388836, + "learning_rate": 1.5361402376422594e-05, + "loss": 0.7363, + "step": 4869 + }, + { + "epoch": 1.0211784441182639, + "grad_norm": 5.782538980568519, + "learning_rate": 1.5359491228573385e-05, + "loss": 0.6333, + "step": 4870 + }, + { + "epoch": 1.0213881316837912, + "grad_norm": 6.8108533350524, + "learning_rate": 1.5357579806040953e-05, + "loss": 0.9333, + "step": 4871 + }, + { + "epoch": 1.0215978192493185, + "grad_norm": 6.561139136800119, + "learning_rate": 1.5355668108923265e-05, + "loss": 0.8766, + "step": 4872 + }, + { + "epoch": 1.0218075068148458, + "grad_norm": 7.207667521170825, + "learning_rate": 1.53537561373183e-05, + "loss": 1.0119, + "step": 4873 + }, + { + "epoch": 1.0220171943803733, + "grad_norm": 6.879222371421032, + "learning_rate": 1.5351843891324045e-05, + "loss": 0.9621, + "step": 4874 + }, + { + "epoch": 1.0222268819459006, + "grad_norm": 8.863317670280777, + "learning_rate": 1.534993137103851e-05, + "loss": 0.891, + "step": 4875 + }, + { + "epoch": 1.022436569511428, + "grad_norm": 7.539894066789852, + "learning_rate": 1.5348018576559707e-05, + "loss": 0.7907, + "step": 4876 + }, + { + "epoch": 1.0226462570769554, + "grad_norm": 7.723878179152712, + "learning_rate": 1.534610550798568e-05, + "loss": 0.909, + "step": 4877 + }, + { + "epoch": 1.0228559446424828, + "grad_norm": 6.255393201148589, + "learning_rate": 1.534419216541447e-05, + "loss": 0.655, + "step": 4878 + }, + { + "epoch": 1.02306563220801, + "grad_norm": 5.5256679205084565, + "learning_rate": 1.534227854894415e-05, + "loss": 0.8273, + "step": 4879 + }, + { + "epoch": 1.0232753197735374, + "grad_norm": 8.1422508065986, + "learning_rate": 1.5340364658672787e-05, + "loss": 0.998, + "step": 4880 + }, + { + "epoch": 1.023485007339065, + "grad_norm": 7.899056320304972, + "learning_rate": 1.533845049469847e-05, + "loss": 0.6781, + "step": 4881 + }, + { + "epoch": 1.0236946949045922, + "grad_norm": 5.98551241869694, + "learning_rate": 1.5336536057119307e-05, + "loss": 0.7395, + "step": 4882 + }, + { + "epoch": 1.0239043824701195, + "grad_norm": 7.06497172629204, + "learning_rate": 1.533462134603341e-05, + "loss": 1.0321, + "step": 4883 + }, + { + "epoch": 1.0241140700356468, + "grad_norm": 5.940506712153749, + "learning_rate": 1.5332706361538927e-05, + "loss": 0.7395, + "step": 4884 + }, + { + "epoch": 1.0243237576011743, + "grad_norm": 5.973792579108355, + "learning_rate": 1.5330791103733985e-05, + "loss": 0.7007, + "step": 4885 + }, + { + "epoch": 1.0245334451667016, + "grad_norm": 6.682918348080621, + "learning_rate": 1.532887557271676e-05, + "loss": 0.9773, + "step": 4886 + }, + { + "epoch": 1.024743132732229, + "grad_norm": 6.170331379094081, + "learning_rate": 1.5326959768585413e-05, + "loss": 0.6962, + "step": 4887 + }, + { + "epoch": 1.0249528202977563, + "grad_norm": 8.344971099761546, + "learning_rate": 1.5325043691438142e-05, + "loss": 1.1157, + "step": 4888 + }, + { + "epoch": 1.0251625078632838, + "grad_norm": 6.937852039995552, + "learning_rate": 1.5323127341373146e-05, + "loss": 0.8421, + "step": 4889 + }, + { + "epoch": 1.025372195428811, + "grad_norm": 8.392026438345647, + "learning_rate": 1.5321210718488642e-05, + "loss": 0.9602, + "step": 4890 + }, + { + "epoch": 1.0255818829943384, + "grad_norm": 6.307493919128374, + "learning_rate": 1.531929382288286e-05, + "loss": 1.0047, + "step": 4891 + }, + { + "epoch": 1.0257915705598657, + "grad_norm": 7.381614471042405, + "learning_rate": 1.5317376654654043e-05, + "loss": 1.0528, + "step": 4892 + }, + { + "epoch": 1.0260012581253932, + "grad_norm": 8.015427527495586, + "learning_rate": 1.5315459213900448e-05, + "loss": 0.8527, + "step": 4893 + }, + { + "epoch": 1.0262109456909205, + "grad_norm": 7.079622282020579, + "learning_rate": 1.531354150072035e-05, + "loss": 0.9815, + "step": 4894 + }, + { + "epoch": 1.0264206332564478, + "grad_norm": 5.875508104929852, + "learning_rate": 1.5311623515212036e-05, + "loss": 0.6933, + "step": 4895 + }, + { + "epoch": 1.0266303208219754, + "grad_norm": 8.498549735591824, + "learning_rate": 1.53097052574738e-05, + "loss": 1.0195, + "step": 4896 + }, + { + "epoch": 1.0268400083875027, + "grad_norm": 6.673638914952235, + "learning_rate": 1.5307786727603962e-05, + "loss": 0.7539, + "step": 4897 + }, + { + "epoch": 1.02704969595303, + "grad_norm": 6.593148830091223, + "learning_rate": 1.5305867925700853e-05, + "loss": 0.7908, + "step": 4898 + }, + { + "epoch": 1.0272593835185573, + "grad_norm": 9.085220206714018, + "learning_rate": 1.5303948851862806e-05, + "loss": 1.2864, + "step": 4899 + }, + { + "epoch": 1.0274690710840848, + "grad_norm": 6.637584637932477, + "learning_rate": 1.530202950618818e-05, + "loss": 0.8352, + "step": 4900 + }, + { + "epoch": 1.0276787586496121, + "grad_norm": 6.194410226825367, + "learning_rate": 1.5300109888775348e-05, + "loss": 0.5937, + "step": 4901 + }, + { + "epoch": 1.0278884462151394, + "grad_norm": 7.7949109501524365, + "learning_rate": 1.5298189999722693e-05, + "loss": 0.8478, + "step": 4902 + }, + { + "epoch": 1.0280981337806667, + "grad_norm": 6.422156473358355, + "learning_rate": 1.5296269839128612e-05, + "loss": 0.7847, + "step": 4903 + }, + { + "epoch": 1.0283078213461943, + "grad_norm": 8.206784981133389, + "learning_rate": 1.529434940709151e-05, + "loss": 0.9669, + "step": 4904 + }, + { + "epoch": 1.0285175089117216, + "grad_norm": 8.8965749407282, + "learning_rate": 1.5292428703709825e-05, + "loss": 0.9365, + "step": 4905 + }, + { + "epoch": 1.0287271964772489, + "grad_norm": 6.954027681744386, + "learning_rate": 1.529050772908198e-05, + "loss": 0.7532, + "step": 4906 + }, + { + "epoch": 1.0289368840427762, + "grad_norm": 6.951945879429214, + "learning_rate": 1.528858648330645e-05, + "loss": 0.9188, + "step": 4907 + }, + { + "epoch": 1.0291465716083037, + "grad_norm": 7.285777060926327, + "learning_rate": 1.528666496648168e-05, + "loss": 0.8034, + "step": 4908 + }, + { + "epoch": 1.029356259173831, + "grad_norm": 8.242924612766094, + "learning_rate": 1.5284743178706168e-05, + "loss": 0.9133, + "step": 4909 + }, + { + "epoch": 1.0295659467393583, + "grad_norm": 6.4857903543043225, + "learning_rate": 1.52828211200784e-05, + "loss": 0.7224, + "step": 4910 + }, + { + "epoch": 1.0297756343048856, + "grad_norm": 7.632058495847685, + "learning_rate": 1.5280898790696887e-05, + "loss": 1.0222, + "step": 4911 + }, + { + "epoch": 1.0299853218704131, + "grad_norm": 8.241533774173002, + "learning_rate": 1.5278976190660152e-05, + "loss": 0.9351, + "step": 4912 + }, + { + "epoch": 1.0301950094359404, + "grad_norm": 6.900869870056787, + "learning_rate": 1.527705332006673e-05, + "loss": 0.797, + "step": 4913 + }, + { + "epoch": 1.0304046970014678, + "grad_norm": 7.728788265724365, + "learning_rate": 1.5275130179015173e-05, + "loss": 0.9883, + "step": 4914 + }, + { + "epoch": 1.0306143845669953, + "grad_norm": 7.829491061975332, + "learning_rate": 1.5273206767604047e-05, + "loss": 0.9612, + "step": 4915 + }, + { + "epoch": 1.0308240721325226, + "grad_norm": 6.348711659796047, + "learning_rate": 1.5271283085931927e-05, + "loss": 0.807, + "step": 4916 + }, + { + "epoch": 1.03103375969805, + "grad_norm": 7.119047448839702, + "learning_rate": 1.526935913409741e-05, + "loss": 0.8866, + "step": 4917 + }, + { + "epoch": 1.0312434472635772, + "grad_norm": 7.073621453726036, + "learning_rate": 1.526743491219909e-05, + "loss": 0.7166, + "step": 4918 + }, + { + "epoch": 1.0314531348291047, + "grad_norm": 6.2562974434917695, + "learning_rate": 1.52655104203356e-05, + "loss": 0.6898, + "step": 4919 + }, + { + "epoch": 1.031662822394632, + "grad_norm": 7.45145563132765, + "learning_rate": 1.526358565860557e-05, + "loss": 0.7589, + "step": 4920 + }, + { + "epoch": 1.0318725099601593, + "grad_norm": 7.163729357019478, + "learning_rate": 1.5261660627107643e-05, + "loss": 0.8901, + "step": 4921 + }, + { + "epoch": 1.0320821975256866, + "grad_norm": 7.969289223341199, + "learning_rate": 1.5259735325940482e-05, + "loss": 1.0353, + "step": 4922 + }, + { + "epoch": 1.0322918850912142, + "grad_norm": 8.35151287533368, + "learning_rate": 1.5257809755202766e-05, + "loss": 0.9723, + "step": 4923 + }, + { + "epoch": 1.0325015726567415, + "grad_norm": 8.57390297153983, + "learning_rate": 1.525588391499318e-05, + "loss": 1.0289, + "step": 4924 + }, + { + "epoch": 1.0327112602222688, + "grad_norm": 5.166400977438753, + "learning_rate": 1.5253957805410427e-05, + "loss": 0.6506, + "step": 4925 + }, + { + "epoch": 1.032920947787796, + "grad_norm": 6.28417191108513, + "learning_rate": 1.5252031426553223e-05, + "loss": 0.7086, + "step": 4926 + }, + { + "epoch": 1.0331306353533236, + "grad_norm": 6.016081697449043, + "learning_rate": 1.52501047785203e-05, + "loss": 0.8106, + "step": 4927 + }, + { + "epoch": 1.033340322918851, + "grad_norm": 7.938133083908658, + "learning_rate": 1.5248177861410401e-05, + "loss": 0.9777, + "step": 4928 + }, + { + "epoch": 1.0335500104843782, + "grad_norm": 6.2390497542837915, + "learning_rate": 1.5246250675322282e-05, + "loss": 0.8393, + "step": 4929 + }, + { + "epoch": 1.0337596980499058, + "grad_norm": 7.9152307318907935, + "learning_rate": 1.5244323220354722e-05, + "loss": 1.0305, + "step": 4930 + }, + { + "epoch": 1.033969385615433, + "grad_norm": 6.408397883786978, + "learning_rate": 1.5242395496606497e-05, + "loss": 0.8316, + "step": 4931 + }, + { + "epoch": 1.0341790731809604, + "grad_norm": 8.912013223084658, + "learning_rate": 1.5240467504176413e-05, + "loss": 1.0824, + "step": 4932 + }, + { + "epoch": 1.0343887607464877, + "grad_norm": 6.611544571155025, + "learning_rate": 1.5238539243163277e-05, + "loss": 0.8369, + "step": 4933 + }, + { + "epoch": 1.0345984483120152, + "grad_norm": 7.142271040354011, + "learning_rate": 1.5236610713665922e-05, + "loss": 0.8512, + "step": 4934 + }, + { + "epoch": 1.0348081358775425, + "grad_norm": 7.02304870942379, + "learning_rate": 1.5234681915783182e-05, + "loss": 0.938, + "step": 4935 + }, + { + "epoch": 1.0350178234430698, + "grad_norm": 8.259218607351727, + "learning_rate": 1.5232752849613919e-05, + "loss": 1.0997, + "step": 4936 + }, + { + "epoch": 1.0352275110085971, + "grad_norm": 7.699991968792437, + "learning_rate": 1.5230823515256994e-05, + "loss": 0.9936, + "step": 4937 + }, + { + "epoch": 1.0354371985741246, + "grad_norm": 7.373855366620911, + "learning_rate": 1.5228893912811293e-05, + "loss": 0.7628, + "step": 4938 + }, + { + "epoch": 1.035646886139652, + "grad_norm": 7.3137595967106925, + "learning_rate": 1.522696404237571e-05, + "loss": 0.8644, + "step": 4939 + }, + { + "epoch": 1.0358565737051793, + "grad_norm": 7.19687470729279, + "learning_rate": 1.5225033904049152e-05, + "loss": 0.8256, + "step": 4940 + }, + { + "epoch": 1.0360662612707066, + "grad_norm": 6.3141301400210885, + "learning_rate": 1.5223103497930546e-05, + "loss": 0.636, + "step": 4941 + }, + { + "epoch": 1.036275948836234, + "grad_norm": 7.178556473291898, + "learning_rate": 1.5221172824118827e-05, + "loss": 0.9722, + "step": 4942 + }, + { + "epoch": 1.0364856364017614, + "grad_norm": 6.369711047600969, + "learning_rate": 1.521924188271294e-05, + "loss": 0.828, + "step": 4943 + }, + { + "epoch": 1.0366953239672887, + "grad_norm": 7.155809870583605, + "learning_rate": 1.521731067381186e-05, + "loss": 0.8976, + "step": 4944 + }, + { + "epoch": 1.036905011532816, + "grad_norm": 5.71419247433343, + "learning_rate": 1.5215379197514559e-05, + "loss": 0.6153, + "step": 4945 + }, + { + "epoch": 1.0371146990983435, + "grad_norm": 7.841874080434669, + "learning_rate": 1.5213447453920024e-05, + "loss": 1.0807, + "step": 4946 + }, + { + "epoch": 1.0373243866638708, + "grad_norm": 6.8902263484140125, + "learning_rate": 1.5211515443127268e-05, + "loss": 0.8236, + "step": 4947 + }, + { + "epoch": 1.0375340742293981, + "grad_norm": 5.459616391100047, + "learning_rate": 1.5209583165235307e-05, + "loss": 0.6613, + "step": 4948 + }, + { + "epoch": 1.0377437617949257, + "grad_norm": 8.009252446666222, + "learning_rate": 1.5207650620343173e-05, + "loss": 1.0604, + "step": 4949 + }, + { + "epoch": 1.037953449360453, + "grad_norm": 6.631908663650191, + "learning_rate": 1.5205717808549908e-05, + "loss": 0.8408, + "step": 4950 + }, + { + "epoch": 1.0381631369259803, + "grad_norm": 7.460614510230662, + "learning_rate": 1.5203784729954584e-05, + "loss": 1.0238, + "step": 4951 + }, + { + "epoch": 1.0383728244915076, + "grad_norm": 7.773727793281137, + "learning_rate": 1.520185138465626e-05, + "loss": 0.978, + "step": 4952 + }, + { + "epoch": 1.0385825120570351, + "grad_norm": 6.929464349115824, + "learning_rate": 1.5199917772754036e-05, + "loss": 0.8351, + "step": 4953 + }, + { + "epoch": 1.0387921996225624, + "grad_norm": 7.353387446001933, + "learning_rate": 1.5197983894347004e-05, + "loss": 0.9884, + "step": 4954 + }, + { + "epoch": 1.0390018871880897, + "grad_norm": 7.122582879420277, + "learning_rate": 1.5196049749534285e-05, + "loss": 1.0697, + "step": 4955 + }, + { + "epoch": 1.039211574753617, + "grad_norm": 7.801533730077766, + "learning_rate": 1.5194115338415e-05, + "loss": 0.8123, + "step": 4956 + }, + { + "epoch": 1.0394212623191446, + "grad_norm": 6.425233416553472, + "learning_rate": 1.51921806610883e-05, + "loss": 0.6701, + "step": 4957 + }, + { + "epoch": 1.0396309498846719, + "grad_norm": 7.047847522101738, + "learning_rate": 1.5190245717653332e-05, + "loss": 0.8372, + "step": 4958 + }, + { + "epoch": 1.0398406374501992, + "grad_norm": 6.395206387957378, + "learning_rate": 1.5188310508209273e-05, + "loss": 0.8802, + "step": 4959 + }, + { + "epoch": 1.0400503250157265, + "grad_norm": 5.875925723213028, + "learning_rate": 1.5186375032855296e-05, + "loss": 0.8079, + "step": 4960 + }, + { + "epoch": 1.040260012581254, + "grad_norm": 6.694159306339121, + "learning_rate": 1.5184439291690606e-05, + "loss": 0.7786, + "step": 4961 + }, + { + "epoch": 1.0404697001467813, + "grad_norm": 6.794040026490486, + "learning_rate": 1.5182503284814409e-05, + "loss": 0.9839, + "step": 4962 + }, + { + "epoch": 1.0406793877123086, + "grad_norm": 5.94054012017211, + "learning_rate": 1.5180567012325935e-05, + "loss": 0.8323, + "step": 4963 + }, + { + "epoch": 1.040889075277836, + "grad_norm": 5.641050102965108, + "learning_rate": 1.517863047432441e-05, + "loss": 0.7168, + "step": 4964 + }, + { + "epoch": 1.0410987628433634, + "grad_norm": 8.21345482293305, + "learning_rate": 1.5176693670909096e-05, + "loss": 0.9992, + "step": 4965 + }, + { + "epoch": 1.0413084504088908, + "grad_norm": 7.375461630597876, + "learning_rate": 1.517475660217925e-05, + "loss": 0.787, + "step": 4966 + }, + { + "epoch": 1.041518137974418, + "grad_norm": 6.789936193276466, + "learning_rate": 1.5172819268234155e-05, + "loss": 0.8067, + "step": 4967 + }, + { + "epoch": 1.0417278255399456, + "grad_norm": 6.168768720045257, + "learning_rate": 1.5170881669173102e-05, + "loss": 0.7556, + "step": 4968 + }, + { + "epoch": 1.041937513105473, + "grad_norm": 7.442692624941551, + "learning_rate": 1.5168943805095391e-05, + "loss": 0.8909, + "step": 4969 + }, + { + "epoch": 1.0421472006710002, + "grad_norm": 6.987253192762509, + "learning_rate": 1.5167005676100346e-05, + "loss": 0.9609, + "step": 4970 + }, + { + "epoch": 1.0423568882365275, + "grad_norm": 6.68226057430912, + "learning_rate": 1.51650672822873e-05, + "loss": 0.8657, + "step": 4971 + }, + { + "epoch": 1.042566575802055, + "grad_norm": 10.187132487220923, + "learning_rate": 1.5163128623755596e-05, + "loss": 1.0124, + "step": 4972 + }, + { + "epoch": 1.0427762633675823, + "grad_norm": 6.933568706735866, + "learning_rate": 1.5161189700604598e-05, + "loss": 0.8859, + "step": 4973 + }, + { + "epoch": 1.0429859509331096, + "grad_norm": 6.709281871479803, + "learning_rate": 1.5159250512933672e-05, + "loss": 0.5756, + "step": 4974 + }, + { + "epoch": 1.043195638498637, + "grad_norm": 6.880010295983627, + "learning_rate": 1.515731106084221e-05, + "loss": 0.7291, + "step": 4975 + }, + { + "epoch": 1.0434053260641645, + "grad_norm": 7.2003832373081655, + "learning_rate": 1.5155371344429612e-05, + "loss": 0.6924, + "step": 4976 + }, + { + "epoch": 1.0436150136296918, + "grad_norm": 6.840324106674076, + "learning_rate": 1.5153431363795292e-05, + "loss": 0.806, + "step": 4977 + }, + { + "epoch": 1.043824701195219, + "grad_norm": 8.050995520785525, + "learning_rate": 1.5151491119038672e-05, + "loss": 0.8851, + "step": 4978 + }, + { + "epoch": 1.0440343887607464, + "grad_norm": 8.003452055421375, + "learning_rate": 1.5149550610259203e-05, + "loss": 1.0227, + "step": 4979 + }, + { + "epoch": 1.044244076326274, + "grad_norm": 8.623397406152913, + "learning_rate": 1.5147609837556334e-05, + "loss": 1.0489, + "step": 4980 + }, + { + "epoch": 1.0444537638918012, + "grad_norm": 7.591865887265669, + "learning_rate": 1.5145668801029528e-05, + "loss": 0.8448, + "step": 4981 + }, + { + "epoch": 1.0446634514573285, + "grad_norm": 6.125270430642908, + "learning_rate": 1.5143727500778273e-05, + "loss": 0.7292, + "step": 4982 + }, + { + "epoch": 1.0448731390228558, + "grad_norm": 9.900133971716045, + "learning_rate": 1.5141785936902062e-05, + "loss": 0.9333, + "step": 4983 + }, + { + "epoch": 1.0450828265883834, + "grad_norm": 6.635919598445679, + "learning_rate": 1.5139844109500405e-05, + "loss": 0.8179, + "step": 4984 + }, + { + "epoch": 1.0452925141539107, + "grad_norm": 7.8794737090461995, + "learning_rate": 1.5137902018672823e-05, + "loss": 0.8866, + "step": 4985 + }, + { + "epoch": 1.045502201719438, + "grad_norm": 7.954706874316916, + "learning_rate": 1.5135959664518853e-05, + "loss": 1.1873, + "step": 4986 + }, + { + "epoch": 1.0457118892849655, + "grad_norm": 6.260871602935669, + "learning_rate": 1.5134017047138038e-05, + "loss": 0.7074, + "step": 4987 + }, + { + "epoch": 1.0459215768504928, + "grad_norm": 7.47829953257803, + "learning_rate": 1.5132074166629949e-05, + "loss": 0.834, + "step": 4988 + }, + { + "epoch": 1.0461312644160201, + "grad_norm": 7.103226589199328, + "learning_rate": 1.5130131023094154e-05, + "loss": 0.7806, + "step": 4989 + }, + { + "epoch": 1.0463409519815474, + "grad_norm": 7.026415806032829, + "learning_rate": 1.5128187616630248e-05, + "loss": 0.7655, + "step": 4990 + }, + { + "epoch": 1.046550639547075, + "grad_norm": 7.111847419815197, + "learning_rate": 1.5126243947337834e-05, + "loss": 0.8572, + "step": 4991 + }, + { + "epoch": 1.0467603271126023, + "grad_norm": 7.387592909986293, + "learning_rate": 1.5124300015316524e-05, + "loss": 0.7374, + "step": 4992 + }, + { + "epoch": 1.0469700146781296, + "grad_norm": 8.384145261677098, + "learning_rate": 1.512235582066595e-05, + "loss": 0.8726, + "step": 4993 + }, + { + "epoch": 1.0471797022436569, + "grad_norm": 10.48463549979645, + "learning_rate": 1.5120411363485758e-05, + "loss": 1.1321, + "step": 4994 + }, + { + "epoch": 1.0473893898091844, + "grad_norm": 7.726540855829647, + "learning_rate": 1.5118466643875603e-05, + "loss": 0.6866, + "step": 4995 + }, + { + "epoch": 1.0475990773747117, + "grad_norm": 7.58518523649189, + "learning_rate": 1.5116521661935153e-05, + "loss": 0.9859, + "step": 4996 + }, + { + "epoch": 1.047808764940239, + "grad_norm": 5.823361326959982, + "learning_rate": 1.5114576417764095e-05, + "loss": 0.6076, + "step": 4997 + }, + { + "epoch": 1.0480184525057663, + "grad_norm": 6.635570726572421, + "learning_rate": 1.5112630911462124e-05, + "loss": 0.7668, + "step": 4998 + }, + { + "epoch": 1.0482281400712938, + "grad_norm": 7.485108338377767, + "learning_rate": 1.511068514312895e-05, + "loss": 0.8497, + "step": 4999 + }, + { + "epoch": 1.0484378276368211, + "grad_norm": 7.858796476170127, + "learning_rate": 1.5108739112864301e-05, + "loss": 0.9937, + "step": 5000 + }, + { + "epoch": 1.0486475152023484, + "grad_norm": 6.761089314386591, + "learning_rate": 1.510679282076791e-05, + "loss": 0.736, + "step": 5001 + }, + { + "epoch": 1.0488572027678758, + "grad_norm": 8.09648809458204, + "learning_rate": 1.5104846266939532e-05, + "loss": 0.9433, + "step": 5002 + }, + { + "epoch": 1.0490668903334033, + "grad_norm": 5.799159783945289, + "learning_rate": 1.510289945147893e-05, + "loss": 0.7523, + "step": 5003 + }, + { + "epoch": 1.0492765778989306, + "grad_norm": 7.207518715900732, + "learning_rate": 1.5100952374485877e-05, + "loss": 0.9175, + "step": 5004 + }, + { + "epoch": 1.049486265464458, + "grad_norm": 6.233520566678287, + "learning_rate": 1.509900503606017e-05, + "loss": 0.849, + "step": 5005 + }, + { + "epoch": 1.0496959530299854, + "grad_norm": 6.7338206586902265, + "learning_rate": 1.509705743630161e-05, + "loss": 0.6924, + "step": 5006 + }, + { + "epoch": 1.0499056405955127, + "grad_norm": 6.656545986727053, + "learning_rate": 1.5095109575310014e-05, + "loss": 0.9097, + "step": 5007 + }, + { + "epoch": 1.05011532816104, + "grad_norm": 6.4659190885520355, + "learning_rate": 1.5093161453185219e-05, + "loss": 0.8672, + "step": 5008 + }, + { + "epoch": 1.0503250157265673, + "grad_norm": 7.655913405658616, + "learning_rate": 1.5091213070027064e-05, + "loss": 0.9757, + "step": 5009 + }, + { + "epoch": 1.0505347032920949, + "grad_norm": 6.638035837193669, + "learning_rate": 1.5089264425935408e-05, + "loss": 0.8311, + "step": 5010 + }, + { + "epoch": 1.0507443908576222, + "grad_norm": 7.331557028874784, + "learning_rate": 1.5087315521010127e-05, + "loss": 0.9073, + "step": 5011 + }, + { + "epoch": 1.0509540784231495, + "grad_norm": 6.625284771376989, + "learning_rate": 1.5085366355351099e-05, + "loss": 0.7779, + "step": 5012 + }, + { + "epoch": 1.0511637659886768, + "grad_norm": 6.222845317999916, + "learning_rate": 1.5083416929058223e-05, + "loss": 0.8391, + "step": 5013 + }, + { + "epoch": 1.0513734535542043, + "grad_norm": 7.473935522760192, + "learning_rate": 1.5081467242231415e-05, + "loss": 0.8763, + "step": 5014 + }, + { + "epoch": 1.0515831411197316, + "grad_norm": 6.795897657731762, + "learning_rate": 1.5079517294970601e-05, + "loss": 0.7929, + "step": 5015 + }, + { + "epoch": 1.051792828685259, + "grad_norm": 6.486017232070313, + "learning_rate": 1.5077567087375708e-05, + "loss": 0.939, + "step": 5016 + }, + { + "epoch": 1.0520025162507862, + "grad_norm": 7.552929790130162, + "learning_rate": 1.5075616619546702e-05, + "loss": 0.9303, + "step": 5017 + }, + { + "epoch": 1.0522122038163138, + "grad_norm": 7.005488479592218, + "learning_rate": 1.5073665891583538e-05, + "loss": 0.8501, + "step": 5018 + }, + { + "epoch": 1.052421891381841, + "grad_norm": 6.447850865274571, + "learning_rate": 1.5071714903586196e-05, + "loss": 0.9588, + "step": 5019 + }, + { + "epoch": 1.0526315789473684, + "grad_norm": 7.881188345711636, + "learning_rate": 1.5069763655654672e-05, + "loss": 0.9151, + "step": 5020 + }, + { + "epoch": 1.052841266512896, + "grad_norm": 8.856306952193773, + "learning_rate": 1.5067812147888967e-05, + "loss": 0.9757, + "step": 5021 + }, + { + "epoch": 1.0530509540784232, + "grad_norm": 7.510488000659729, + "learning_rate": 1.50658603803891e-05, + "loss": 0.8956, + "step": 5022 + }, + { + "epoch": 1.0532606416439505, + "grad_norm": 6.994093836212675, + "learning_rate": 1.50639083532551e-05, + "loss": 0.9696, + "step": 5023 + }, + { + "epoch": 1.0534703292094778, + "grad_norm": 8.651994042616476, + "learning_rate": 1.5061956066587015e-05, + "loss": 1.1563, + "step": 5024 + }, + { + "epoch": 1.0536800167750053, + "grad_norm": 7.38758974472631, + "learning_rate": 1.5060003520484903e-05, + "loss": 0.8339, + "step": 5025 + }, + { + "epoch": 1.0538897043405326, + "grad_norm": 7.112098953810162, + "learning_rate": 1.5058050715048831e-05, + "loss": 0.8215, + "step": 5026 + }, + { + "epoch": 1.05409939190606, + "grad_norm": 6.159410725172554, + "learning_rate": 1.5056097650378888e-05, + "loss": 0.663, + "step": 5027 + }, + { + "epoch": 1.0543090794715873, + "grad_norm": 6.532512547897546, + "learning_rate": 1.5054144326575175e-05, + "loss": 0.6908, + "step": 5028 + }, + { + "epoch": 1.0545187670371148, + "grad_norm": 7.456324328260314, + "learning_rate": 1.5052190743737795e-05, + "loss": 0.785, + "step": 5029 + }, + { + "epoch": 1.054728454602642, + "grad_norm": 7.585509049714016, + "learning_rate": 1.5050236901966883e-05, + "loss": 0.8609, + "step": 5030 + }, + { + "epoch": 1.0549381421681694, + "grad_norm": 7.14029506906107, + "learning_rate": 1.5048282801362564e-05, + "loss": 0.8689, + "step": 5031 + }, + { + "epoch": 1.0551478297336967, + "grad_norm": 6.972872239996281, + "learning_rate": 1.5046328442025e-05, + "loss": 0.9703, + "step": 5032 + }, + { + "epoch": 1.0553575172992242, + "grad_norm": 6.417710484806855, + "learning_rate": 1.5044373824054348e-05, + "loss": 0.6173, + "step": 5033 + }, + { + "epoch": 1.0555672048647515, + "grad_norm": 7.509771623100593, + "learning_rate": 1.504241894755079e-05, + "loss": 1.0538, + "step": 5034 + }, + { + "epoch": 1.0557768924302788, + "grad_norm": 6.664900168344534, + "learning_rate": 1.5040463812614514e-05, + "loss": 0.7029, + "step": 5035 + }, + { + "epoch": 1.0559865799958064, + "grad_norm": 8.417358870883497, + "learning_rate": 1.5038508419345726e-05, + "loss": 0.9586, + "step": 5036 + }, + { + "epoch": 1.0561962675613337, + "grad_norm": 8.533497464112534, + "learning_rate": 1.5036552767844641e-05, + "loss": 0.9426, + "step": 5037 + }, + { + "epoch": 1.056405955126861, + "grad_norm": 8.999010077149382, + "learning_rate": 1.5034596858211496e-05, + "loss": 1.0272, + "step": 5038 + }, + { + "epoch": 1.0566156426923883, + "grad_norm": 7.897554840193791, + "learning_rate": 1.5032640690546524e-05, + "loss": 0.9727, + "step": 5039 + }, + { + "epoch": 1.0568253302579158, + "grad_norm": 6.860248678936019, + "learning_rate": 1.5030684264949993e-05, + "loss": 0.9517, + "step": 5040 + }, + { + "epoch": 1.0570350178234431, + "grad_norm": 10.069039612628313, + "learning_rate": 1.5028727581522167e-05, + "loss": 1.3065, + "step": 5041 + }, + { + "epoch": 1.0572447053889704, + "grad_norm": 6.314865799990083, + "learning_rate": 1.5026770640363328e-05, + "loss": 0.7154, + "step": 5042 + }, + { + "epoch": 1.0574543929544977, + "grad_norm": 8.158003975188912, + "learning_rate": 1.5024813441573777e-05, + "loss": 1.1046, + "step": 5043 + }, + { + "epoch": 1.0576640805200253, + "grad_norm": 6.532942462505225, + "learning_rate": 1.5022855985253823e-05, + "loss": 0.8287, + "step": 5044 + }, + { + "epoch": 1.0578737680855526, + "grad_norm": 8.8281936400201, + "learning_rate": 1.5020898271503785e-05, + "loss": 1.0699, + "step": 5045 + }, + { + "epoch": 1.0580834556510799, + "grad_norm": 7.351426957154842, + "learning_rate": 1.5018940300424005e-05, + "loss": 0.8466, + "step": 5046 + }, + { + "epoch": 1.0582931432166072, + "grad_norm": 8.57503969611497, + "learning_rate": 1.501698207211483e-05, + "loss": 1.0748, + "step": 5047 + }, + { + "epoch": 1.0585028307821347, + "grad_norm": 5.502480170194385, + "learning_rate": 1.5015023586676621e-05, + "loss": 0.6201, + "step": 5048 + }, + { + "epoch": 1.058712518347662, + "grad_norm": 6.358821766921522, + "learning_rate": 1.5013064844209754e-05, + "loss": 0.8283, + "step": 5049 + }, + { + "epoch": 1.0589222059131893, + "grad_norm": 6.7537490273196426, + "learning_rate": 1.5011105844814622e-05, + "loss": 0.7816, + "step": 5050 + }, + { + "epoch": 1.0591318934787166, + "grad_norm": 6.905860820329714, + "learning_rate": 1.5009146588591619e-05, + "loss": 1.0077, + "step": 5051 + }, + { + "epoch": 1.0593415810442441, + "grad_norm": 6.843285441417043, + "learning_rate": 1.5007187075641172e-05, + "loss": 0.8985, + "step": 5052 + }, + { + "epoch": 1.0595512686097714, + "grad_norm": 6.686383296745393, + "learning_rate": 1.5005227306063699e-05, + "loss": 0.9319, + "step": 5053 + }, + { + "epoch": 1.0597609561752988, + "grad_norm": 7.239350534445047, + "learning_rate": 1.5003267279959646e-05, + "loss": 0.7826, + "step": 5054 + }, + { + "epoch": 1.0599706437408263, + "grad_norm": 7.479503429119001, + "learning_rate": 1.5001306997429467e-05, + "loss": 0.8067, + "step": 5055 + }, + { + "epoch": 1.0601803313063536, + "grad_norm": 8.055862482582052, + "learning_rate": 1.4999346458573629e-05, + "loss": 0.7961, + "step": 5056 + }, + { + "epoch": 1.060390018871881, + "grad_norm": 6.9712515815905265, + "learning_rate": 1.4997385663492616e-05, + "loss": 0.7604, + "step": 5057 + }, + { + "epoch": 1.0605997064374082, + "grad_norm": 7.230681709654834, + "learning_rate": 1.499542461228692e-05, + "loss": 1.0907, + "step": 5058 + }, + { + "epoch": 1.0608093940029357, + "grad_norm": 7.024082197874813, + "learning_rate": 1.4993463305057048e-05, + "loss": 0.8435, + "step": 5059 + }, + { + "epoch": 1.061019081568463, + "grad_norm": 6.697058648888024, + "learning_rate": 1.4991501741903519e-05, + "loss": 0.7056, + "step": 5060 + }, + { + "epoch": 1.0612287691339903, + "grad_norm": 6.932821685224161, + "learning_rate": 1.4989539922926872e-05, + "loss": 0.8605, + "step": 5061 + }, + { + "epoch": 1.0614384566995176, + "grad_norm": 6.357173665353515, + "learning_rate": 1.4987577848227646e-05, + "loss": 0.7406, + "step": 5062 + }, + { + "epoch": 1.0616481442650452, + "grad_norm": 6.3688478718411945, + "learning_rate": 1.4985615517906409e-05, + "loss": 0.599, + "step": 5063 + }, + { + "epoch": 1.0618578318305725, + "grad_norm": 6.7535861735399285, + "learning_rate": 1.4983652932063728e-05, + "loss": 0.7684, + "step": 5064 + }, + { + "epoch": 1.0620675193960998, + "grad_norm": 7.118259871648339, + "learning_rate": 1.4981690090800189e-05, + "loss": 0.8294, + "step": 5065 + }, + { + "epoch": 1.062277206961627, + "grad_norm": 7.4079827978135295, + "learning_rate": 1.4979726994216395e-05, + "loss": 0.8515, + "step": 5066 + }, + { + "epoch": 1.0624868945271546, + "grad_norm": 5.6860573798892275, + "learning_rate": 1.4977763642412955e-05, + "loss": 0.7886, + "step": 5067 + }, + { + "epoch": 1.062696582092682, + "grad_norm": 7.193377330104058, + "learning_rate": 1.4975800035490492e-05, + "loss": 0.9114, + "step": 5068 + }, + { + "epoch": 1.0629062696582092, + "grad_norm": 6.107500476685139, + "learning_rate": 1.4973836173549652e-05, + "loss": 0.902, + "step": 5069 + }, + { + "epoch": 1.0631159572237365, + "grad_norm": 6.679559077595515, + "learning_rate": 1.497187205669108e-05, + "loss": 0.7374, + "step": 5070 + }, + { + "epoch": 1.063325644789264, + "grad_norm": 7.884939057637599, + "learning_rate": 1.4969907685015439e-05, + "loss": 0.9913, + "step": 5071 + }, + { + "epoch": 1.0635353323547914, + "grad_norm": 7.17073178191822, + "learning_rate": 1.4967943058623412e-05, + "loss": 0.8556, + "step": 5072 + }, + { + "epoch": 1.0637450199203187, + "grad_norm": 8.670535533144832, + "learning_rate": 1.4965978177615686e-05, + "loss": 0.9672, + "step": 5073 + }, + { + "epoch": 1.0639547074858462, + "grad_norm": 7.605005697878998, + "learning_rate": 1.4964013042092962e-05, + "loss": 0.7535, + "step": 5074 + }, + { + "epoch": 1.0641643950513735, + "grad_norm": 7.131441873556847, + "learning_rate": 1.4962047652155964e-05, + "loss": 0.916, + "step": 5075 + }, + { + "epoch": 1.0643740826169008, + "grad_norm": 6.243887917670286, + "learning_rate": 1.4960082007905416e-05, + "loss": 0.8018, + "step": 5076 + }, + { + "epoch": 1.0645837701824281, + "grad_norm": 8.721087733050263, + "learning_rate": 1.4958116109442063e-05, + "loss": 1.2262, + "step": 5077 + }, + { + "epoch": 1.0647934577479556, + "grad_norm": 6.6928629166504505, + "learning_rate": 1.4956149956866659e-05, + "loss": 0.9322, + "step": 5078 + }, + { + "epoch": 1.065003145313483, + "grad_norm": 8.22266428509861, + "learning_rate": 1.4954183550279975e-05, + "loss": 0.901, + "step": 5079 + }, + { + "epoch": 1.0652128328790103, + "grad_norm": 9.054091447716381, + "learning_rate": 1.4952216889782791e-05, + "loss": 0.9547, + "step": 5080 + }, + { + "epoch": 1.0654225204445376, + "grad_norm": 6.729182647674046, + "learning_rate": 1.4950249975475904e-05, + "loss": 0.7237, + "step": 5081 + }, + { + "epoch": 1.065632208010065, + "grad_norm": 7.454025735448186, + "learning_rate": 1.4948282807460116e-05, + "loss": 0.9305, + "step": 5082 + }, + { + "epoch": 1.0658418955755924, + "grad_norm": 8.24670161997894, + "learning_rate": 1.4946315385836256e-05, + "loss": 0.9521, + "step": 5083 + }, + { + "epoch": 1.0660515831411197, + "grad_norm": 6.56928389236455, + "learning_rate": 1.4944347710705151e-05, + "loss": 0.797, + "step": 5084 + }, + { + "epoch": 1.066261270706647, + "grad_norm": 7.482313168282947, + "learning_rate": 1.494237978216765e-05, + "loss": 0.9292, + "step": 5085 + }, + { + "epoch": 1.0664709582721745, + "grad_norm": 8.32617495145479, + "learning_rate": 1.4940411600324613e-05, + "loss": 0.9727, + "step": 5086 + }, + { + "epoch": 1.0666806458377018, + "grad_norm": 7.968060626336125, + "learning_rate": 1.4938443165276914e-05, + "loss": 0.7656, + "step": 5087 + }, + { + "epoch": 1.0668903334032291, + "grad_norm": 6.859134346845168, + "learning_rate": 1.4936474477125438e-05, + "loss": 0.8678, + "step": 5088 + }, + { + "epoch": 1.0671000209687564, + "grad_norm": 7.6585384264847445, + "learning_rate": 1.493450553597108e-05, + "loss": 0.9392, + "step": 5089 + }, + { + "epoch": 1.067309708534284, + "grad_norm": 7.289850863910213, + "learning_rate": 1.4932536341914761e-05, + "loss": 0.7779, + "step": 5090 + }, + { + "epoch": 1.0675193960998113, + "grad_norm": 7.924613416317872, + "learning_rate": 1.4930566895057394e-05, + "loss": 0.9352, + "step": 5091 + }, + { + "epoch": 1.0677290836653386, + "grad_norm": 5.74181356672226, + "learning_rate": 1.4928597195499924e-05, + "loss": 0.644, + "step": 5092 + }, + { + "epoch": 1.0679387712308661, + "grad_norm": 7.874459601254827, + "learning_rate": 1.4926627243343297e-05, + "loss": 0.8588, + "step": 5093 + }, + { + "epoch": 1.0681484587963934, + "grad_norm": 7.289917142427254, + "learning_rate": 1.4924657038688481e-05, + "loss": 0.8362, + "step": 5094 + }, + { + "epoch": 1.0683581463619207, + "grad_norm": 8.110344871274794, + "learning_rate": 1.492268658163645e-05, + "loss": 0.9563, + "step": 5095 + }, + { + "epoch": 1.068567833927448, + "grad_norm": 7.476859412680474, + "learning_rate": 1.4920715872288195e-05, + "loss": 0.8758, + "step": 5096 + }, + { + "epoch": 1.0687775214929756, + "grad_norm": 7.441152927512292, + "learning_rate": 1.4918744910744716e-05, + "loss": 0.9042, + "step": 5097 + }, + { + "epoch": 1.0689872090585029, + "grad_norm": 8.915858789424579, + "learning_rate": 1.491677369710703e-05, + "loss": 0.9766, + "step": 5098 + }, + { + "epoch": 1.0691968966240302, + "grad_norm": 6.306695900654438, + "learning_rate": 1.4914802231476161e-05, + "loss": 0.6519, + "step": 5099 + }, + { + "epoch": 1.0694065841895575, + "grad_norm": 6.920464718882883, + "learning_rate": 1.4912830513953158e-05, + "loss": 0.7845, + "step": 5100 + }, + { + "epoch": 1.069616271755085, + "grad_norm": 6.738444734608047, + "learning_rate": 1.4910858544639068e-05, + "loss": 0.6977, + "step": 5101 + }, + { + "epoch": 1.0698259593206123, + "grad_norm": 6.580371801305737, + "learning_rate": 1.490888632363496e-05, + "loss": 0.8193, + "step": 5102 + }, + { + "epoch": 1.0700356468861396, + "grad_norm": 7.450143709152783, + "learning_rate": 1.4906913851041914e-05, + "loss": 1.0196, + "step": 5103 + }, + { + "epoch": 1.070245334451667, + "grad_norm": 6.420303412015791, + "learning_rate": 1.4904941126961024e-05, + "loss": 0.7946, + "step": 5104 + }, + { + "epoch": 1.0704550220171944, + "grad_norm": 7.265128863222921, + "learning_rate": 1.4902968151493391e-05, + "loss": 0.929, + "step": 5105 + }, + { + "epoch": 1.0706647095827218, + "grad_norm": 6.271818568206735, + "learning_rate": 1.490099492474014e-05, + "loss": 0.7703, + "step": 5106 + }, + { + "epoch": 1.070874397148249, + "grad_norm": 7.565624944595501, + "learning_rate": 1.4899021446802395e-05, + "loss": 0.8648, + "step": 5107 + }, + { + "epoch": 1.0710840847137764, + "grad_norm": 7.460818308482355, + "learning_rate": 1.4897047717781305e-05, + "loss": 0.8472, + "step": 5108 + }, + { + "epoch": 1.071293772279304, + "grad_norm": 6.338633125848731, + "learning_rate": 1.4895073737778027e-05, + "loss": 0.8262, + "step": 5109 + }, + { + "epoch": 1.0715034598448312, + "grad_norm": 6.326284774133242, + "learning_rate": 1.4893099506893732e-05, + "loss": 0.5766, + "step": 5110 + }, + { + "epoch": 1.0717131474103585, + "grad_norm": 7.07204757708267, + "learning_rate": 1.4891125025229599e-05, + "loss": 0.8925, + "step": 5111 + }, + { + "epoch": 1.071922834975886, + "grad_norm": 6.449350049572894, + "learning_rate": 1.4889150292886825e-05, + "loss": 0.8095, + "step": 5112 + }, + { + "epoch": 1.0721325225414133, + "grad_norm": 6.610053039777937, + "learning_rate": 1.4887175309966617e-05, + "loss": 0.9039, + "step": 5113 + }, + { + "epoch": 1.0723422101069406, + "grad_norm": 6.804611416538084, + "learning_rate": 1.4885200076570199e-05, + "loss": 0.6904, + "step": 5114 + }, + { + "epoch": 1.072551897672468, + "grad_norm": 7.006213050829078, + "learning_rate": 1.4883224592798804e-05, + "loss": 0.8052, + "step": 5115 + }, + { + "epoch": 1.0727615852379955, + "grad_norm": 8.352361355838735, + "learning_rate": 1.4881248858753678e-05, + "loss": 1.0357, + "step": 5116 + }, + { + "epoch": 1.0729712728035228, + "grad_norm": 8.017781821206086, + "learning_rate": 1.4879272874536085e-05, + "loss": 0.8296, + "step": 5117 + }, + { + "epoch": 1.07318096036905, + "grad_norm": 7.85931514841387, + "learning_rate": 1.4877296640247292e-05, + "loss": 0.9281, + "step": 5118 + }, + { + "epoch": 1.0733906479345774, + "grad_norm": 6.351376863936367, + "learning_rate": 1.4875320155988591e-05, + "loss": 0.6532, + "step": 5119 + }, + { + "epoch": 1.073600335500105, + "grad_norm": 6.010153268245333, + "learning_rate": 1.4873343421861273e-05, + "loss": 0.7705, + "step": 5120 + }, + { + "epoch": 1.0738100230656322, + "grad_norm": 7.110124894825287, + "learning_rate": 1.487136643796665e-05, + "loss": 0.7884, + "step": 5121 + }, + { + "epoch": 1.0740197106311595, + "grad_norm": 6.640811958766606, + "learning_rate": 1.486938920440605e-05, + "loss": 0.7357, + "step": 5122 + }, + { + "epoch": 1.0742293981966868, + "grad_norm": 6.808465627601769, + "learning_rate": 1.4867411721280809e-05, + "loss": 0.7428, + "step": 5123 + }, + { + "epoch": 1.0744390857622144, + "grad_norm": 8.396247758169004, + "learning_rate": 1.4865433988692272e-05, + "loss": 0.9646, + "step": 5124 + }, + { + "epoch": 1.0746487733277417, + "grad_norm": 6.42436686286732, + "learning_rate": 1.486345600674181e-05, + "loss": 0.6844, + "step": 5125 + }, + { + "epoch": 1.074858460893269, + "grad_norm": 8.832085868459043, + "learning_rate": 1.4861477775530788e-05, + "loss": 1.271, + "step": 5126 + }, + { + "epoch": 1.0750681484587963, + "grad_norm": 7.417368101446979, + "learning_rate": 1.48594992951606e-05, + "loss": 0.8365, + "step": 5127 + }, + { + "epoch": 1.0752778360243238, + "grad_norm": 6.507268527853079, + "learning_rate": 1.4857520565732644e-05, + "loss": 0.8939, + "step": 5128 + }, + { + "epoch": 1.0754875235898511, + "grad_norm": 9.277868070944484, + "learning_rate": 1.4855541587348335e-05, + "loss": 1.1368, + "step": 5129 + }, + { + "epoch": 1.0756972111553784, + "grad_norm": 8.11034903735834, + "learning_rate": 1.4853562360109096e-05, + "loss": 0.7677, + "step": 5130 + }, + { + "epoch": 1.075906898720906, + "grad_norm": 8.258372820364784, + "learning_rate": 1.4851582884116371e-05, + "loss": 0.9032, + "step": 5131 + }, + { + "epoch": 1.0761165862864333, + "grad_norm": 7.014858438406556, + "learning_rate": 1.4849603159471604e-05, + "loss": 0.9535, + "step": 5132 + }, + { + "epoch": 1.0763262738519606, + "grad_norm": 7.145536286340474, + "learning_rate": 1.484762318627627e-05, + "loss": 0.999, + "step": 5133 + }, + { + "epoch": 1.0765359614174879, + "grad_norm": 8.180629374728934, + "learning_rate": 1.4845642964631834e-05, + "loss": 1.1019, + "step": 5134 + }, + { + "epoch": 1.0767456489830154, + "grad_norm": 6.21694311802288, + "learning_rate": 1.4843662494639795e-05, + "loss": 0.542, + "step": 5135 + }, + { + "epoch": 1.0769553365485427, + "grad_norm": 8.549348868479944, + "learning_rate": 1.484168177640165e-05, + "loss": 1.0476, + "step": 5136 + }, + { + "epoch": 1.07716502411407, + "grad_norm": 7.102699427719841, + "learning_rate": 1.483970081001892e-05, + "loss": 0.9204, + "step": 5137 + }, + { + "epoch": 1.0773747116795973, + "grad_norm": 7.855004380346779, + "learning_rate": 1.4837719595593125e-05, + "loss": 0.9728, + "step": 5138 + }, + { + "epoch": 1.0775843992451248, + "grad_norm": 6.527366581356953, + "learning_rate": 1.4835738133225814e-05, + "loss": 0.6538, + "step": 5139 + }, + { + "epoch": 1.0777940868106521, + "grad_norm": 7.2937101877069015, + "learning_rate": 1.4833756423018537e-05, + "loss": 0.9038, + "step": 5140 + }, + { + "epoch": 1.0780037743761794, + "grad_norm": 6.446587503930849, + "learning_rate": 1.4831774465072855e-05, + "loss": 0.6871, + "step": 5141 + }, + { + "epoch": 1.078213461941707, + "grad_norm": 6.208220676098961, + "learning_rate": 1.4829792259490356e-05, + "loss": 0.8927, + "step": 5142 + }, + { + "epoch": 1.0784231495072343, + "grad_norm": 6.886790982471816, + "learning_rate": 1.4827809806372623e-05, + "loss": 0.8756, + "step": 5143 + }, + { + "epoch": 1.0786328370727616, + "grad_norm": 5.6377185807506995, + "learning_rate": 1.4825827105821266e-05, + "loss": 0.6492, + "step": 5144 + }, + { + "epoch": 1.078842524638289, + "grad_norm": 6.048395549560448, + "learning_rate": 1.4823844157937901e-05, + "loss": 0.7291, + "step": 5145 + }, + { + "epoch": 1.0790522122038162, + "grad_norm": 7.061994221181714, + "learning_rate": 1.4821860962824157e-05, + "loss": 0.8241, + "step": 5146 + }, + { + "epoch": 1.0792618997693437, + "grad_norm": 6.403883603231528, + "learning_rate": 1.4819877520581671e-05, + "loss": 0.6336, + "step": 5147 + }, + { + "epoch": 1.079471587334871, + "grad_norm": 8.089186320721302, + "learning_rate": 1.4817893831312107e-05, + "loss": 0.9253, + "step": 5148 + }, + { + "epoch": 1.0796812749003983, + "grad_norm": 7.199145283844229, + "learning_rate": 1.4815909895117127e-05, + "loss": 0.7524, + "step": 5149 + }, + { + "epoch": 1.0798909624659259, + "grad_norm": 6.664254077754818, + "learning_rate": 1.4813925712098409e-05, + "loss": 0.6835, + "step": 5150 + }, + { + "epoch": 1.0801006500314532, + "grad_norm": 7.2141565789241, + "learning_rate": 1.4811941282357652e-05, + "loss": 0.7738, + "step": 5151 + }, + { + "epoch": 1.0803103375969805, + "grad_norm": 6.6925461600405365, + "learning_rate": 1.4809956605996561e-05, + "loss": 0.8691, + "step": 5152 + }, + { + "epoch": 1.0805200251625078, + "grad_norm": 7.709772242966372, + "learning_rate": 1.4807971683116844e-05, + "loss": 0.8742, + "step": 5153 + }, + { + "epoch": 1.0807297127280353, + "grad_norm": 7.215700190095787, + "learning_rate": 1.4805986513820245e-05, + "loss": 0.7748, + "step": 5154 + }, + { + "epoch": 1.0809394002935626, + "grad_norm": 6.712058906762827, + "learning_rate": 1.4804001098208498e-05, + "loss": 0.6353, + "step": 5155 + }, + { + "epoch": 1.08114908785909, + "grad_norm": 7.729957184748364, + "learning_rate": 1.4802015436383366e-05, + "loss": 0.9433, + "step": 5156 + }, + { + "epoch": 1.0813587754246172, + "grad_norm": 6.706142351914358, + "learning_rate": 1.4800029528446614e-05, + "loss": 0.7706, + "step": 5157 + }, + { + "epoch": 1.0815684629901448, + "grad_norm": 6.56618578158904, + "learning_rate": 1.479804337450002e-05, + "loss": 0.6999, + "step": 5158 + }, + { + "epoch": 1.081778150555672, + "grad_norm": 8.318220758100463, + "learning_rate": 1.4796056974645383e-05, + "loss": 0.9899, + "step": 5159 + }, + { + "epoch": 1.0819878381211994, + "grad_norm": 8.57235735455009, + "learning_rate": 1.479407032898451e-05, + "loss": 0.9584, + "step": 5160 + }, + { + "epoch": 1.082197525686727, + "grad_norm": 7.918182685146835, + "learning_rate": 1.4792083437619213e-05, + "loss": 0.8264, + "step": 5161 + }, + { + "epoch": 1.0824072132522542, + "grad_norm": 8.221571827235467, + "learning_rate": 1.4790096300651333e-05, + "loss": 0.9761, + "step": 5162 + }, + { + "epoch": 1.0826169008177815, + "grad_norm": 6.368757183741934, + "learning_rate": 1.4788108918182709e-05, + "loss": 0.6081, + "step": 5163 + }, + { + "epoch": 1.0828265883833088, + "grad_norm": 7.362813489763786, + "learning_rate": 1.4786121290315196e-05, + "loss": 0.9601, + "step": 5164 + }, + { + "epoch": 1.0830362759488363, + "grad_norm": 7.5589034034327325, + "learning_rate": 1.4784133417150669e-05, + "loss": 0.9928, + "step": 5165 + }, + { + "epoch": 1.0832459635143636, + "grad_norm": 7.269786994971708, + "learning_rate": 1.4782145298791003e-05, + "loss": 0.8423, + "step": 5166 + }, + { + "epoch": 1.083455651079891, + "grad_norm": 7.652543355106625, + "learning_rate": 1.4780156935338096e-05, + "loss": 1.0243, + "step": 5167 + }, + { + "epoch": 1.0836653386454183, + "grad_norm": 7.418779390896711, + "learning_rate": 1.4778168326893858e-05, + "loss": 0.7405, + "step": 5168 + }, + { + "epoch": 1.0838750262109458, + "grad_norm": 9.027682976846481, + "learning_rate": 1.4776179473560205e-05, + "loss": 1.1147, + "step": 5169 + }, + { + "epoch": 1.084084713776473, + "grad_norm": 7.144663366302767, + "learning_rate": 1.477419037543907e-05, + "loss": 0.8702, + "step": 5170 + }, + { + "epoch": 1.0842944013420004, + "grad_norm": 6.30187481438015, + "learning_rate": 1.4772201032632398e-05, + "loss": 0.91, + "step": 5171 + }, + { + "epoch": 1.0845040889075277, + "grad_norm": 8.886982425089645, + "learning_rate": 1.4770211445242147e-05, + "loss": 1.0972, + "step": 5172 + }, + { + "epoch": 1.0847137764730552, + "grad_norm": 6.527057871445205, + "learning_rate": 1.4768221613370285e-05, + "loss": 0.691, + "step": 5173 + }, + { + "epoch": 1.0849234640385825, + "grad_norm": 7.045321711506494, + "learning_rate": 1.4766231537118795e-05, + "loss": 0.7956, + "step": 5174 + }, + { + "epoch": 1.0851331516041098, + "grad_norm": 7.2863804417603575, + "learning_rate": 1.4764241216589674e-05, + "loss": 0.7647, + "step": 5175 + }, + { + "epoch": 1.0853428391696371, + "grad_norm": 6.838303966036686, + "learning_rate": 1.4762250651884924e-05, + "loss": 0.7897, + "step": 5176 + }, + { + "epoch": 1.0855525267351647, + "grad_norm": 7.547915679396741, + "learning_rate": 1.4760259843106573e-05, + "loss": 0.9521, + "step": 5177 + }, + { + "epoch": 1.085762214300692, + "grad_norm": 7.7508941531621645, + "learning_rate": 1.4758268790356644e-05, + "loss": 0.8405, + "step": 5178 + }, + { + "epoch": 1.0859719018662193, + "grad_norm": 6.39595317702409, + "learning_rate": 1.475627749373719e-05, + "loss": 0.8321, + "step": 5179 + }, + { + "epoch": 1.0861815894317468, + "grad_norm": 7.735831365356152, + "learning_rate": 1.4754285953350262e-05, + "loss": 0.878, + "step": 5180 + }, + { + "epoch": 1.0863912769972741, + "grad_norm": 6.805298123078015, + "learning_rate": 1.4752294169297938e-05, + "loss": 0.9383, + "step": 5181 + }, + { + "epoch": 1.0866009645628014, + "grad_norm": 6.266710651964672, + "learning_rate": 1.4750302141682292e-05, + "loss": 0.8925, + "step": 5182 + }, + { + "epoch": 1.0868106521283287, + "grad_norm": 6.362613932046409, + "learning_rate": 1.4748309870605422e-05, + "loss": 0.905, + "step": 5183 + }, + { + "epoch": 1.0870203396938563, + "grad_norm": 8.217743508051532, + "learning_rate": 1.4746317356169436e-05, + "loss": 1.0756, + "step": 5184 + }, + { + "epoch": 1.0872300272593836, + "grad_norm": 7.1576847255600615, + "learning_rate": 1.4744324598476454e-05, + "loss": 0.8169, + "step": 5185 + }, + { + "epoch": 1.0874397148249109, + "grad_norm": 6.3637126532829775, + "learning_rate": 1.4742331597628608e-05, + "loss": 0.854, + "step": 5186 + }, + { + "epoch": 1.0876494023904382, + "grad_norm": 6.6180057624909505, + "learning_rate": 1.474033835372804e-05, + "loss": 0.7927, + "step": 5187 + }, + { + "epoch": 1.0878590899559657, + "grad_norm": 7.571565965563071, + "learning_rate": 1.4738344866876913e-05, + "loss": 0.8628, + "step": 5188 + }, + { + "epoch": 1.088068777521493, + "grad_norm": 7.450952980155977, + "learning_rate": 1.4736351137177392e-05, + "loss": 0.8944, + "step": 5189 + }, + { + "epoch": 1.0882784650870203, + "grad_norm": 7.1631545659673606, + "learning_rate": 1.4734357164731659e-05, + "loss": 0.8991, + "step": 5190 + }, + { + "epoch": 1.0884881526525476, + "grad_norm": 5.818817333056958, + "learning_rate": 1.4732362949641911e-05, + "loss": 0.6424, + "step": 5191 + }, + { + "epoch": 1.0886978402180751, + "grad_norm": 7.154352660564363, + "learning_rate": 1.4730368492010354e-05, + "loss": 0.8967, + "step": 5192 + }, + { + "epoch": 1.0889075277836024, + "grad_norm": 7.427578493378902, + "learning_rate": 1.472837379193921e-05, + "loss": 0.8014, + "step": 5193 + }, + { + "epoch": 1.0891172153491298, + "grad_norm": 5.973005129926664, + "learning_rate": 1.4726378849530705e-05, + "loss": 0.7712, + "step": 5194 + }, + { + "epoch": 1.089326902914657, + "grad_norm": 6.170650200998096, + "learning_rate": 1.4724383664887088e-05, + "loss": 0.7955, + "step": 5195 + }, + { + "epoch": 1.0895365904801846, + "grad_norm": 8.53144146762034, + "learning_rate": 1.4722388238110611e-05, + "loss": 1.2179, + "step": 5196 + }, + { + "epoch": 1.089746278045712, + "grad_norm": 7.709237393908878, + "learning_rate": 1.472039256930355e-05, + "loss": 0.9327, + "step": 5197 + }, + { + "epoch": 1.0899559656112392, + "grad_norm": 6.428577573680235, + "learning_rate": 1.4718396658568181e-05, + "loss": 0.8474, + "step": 5198 + }, + { + "epoch": 1.0901656531767667, + "grad_norm": 8.160447321823169, + "learning_rate": 1.4716400506006798e-05, + "loss": 0.9434, + "step": 5199 + }, + { + "epoch": 1.090375340742294, + "grad_norm": 8.223173609969216, + "learning_rate": 1.4714404111721715e-05, + "loss": 0.9066, + "step": 5200 + }, + { + "epoch": 1.0905850283078213, + "grad_norm": 6.716984493450535, + "learning_rate": 1.4712407475815238e-05, + "loss": 0.8658, + "step": 5201 + }, + { + "epoch": 1.0907947158733486, + "grad_norm": 6.13662592434018, + "learning_rate": 1.4710410598389708e-05, + "loss": 0.7111, + "step": 5202 + }, + { + "epoch": 1.0910044034388762, + "grad_norm": 7.939494952599585, + "learning_rate": 1.4708413479547463e-05, + "loss": 0.9595, + "step": 5203 + }, + { + "epoch": 1.0912140910044035, + "grad_norm": 8.844054160009875, + "learning_rate": 1.4706416119390863e-05, + "loss": 1.0835, + "step": 5204 + }, + { + "epoch": 1.0914237785699308, + "grad_norm": 7.118598590697339, + "learning_rate": 1.4704418518022273e-05, + "loss": 0.8725, + "step": 5205 + }, + { + "epoch": 1.091633466135458, + "grad_norm": 6.270520499430026, + "learning_rate": 1.4702420675544073e-05, + "loss": 0.8771, + "step": 5206 + }, + { + "epoch": 1.0918431537009856, + "grad_norm": 5.8874066641634135, + "learning_rate": 1.470042259205866e-05, + "loss": 0.9082, + "step": 5207 + }, + { + "epoch": 1.092052841266513, + "grad_norm": 8.380986299400435, + "learning_rate": 1.4698424267668436e-05, + "loss": 1.1283, + "step": 5208 + }, + { + "epoch": 1.0922625288320402, + "grad_norm": 6.494949379374786, + "learning_rate": 1.4696425702475818e-05, + "loss": 0.7135, + "step": 5209 + }, + { + "epoch": 1.0924722163975675, + "grad_norm": 8.693236211299794, + "learning_rate": 1.4694426896583239e-05, + "loss": 1.0204, + "step": 5210 + }, + { + "epoch": 1.092681903963095, + "grad_norm": 7.31116612227341, + "learning_rate": 1.4692427850093136e-05, + "loss": 0.8029, + "step": 5211 + }, + { + "epoch": 1.0928915915286224, + "grad_norm": 7.053077825536181, + "learning_rate": 1.4690428563107972e-05, + "loss": 0.7885, + "step": 5212 + }, + { + "epoch": 1.0931012790941497, + "grad_norm": 7.103941672845228, + "learning_rate": 1.4688429035730207e-05, + "loss": 0.784, + "step": 5213 + }, + { + "epoch": 1.093310966659677, + "grad_norm": 7.184952343470437, + "learning_rate": 1.4686429268062321e-05, + "loss": 1.0048, + "step": 5214 + }, + { + "epoch": 1.0935206542252045, + "grad_norm": 8.715002074528302, + "learning_rate": 1.4684429260206808e-05, + "loss": 1.1955, + "step": 5215 + }, + { + "epoch": 1.0937303417907318, + "grad_norm": 7.76117673401112, + "learning_rate": 1.468242901226617e-05, + "loss": 1.0695, + "step": 5216 + }, + { + "epoch": 1.0939400293562591, + "grad_norm": 7.31097933631295, + "learning_rate": 1.4680428524342924e-05, + "loss": 0.9259, + "step": 5217 + }, + { + "epoch": 1.0941497169217866, + "grad_norm": 7.761901331848131, + "learning_rate": 1.4678427796539598e-05, + "loss": 1.0372, + "step": 5218 + }, + { + "epoch": 1.094359404487314, + "grad_norm": 6.255989377055225, + "learning_rate": 1.4676426828958735e-05, + "loss": 0.8217, + "step": 5219 + }, + { + "epoch": 1.0945690920528413, + "grad_norm": 6.757370022526783, + "learning_rate": 1.4674425621702884e-05, + "loss": 0.7647, + "step": 5220 + }, + { + "epoch": 1.0947787796183686, + "grad_norm": 7.736046708190815, + "learning_rate": 1.4672424174874613e-05, + "loss": 1.0537, + "step": 5221 + }, + { + "epoch": 1.094988467183896, + "grad_norm": 6.099275543820175, + "learning_rate": 1.4670422488576501e-05, + "loss": 0.8327, + "step": 5222 + }, + { + "epoch": 1.0951981547494234, + "grad_norm": 7.505195641683392, + "learning_rate": 1.4668420562911134e-05, + "loss": 0.9174, + "step": 5223 + }, + { + "epoch": 1.0954078423149507, + "grad_norm": 5.96454124723484, + "learning_rate": 1.4666418397981117e-05, + "loss": 0.7314, + "step": 5224 + }, + { + "epoch": 1.095617529880478, + "grad_norm": 6.144113544639926, + "learning_rate": 1.4664415993889062e-05, + "loss": 0.8219, + "step": 5225 + }, + { + "epoch": 1.0958272174460055, + "grad_norm": 8.483985177326398, + "learning_rate": 1.4662413350737598e-05, + "loss": 1.0613, + "step": 5226 + }, + { + "epoch": 1.0960369050115328, + "grad_norm": 6.104065812146913, + "learning_rate": 1.4660410468629367e-05, + "loss": 0.7293, + "step": 5227 + }, + { + "epoch": 1.0962465925770601, + "grad_norm": 9.169389924208907, + "learning_rate": 1.465840734766701e-05, + "loss": 1.052, + "step": 5228 + }, + { + "epoch": 1.0964562801425874, + "grad_norm": 7.41990455922644, + "learning_rate": 1.4656403987953203e-05, + "loss": 0.8779, + "step": 5229 + }, + { + "epoch": 1.096665967708115, + "grad_norm": 6.716426703282258, + "learning_rate": 1.465440038959061e-05, + "loss": 0.9721, + "step": 5230 + }, + { + "epoch": 1.0968756552736423, + "grad_norm": 8.3008178030219, + "learning_rate": 1.4652396552681927e-05, + "loss": 0.9549, + "step": 5231 + }, + { + "epoch": 1.0970853428391696, + "grad_norm": 7.444636334720324, + "learning_rate": 1.4650392477329853e-05, + "loss": 0.9387, + "step": 5232 + }, + { + "epoch": 1.097295030404697, + "grad_norm": 6.770186121830977, + "learning_rate": 1.46483881636371e-05, + "loss": 0.8118, + "step": 5233 + }, + { + "epoch": 1.0975047179702244, + "grad_norm": 6.425003740410699, + "learning_rate": 1.4646383611706386e-05, + "loss": 0.8559, + "step": 5234 + }, + { + "epoch": 1.0977144055357517, + "grad_norm": 6.9036210920477385, + "learning_rate": 1.4644378821640458e-05, + "loss": 0.833, + "step": 5235 + }, + { + "epoch": 1.097924093101279, + "grad_norm": 6.975999700655976, + "learning_rate": 1.4642373793542058e-05, + "loss": 0.9836, + "step": 5236 + }, + { + "epoch": 1.0981337806668066, + "grad_norm": 6.999536986183994, + "learning_rate": 1.4640368527513946e-05, + "loss": 0.728, + "step": 5237 + }, + { + "epoch": 1.0983434682323339, + "grad_norm": 7.368083529477797, + "learning_rate": 1.4638363023658903e-05, + "loss": 0.9781, + "step": 5238 + }, + { + "epoch": 1.0985531557978612, + "grad_norm": 6.90949812596724, + "learning_rate": 1.463635728207971e-05, + "loss": 0.6899, + "step": 5239 + }, + { + "epoch": 1.0987628433633885, + "grad_norm": 6.3207946313184955, + "learning_rate": 1.4634351302879162e-05, + "loss": 0.9654, + "step": 5240 + }, + { + "epoch": 1.098972530928916, + "grad_norm": 6.780499414301449, + "learning_rate": 1.4632345086160076e-05, + "loss": 0.8072, + "step": 5241 + }, + { + "epoch": 1.0991822184944433, + "grad_norm": 8.34826574657584, + "learning_rate": 1.4630338632025267e-05, + "loss": 0.9392, + "step": 5242 + }, + { + "epoch": 1.0993919060599706, + "grad_norm": 7.0572576226695425, + "learning_rate": 1.4628331940577574e-05, + "loss": 0.794, + "step": 5243 + }, + { + "epoch": 1.099601593625498, + "grad_norm": 8.290967643697805, + "learning_rate": 1.462632501191984e-05, + "loss": 0.8568, + "step": 5244 + }, + { + "epoch": 1.0998112811910254, + "grad_norm": 8.058207723623731, + "learning_rate": 1.4624317846154924e-05, + "loss": 0.8509, + "step": 5245 + }, + { + "epoch": 1.1000209687565528, + "grad_norm": 5.431495949033383, + "learning_rate": 1.4622310443385701e-05, + "loss": 0.6222, + "step": 5246 + }, + { + "epoch": 1.10023065632208, + "grad_norm": 9.151745998922761, + "learning_rate": 1.4620302803715049e-05, + "loss": 0.9728, + "step": 5247 + }, + { + "epoch": 1.1004403438876074, + "grad_norm": 8.405932375249979, + "learning_rate": 1.4618294927245863e-05, + "loss": 1.1251, + "step": 5248 + }, + { + "epoch": 1.100650031453135, + "grad_norm": 6.491227088105822, + "learning_rate": 1.4616286814081058e-05, + "loss": 0.7245, + "step": 5249 + }, + { + "epoch": 1.1008597190186622, + "grad_norm": 6.776466609688915, + "learning_rate": 1.4614278464323544e-05, + "loss": 0.9516, + "step": 5250 + }, + { + "epoch": 1.1010694065841895, + "grad_norm": 7.034400417074577, + "learning_rate": 1.4612269878076253e-05, + "loss": 0.9885, + "step": 5251 + }, + { + "epoch": 1.1012790941497168, + "grad_norm": 6.616594478849876, + "learning_rate": 1.4610261055442135e-05, + "loss": 0.751, + "step": 5252 + }, + { + "epoch": 1.1014887817152443, + "grad_norm": 8.52526546355809, + "learning_rate": 1.4608251996524138e-05, + "loss": 0.746, + "step": 5253 + }, + { + "epoch": 1.1016984692807716, + "grad_norm": 6.92929062894921, + "learning_rate": 1.4606242701425237e-05, + "loss": 1.0006, + "step": 5254 + }, + { + "epoch": 1.101908156846299, + "grad_norm": 7.675093962691405, + "learning_rate": 1.4604233170248408e-05, + "loss": 0.8741, + "step": 5255 + }, + { + "epoch": 1.1021178444118265, + "grad_norm": 6.685026314885749, + "learning_rate": 1.4602223403096642e-05, + "loss": 0.8539, + "step": 5256 + }, + { + "epoch": 1.1023275319773538, + "grad_norm": 7.414838832903433, + "learning_rate": 1.4600213400072945e-05, + "loss": 0.8243, + "step": 5257 + }, + { + "epoch": 1.102537219542881, + "grad_norm": 7.188339848015743, + "learning_rate": 1.4598203161280333e-05, + "loss": 0.5663, + "step": 5258 + }, + { + "epoch": 1.1027469071084084, + "grad_norm": 5.8988011985400695, + "learning_rate": 1.4596192686821833e-05, + "loss": 0.7032, + "step": 5259 + }, + { + "epoch": 1.102956594673936, + "grad_norm": 7.410839787791522, + "learning_rate": 1.4594181976800488e-05, + "loss": 0.8251, + "step": 5260 + }, + { + "epoch": 1.1031662822394632, + "grad_norm": 7.003744012721617, + "learning_rate": 1.4592171031319344e-05, + "loss": 0.8218, + "step": 5261 + }, + { + "epoch": 1.1033759698049905, + "grad_norm": 7.960675257705126, + "learning_rate": 1.4590159850481475e-05, + "loss": 0.965, + "step": 5262 + }, + { + "epoch": 1.1035856573705178, + "grad_norm": 6.403035492792724, + "learning_rate": 1.4588148434389949e-05, + "loss": 0.6962, + "step": 5263 + }, + { + "epoch": 1.1037953449360454, + "grad_norm": 5.879248042642327, + "learning_rate": 1.4586136783147862e-05, + "loss": 0.7019, + "step": 5264 + }, + { + "epoch": 1.1040050325015727, + "grad_norm": 5.625515725249563, + "learning_rate": 1.4584124896858307e-05, + "loss": 0.5377, + "step": 5265 + }, + { + "epoch": 1.1042147200671, + "grad_norm": 6.881322513524374, + "learning_rate": 1.4582112775624402e-05, + "loss": 0.8795, + "step": 5266 + }, + { + "epoch": 1.1044244076326275, + "grad_norm": 6.620952160764525, + "learning_rate": 1.4580100419549267e-05, + "loss": 0.7854, + "step": 5267 + }, + { + "epoch": 1.1046340951981548, + "grad_norm": 6.3558135097406225, + "learning_rate": 1.4578087828736044e-05, + "loss": 0.7925, + "step": 5268 + }, + { + "epoch": 1.1048437827636821, + "grad_norm": 7.194166795475136, + "learning_rate": 1.457607500328788e-05, + "loss": 0.9547, + "step": 5269 + }, + { + "epoch": 1.1050534703292094, + "grad_norm": 7.974793152439824, + "learning_rate": 1.4574061943307935e-05, + "loss": 1.1181, + "step": 5270 + }, + { + "epoch": 1.1052631578947367, + "grad_norm": 7.298391176730906, + "learning_rate": 1.4572048648899381e-05, + "loss": 0.7858, + "step": 5271 + }, + { + "epoch": 1.1054728454602643, + "grad_norm": 7.120882100733394, + "learning_rate": 1.4570035120165407e-05, + "loss": 1.0643, + "step": 5272 + }, + { + "epoch": 1.1056825330257916, + "grad_norm": 6.205183015980851, + "learning_rate": 1.4568021357209204e-05, + "loss": 0.8022, + "step": 5273 + }, + { + "epoch": 1.1058922205913189, + "grad_norm": 6.7955672747613, + "learning_rate": 1.4566007360133982e-05, + "loss": 0.8241, + "step": 5274 + }, + { + "epoch": 1.1061019081568464, + "grad_norm": 6.68587754134207, + "learning_rate": 1.4563993129042966e-05, + "loss": 0.8863, + "step": 5275 + }, + { + "epoch": 1.1063115957223737, + "grad_norm": 8.125446052411736, + "learning_rate": 1.4561978664039384e-05, + "loss": 0.8918, + "step": 5276 + }, + { + "epoch": 1.106521283287901, + "grad_norm": 7.01196137170565, + "learning_rate": 1.4559963965226486e-05, + "loss": 0.6808, + "step": 5277 + }, + { + "epoch": 1.1067309708534283, + "grad_norm": 6.267591726508855, + "learning_rate": 1.4557949032707526e-05, + "loss": 0.855, + "step": 5278 + }, + { + "epoch": 1.1069406584189558, + "grad_norm": 7.2587708398946145, + "learning_rate": 1.4555933866585772e-05, + "loss": 0.7654, + "step": 5279 + }, + { + "epoch": 1.1071503459844831, + "grad_norm": 6.799538260492512, + "learning_rate": 1.4553918466964502e-05, + "loss": 0.7683, + "step": 5280 + }, + { + "epoch": 1.1073600335500104, + "grad_norm": 7.159715032224461, + "learning_rate": 1.4551902833947018e-05, + "loss": 0.9596, + "step": 5281 + }, + { + "epoch": 1.1075697211155378, + "grad_norm": 7.685335171896852, + "learning_rate": 1.4549886967636616e-05, + "loss": 0.8096, + "step": 5282 + }, + { + "epoch": 1.1077794086810653, + "grad_norm": 6.95621283818804, + "learning_rate": 1.4547870868136615e-05, + "loss": 0.8672, + "step": 5283 + }, + { + "epoch": 1.1079890962465926, + "grad_norm": 6.845828659223529, + "learning_rate": 1.4545854535550345e-05, + "loss": 0.8717, + "step": 5284 + }, + { + "epoch": 1.10819878381212, + "grad_norm": 7.405655499532415, + "learning_rate": 1.4543837969981147e-05, + "loss": 0.8813, + "step": 5285 + }, + { + "epoch": 1.1084084713776474, + "grad_norm": 6.038162646175789, + "learning_rate": 1.4541821171532369e-05, + "loss": 0.6046, + "step": 5286 + }, + { + "epoch": 1.1086181589431747, + "grad_norm": 7.125991913803899, + "learning_rate": 1.4539804140307381e-05, + "loss": 0.8772, + "step": 5287 + }, + { + "epoch": 1.108827846508702, + "grad_norm": 7.057075598207975, + "learning_rate": 1.4537786876409556e-05, + "loss": 0.8591, + "step": 5288 + }, + { + "epoch": 1.1090375340742293, + "grad_norm": 6.4348006207981445, + "learning_rate": 1.4535769379942284e-05, + "loss": 0.6179, + "step": 5289 + }, + { + "epoch": 1.1092472216397569, + "grad_norm": 7.987256252031424, + "learning_rate": 1.4533751651008963e-05, + "loss": 0.9093, + "step": 5290 + }, + { + "epoch": 1.1094569092052842, + "grad_norm": 6.985077084092817, + "learning_rate": 1.4531733689713008e-05, + "loss": 0.8381, + "step": 5291 + }, + { + "epoch": 1.1096665967708115, + "grad_norm": 7.102520691559278, + "learning_rate": 1.452971549615784e-05, + "loss": 0.829, + "step": 5292 + }, + { + "epoch": 1.1098762843363388, + "grad_norm": 8.727475404800678, + "learning_rate": 1.4527697070446898e-05, + "loss": 1.057, + "step": 5293 + }, + { + "epoch": 1.1100859719018663, + "grad_norm": 7.131393791736429, + "learning_rate": 1.4525678412683627e-05, + "loss": 0.7068, + "step": 5294 + }, + { + "epoch": 1.1102956594673936, + "grad_norm": 6.492703477551265, + "learning_rate": 1.4523659522971487e-05, + "loss": 0.6083, + "step": 5295 + }, + { + "epoch": 1.110505347032921, + "grad_norm": 6.904946199420258, + "learning_rate": 1.4521640401413953e-05, + "loss": 0.6057, + "step": 5296 + }, + { + "epoch": 1.1107150345984482, + "grad_norm": 7.197071066598601, + "learning_rate": 1.4519621048114505e-05, + "loss": 0.8344, + "step": 5297 + }, + { + "epoch": 1.1109247221639758, + "grad_norm": 7.766300606169771, + "learning_rate": 1.4517601463176639e-05, + "loss": 0.9885, + "step": 5298 + }, + { + "epoch": 1.111134409729503, + "grad_norm": 6.980842010656182, + "learning_rate": 1.4515581646703862e-05, + "loss": 0.928, + "step": 5299 + }, + { + "epoch": 1.1113440972950304, + "grad_norm": 10.098313373677547, + "learning_rate": 1.4513561598799697e-05, + "loss": 0.9939, + "step": 5300 + }, + { + "epoch": 1.1115537848605577, + "grad_norm": 7.671926527131913, + "learning_rate": 1.4511541319567667e-05, + "loss": 0.8654, + "step": 5301 + }, + { + "epoch": 1.1117634724260852, + "grad_norm": 7.036528098010894, + "learning_rate": 1.4509520809111323e-05, + "loss": 0.7219, + "step": 5302 + }, + { + "epoch": 1.1119731599916125, + "grad_norm": 7.373465532454491, + "learning_rate": 1.4507500067534214e-05, + "loss": 0.8878, + "step": 5303 + }, + { + "epoch": 1.1121828475571398, + "grad_norm": 7.247153458904141, + "learning_rate": 1.4505479094939911e-05, + "loss": 0.8321, + "step": 5304 + }, + { + "epoch": 1.1123925351226673, + "grad_norm": 7.82869427456522, + "learning_rate": 1.4503457891431988e-05, + "loss": 0.865, + "step": 5305 + }, + { + "epoch": 1.1126022226881946, + "grad_norm": 7.936958513929337, + "learning_rate": 1.4501436457114037e-05, + "loss": 0.8269, + "step": 5306 + }, + { + "epoch": 1.112811910253722, + "grad_norm": 6.087566469292746, + "learning_rate": 1.4499414792089661e-05, + "loss": 0.7164, + "step": 5307 + }, + { + "epoch": 1.1130215978192493, + "grad_norm": 5.646810120677337, + "learning_rate": 1.4497392896462477e-05, + "loss": 0.6004, + "step": 5308 + }, + { + "epoch": 1.1132312853847768, + "grad_norm": 6.788673545298981, + "learning_rate": 1.4495370770336099e-05, + "loss": 0.8651, + "step": 5309 + }, + { + "epoch": 1.113440972950304, + "grad_norm": 7.420968616252654, + "learning_rate": 1.4493348413814179e-05, + "loss": 0.9926, + "step": 5310 + }, + { + "epoch": 1.1136506605158314, + "grad_norm": 6.840141496735218, + "learning_rate": 1.4491325827000355e-05, + "loss": 0.7206, + "step": 5311 + }, + { + "epoch": 1.1138603480813587, + "grad_norm": 6.225287715575675, + "learning_rate": 1.4489303009998294e-05, + "loss": 0.6719, + "step": 5312 + }, + { + "epoch": 1.1140700356468862, + "grad_norm": 6.826950050988913, + "learning_rate": 1.4487279962911666e-05, + "loss": 0.8167, + "step": 5313 + }, + { + "epoch": 1.1142797232124135, + "grad_norm": 7.252652795903682, + "learning_rate": 1.4485256685844163e-05, + "loss": 0.9985, + "step": 5314 + }, + { + "epoch": 1.1144894107779408, + "grad_norm": 7.536198834600522, + "learning_rate": 1.4483233178899468e-05, + "loss": 1.0543, + "step": 5315 + }, + { + "epoch": 1.1146990983434681, + "grad_norm": 7.116797576538685, + "learning_rate": 1.4481209442181301e-05, + "loss": 0.7723, + "step": 5316 + }, + { + "epoch": 1.1149087859089957, + "grad_norm": 6.140798938753212, + "learning_rate": 1.4479185475793378e-05, + "loss": 0.6864, + "step": 5317 + }, + { + "epoch": 1.115118473474523, + "grad_norm": 6.785531510740913, + "learning_rate": 1.4477161279839429e-05, + "loss": 0.7987, + "step": 5318 + }, + { + "epoch": 1.1153281610400503, + "grad_norm": 6.927082267142096, + "learning_rate": 1.44751368544232e-05, + "loss": 0.836, + "step": 5319 + }, + { + "epoch": 1.1155378486055776, + "grad_norm": 5.327226625807522, + "learning_rate": 1.4473112199648447e-05, + "loss": 0.6378, + "step": 5320 + }, + { + "epoch": 1.1157475361711051, + "grad_norm": 7.829926846860998, + "learning_rate": 1.447108731561893e-05, + "loss": 1.0554, + "step": 5321 + }, + { + "epoch": 1.1159572237366324, + "grad_norm": 7.559542275088372, + "learning_rate": 1.446906220243844e-05, + "loss": 1.0161, + "step": 5322 + }, + { + "epoch": 1.1161669113021597, + "grad_norm": 9.054593005111048, + "learning_rate": 1.4467036860210756e-05, + "loss": 0.91, + "step": 5323 + }, + { + "epoch": 1.1163765988676873, + "grad_norm": 7.6273218046261615, + "learning_rate": 1.4465011289039688e-05, + "loss": 0.885, + "step": 5324 + }, + { + "epoch": 1.1165862864332146, + "grad_norm": 6.8352987778493235, + "learning_rate": 1.4462985489029047e-05, + "loss": 0.8547, + "step": 5325 + }, + { + "epoch": 1.1167959739987419, + "grad_norm": 8.296011463134029, + "learning_rate": 1.4460959460282657e-05, + "loss": 0.9299, + "step": 5326 + }, + { + "epoch": 1.1170056615642692, + "grad_norm": 8.584075025442193, + "learning_rate": 1.4458933202904359e-05, + "loss": 1.1658, + "step": 5327 + }, + { + "epoch": 1.1172153491297967, + "grad_norm": 6.960218398308803, + "learning_rate": 1.4456906716998001e-05, + "loss": 0.9651, + "step": 5328 + }, + { + "epoch": 1.117425036695324, + "grad_norm": 7.031369510005853, + "learning_rate": 1.4454880002667447e-05, + "loss": 0.7441, + "step": 5329 + }, + { + "epoch": 1.1176347242608513, + "grad_norm": 8.01794853655793, + "learning_rate": 1.4452853060016558e-05, + "loss": 0.7905, + "step": 5330 + }, + { + "epoch": 1.1178444118263786, + "grad_norm": 7.262109132423744, + "learning_rate": 1.4450825889149234e-05, + "loss": 0.7935, + "step": 5331 + }, + { + "epoch": 1.1180540993919061, + "grad_norm": 6.354503037423062, + "learning_rate": 1.444879849016936e-05, + "loss": 0.7458, + "step": 5332 + }, + { + "epoch": 1.1182637869574334, + "grad_norm": 7.212829643299545, + "learning_rate": 1.4446770863180848e-05, + "loss": 0.7553, + "step": 5333 + }, + { + "epoch": 1.1184734745229608, + "grad_norm": 7.188151678701964, + "learning_rate": 1.4444743008287616e-05, + "loss": 0.7569, + "step": 5334 + }, + { + "epoch": 1.118683162088488, + "grad_norm": 8.4531894189342, + "learning_rate": 1.44427149255936e-05, + "loss": 1.016, + "step": 5335 + }, + { + "epoch": 1.1188928496540156, + "grad_norm": 7.2243409701215064, + "learning_rate": 1.4440686615202732e-05, + "loss": 1.0894, + "step": 5336 + }, + { + "epoch": 1.119102537219543, + "grad_norm": 7.06649757413754, + "learning_rate": 1.4438658077218979e-05, + "loss": 0.7815, + "step": 5337 + }, + { + "epoch": 1.1193122247850702, + "grad_norm": 6.367088561664283, + "learning_rate": 1.4436629311746295e-05, + "loss": 0.6352, + "step": 5338 + }, + { + "epoch": 1.1195219123505975, + "grad_norm": 5.582258342998007, + "learning_rate": 1.443460031888867e-05, + "loss": 0.6694, + "step": 5339 + }, + { + "epoch": 1.119731599916125, + "grad_norm": 9.142133339711851, + "learning_rate": 1.4432571098750083e-05, + "loss": 1.0844, + "step": 5340 + }, + { + "epoch": 1.1199412874816523, + "grad_norm": 7.036624129810012, + "learning_rate": 1.4430541651434541e-05, + "loss": 0.7608, + "step": 5341 + }, + { + "epoch": 1.1201509750471796, + "grad_norm": 6.268651703016797, + "learning_rate": 1.4428511977046056e-05, + "loss": 0.786, + "step": 5342 + }, + { + "epoch": 1.1203606626127072, + "grad_norm": 7.508689757916, + "learning_rate": 1.442648207568865e-05, + "loss": 0.917, + "step": 5343 + }, + { + "epoch": 1.1205703501782345, + "grad_norm": 7.450447352270454, + "learning_rate": 1.442445194746636e-05, + "loss": 0.7783, + "step": 5344 + }, + { + "epoch": 1.1207800377437618, + "grad_norm": 8.498621452351479, + "learning_rate": 1.4422421592483236e-05, + "loss": 1.022, + "step": 5345 + }, + { + "epoch": 1.120989725309289, + "grad_norm": 7.928145217775752, + "learning_rate": 1.4420391010843333e-05, + "loss": 1.0667, + "step": 5346 + }, + { + "epoch": 1.1211994128748166, + "grad_norm": 6.888034379791948, + "learning_rate": 1.4418360202650723e-05, + "loss": 0.9739, + "step": 5347 + }, + { + "epoch": 1.121409100440344, + "grad_norm": 7.147667928868137, + "learning_rate": 1.441632916800949e-05, + "loss": 0.7517, + "step": 5348 + }, + { + "epoch": 1.1216187880058712, + "grad_norm": 6.788044712192266, + "learning_rate": 1.4414297907023729e-05, + "loss": 0.8655, + "step": 5349 + }, + { + "epoch": 1.1218284755713985, + "grad_norm": 7.312882044839968, + "learning_rate": 1.4412266419797543e-05, + "loss": 0.8724, + "step": 5350 + }, + { + "epoch": 1.122038163136926, + "grad_norm": 6.9142472127982595, + "learning_rate": 1.441023470643505e-05, + "loss": 0.7792, + "step": 5351 + }, + { + "epoch": 1.1222478507024534, + "grad_norm": 7.841383018594161, + "learning_rate": 1.440820276704038e-05, + "loss": 0.7208, + "step": 5352 + }, + { + "epoch": 1.1224575382679807, + "grad_norm": 6.9939743963396035, + "learning_rate": 1.4406170601717673e-05, + "loss": 0.7733, + "step": 5353 + }, + { + "epoch": 1.122667225833508, + "grad_norm": 6.554386105274742, + "learning_rate": 1.4404138210571077e-05, + "loss": 0.6831, + "step": 5354 + }, + { + "epoch": 1.1228769133990355, + "grad_norm": 7.098724082310701, + "learning_rate": 1.4402105593704762e-05, + "loss": 0.7546, + "step": 5355 + }, + { + "epoch": 1.1230866009645628, + "grad_norm": 7.53571305583368, + "learning_rate": 1.44000727512229e-05, + "loss": 0.8781, + "step": 5356 + }, + { + "epoch": 1.1232962885300901, + "grad_norm": 7.882009548978248, + "learning_rate": 1.4398039683229677e-05, + "loss": 0.8757, + "step": 5357 + }, + { + "epoch": 1.1235059760956174, + "grad_norm": 6.815514278255862, + "learning_rate": 1.4396006389829296e-05, + "loss": 0.8982, + "step": 5358 + }, + { + "epoch": 1.123715663661145, + "grad_norm": 7.538810848325119, + "learning_rate": 1.4393972871125956e-05, + "loss": 0.8587, + "step": 5359 + }, + { + "epoch": 1.1239253512266723, + "grad_norm": 8.58280638530999, + "learning_rate": 1.4391939127223893e-05, + "loss": 0.884, + "step": 5360 + }, + { + "epoch": 1.1241350387921996, + "grad_norm": 7.060709874657798, + "learning_rate": 1.4389905158227328e-05, + "loss": 0.7607, + "step": 5361 + }, + { + "epoch": 1.124344726357727, + "grad_norm": 10.108765781270456, + "learning_rate": 1.438787096424051e-05, + "loss": 1.2638, + "step": 5362 + }, + { + "epoch": 1.1245544139232544, + "grad_norm": 6.281611799856333, + "learning_rate": 1.4385836545367695e-05, + "loss": 0.8481, + "step": 5363 + }, + { + "epoch": 1.1247641014887817, + "grad_norm": 7.298175324749741, + "learning_rate": 1.4383801901713153e-05, + "loss": 0.9415, + "step": 5364 + }, + { + "epoch": 1.124973789054309, + "grad_norm": 8.496575626212818, + "learning_rate": 1.4381767033381155e-05, + "loss": 1.0822, + "step": 5365 + }, + { + "epoch": 1.1251834766198365, + "grad_norm": 9.117897448564515, + "learning_rate": 1.4379731940476001e-05, + "loss": 0.7974, + "step": 5366 + }, + { + "epoch": 1.1253931641853638, + "grad_norm": 5.7202721694242475, + "learning_rate": 1.4377696623101985e-05, + "loss": 0.6565, + "step": 5367 + }, + { + "epoch": 1.1256028517508911, + "grad_norm": 7.9782853792509, + "learning_rate": 1.4375661081363427e-05, + "loss": 0.9567, + "step": 5368 + }, + { + "epoch": 1.1258125393164184, + "grad_norm": 6.643879972056351, + "learning_rate": 1.437362531536465e-05, + "loss": 0.7296, + "step": 5369 + }, + { + "epoch": 1.126022226881946, + "grad_norm": 8.583335207889562, + "learning_rate": 1.4371589325209989e-05, + "loss": 0.9101, + "step": 5370 + }, + { + "epoch": 1.1262319144474733, + "grad_norm": 6.836243950650538, + "learning_rate": 1.4369553111003792e-05, + "loss": 0.9839, + "step": 5371 + }, + { + "epoch": 1.1264416020130006, + "grad_norm": 7.320035137857756, + "learning_rate": 1.4367516672850422e-05, + "loss": 0.9172, + "step": 5372 + }, + { + "epoch": 1.1266512895785281, + "grad_norm": 8.798982808132612, + "learning_rate": 1.4365480010854245e-05, + "loss": 0.8326, + "step": 5373 + }, + { + "epoch": 1.1268609771440554, + "grad_norm": 6.526478788559566, + "learning_rate": 1.4363443125119648e-05, + "loss": 0.6422, + "step": 5374 + }, + { + "epoch": 1.1270706647095827, + "grad_norm": 6.707836526018642, + "learning_rate": 1.436140601575102e-05, + "loss": 0.8116, + "step": 5375 + }, + { + "epoch": 1.12728035227511, + "grad_norm": 7.48048492036471, + "learning_rate": 1.435936868285277e-05, + "loss": 1.0693, + "step": 5376 + }, + { + "epoch": 1.1274900398406373, + "grad_norm": 6.426645297364127, + "learning_rate": 1.4357331126529314e-05, + "loss": 0.7899, + "step": 5377 + }, + { + "epoch": 1.1276997274061649, + "grad_norm": 7.930337098484812, + "learning_rate": 1.4355293346885081e-05, + "loss": 1.1519, + "step": 5378 + }, + { + "epoch": 1.1279094149716922, + "grad_norm": 8.089741755707838, + "learning_rate": 1.435325534402451e-05, + "loss": 0.8626, + "step": 5379 + }, + { + "epoch": 1.1281191025372195, + "grad_norm": 9.137835228507097, + "learning_rate": 1.4351217118052055e-05, + "loss": 1.113, + "step": 5380 + }, + { + "epoch": 1.128328790102747, + "grad_norm": 7.1065101296877735, + "learning_rate": 1.4349178669072172e-05, + "loss": 0.9178, + "step": 5381 + }, + { + "epoch": 1.1285384776682743, + "grad_norm": 7.4689392343693255, + "learning_rate": 1.4347139997189341e-05, + "loss": 1.0464, + "step": 5382 + }, + { + "epoch": 1.1287481652338016, + "grad_norm": 6.187891644006432, + "learning_rate": 1.4345101102508046e-05, + "loss": 0.6993, + "step": 5383 + }, + { + "epoch": 1.128957852799329, + "grad_norm": 6.240077443306164, + "learning_rate": 1.434306198513278e-05, + "loss": 0.6045, + "step": 5384 + }, + { + "epoch": 1.1291675403648564, + "grad_norm": 5.174570112136311, + "learning_rate": 1.4341022645168058e-05, + "loss": 0.7004, + "step": 5385 + }, + { + "epoch": 1.1293772279303838, + "grad_norm": 8.060676888391821, + "learning_rate": 1.4338983082718395e-05, + "loss": 0.9166, + "step": 5386 + }, + { + "epoch": 1.129586915495911, + "grad_norm": 6.777680341403954, + "learning_rate": 1.4336943297888324e-05, + "loss": 0.8038, + "step": 5387 + }, + { + "epoch": 1.1297966030614384, + "grad_norm": 7.80021177986887, + "learning_rate": 1.4334903290782384e-05, + "loss": 1.0197, + "step": 5388 + }, + { + "epoch": 1.130006290626966, + "grad_norm": 8.0601547893169, + "learning_rate": 1.4332863061505136e-05, + "loss": 1.1241, + "step": 5389 + }, + { + "epoch": 1.1302159781924932, + "grad_norm": 5.897357256287947, + "learning_rate": 1.433082261016114e-05, + "loss": 0.7745, + "step": 5390 + }, + { + "epoch": 1.1304256657580205, + "grad_norm": 7.275952764466097, + "learning_rate": 1.4328781936854973e-05, + "loss": 0.9076, + "step": 5391 + }, + { + "epoch": 1.130635353323548, + "grad_norm": 7.298029037435003, + "learning_rate": 1.4326741041691221e-05, + "loss": 0.8781, + "step": 5392 + }, + { + "epoch": 1.1308450408890753, + "grad_norm": 6.690345458777128, + "learning_rate": 1.4324699924774493e-05, + "loss": 0.9666, + "step": 5393 + }, + { + "epoch": 1.1310547284546026, + "grad_norm": 8.691288924121832, + "learning_rate": 1.4322658586209386e-05, + "loss": 0.8209, + "step": 5394 + }, + { + "epoch": 1.13126441602013, + "grad_norm": 8.053075990640624, + "learning_rate": 1.4320617026100534e-05, + "loss": 0.8565, + "step": 5395 + }, + { + "epoch": 1.1314741035856573, + "grad_norm": 7.633574091009824, + "learning_rate": 1.4318575244552564e-05, + "loss": 0.7277, + "step": 5396 + }, + { + "epoch": 1.1316837911511848, + "grad_norm": 8.938959265792771, + "learning_rate": 1.4316533241670122e-05, + "loss": 1.2485, + "step": 5397 + }, + { + "epoch": 1.131893478716712, + "grad_norm": 7.868265820560508, + "learning_rate": 1.4314491017557865e-05, + "loss": 0.8427, + "step": 5398 + }, + { + "epoch": 1.1321031662822394, + "grad_norm": 7.3085721225612215, + "learning_rate": 1.431244857232046e-05, + "loss": 0.8427, + "step": 5399 + }, + { + "epoch": 1.132312853847767, + "grad_norm": 7.119899870962551, + "learning_rate": 1.4310405906062585e-05, + "loss": 0.7924, + "step": 5400 + }, + { + "epoch": 1.1325225414132942, + "grad_norm": 8.093826268501388, + "learning_rate": 1.4308363018888937e-05, + "loss": 0.8938, + "step": 5401 + }, + { + "epoch": 1.1327322289788215, + "grad_norm": 7.320233829334078, + "learning_rate": 1.4306319910904204e-05, + "loss": 0.9461, + "step": 5402 + }, + { + "epoch": 1.1329419165443488, + "grad_norm": 7.880379328471395, + "learning_rate": 1.4304276582213113e-05, + "loss": 0.8792, + "step": 5403 + }, + { + "epoch": 1.1331516041098764, + "grad_norm": 7.882877956852063, + "learning_rate": 1.4302233032920378e-05, + "loss": 0.8881, + "step": 5404 + }, + { + "epoch": 1.1333612916754037, + "grad_norm": 6.806324842836676, + "learning_rate": 1.430018926313074e-05, + "loss": 0.8469, + "step": 5405 + }, + { + "epoch": 1.133570979240931, + "grad_norm": 7.450142492081317, + "learning_rate": 1.4298145272948945e-05, + "loss": 0.7573, + "step": 5406 + }, + { + "epoch": 1.1337806668064583, + "grad_norm": 7.566267394090057, + "learning_rate": 1.4296101062479745e-05, + "loss": 0.7798, + "step": 5407 + }, + { + "epoch": 1.1339903543719858, + "grad_norm": 7.125806302742905, + "learning_rate": 1.4294056631827919e-05, + "loss": 0.7141, + "step": 5408 + }, + { + "epoch": 1.1342000419375131, + "grad_norm": 8.495944212331963, + "learning_rate": 1.4292011981098242e-05, + "loss": 1.0081, + "step": 5409 + }, + { + "epoch": 1.1344097295030404, + "grad_norm": 7.441589839803221, + "learning_rate": 1.4289967110395507e-05, + "loss": 0.8588, + "step": 5410 + }, + { + "epoch": 1.134619417068568, + "grad_norm": 6.575428153209504, + "learning_rate": 1.4287922019824516e-05, + "loss": 0.7162, + "step": 5411 + }, + { + "epoch": 1.1348291046340953, + "grad_norm": 7.176187410836672, + "learning_rate": 1.4285876709490085e-05, + "loss": 1.0302, + "step": 5412 + }, + { + "epoch": 1.1350387921996226, + "grad_norm": 8.880645963740191, + "learning_rate": 1.4283831179497039e-05, + "loss": 1.0544, + "step": 5413 + }, + { + "epoch": 1.1352484797651499, + "grad_norm": 6.43805237999645, + "learning_rate": 1.4281785429950214e-05, + "loss": 0.7447, + "step": 5414 + }, + { + "epoch": 1.1354581673306772, + "grad_norm": 7.11316350625244, + "learning_rate": 1.4279739460954462e-05, + "loss": 0.8329, + "step": 5415 + }, + { + "epoch": 1.1356678548962047, + "grad_norm": 7.289649089850713, + "learning_rate": 1.4277693272614638e-05, + "loss": 0.892, + "step": 5416 + }, + { + "epoch": 1.135877542461732, + "grad_norm": 7.6424405921300576, + "learning_rate": 1.4275646865035615e-05, + "loss": 0.8431, + "step": 5417 + }, + { + "epoch": 1.1360872300272593, + "grad_norm": 6.440815832464622, + "learning_rate": 1.4273600238322276e-05, + "loss": 0.641, + "step": 5418 + }, + { + "epoch": 1.1362969175927868, + "grad_norm": 7.600968965507533, + "learning_rate": 1.427155339257951e-05, + "loss": 0.8979, + "step": 5419 + }, + { + "epoch": 1.1365066051583141, + "grad_norm": 7.737941670705575, + "learning_rate": 1.4269506327912223e-05, + "loss": 1.13, + "step": 5420 + }, + { + "epoch": 1.1367162927238414, + "grad_norm": 7.535933877638378, + "learning_rate": 1.4267459044425336e-05, + "loss": 0.9126, + "step": 5421 + }, + { + "epoch": 1.1369259802893688, + "grad_norm": 7.42295276144768, + "learning_rate": 1.4265411542223771e-05, + "loss": 0.7701, + "step": 5422 + }, + { + "epoch": 1.1371356678548963, + "grad_norm": 8.979993618462947, + "learning_rate": 1.426336382141246e-05, + "loss": 0.9879, + "step": 5423 + }, + { + "epoch": 1.1373453554204236, + "grad_norm": 7.87568916179552, + "learning_rate": 1.4261315882096365e-05, + "loss": 0.8786, + "step": 5424 + }, + { + "epoch": 1.137555042985951, + "grad_norm": 6.525528916586531, + "learning_rate": 1.425926772438044e-05, + "loss": 0.7614, + "step": 5425 + }, + { + "epoch": 1.1377647305514782, + "grad_norm": 6.422331421067672, + "learning_rate": 1.4257219348369655e-05, + "loss": 0.6883, + "step": 5426 + }, + { + "epoch": 1.1379744181170057, + "grad_norm": 7.327133505357456, + "learning_rate": 1.4255170754168991e-05, + "loss": 0.8042, + "step": 5427 + }, + { + "epoch": 1.138184105682533, + "grad_norm": 6.436046421043193, + "learning_rate": 1.4253121941883453e-05, + "loss": 0.7845, + "step": 5428 + }, + { + "epoch": 1.1383937932480603, + "grad_norm": 7.847622278661194, + "learning_rate": 1.4251072911618034e-05, + "loss": 0.6945, + "step": 5429 + }, + { + "epoch": 1.1386034808135879, + "grad_norm": 8.087800338874967, + "learning_rate": 1.4249023663477756e-05, + "loss": 0.9851, + "step": 5430 + }, + { + "epoch": 1.1388131683791152, + "grad_norm": 6.861010799839549, + "learning_rate": 1.4246974197567646e-05, + "loss": 0.5995, + "step": 5431 + }, + { + "epoch": 1.1390228559446425, + "grad_norm": 6.665686617124723, + "learning_rate": 1.4244924513992744e-05, + "loss": 0.9058, + "step": 5432 + }, + { + "epoch": 1.1392325435101698, + "grad_norm": 7.153367845813502, + "learning_rate": 1.4242874612858095e-05, + "loss": 0.7087, + "step": 5433 + }, + { + "epoch": 1.139442231075697, + "grad_norm": 7.148266461937071, + "learning_rate": 1.4240824494268766e-05, + "loss": 0.7308, + "step": 5434 + }, + { + "epoch": 1.1396519186412246, + "grad_norm": 6.460811251948656, + "learning_rate": 1.4238774158329825e-05, + "loss": 0.7916, + "step": 5435 + }, + { + "epoch": 1.139861606206752, + "grad_norm": 7.631437728851122, + "learning_rate": 1.4236723605146357e-05, + "loss": 0.9981, + "step": 5436 + }, + { + "epoch": 1.1400712937722792, + "grad_norm": 8.402694519254885, + "learning_rate": 1.4234672834823455e-05, + "loss": 1.0714, + "step": 5437 + }, + { + "epoch": 1.1402809813378068, + "grad_norm": 7.574711241112634, + "learning_rate": 1.4232621847466226e-05, + "loss": 0.8949, + "step": 5438 + }, + { + "epoch": 1.140490668903334, + "grad_norm": 6.289394414033147, + "learning_rate": 1.4230570643179788e-05, + "loss": 0.6595, + "step": 5439 + }, + { + "epoch": 1.1407003564688614, + "grad_norm": 8.17520562098901, + "learning_rate": 1.4228519222069263e-05, + "loss": 0.7566, + "step": 5440 + }, + { + "epoch": 1.1409100440343887, + "grad_norm": 6.496316717579892, + "learning_rate": 1.4226467584239798e-05, + "loss": 0.802, + "step": 5441 + }, + { + "epoch": 1.1411197315999162, + "grad_norm": 7.5708734055535265, + "learning_rate": 1.4224415729796537e-05, + "loss": 0.7837, + "step": 5442 + }, + { + "epoch": 1.1413294191654435, + "grad_norm": 7.642492943105506, + "learning_rate": 1.4222363658844644e-05, + "loss": 0.9662, + "step": 5443 + }, + { + "epoch": 1.1415391067309708, + "grad_norm": 8.61439368358466, + "learning_rate": 1.4220311371489288e-05, + "loss": 1.0253, + "step": 5444 + }, + { + "epoch": 1.1417487942964981, + "grad_norm": 7.218490475421361, + "learning_rate": 1.4218258867835659e-05, + "loss": 0.9208, + "step": 5445 + }, + { + "epoch": 1.1419584818620256, + "grad_norm": 7.367251894731128, + "learning_rate": 1.421620614798894e-05, + "loss": 0.8471, + "step": 5446 + }, + { + "epoch": 1.142168169427553, + "grad_norm": 7.797885488867483, + "learning_rate": 1.421415321205435e-05, + "loss": 0.9664, + "step": 5447 + }, + { + "epoch": 1.1423778569930803, + "grad_norm": 6.8906410375503615, + "learning_rate": 1.4212100060137095e-05, + "loss": 0.8608, + "step": 5448 + }, + { + "epoch": 1.1425875445586078, + "grad_norm": 7.2203249763958, + "learning_rate": 1.4210046692342407e-05, + "loss": 0.9089, + "step": 5449 + }, + { + "epoch": 1.142797232124135, + "grad_norm": 7.503569062454911, + "learning_rate": 1.4207993108775525e-05, + "loss": 0.9927, + "step": 5450 + }, + { + "epoch": 1.1430069196896624, + "grad_norm": 6.232991054800343, + "learning_rate": 1.4205939309541696e-05, + "loss": 0.6247, + "step": 5451 + }, + { + "epoch": 1.1432166072551897, + "grad_norm": 6.677568354945869, + "learning_rate": 1.4203885294746182e-05, + "loss": 0.7053, + "step": 5452 + }, + { + "epoch": 1.1434262948207172, + "grad_norm": 8.240647765606626, + "learning_rate": 1.420183106449426e-05, + "loss": 1.0007, + "step": 5453 + }, + { + "epoch": 1.1436359823862445, + "grad_norm": 6.646773335539556, + "learning_rate": 1.4199776618891203e-05, + "loss": 0.9256, + "step": 5454 + }, + { + "epoch": 1.1438456699517718, + "grad_norm": 6.482277637266567, + "learning_rate": 1.4197721958042314e-05, + "loss": 0.8001, + "step": 5455 + }, + { + "epoch": 1.1440553575172991, + "grad_norm": 7.002879712712853, + "learning_rate": 1.4195667082052891e-05, + "loss": 0.9192, + "step": 5456 + }, + { + "epoch": 1.1442650450828267, + "grad_norm": 7.143624612409936, + "learning_rate": 1.4193611991028256e-05, + "loss": 0.8597, + "step": 5457 + }, + { + "epoch": 1.144474732648354, + "grad_norm": 7.338617162956875, + "learning_rate": 1.419155668507373e-05, + "loss": 0.7998, + "step": 5458 + }, + { + "epoch": 1.1446844202138813, + "grad_norm": 7.6422898013689435, + "learning_rate": 1.4189501164294657e-05, + "loss": 1.0, + "step": 5459 + }, + { + "epoch": 1.1448941077794086, + "grad_norm": 9.784501662221857, + "learning_rate": 1.4187445428796381e-05, + "loss": 1.2785, + "step": 5460 + }, + { + "epoch": 1.1451037953449361, + "grad_norm": 6.850346504508089, + "learning_rate": 1.4185389478684264e-05, + "loss": 0.7997, + "step": 5461 + }, + { + "epoch": 1.1453134829104634, + "grad_norm": 7.633463296409012, + "learning_rate": 1.4183333314063678e-05, + "loss": 0.9655, + "step": 5462 + }, + { + "epoch": 1.1455231704759907, + "grad_norm": 6.754837321516644, + "learning_rate": 1.4181276935040003e-05, + "loss": 0.9435, + "step": 5463 + }, + { + "epoch": 1.145732858041518, + "grad_norm": 6.987679325512715, + "learning_rate": 1.4179220341718633e-05, + "loss": 0.8839, + "step": 5464 + }, + { + "epoch": 1.1459425456070456, + "grad_norm": 6.070002473673756, + "learning_rate": 1.4177163534204971e-05, + "loss": 0.6228, + "step": 5465 + }, + { + "epoch": 1.1461522331725729, + "grad_norm": 7.71743445030266, + "learning_rate": 1.4175106512604435e-05, + "loss": 0.9694, + "step": 5466 + }, + { + "epoch": 1.1463619207381002, + "grad_norm": 6.2618518902646025, + "learning_rate": 1.4173049277022448e-05, + "loss": 0.8574, + "step": 5467 + }, + { + "epoch": 1.1465716083036277, + "grad_norm": 7.380993351776506, + "learning_rate": 1.4170991827564451e-05, + "loss": 1.0179, + "step": 5468 + }, + { + "epoch": 1.146781295869155, + "grad_norm": 7.817066112586511, + "learning_rate": 1.4168934164335883e-05, + "loss": 0.9811, + "step": 5469 + }, + { + "epoch": 1.1469909834346823, + "grad_norm": 7.299111753765806, + "learning_rate": 1.4166876287442212e-05, + "loss": 0.993, + "step": 5470 + }, + { + "epoch": 1.1472006710002096, + "grad_norm": 8.375354497127166, + "learning_rate": 1.4164818196988902e-05, + "loss": 1.2464, + "step": 5471 + }, + { + "epoch": 1.1474103585657371, + "grad_norm": 6.314364313934077, + "learning_rate": 1.416275989308144e-05, + "loss": 0.8252, + "step": 5472 + }, + { + "epoch": 1.1476200461312644, + "grad_norm": 8.534205740589762, + "learning_rate": 1.416070137582531e-05, + "loss": 1.0242, + "step": 5473 + }, + { + "epoch": 1.1478297336967918, + "grad_norm": 6.490236419386976, + "learning_rate": 1.4158642645326022e-05, + "loss": 0.7404, + "step": 5474 + }, + { + "epoch": 1.148039421262319, + "grad_norm": 9.556320277445884, + "learning_rate": 1.415658370168908e-05, + "loss": 1.0618, + "step": 5475 + }, + { + "epoch": 1.1482491088278466, + "grad_norm": 9.629835816676714, + "learning_rate": 1.4154524545020016e-05, + "loss": 1.0136, + "step": 5476 + }, + { + "epoch": 1.148458796393374, + "grad_norm": 7.8685707630138415, + "learning_rate": 1.4152465175424364e-05, + "loss": 0.9567, + "step": 5477 + }, + { + "epoch": 1.1486684839589012, + "grad_norm": 6.944827705966579, + "learning_rate": 1.415040559300767e-05, + "loss": 0.7847, + "step": 5478 + }, + { + "epoch": 1.1488781715244287, + "grad_norm": 6.777523453800765, + "learning_rate": 1.4148345797875491e-05, + "loss": 0.7664, + "step": 5479 + }, + { + "epoch": 1.149087859089956, + "grad_norm": 8.704783105333753, + "learning_rate": 1.4146285790133397e-05, + "loss": 1.074, + "step": 5480 + }, + { + "epoch": 1.1492975466554833, + "grad_norm": 7.09972048807626, + "learning_rate": 1.4144225569886958e-05, + "loss": 1.1704, + "step": 5481 + }, + { + "epoch": 1.1495072342210106, + "grad_norm": 7.594352756869436, + "learning_rate": 1.4142165137241776e-05, + "loss": 0.9961, + "step": 5482 + }, + { + "epoch": 1.149716921786538, + "grad_norm": 8.34161196552639, + "learning_rate": 1.4140104492303448e-05, + "loss": 0.9647, + "step": 5483 + }, + { + "epoch": 1.1499266093520655, + "grad_norm": 7.735172609181506, + "learning_rate": 1.4138043635177581e-05, + "loss": 0.791, + "step": 5484 + }, + { + "epoch": 1.1501362969175928, + "grad_norm": 6.786035346943688, + "learning_rate": 1.4135982565969801e-05, + "loss": 0.8098, + "step": 5485 + }, + { + "epoch": 1.15034598448312, + "grad_norm": 7.4437442520881705, + "learning_rate": 1.413392128478574e-05, + "loss": 0.7806, + "step": 5486 + }, + { + "epoch": 1.1505556720486476, + "grad_norm": 6.819171280766591, + "learning_rate": 1.4131859791731044e-05, + "loss": 0.8952, + "step": 5487 + }, + { + "epoch": 1.150765359614175, + "grad_norm": 6.582856210208786, + "learning_rate": 1.4129798086911368e-05, + "loss": 0.7943, + "step": 5488 + }, + { + "epoch": 1.1509750471797022, + "grad_norm": 7.271065259561847, + "learning_rate": 1.4127736170432377e-05, + "loss": 0.7505, + "step": 5489 + }, + { + "epoch": 1.1511847347452295, + "grad_norm": 6.887401926933448, + "learning_rate": 1.412567404239975e-05, + "loss": 0.9178, + "step": 5490 + }, + { + "epoch": 1.151394422310757, + "grad_norm": 6.193309009302699, + "learning_rate": 1.4123611702919171e-05, + "loss": 0.8851, + "step": 5491 + }, + { + "epoch": 1.1516041098762844, + "grad_norm": 6.81822685477887, + "learning_rate": 1.412154915209634e-05, + "loss": 0.8567, + "step": 5492 + }, + { + "epoch": 1.1518137974418117, + "grad_norm": 8.329740290258938, + "learning_rate": 1.4119486390036965e-05, + "loss": 0.9356, + "step": 5493 + }, + { + "epoch": 1.152023485007339, + "grad_norm": 7.950447617615511, + "learning_rate": 1.411742341684677e-05, + "loss": 0.9432, + "step": 5494 + }, + { + "epoch": 1.1522331725728665, + "grad_norm": 7.8955176151099185, + "learning_rate": 1.4115360232631483e-05, + "loss": 0.936, + "step": 5495 + }, + { + "epoch": 1.1524428601383938, + "grad_norm": 7.181955443057608, + "learning_rate": 1.4113296837496845e-05, + "loss": 0.9743, + "step": 5496 + }, + { + "epoch": 1.1526525477039211, + "grad_norm": 7.175857646589454, + "learning_rate": 1.4111233231548613e-05, + "loss": 0.7054, + "step": 5497 + }, + { + "epoch": 1.1528622352694486, + "grad_norm": 6.605911814145561, + "learning_rate": 1.4109169414892542e-05, + "loss": 0.7528, + "step": 5498 + }, + { + "epoch": 1.153071922834976, + "grad_norm": 6.698349629541683, + "learning_rate": 1.410710538763442e-05, + "loss": 0.8278, + "step": 5499 + }, + { + "epoch": 1.1532816104005033, + "grad_norm": 7.1362884474757, + "learning_rate": 1.4105041149880017e-05, + "loss": 0.9904, + "step": 5500 + }, + { + "epoch": 1.1534912979660306, + "grad_norm": 7.4270664036861245, + "learning_rate": 1.4102976701735136e-05, + "loss": 0.8968, + "step": 5501 + }, + { + "epoch": 1.1537009855315579, + "grad_norm": 8.133383293898993, + "learning_rate": 1.4100912043305583e-05, + "loss": 1.0909, + "step": 5502 + }, + { + "epoch": 1.1539106730970854, + "grad_norm": 8.679252677141207, + "learning_rate": 1.4098847174697177e-05, + "loss": 1.153, + "step": 5503 + }, + { + "epoch": 1.1541203606626127, + "grad_norm": 6.3730700554106114, + "learning_rate": 1.409678209601574e-05, + "loss": 0.7648, + "step": 5504 + }, + { + "epoch": 1.15433004822814, + "grad_norm": 6.911199352827545, + "learning_rate": 1.4094716807367119e-05, + "loss": 0.7541, + "step": 5505 + }, + { + "epoch": 1.1545397357936675, + "grad_norm": 7.3096013355807195, + "learning_rate": 1.409265130885716e-05, + "loss": 0.9996, + "step": 5506 + }, + { + "epoch": 1.1547494233591948, + "grad_norm": 9.500041494545538, + "learning_rate": 1.4090585600591719e-05, + "loss": 1.2379, + "step": 5507 + }, + { + "epoch": 1.1549591109247221, + "grad_norm": 8.337608550441358, + "learning_rate": 1.4088519682676676e-05, + "loss": 1.0607, + "step": 5508 + }, + { + "epoch": 1.1551687984902494, + "grad_norm": 7.315950088466112, + "learning_rate": 1.4086453555217908e-05, + "loss": 0.9575, + "step": 5509 + }, + { + "epoch": 1.155378486055777, + "grad_norm": 5.79010133280334, + "learning_rate": 1.4084387218321302e-05, + "loss": 0.5228, + "step": 5510 + }, + { + "epoch": 1.1555881736213043, + "grad_norm": 6.080233663676579, + "learning_rate": 1.4082320672092773e-05, + "loss": 0.8593, + "step": 5511 + }, + { + "epoch": 1.1557978611868316, + "grad_norm": 8.448473847480178, + "learning_rate": 1.4080253916638226e-05, + "loss": 0.8149, + "step": 5512 + }, + { + "epoch": 1.156007548752359, + "grad_norm": 7.138541019769388, + "learning_rate": 1.4078186952063588e-05, + "loss": 0.7999, + "step": 5513 + }, + { + "epoch": 1.1562172363178864, + "grad_norm": 8.235108656836639, + "learning_rate": 1.4076119778474798e-05, + "loss": 0.973, + "step": 5514 + }, + { + "epoch": 1.1564269238834137, + "grad_norm": 6.83693505093105, + "learning_rate": 1.4074052395977798e-05, + "loss": 0.777, + "step": 5515 + }, + { + "epoch": 1.156636611448941, + "grad_norm": 7.713760104790157, + "learning_rate": 1.4071984804678547e-05, + "loss": 1.0136, + "step": 5516 + }, + { + "epoch": 1.1568462990144686, + "grad_norm": 7.964724850386856, + "learning_rate": 1.4069917004683014e-05, + "loss": 1.0548, + "step": 5517 + }, + { + "epoch": 1.1570559865799959, + "grad_norm": 8.074774986415274, + "learning_rate": 1.4067848996097176e-05, + "loss": 1.0524, + "step": 5518 + }, + { + "epoch": 1.1572656741455232, + "grad_norm": 5.786259214412862, + "learning_rate": 1.4065780779027017e-05, + "loss": 0.674, + "step": 5519 + }, + { + "epoch": 1.1574753617110505, + "grad_norm": 6.6902506177516505, + "learning_rate": 1.4063712353578545e-05, + "loss": 0.7824, + "step": 5520 + }, + { + "epoch": 1.1576850492765778, + "grad_norm": 5.974558650783818, + "learning_rate": 1.4061643719857764e-05, + "loss": 0.785, + "step": 5521 + }, + { + "epoch": 1.1578947368421053, + "grad_norm": 8.159967530096615, + "learning_rate": 1.4059574877970701e-05, + "loss": 1.0235, + "step": 5522 + }, + { + "epoch": 1.1581044244076326, + "grad_norm": 8.013114604405413, + "learning_rate": 1.4057505828023384e-05, + "loss": 1.118, + "step": 5523 + }, + { + "epoch": 1.15831411197316, + "grad_norm": 7.218966011225552, + "learning_rate": 1.4055436570121854e-05, + "loss": 0.8187, + "step": 5524 + }, + { + "epoch": 1.1585237995386874, + "grad_norm": 6.347011089257012, + "learning_rate": 1.4053367104372169e-05, + "loss": 0.675, + "step": 5525 + }, + { + "epoch": 1.1587334871042148, + "grad_norm": 7.232725412116838, + "learning_rate": 1.4051297430880389e-05, + "loss": 0.97, + "step": 5526 + }, + { + "epoch": 1.158943174669742, + "grad_norm": 7.367964054879141, + "learning_rate": 1.4049227549752584e-05, + "loss": 0.7577, + "step": 5527 + }, + { + "epoch": 1.1591528622352694, + "grad_norm": 6.3463847016507975, + "learning_rate": 1.4047157461094849e-05, + "loss": 0.83, + "step": 5528 + }, + { + "epoch": 1.159362549800797, + "grad_norm": 8.202100298322755, + "learning_rate": 1.4045087165013275e-05, + "loss": 0.8041, + "step": 5529 + }, + { + "epoch": 1.1595722373663242, + "grad_norm": 6.54470561602857, + "learning_rate": 1.4043016661613967e-05, + "loss": 0.6681, + "step": 5530 + }, + { + "epoch": 1.1597819249318515, + "grad_norm": 7.096689065706821, + "learning_rate": 1.404094595100304e-05, + "loss": 1.0322, + "step": 5531 + }, + { + "epoch": 1.1599916124973788, + "grad_norm": 7.382077728774741, + "learning_rate": 1.403887503328663e-05, + "loss": 0.7374, + "step": 5532 + }, + { + "epoch": 1.1602013000629063, + "grad_norm": 6.215511067107648, + "learning_rate": 1.4036803908570864e-05, + "loss": 0.8929, + "step": 5533 + }, + { + "epoch": 1.1604109876284336, + "grad_norm": 6.627689878403522, + "learning_rate": 1.40347325769619e-05, + "loss": 0.8195, + "step": 5534 + }, + { + "epoch": 1.160620675193961, + "grad_norm": 8.518418074280659, + "learning_rate": 1.4032661038565892e-05, + "loss": 1.1725, + "step": 5535 + }, + { + "epoch": 1.1608303627594885, + "grad_norm": 7.432250519187269, + "learning_rate": 1.4030589293489012e-05, + "loss": 1.0548, + "step": 5536 + }, + { + "epoch": 1.1610400503250158, + "grad_norm": 7.426710246257958, + "learning_rate": 1.4028517341837438e-05, + "loss": 0.8218, + "step": 5537 + }, + { + "epoch": 1.161249737890543, + "grad_norm": 7.370784988061927, + "learning_rate": 1.4026445183717367e-05, + "loss": 0.9162, + "step": 5538 + }, + { + "epoch": 1.1614594254560704, + "grad_norm": 7.189066173863027, + "learning_rate": 1.4024372819234992e-05, + "loss": 0.7991, + "step": 5539 + }, + { + "epoch": 1.1616691130215977, + "grad_norm": 6.5924398303649445, + "learning_rate": 1.4022300248496534e-05, + "loss": 0.8224, + "step": 5540 + }, + { + "epoch": 1.1618788005871252, + "grad_norm": 6.314382383975065, + "learning_rate": 1.402022747160821e-05, + "loss": 0.7187, + "step": 5541 + }, + { + "epoch": 1.1620884881526525, + "grad_norm": 8.031428610173755, + "learning_rate": 1.4018154488676255e-05, + "loss": 0.9767, + "step": 5542 + }, + { + "epoch": 1.1622981757181798, + "grad_norm": 7.652719003610233, + "learning_rate": 1.4016081299806914e-05, + "loss": 0.9668, + "step": 5543 + }, + { + "epoch": 1.1625078632837074, + "grad_norm": 8.804170511260384, + "learning_rate": 1.4014007905106442e-05, + "loss": 1.0849, + "step": 5544 + }, + { + "epoch": 1.1627175508492347, + "grad_norm": 6.260027568115352, + "learning_rate": 1.40119343046811e-05, + "loss": 0.708, + "step": 5545 + }, + { + "epoch": 1.162927238414762, + "grad_norm": 6.406260400129809, + "learning_rate": 1.400986049863717e-05, + "loss": 0.5666, + "step": 5546 + }, + { + "epoch": 1.1631369259802893, + "grad_norm": 7.281576327350118, + "learning_rate": 1.4007786487080935e-05, + "loss": 0.7486, + "step": 5547 + }, + { + "epoch": 1.1633466135458168, + "grad_norm": 7.2884372167514035, + "learning_rate": 1.4005712270118685e-05, + "loss": 0.8374, + "step": 5548 + }, + { + "epoch": 1.1635563011113441, + "grad_norm": 8.07932108328747, + "learning_rate": 1.4003637847856741e-05, + "loss": 1.0004, + "step": 5549 + }, + { + "epoch": 1.1637659886768714, + "grad_norm": 7.502899606791628, + "learning_rate": 1.4001563220401409e-05, + "loss": 0.8037, + "step": 5550 + }, + { + "epoch": 1.1639756762423987, + "grad_norm": 7.1295667896213875, + "learning_rate": 1.3999488387859023e-05, + "loss": 0.7517, + "step": 5551 + }, + { + "epoch": 1.1641853638079263, + "grad_norm": 7.92445807553929, + "learning_rate": 1.3997413350335922e-05, + "loss": 0.8036, + "step": 5552 + }, + { + "epoch": 1.1643950513734536, + "grad_norm": 6.495541696637586, + "learning_rate": 1.399533810793845e-05, + "loss": 0.7468, + "step": 5553 + }, + { + "epoch": 1.1646047389389809, + "grad_norm": 7.10402048255727, + "learning_rate": 1.3993262660772973e-05, + "loss": 1.0148, + "step": 5554 + }, + { + "epoch": 1.1648144265045084, + "grad_norm": 8.104166012369038, + "learning_rate": 1.3991187008945859e-05, + "loss": 0.9111, + "step": 5555 + }, + { + "epoch": 1.1650241140700357, + "grad_norm": 8.564299309979349, + "learning_rate": 1.3989111152563486e-05, + "loss": 0.7724, + "step": 5556 + }, + { + "epoch": 1.165233801635563, + "grad_norm": 8.384485187190366, + "learning_rate": 1.3987035091732251e-05, + "loss": 0.843, + "step": 5557 + }, + { + "epoch": 1.1654434892010903, + "grad_norm": 7.8049558778581165, + "learning_rate": 1.398495882655855e-05, + "loss": 0.9738, + "step": 5558 + }, + { + "epoch": 1.1656531767666176, + "grad_norm": 7.208790850154003, + "learning_rate": 1.3982882357148796e-05, + "loss": 0.902, + "step": 5559 + }, + { + "epoch": 1.1658628643321451, + "grad_norm": 6.558483501709594, + "learning_rate": 1.3980805683609418e-05, + "loss": 0.7199, + "step": 5560 + }, + { + "epoch": 1.1660725518976724, + "grad_norm": 6.858139462030764, + "learning_rate": 1.3978728806046845e-05, + "loss": 0.8666, + "step": 5561 + }, + { + "epoch": 1.1662822394631998, + "grad_norm": 8.262889832043188, + "learning_rate": 1.3976651724567511e-05, + "loss": 1.0382, + "step": 5562 + }, + { + "epoch": 1.1664919270287273, + "grad_norm": 8.028379588806763, + "learning_rate": 1.3974574439277889e-05, + "loss": 0.9271, + "step": 5563 + }, + { + "epoch": 1.1667016145942546, + "grad_norm": 7.218026748875909, + "learning_rate": 1.3972496950284427e-05, + "loss": 0.8656, + "step": 5564 + }, + { + "epoch": 1.166911302159782, + "grad_norm": 8.720782145476438, + "learning_rate": 1.3970419257693609e-05, + "loss": 0.998, + "step": 5565 + }, + { + "epoch": 1.1671209897253092, + "grad_norm": 6.810971121596743, + "learning_rate": 1.3968341361611916e-05, + "loss": 0.8755, + "step": 5566 + }, + { + "epoch": 1.1673306772908367, + "grad_norm": 8.250873105852879, + "learning_rate": 1.3966263262145848e-05, + "loss": 0.9699, + "step": 5567 + }, + { + "epoch": 1.167540364856364, + "grad_norm": 7.035244909248895, + "learning_rate": 1.3964184959401906e-05, + "loss": 0.7971, + "step": 5568 + }, + { + "epoch": 1.1677500524218913, + "grad_norm": 7.776495239408806, + "learning_rate": 1.3962106453486613e-05, + "loss": 1.1185, + "step": 5569 + }, + { + "epoch": 1.1679597399874186, + "grad_norm": 6.995030796561895, + "learning_rate": 1.3960027744506488e-05, + "loss": 0.7593, + "step": 5570 + }, + { + "epoch": 1.1681694275529462, + "grad_norm": 9.033657207726348, + "learning_rate": 1.3957948832568076e-05, + "loss": 0.833, + "step": 5571 + }, + { + "epoch": 1.1683791151184735, + "grad_norm": 6.641138610331567, + "learning_rate": 1.3955869717777918e-05, + "loss": 0.8655, + "step": 5572 + }, + { + "epoch": 1.1685888026840008, + "grad_norm": 7.708460540892471, + "learning_rate": 1.395379040024258e-05, + "loss": 0.8105, + "step": 5573 + }, + { + "epoch": 1.1687984902495283, + "grad_norm": 7.373430849396879, + "learning_rate": 1.3951710880068623e-05, + "loss": 0.9267, + "step": 5574 + }, + { + "epoch": 1.1690081778150556, + "grad_norm": 7.124819158258448, + "learning_rate": 1.394963115736263e-05, + "loss": 0.8531, + "step": 5575 + }, + { + "epoch": 1.169217865380583, + "grad_norm": 7.59226208133222, + "learning_rate": 1.3947551232231194e-05, + "loss": 0.9472, + "step": 5576 + }, + { + "epoch": 1.1694275529461102, + "grad_norm": 5.888968694301112, + "learning_rate": 1.3945471104780906e-05, + "loss": 0.745, + "step": 5577 + }, + { + "epoch": 1.1696372405116378, + "grad_norm": 7.268625721218548, + "learning_rate": 1.3943390775118383e-05, + "loss": 0.8901, + "step": 5578 + }, + { + "epoch": 1.169846928077165, + "grad_norm": 7.380352274636955, + "learning_rate": 1.3941310243350244e-05, + "loss": 0.6978, + "step": 5579 + }, + { + "epoch": 1.1700566156426924, + "grad_norm": 7.473076522652288, + "learning_rate": 1.3939229509583117e-05, + "loss": 0.8239, + "step": 5580 + }, + { + "epoch": 1.1702663032082197, + "grad_norm": 6.80577748516548, + "learning_rate": 1.3937148573923647e-05, + "loss": 0.7917, + "step": 5581 + }, + { + "epoch": 1.1704759907737472, + "grad_norm": 6.2537552565444, + "learning_rate": 1.3935067436478483e-05, + "loss": 0.7429, + "step": 5582 + }, + { + "epoch": 1.1706856783392745, + "grad_norm": 9.183670258619124, + "learning_rate": 1.3932986097354288e-05, + "loss": 1.0381, + "step": 5583 + }, + { + "epoch": 1.1708953659048018, + "grad_norm": 5.916981098487638, + "learning_rate": 1.3930904556657735e-05, + "loss": 0.5183, + "step": 5584 + }, + { + "epoch": 1.1711050534703291, + "grad_norm": 6.4062889992003695, + "learning_rate": 1.3928822814495504e-05, + "loss": 0.8093, + "step": 5585 + }, + { + "epoch": 1.1713147410358566, + "grad_norm": 6.652514843020139, + "learning_rate": 1.392674087097429e-05, + "loss": 0.9062, + "step": 5586 + }, + { + "epoch": 1.171524428601384, + "grad_norm": 6.97181508790011, + "learning_rate": 1.3924658726200798e-05, + "loss": 0.851, + "step": 5587 + }, + { + "epoch": 1.1717341161669113, + "grad_norm": 8.499517112840234, + "learning_rate": 1.3922576380281739e-05, + "loss": 1.073, + "step": 5588 + }, + { + "epoch": 1.1719438037324386, + "grad_norm": 6.077591549342404, + "learning_rate": 1.3920493833323833e-05, + "loss": 0.8214, + "step": 5589 + }, + { + "epoch": 1.172153491297966, + "grad_norm": 5.993412622136807, + "learning_rate": 1.3918411085433822e-05, + "loss": 0.7873, + "step": 5590 + }, + { + "epoch": 1.1723631788634934, + "grad_norm": 6.697136904012644, + "learning_rate": 1.3916328136718444e-05, + "loss": 0.7363, + "step": 5591 + }, + { + "epoch": 1.1725728664290207, + "grad_norm": 6.975223740668854, + "learning_rate": 1.391424498728446e-05, + "loss": 0.8073, + "step": 5592 + }, + { + "epoch": 1.1727825539945482, + "grad_norm": 6.302969948249736, + "learning_rate": 1.391216163723863e-05, + "loss": 0.8286, + "step": 5593 + }, + { + "epoch": 1.1729922415600755, + "grad_norm": 6.3999766222298184, + "learning_rate": 1.3910078086687731e-05, + "loss": 0.7968, + "step": 5594 + }, + { + "epoch": 1.1732019291256028, + "grad_norm": 7.694049833517075, + "learning_rate": 1.3907994335738546e-05, + "loss": 0.7568, + "step": 5595 + }, + { + "epoch": 1.1734116166911301, + "grad_norm": 9.462558731463119, + "learning_rate": 1.3905910384497876e-05, + "loss": 0.7763, + "step": 5596 + }, + { + "epoch": 1.1736213042566577, + "grad_norm": 7.8162239671601075, + "learning_rate": 1.3903826233072523e-05, + "loss": 1.0712, + "step": 5597 + }, + { + "epoch": 1.173830991822185, + "grad_norm": 7.528316198425317, + "learning_rate": 1.3901741881569306e-05, + "loss": 1.0436, + "step": 5598 + }, + { + "epoch": 1.1740406793877123, + "grad_norm": 8.156716409452379, + "learning_rate": 1.389965733009505e-05, + "loss": 0.901, + "step": 5599 + }, + { + "epoch": 1.1742503669532396, + "grad_norm": 7.7806969009443, + "learning_rate": 1.389757257875659e-05, + "loss": 0.8909, + "step": 5600 + }, + { + "epoch": 1.1744600545187671, + "grad_norm": 6.747483240429086, + "learning_rate": 1.389548762766078e-05, + "loss": 0.8724, + "step": 5601 + }, + { + "epoch": 1.1746697420842944, + "grad_norm": 7.882417573478209, + "learning_rate": 1.3893402476914468e-05, + "loss": 0.8693, + "step": 5602 + }, + { + "epoch": 1.1748794296498217, + "grad_norm": 6.231588148887154, + "learning_rate": 1.3891317126624528e-05, + "loss": 0.8189, + "step": 5603 + }, + { + "epoch": 1.175089117215349, + "grad_norm": 7.172811633752633, + "learning_rate": 1.3889231576897837e-05, + "loss": 0.9693, + "step": 5604 + }, + { + "epoch": 1.1752988047808766, + "grad_norm": 8.951661442173059, + "learning_rate": 1.3887145827841282e-05, + "loss": 0.7747, + "step": 5605 + }, + { + "epoch": 1.1755084923464039, + "grad_norm": 7.818523917704988, + "learning_rate": 1.388505987956176e-05, + "loss": 0.7961, + "step": 5606 + }, + { + "epoch": 1.1757181799119312, + "grad_norm": 6.471893346882293, + "learning_rate": 1.3882973732166183e-05, + "loss": 0.763, + "step": 5607 + }, + { + "epoch": 1.1759278674774585, + "grad_norm": 8.243674209259343, + "learning_rate": 1.3880887385761466e-05, + "loss": 0.9605, + "step": 5608 + }, + { + "epoch": 1.176137555042986, + "grad_norm": 7.6602389527154955, + "learning_rate": 1.3878800840454539e-05, + "loss": 0.8205, + "step": 5609 + }, + { + "epoch": 1.1763472426085133, + "grad_norm": 6.995182514484757, + "learning_rate": 1.3876714096352341e-05, + "loss": 0.887, + "step": 5610 + }, + { + "epoch": 1.1765569301740406, + "grad_norm": 7.104625402296216, + "learning_rate": 1.3874627153561822e-05, + "loss": 0.8944, + "step": 5611 + }, + { + "epoch": 1.1767666177395681, + "grad_norm": 7.880185866381647, + "learning_rate": 1.3872540012189943e-05, + "loss": 0.7541, + "step": 5612 + }, + { + "epoch": 1.1769763053050954, + "grad_norm": 5.416838954782173, + "learning_rate": 1.3870452672343673e-05, + "loss": 0.566, + "step": 5613 + }, + { + "epoch": 1.1771859928706228, + "grad_norm": 7.566946464813779, + "learning_rate": 1.3868365134129987e-05, + "loss": 0.9677, + "step": 5614 + }, + { + "epoch": 1.17739568043615, + "grad_norm": 8.072919877677197, + "learning_rate": 1.3866277397655879e-05, + "loss": 1.0179, + "step": 5615 + }, + { + "epoch": 1.1776053680016776, + "grad_norm": 6.402403958045082, + "learning_rate": 1.386418946302835e-05, + "loss": 0.8271, + "step": 5616 + }, + { + "epoch": 1.177815055567205, + "grad_norm": 5.880345512606271, + "learning_rate": 1.3862101330354406e-05, + "loss": 0.6467, + "step": 5617 + }, + { + "epoch": 1.1780247431327322, + "grad_norm": 7.154139220718624, + "learning_rate": 1.3860012999741076e-05, + "loss": 0.8529, + "step": 5618 + }, + { + "epoch": 1.1782344306982595, + "grad_norm": 6.169033575279547, + "learning_rate": 1.3857924471295384e-05, + "loss": 0.7324, + "step": 5619 + }, + { + "epoch": 1.178444118263787, + "grad_norm": 7.798326766380699, + "learning_rate": 1.3855835745124365e-05, + "loss": 0.8376, + "step": 5620 + }, + { + "epoch": 1.1786538058293143, + "grad_norm": 8.129685071081152, + "learning_rate": 1.3853746821335084e-05, + "loss": 0.9424, + "step": 5621 + }, + { + "epoch": 1.1788634933948416, + "grad_norm": 7.068441554158626, + "learning_rate": 1.3851657700034591e-05, + "loss": 0.8675, + "step": 5622 + }, + { + "epoch": 1.1790731809603692, + "grad_norm": 7.788652169860666, + "learning_rate": 1.3849568381329958e-05, + "loss": 0.8562, + "step": 5623 + }, + { + "epoch": 1.1792828685258965, + "grad_norm": 6.334949374905051, + "learning_rate": 1.3847478865328272e-05, + "loss": 0.6565, + "step": 5624 + }, + { + "epoch": 1.1794925560914238, + "grad_norm": 8.790791662719936, + "learning_rate": 1.384538915213662e-05, + "loss": 0.8423, + "step": 5625 + }, + { + "epoch": 1.179702243656951, + "grad_norm": 7.3620297013365406, + "learning_rate": 1.3843299241862105e-05, + "loss": 0.8906, + "step": 5626 + }, + { + "epoch": 1.1799119312224784, + "grad_norm": 8.371907005332536, + "learning_rate": 1.384120913461184e-05, + "loss": 1.0709, + "step": 5627 + }, + { + "epoch": 1.180121618788006, + "grad_norm": 8.081840045947917, + "learning_rate": 1.3839118830492942e-05, + "loss": 0.8053, + "step": 5628 + }, + { + "epoch": 1.1803313063535332, + "grad_norm": 7.474865402390149, + "learning_rate": 1.3837028329612542e-05, + "loss": 0.9166, + "step": 5629 + }, + { + "epoch": 1.1805409939190605, + "grad_norm": 7.663703695457021, + "learning_rate": 1.3834937632077794e-05, + "loss": 0.8424, + "step": 5630 + }, + { + "epoch": 1.180750681484588, + "grad_norm": 8.86509457205034, + "learning_rate": 1.3832846737995836e-05, + "loss": 0.8927, + "step": 5631 + }, + { + "epoch": 1.1809603690501154, + "grad_norm": 5.723959602838004, + "learning_rate": 1.3830755647473835e-05, + "loss": 0.7921, + "step": 5632 + }, + { + "epoch": 1.1811700566156427, + "grad_norm": 6.770830196951483, + "learning_rate": 1.382866436061896e-05, + "loss": 0.7365, + "step": 5633 + }, + { + "epoch": 1.18137974418117, + "grad_norm": 6.528917255086503, + "learning_rate": 1.3826572877538403e-05, + "loss": 0.847, + "step": 5634 + }, + { + "epoch": 1.1815894317466975, + "grad_norm": 7.008963864291205, + "learning_rate": 1.3824481198339342e-05, + "loss": 0.6544, + "step": 5635 + }, + { + "epoch": 1.1817991193122248, + "grad_norm": 9.751552006833972, + "learning_rate": 1.382238932312899e-05, + "loss": 1.0352, + "step": 5636 + }, + { + "epoch": 1.1820088068777521, + "grad_norm": 7.27222110116671, + "learning_rate": 1.3820297252014556e-05, + "loss": 0.9215, + "step": 5637 + }, + { + "epoch": 1.1822184944432794, + "grad_norm": 6.613857179429183, + "learning_rate": 1.381820498510326e-05, + "loss": 0.8005, + "step": 5638 + }, + { + "epoch": 1.182428182008807, + "grad_norm": 7.171706679961983, + "learning_rate": 1.3816112522502335e-05, + "loss": 0.8353, + "step": 5639 + }, + { + "epoch": 1.1826378695743343, + "grad_norm": 9.291744305245487, + "learning_rate": 1.381401986431903e-05, + "loss": 0.8169, + "step": 5640 + }, + { + "epoch": 1.1828475571398616, + "grad_norm": 6.493545370857968, + "learning_rate": 1.3811927010660587e-05, + "loss": 0.7442, + "step": 5641 + }, + { + "epoch": 1.183057244705389, + "grad_norm": 6.996088691690428, + "learning_rate": 1.3809833961634275e-05, + "loss": 0.8455, + "step": 5642 + }, + { + "epoch": 1.1832669322709164, + "grad_norm": 9.47597907787096, + "learning_rate": 1.3807740717347362e-05, + "loss": 0.913, + "step": 5643 + }, + { + "epoch": 1.1834766198364437, + "grad_norm": 7.3598284589264695, + "learning_rate": 1.3805647277907137e-05, + "loss": 0.8749, + "step": 5644 + }, + { + "epoch": 1.183686307401971, + "grad_norm": 7.325876382295701, + "learning_rate": 1.3803553643420888e-05, + "loss": 0.9871, + "step": 5645 + }, + { + "epoch": 1.1838959949674983, + "grad_norm": 8.405622361817615, + "learning_rate": 1.3801459813995917e-05, + "loss": 1.0752, + "step": 5646 + }, + { + "epoch": 1.1841056825330258, + "grad_norm": 8.296161667417985, + "learning_rate": 1.3799365789739537e-05, + "loss": 0.712, + "step": 5647 + }, + { + "epoch": 1.1843153700985531, + "grad_norm": 6.877579577069931, + "learning_rate": 1.3797271570759072e-05, + "loss": 0.8776, + "step": 5648 + }, + { + "epoch": 1.1845250576640804, + "grad_norm": 7.899024418042107, + "learning_rate": 1.3795177157161849e-05, + "loss": 0.8103, + "step": 5649 + }, + { + "epoch": 1.184734745229608, + "grad_norm": 7.597763211053912, + "learning_rate": 1.379308254905522e-05, + "loss": 0.9938, + "step": 5650 + }, + { + "epoch": 1.1849444327951353, + "grad_norm": 6.456469495670153, + "learning_rate": 1.3790987746546527e-05, + "loss": 0.7621, + "step": 5651 + }, + { + "epoch": 1.1851541203606626, + "grad_norm": 8.15718963114159, + "learning_rate": 1.3788892749743137e-05, + "loss": 0.8181, + "step": 5652 + }, + { + "epoch": 1.18536380792619, + "grad_norm": 7.2348404563641076, + "learning_rate": 1.3786797558752424e-05, + "loss": 0.962, + "step": 5653 + }, + { + "epoch": 1.1855734954917174, + "grad_norm": 7.1112588427510115, + "learning_rate": 1.3784702173681767e-05, + "loss": 0.8297, + "step": 5654 + }, + { + "epoch": 1.1857831830572447, + "grad_norm": 7.453227357966235, + "learning_rate": 1.378260659463856e-05, + "loss": 0.9245, + "step": 5655 + }, + { + "epoch": 1.185992870622772, + "grad_norm": 6.894858060769164, + "learning_rate": 1.3780510821730205e-05, + "loss": 0.7762, + "step": 5656 + }, + { + "epoch": 1.1862025581882993, + "grad_norm": 7.9072020500757665, + "learning_rate": 1.3778414855064112e-05, + "loss": 0.8422, + "step": 5657 + }, + { + "epoch": 1.1864122457538269, + "grad_norm": 5.900324071668547, + "learning_rate": 1.3776318694747704e-05, + "loss": 0.6924, + "step": 5658 + }, + { + "epoch": 1.1866219333193542, + "grad_norm": 7.368427903878405, + "learning_rate": 1.3774222340888418e-05, + "loss": 0.7939, + "step": 5659 + }, + { + "epoch": 1.1868316208848815, + "grad_norm": 7.423932904741232, + "learning_rate": 1.3772125793593684e-05, + "loss": 0.9915, + "step": 5660 + }, + { + "epoch": 1.187041308450409, + "grad_norm": 7.6574107222029975, + "learning_rate": 1.3770029052970965e-05, + "loss": 0.9254, + "step": 5661 + }, + { + "epoch": 1.1872509960159363, + "grad_norm": 6.7788323944186155, + "learning_rate": 1.3767932119127717e-05, + "loss": 0.8037, + "step": 5662 + }, + { + "epoch": 1.1874606835814636, + "grad_norm": 7.6923952980322765, + "learning_rate": 1.3765834992171413e-05, + "loss": 0.8313, + "step": 5663 + }, + { + "epoch": 1.187670371146991, + "grad_norm": 6.733331331135807, + "learning_rate": 1.3763737672209534e-05, + "loss": 0.6348, + "step": 5664 + }, + { + "epoch": 1.1878800587125182, + "grad_norm": 6.205615802015711, + "learning_rate": 1.3761640159349572e-05, + "loss": 0.7039, + "step": 5665 + }, + { + "epoch": 1.1880897462780458, + "grad_norm": 7.551457945344353, + "learning_rate": 1.3759542453699028e-05, + "loss": 0.7845, + "step": 5666 + }, + { + "epoch": 1.188299433843573, + "grad_norm": 7.362289169878077, + "learning_rate": 1.375744455536541e-05, + "loss": 0.9324, + "step": 5667 + }, + { + "epoch": 1.1885091214091004, + "grad_norm": 8.133515370656067, + "learning_rate": 1.3755346464456245e-05, + "loss": 0.7779, + "step": 5668 + }, + { + "epoch": 1.188718808974628, + "grad_norm": 7.999635896276008, + "learning_rate": 1.375324818107906e-05, + "loss": 1.2023, + "step": 5669 + }, + { + "epoch": 1.1889284965401552, + "grad_norm": 7.410687155600046, + "learning_rate": 1.3751149705341393e-05, + "loss": 1.1157, + "step": 5670 + }, + { + "epoch": 1.1891381841056825, + "grad_norm": 8.057169250716472, + "learning_rate": 1.37490510373508e-05, + "loss": 0.906, + "step": 5671 + }, + { + "epoch": 1.1893478716712098, + "grad_norm": 6.73665642484448, + "learning_rate": 1.3746952177214838e-05, + "loss": 0.863, + "step": 5672 + }, + { + "epoch": 1.1895575592367373, + "grad_norm": 7.778994155451082, + "learning_rate": 1.374485312504108e-05, + "loss": 0.8987, + "step": 5673 + }, + { + "epoch": 1.1897672468022646, + "grad_norm": 8.323020373750847, + "learning_rate": 1.3742753880937101e-05, + "loss": 0.8837, + "step": 5674 + }, + { + "epoch": 1.189976934367792, + "grad_norm": 8.013136007869543, + "learning_rate": 1.3740654445010497e-05, + "loss": 0.8605, + "step": 5675 + }, + { + "epoch": 1.1901866219333193, + "grad_norm": 6.096103210483433, + "learning_rate": 1.3738554817368862e-05, + "loss": 0.829, + "step": 5676 + }, + { + "epoch": 1.1903963094988468, + "grad_norm": 7.060517636841401, + "learning_rate": 1.373645499811981e-05, + "loss": 0.8782, + "step": 5677 + }, + { + "epoch": 1.190605997064374, + "grad_norm": 7.157504180440596, + "learning_rate": 1.373435498737096e-05, + "loss": 0.9056, + "step": 5678 + }, + { + "epoch": 1.1908156846299014, + "grad_norm": 7.846248401214299, + "learning_rate": 1.3732254785229937e-05, + "loss": 0.9993, + "step": 5679 + }, + { + "epoch": 1.191025372195429, + "grad_norm": 7.739747852972373, + "learning_rate": 1.3730154391804387e-05, + "loss": 0.9501, + "step": 5680 + }, + { + "epoch": 1.1912350597609562, + "grad_norm": 5.990113444486688, + "learning_rate": 1.3728053807201953e-05, + "loss": 0.7729, + "step": 5681 + }, + { + "epoch": 1.1914447473264835, + "grad_norm": 7.768612059040285, + "learning_rate": 1.3725953031530295e-05, + "loss": 1.0722, + "step": 5682 + }, + { + "epoch": 1.1916544348920108, + "grad_norm": 7.3697195910017275, + "learning_rate": 1.372385206489708e-05, + "loss": 0.867, + "step": 5683 + }, + { + "epoch": 1.1918641224575381, + "grad_norm": 6.991360739360432, + "learning_rate": 1.3721750907409989e-05, + "loss": 0.8595, + "step": 5684 + }, + { + "epoch": 1.1920738100230657, + "grad_norm": 7.749075191615825, + "learning_rate": 1.371964955917671e-05, + "loss": 0.7957, + "step": 5685 + }, + { + "epoch": 1.192283497588593, + "grad_norm": 6.2818291643835185, + "learning_rate": 1.3717548020304943e-05, + "loss": 0.7302, + "step": 5686 + }, + { + "epoch": 1.1924931851541203, + "grad_norm": 7.270376324921736, + "learning_rate": 1.3715446290902385e-05, + "loss": 0.805, + "step": 5687 + }, + { + "epoch": 1.1927028727196478, + "grad_norm": 5.963688055820944, + "learning_rate": 1.3713344371076767e-05, + "loss": 0.7588, + "step": 5688 + }, + { + "epoch": 1.1929125602851751, + "grad_norm": 5.788906936865552, + "learning_rate": 1.3711242260935806e-05, + "loss": 0.6473, + "step": 5689 + }, + { + "epoch": 1.1931222478507024, + "grad_norm": 7.528176276733837, + "learning_rate": 1.3709139960587244e-05, + "loss": 1.0728, + "step": 5690 + }, + { + "epoch": 1.1933319354162297, + "grad_norm": 6.537163593453765, + "learning_rate": 1.3707037470138823e-05, + "loss": 0.7303, + "step": 5691 + }, + { + "epoch": 1.1935416229817573, + "grad_norm": 5.456843612546987, + "learning_rate": 1.3704934789698305e-05, + "loss": 0.642, + "step": 5692 + }, + { + "epoch": 1.1937513105472846, + "grad_norm": 6.972496139600782, + "learning_rate": 1.3702831919373449e-05, + "loss": 0.867, + "step": 5693 + }, + { + "epoch": 1.1939609981128119, + "grad_norm": 6.572162711965942, + "learning_rate": 1.370072885927204e-05, + "loss": 0.705, + "step": 5694 + }, + { + "epoch": 1.1941706856783392, + "grad_norm": 7.778179095339917, + "learning_rate": 1.3698625609501857e-05, + "loss": 0.944, + "step": 5695 + }, + { + "epoch": 1.1943803732438667, + "grad_norm": 6.357895186306804, + "learning_rate": 1.3696522170170696e-05, + "loss": 0.697, + "step": 5696 + }, + { + "epoch": 1.194590060809394, + "grad_norm": 5.8972221361853485, + "learning_rate": 1.3694418541386361e-05, + "loss": 0.5594, + "step": 5697 + }, + { + "epoch": 1.1947997483749213, + "grad_norm": 8.618190455193671, + "learning_rate": 1.3692314723256672e-05, + "loss": 0.8786, + "step": 5698 + }, + { + "epoch": 1.1950094359404488, + "grad_norm": 6.7017434015992645, + "learning_rate": 1.3690210715889445e-05, + "loss": 0.7151, + "step": 5699 + }, + { + "epoch": 1.1952191235059761, + "grad_norm": 6.059835278007399, + "learning_rate": 1.3688106519392523e-05, + "loss": 0.7323, + "step": 5700 + }, + { + "epoch": 1.1954288110715034, + "grad_norm": 8.125017227328817, + "learning_rate": 1.3686002133873742e-05, + "loss": 1.0936, + "step": 5701 + }, + { + "epoch": 1.1956384986370308, + "grad_norm": 9.414138666612521, + "learning_rate": 1.368389755944096e-05, + "loss": 1.0402, + "step": 5702 + }, + { + "epoch": 1.1958481862025583, + "grad_norm": 8.127317288872796, + "learning_rate": 1.3681792796202036e-05, + "loss": 1.0002, + "step": 5703 + }, + { + "epoch": 1.1960578737680856, + "grad_norm": 6.777870999255757, + "learning_rate": 1.3679687844264847e-05, + "loss": 0.5327, + "step": 5704 + }, + { + "epoch": 1.196267561333613, + "grad_norm": 7.790669362522949, + "learning_rate": 1.3677582703737273e-05, + "loss": 0.8775, + "step": 5705 + }, + { + "epoch": 1.1964772488991402, + "grad_norm": 8.645957764865734, + "learning_rate": 1.3675477374727208e-05, + "loss": 1.1264, + "step": 5706 + }, + { + "epoch": 1.1966869364646677, + "grad_norm": 7.03945421016497, + "learning_rate": 1.3673371857342552e-05, + "loss": 0.8234, + "step": 5707 + }, + { + "epoch": 1.196896624030195, + "grad_norm": 6.001833355127882, + "learning_rate": 1.3671266151691218e-05, + "loss": 0.8453, + "step": 5708 + }, + { + "epoch": 1.1971063115957223, + "grad_norm": 6.521835098649482, + "learning_rate": 1.3669160257881129e-05, + "loss": 0.7212, + "step": 5709 + }, + { + "epoch": 1.1973159991612496, + "grad_norm": 6.650573122010452, + "learning_rate": 1.3667054176020208e-05, + "loss": 0.6899, + "step": 5710 + }, + { + "epoch": 1.1975256867267772, + "grad_norm": 6.268032820509766, + "learning_rate": 1.3664947906216405e-05, + "loss": 0.826, + "step": 5711 + }, + { + "epoch": 1.1977353742923045, + "grad_norm": 8.564742689990075, + "learning_rate": 1.3662841448577662e-05, + "loss": 0.9765, + "step": 5712 + }, + { + "epoch": 1.1979450618578318, + "grad_norm": 6.21533456164436, + "learning_rate": 1.3660734803211944e-05, + "loss": 0.6996, + "step": 5713 + }, + { + "epoch": 1.198154749423359, + "grad_norm": 7.504434566476642, + "learning_rate": 1.3658627970227217e-05, + "loss": 0.8581, + "step": 5714 + }, + { + "epoch": 1.1983644369888866, + "grad_norm": 7.1183965613350475, + "learning_rate": 1.3656520949731462e-05, + "loss": 0.6891, + "step": 5715 + }, + { + "epoch": 1.198574124554414, + "grad_norm": 5.930331315207084, + "learning_rate": 1.3654413741832661e-05, + "loss": 0.6662, + "step": 5716 + }, + { + "epoch": 1.1987838121199412, + "grad_norm": 6.493826608451485, + "learning_rate": 1.3652306346638824e-05, + "loss": 0.8084, + "step": 5717 + }, + { + "epoch": 1.1989934996854688, + "grad_norm": 7.326420822225876, + "learning_rate": 1.365019876425795e-05, + "loss": 0.9489, + "step": 5718 + }, + { + "epoch": 1.199203187250996, + "grad_norm": 8.382636046841132, + "learning_rate": 1.3648090994798058e-05, + "loss": 0.9105, + "step": 5719 + }, + { + "epoch": 1.1994128748165234, + "grad_norm": 6.6749204942302685, + "learning_rate": 1.3645983038367177e-05, + "loss": 0.6622, + "step": 5720 + }, + { + "epoch": 1.1996225623820507, + "grad_norm": 7.020075640003441, + "learning_rate": 1.3643874895073341e-05, + "loss": 0.7872, + "step": 5721 + }, + { + "epoch": 1.1998322499475782, + "grad_norm": 8.708283237671763, + "learning_rate": 1.3641766565024592e-05, + "loss": 1.0609, + "step": 5722 + }, + { + "epoch": 1.2000419375131055, + "grad_norm": 8.026461588750792, + "learning_rate": 1.3639658048328997e-05, + "loss": 1.0342, + "step": 5723 + }, + { + "epoch": 1.2002516250786328, + "grad_norm": 7.397692312896382, + "learning_rate": 1.3637549345094611e-05, + "loss": 0.9278, + "step": 5724 + }, + { + "epoch": 1.2004613126441601, + "grad_norm": 6.716244304076115, + "learning_rate": 1.3635440455429512e-05, + "loss": 0.7962, + "step": 5725 + }, + { + "epoch": 1.2006710002096876, + "grad_norm": 8.754467628706172, + "learning_rate": 1.3633331379441784e-05, + "loss": 0.8922, + "step": 5726 + }, + { + "epoch": 1.200880687775215, + "grad_norm": 7.589019317339162, + "learning_rate": 1.3631222117239522e-05, + "loss": 0.9809, + "step": 5727 + }, + { + "epoch": 1.2010903753407423, + "grad_norm": 6.398821724311576, + "learning_rate": 1.3629112668930829e-05, + "loss": 0.6275, + "step": 5728 + }, + { + "epoch": 1.2013000629062696, + "grad_norm": 7.626480589037755, + "learning_rate": 1.3627003034623819e-05, + "loss": 1.0377, + "step": 5729 + }, + { + "epoch": 1.201509750471797, + "grad_norm": 8.479275665629238, + "learning_rate": 1.3624893214426607e-05, + "loss": 0.9998, + "step": 5730 + }, + { + "epoch": 1.2017194380373244, + "grad_norm": 8.001897589863493, + "learning_rate": 1.3622783208447334e-05, + "loss": 0.8685, + "step": 5731 + }, + { + "epoch": 1.2019291256028517, + "grad_norm": 7.297073278857654, + "learning_rate": 1.3620673016794139e-05, + "loss": 0.7937, + "step": 5732 + }, + { + "epoch": 1.202138813168379, + "grad_norm": 6.839904216823691, + "learning_rate": 1.3618562639575171e-05, + "loss": 0.8639, + "step": 5733 + }, + { + "epoch": 1.2023485007339065, + "grad_norm": 6.8544338934169975, + "learning_rate": 1.3616452076898592e-05, + "loss": 0.8296, + "step": 5734 + }, + { + "epoch": 1.2025581882994338, + "grad_norm": 6.562994680606597, + "learning_rate": 1.3614341328872571e-05, + "loss": 0.6985, + "step": 5735 + }, + { + "epoch": 1.2027678758649611, + "grad_norm": 8.827253288509787, + "learning_rate": 1.3612230395605288e-05, + "loss": 1.0727, + "step": 5736 + }, + { + "epoch": 1.2029775634304887, + "grad_norm": 8.029930089777471, + "learning_rate": 1.3610119277204934e-05, + "loss": 1.0457, + "step": 5737 + }, + { + "epoch": 1.203187250996016, + "grad_norm": 7.329344568919264, + "learning_rate": 1.3608007973779707e-05, + "loss": 0.8419, + "step": 5738 + }, + { + "epoch": 1.2033969385615433, + "grad_norm": 6.168357207548109, + "learning_rate": 1.3605896485437806e-05, + "loss": 0.5288, + "step": 5739 + }, + { + "epoch": 1.2036066261270706, + "grad_norm": 6.396981215655067, + "learning_rate": 1.3603784812287464e-05, + "loss": 0.7379, + "step": 5740 + }, + { + "epoch": 1.2038163136925981, + "grad_norm": 7.3451230911238055, + "learning_rate": 1.3601672954436897e-05, + "loss": 0.8139, + "step": 5741 + }, + { + "epoch": 1.2040260012581254, + "grad_norm": 7.6025934306842435, + "learning_rate": 1.3599560911994343e-05, + "loss": 0.9753, + "step": 5742 + }, + { + "epoch": 1.2042356888236527, + "grad_norm": 6.512066890172266, + "learning_rate": 1.3597448685068051e-05, + "loss": 0.7649, + "step": 5743 + }, + { + "epoch": 1.20444537638918, + "grad_norm": 6.902733762901995, + "learning_rate": 1.3595336273766275e-05, + "loss": 0.8065, + "step": 5744 + }, + { + "epoch": 1.2046550639547076, + "grad_norm": 6.294020155438753, + "learning_rate": 1.3593223678197277e-05, + "loss": 0.5302, + "step": 5745 + }, + { + "epoch": 1.2048647515202349, + "grad_norm": 8.148232140380923, + "learning_rate": 1.3591110898469335e-05, + "loss": 0.9913, + "step": 5746 + }, + { + "epoch": 1.2050744390857622, + "grad_norm": 7.991627440028552, + "learning_rate": 1.358899793469073e-05, + "loss": 1.1258, + "step": 5747 + }, + { + "epoch": 1.2052841266512897, + "grad_norm": 7.200093339637468, + "learning_rate": 1.3586884786969756e-05, + "loss": 0.8399, + "step": 5748 + }, + { + "epoch": 1.205493814216817, + "grad_norm": 7.7696380929135875, + "learning_rate": 1.3584771455414716e-05, + "loss": 0.8772, + "step": 5749 + }, + { + "epoch": 1.2057035017823443, + "grad_norm": 6.841433987534582, + "learning_rate": 1.3582657940133923e-05, + "loss": 0.7387, + "step": 5750 + }, + { + "epoch": 1.2059131893478716, + "grad_norm": 8.138228327916083, + "learning_rate": 1.3580544241235693e-05, + "loss": 1.0431, + "step": 5751 + }, + { + "epoch": 1.206122876913399, + "grad_norm": 6.915714214525624, + "learning_rate": 1.3578430358828365e-05, + "loss": 0.844, + "step": 5752 + }, + { + "epoch": 1.2063325644789265, + "grad_norm": 6.3925473917296065, + "learning_rate": 1.3576316293020273e-05, + "loss": 0.8788, + "step": 5753 + }, + { + "epoch": 1.2065422520444538, + "grad_norm": 6.438985375449511, + "learning_rate": 1.3574202043919766e-05, + "loss": 0.713, + "step": 5754 + }, + { + "epoch": 1.206751939609981, + "grad_norm": 6.835101481405855, + "learning_rate": 1.3572087611635208e-05, + "loss": 0.675, + "step": 5755 + }, + { + "epoch": 1.2069616271755086, + "grad_norm": 6.983494548832377, + "learning_rate": 1.356997299627496e-05, + "loss": 0.7774, + "step": 5756 + }, + { + "epoch": 1.207171314741036, + "grad_norm": 7.442430330131553, + "learning_rate": 1.356785819794741e-05, + "loss": 0.951, + "step": 5757 + }, + { + "epoch": 1.2073810023065632, + "grad_norm": 5.9332606574959845, + "learning_rate": 1.3565743216760936e-05, + "loss": 0.6644, + "step": 5758 + }, + { + "epoch": 1.2075906898720905, + "grad_norm": 8.40179973789322, + "learning_rate": 1.356362805282394e-05, + "loss": 1.0526, + "step": 5759 + }, + { + "epoch": 1.207800377437618, + "grad_norm": 6.761574912786429, + "learning_rate": 1.3561512706244821e-05, + "loss": 0.826, + "step": 5760 + }, + { + "epoch": 1.2080100650031453, + "grad_norm": 7.45411495615844, + "learning_rate": 1.3559397177132004e-05, + "loss": 0.8968, + "step": 5761 + }, + { + "epoch": 1.2082197525686726, + "grad_norm": 7.164678889769534, + "learning_rate": 1.3557281465593904e-05, + "loss": 0.7948, + "step": 5762 + }, + { + "epoch": 1.2084294401342, + "grad_norm": 7.213025581436023, + "learning_rate": 1.355516557173896e-05, + "loss": 0.7809, + "step": 5763 + }, + { + "epoch": 1.2086391276997275, + "grad_norm": 6.858731692746869, + "learning_rate": 1.3553049495675615e-05, + "loss": 0.794, + "step": 5764 + }, + { + "epoch": 1.2088488152652548, + "grad_norm": 7.05163035365398, + "learning_rate": 1.355093323751232e-05, + "loss": 0.7515, + "step": 5765 + }, + { + "epoch": 1.209058502830782, + "grad_norm": 7.183114214983822, + "learning_rate": 1.3548816797357538e-05, + "loss": 0.8056, + "step": 5766 + }, + { + "epoch": 1.2092681903963096, + "grad_norm": 7.364000561030981, + "learning_rate": 1.3546700175319742e-05, + "loss": 0.778, + "step": 5767 + }, + { + "epoch": 1.209477877961837, + "grad_norm": 7.656695137902211, + "learning_rate": 1.3544583371507406e-05, + "loss": 0.9765, + "step": 5768 + }, + { + "epoch": 1.2096875655273642, + "grad_norm": 7.025927988790655, + "learning_rate": 1.354246638602903e-05, + "loss": 0.9057, + "step": 5769 + }, + { + "epoch": 1.2098972530928915, + "grad_norm": 6.538520109419792, + "learning_rate": 1.3540349218993105e-05, + "loss": 0.8847, + "step": 5770 + }, + { + "epoch": 1.2101069406584188, + "grad_norm": 9.108695219956731, + "learning_rate": 1.3538231870508144e-05, + "loss": 0.9617, + "step": 5771 + }, + { + "epoch": 1.2103166282239464, + "grad_norm": 7.751211697026719, + "learning_rate": 1.353611434068266e-05, + "loss": 0.7606, + "step": 5772 + }, + { + "epoch": 1.2105263157894737, + "grad_norm": 7.588075721721866, + "learning_rate": 1.3533996629625186e-05, + "loss": 0.8311, + "step": 5773 + }, + { + "epoch": 1.210736003355001, + "grad_norm": 7.6830240271379, + "learning_rate": 1.3531878737444251e-05, + "loss": 0.729, + "step": 5774 + }, + { + "epoch": 1.2109456909205285, + "grad_norm": 6.040714517724161, + "learning_rate": 1.3529760664248409e-05, + "loss": 0.6059, + "step": 5775 + }, + { + "epoch": 1.2111553784860558, + "grad_norm": 6.861154118087046, + "learning_rate": 1.3527642410146208e-05, + "loss": 0.8101, + "step": 5776 + }, + { + "epoch": 1.2113650660515831, + "grad_norm": 6.969826662307672, + "learning_rate": 1.3525523975246218e-05, + "loss": 0.8381, + "step": 5777 + }, + { + "epoch": 1.2115747536171104, + "grad_norm": 6.707459184089527, + "learning_rate": 1.3523405359657007e-05, + "loss": 0.7922, + "step": 5778 + }, + { + "epoch": 1.211784441182638, + "grad_norm": 7.115086227233861, + "learning_rate": 1.3521286563487164e-05, + "loss": 0.8251, + "step": 5779 + }, + { + "epoch": 1.2119941287481653, + "grad_norm": 10.161048983402722, + "learning_rate": 1.3519167586845272e-05, + "loss": 1.1279, + "step": 5780 + }, + { + "epoch": 1.2122038163136926, + "grad_norm": 8.65768740740381, + "learning_rate": 1.3517048429839941e-05, + "loss": 0.8399, + "step": 5781 + }, + { + "epoch": 1.2124135038792199, + "grad_norm": 6.7985642857851305, + "learning_rate": 1.3514929092579779e-05, + "loss": 0.8635, + "step": 5782 + }, + { + "epoch": 1.2126231914447474, + "grad_norm": 9.130974509997364, + "learning_rate": 1.35128095751734e-05, + "loss": 1.0259, + "step": 5783 + }, + { + "epoch": 1.2128328790102747, + "grad_norm": 6.8095952310766785, + "learning_rate": 1.351068987772944e-05, + "loss": 0.8597, + "step": 5784 + }, + { + "epoch": 1.213042566575802, + "grad_norm": 6.289008042887556, + "learning_rate": 1.3508570000356536e-05, + "loss": 0.8336, + "step": 5785 + }, + { + "epoch": 1.2132522541413295, + "grad_norm": 7.488260055714972, + "learning_rate": 1.3506449943163334e-05, + "loss": 0.8446, + "step": 5786 + }, + { + "epoch": 1.2134619417068568, + "grad_norm": 7.404368186402006, + "learning_rate": 1.3504329706258487e-05, + "loss": 0.8246, + "step": 5787 + }, + { + "epoch": 1.2136716292723841, + "grad_norm": 5.540675882143761, + "learning_rate": 1.350220928975067e-05, + "loss": 0.6501, + "step": 5788 + }, + { + "epoch": 1.2138813168379115, + "grad_norm": 7.7800573776504, + "learning_rate": 1.3500088693748548e-05, + "loss": 0.9843, + "step": 5789 + }, + { + "epoch": 1.2140910044034388, + "grad_norm": 7.932426653927388, + "learning_rate": 1.3497967918360814e-05, + "loss": 1.0091, + "step": 5790 + }, + { + "epoch": 1.2143006919689663, + "grad_norm": 5.506493200739908, + "learning_rate": 1.3495846963696155e-05, + "loss": 0.5913, + "step": 5791 + }, + { + "epoch": 1.2145103795344936, + "grad_norm": 6.477688603750444, + "learning_rate": 1.3493725829863275e-05, + "loss": 0.8906, + "step": 5792 + }, + { + "epoch": 1.214720067100021, + "grad_norm": 6.654095785939124, + "learning_rate": 1.3491604516970888e-05, + "loss": 0.9041, + "step": 5793 + }, + { + "epoch": 1.2149297546655484, + "grad_norm": 7.760193316581298, + "learning_rate": 1.3489483025127717e-05, + "loss": 0.9124, + "step": 5794 + }, + { + "epoch": 1.2151394422310757, + "grad_norm": 6.618860455220378, + "learning_rate": 1.3487361354442482e-05, + "loss": 0.9493, + "step": 5795 + }, + { + "epoch": 1.215349129796603, + "grad_norm": 8.493228930089309, + "learning_rate": 1.3485239505023934e-05, + "loss": 1.1713, + "step": 5796 + }, + { + "epoch": 1.2155588173621303, + "grad_norm": 9.080628066415878, + "learning_rate": 1.3483117476980813e-05, + "loss": 1.1585, + "step": 5797 + }, + { + "epoch": 1.2157685049276579, + "grad_norm": 7.9495879968555645, + "learning_rate": 1.3480995270421884e-05, + "loss": 0.9118, + "step": 5798 + }, + { + "epoch": 1.2159781924931852, + "grad_norm": 6.435364138855383, + "learning_rate": 1.3478872885455907e-05, + "loss": 0.8682, + "step": 5799 + }, + { + "epoch": 1.2161878800587125, + "grad_norm": 8.46589264710626, + "learning_rate": 1.3476750322191663e-05, + "loss": 1.0381, + "step": 5800 + }, + { + "epoch": 1.2163975676242398, + "grad_norm": 8.427985165070345, + "learning_rate": 1.3474627580737934e-05, + "loss": 0.9015, + "step": 5801 + }, + { + "epoch": 1.2166072551897673, + "grad_norm": 7.1154281494012475, + "learning_rate": 1.3472504661203517e-05, + "loss": 0.8885, + "step": 5802 + }, + { + "epoch": 1.2168169427552946, + "grad_norm": 7.057007463857058, + "learning_rate": 1.3470381563697208e-05, + "loss": 0.836, + "step": 5803 + }, + { + "epoch": 1.217026630320822, + "grad_norm": 6.589679220047355, + "learning_rate": 1.346825828832783e-05, + "loss": 0.6272, + "step": 5804 + }, + { + "epoch": 1.2172363178863495, + "grad_norm": 7.132657349968361, + "learning_rate": 1.3466134835204194e-05, + "loss": 0.9581, + "step": 5805 + }, + { + "epoch": 1.2174460054518768, + "grad_norm": 6.67662517805874, + "learning_rate": 1.3464011204435141e-05, + "loss": 0.8307, + "step": 5806 + }, + { + "epoch": 1.217655693017404, + "grad_norm": 7.334988777728441, + "learning_rate": 1.3461887396129503e-05, + "loss": 1.0547, + "step": 5807 + }, + { + "epoch": 1.2178653805829314, + "grad_norm": 5.914237707344567, + "learning_rate": 1.3459763410396133e-05, + "loss": 0.6177, + "step": 5808 + }, + { + "epoch": 1.2180750681484587, + "grad_norm": 6.9676987454875965, + "learning_rate": 1.3457639247343885e-05, + "loss": 0.9422, + "step": 5809 + }, + { + "epoch": 1.2182847557139862, + "grad_norm": 7.905012194279419, + "learning_rate": 1.3455514907081635e-05, + "loss": 0.9716, + "step": 5810 + }, + { + "epoch": 1.2184944432795135, + "grad_norm": 6.313732395166456, + "learning_rate": 1.3453390389718246e-05, + "loss": 0.8202, + "step": 5811 + }, + { + "epoch": 1.2187041308450408, + "grad_norm": 7.945915293192182, + "learning_rate": 1.3451265695362615e-05, + "loss": 0.7789, + "step": 5812 + }, + { + "epoch": 1.2189138184105683, + "grad_norm": 7.380412735034448, + "learning_rate": 1.3449140824123626e-05, + "loss": 0.7683, + "step": 5813 + }, + { + "epoch": 1.2191235059760956, + "grad_norm": 7.178414898355785, + "learning_rate": 1.344701577611019e-05, + "loss": 0.9125, + "step": 5814 + }, + { + "epoch": 1.219333193541623, + "grad_norm": 7.401430215930856, + "learning_rate": 1.3444890551431215e-05, + "loss": 0.9301, + "step": 5815 + }, + { + "epoch": 1.2195428811071503, + "grad_norm": 8.920824643369071, + "learning_rate": 1.3442765150195627e-05, + "loss": 1.0505, + "step": 5816 + }, + { + "epoch": 1.2197525686726778, + "grad_norm": 6.611875757190905, + "learning_rate": 1.3440639572512355e-05, + "loss": 0.8275, + "step": 5817 + }, + { + "epoch": 1.219962256238205, + "grad_norm": 6.877354754794797, + "learning_rate": 1.3438513818490333e-05, + "loss": 0.8539, + "step": 5818 + }, + { + "epoch": 1.2201719438037324, + "grad_norm": 7.781236188227104, + "learning_rate": 1.343638788823852e-05, + "loss": 0.9723, + "step": 5819 + }, + { + "epoch": 1.2203816313692597, + "grad_norm": 8.605524634137819, + "learning_rate": 1.3434261781865865e-05, + "loss": 0.9765, + "step": 5820 + }, + { + "epoch": 1.2205913189347872, + "grad_norm": 7.901321182569662, + "learning_rate": 1.3432135499481333e-05, + "loss": 1.0103, + "step": 5821 + }, + { + "epoch": 1.2208010065003145, + "grad_norm": 7.516978977100672, + "learning_rate": 1.343000904119391e-05, + "loss": 1.1042, + "step": 5822 + }, + { + "epoch": 1.2210106940658418, + "grad_norm": 7.49216359570243, + "learning_rate": 1.3427882407112574e-05, + "loss": 1.063, + "step": 5823 + }, + { + "epoch": 1.2212203816313694, + "grad_norm": 8.128059586001783, + "learning_rate": 1.3425755597346313e-05, + "loss": 0.9747, + "step": 5824 + }, + { + "epoch": 1.2214300691968967, + "grad_norm": 6.93213545359693, + "learning_rate": 1.3423628612004143e-05, + "loss": 0.6665, + "step": 5825 + }, + { + "epoch": 1.221639756762424, + "grad_norm": 7.433121774727369, + "learning_rate": 1.3421501451195065e-05, + "loss": 0.9259, + "step": 5826 + }, + { + "epoch": 1.2218494443279513, + "grad_norm": 6.5885814651724655, + "learning_rate": 1.3419374115028105e-05, + "loss": 0.7764, + "step": 5827 + }, + { + "epoch": 1.2220591318934788, + "grad_norm": 7.399287103784915, + "learning_rate": 1.341724660361229e-05, + "loss": 0.8835, + "step": 5828 + }, + { + "epoch": 1.2222688194590061, + "grad_norm": 6.947010390574959, + "learning_rate": 1.341511891705666e-05, + "loss": 0.872, + "step": 5829 + }, + { + "epoch": 1.2224785070245334, + "grad_norm": 8.0832772624248, + "learning_rate": 1.3412991055470262e-05, + "loss": 0.9682, + "step": 5830 + }, + { + "epoch": 1.2226881945900607, + "grad_norm": 8.126337344578992, + "learning_rate": 1.3410863018962155e-05, + "loss": 0.8755, + "step": 5831 + }, + { + "epoch": 1.2228978821555883, + "grad_norm": 6.260063604829489, + "learning_rate": 1.3408734807641398e-05, + "loss": 1.0027, + "step": 5832 + }, + { + "epoch": 1.2231075697211156, + "grad_norm": 6.560116287988558, + "learning_rate": 1.3406606421617075e-05, + "loss": 0.5288, + "step": 5833 + }, + { + "epoch": 1.2233172572866429, + "grad_norm": 5.534130860741491, + "learning_rate": 1.340447786099826e-05, + "loss": 0.6164, + "step": 5834 + }, + { + "epoch": 1.2235269448521702, + "grad_norm": 7.241061585545885, + "learning_rate": 1.3402349125894052e-05, + "loss": 0.9813, + "step": 5835 + }, + { + "epoch": 1.2237366324176977, + "grad_norm": 7.313403086803212, + "learning_rate": 1.3400220216413549e-05, + "loss": 0.7714, + "step": 5836 + }, + { + "epoch": 1.223946319983225, + "grad_norm": 8.064592007794385, + "learning_rate": 1.3398091132665863e-05, + "loss": 0.9183, + "step": 5837 + }, + { + "epoch": 1.2241560075487523, + "grad_norm": 7.0345402871027645, + "learning_rate": 1.3395961874760115e-05, + "loss": 0.6848, + "step": 5838 + }, + { + "epoch": 1.2243656951142796, + "grad_norm": 7.094041493020795, + "learning_rate": 1.339383244280543e-05, + "loss": 0.8814, + "step": 5839 + }, + { + "epoch": 1.2245753826798071, + "grad_norm": 7.638651665661071, + "learning_rate": 1.3391702836910946e-05, + "loss": 0.9442, + "step": 5840 + }, + { + "epoch": 1.2247850702453345, + "grad_norm": 6.784243208199457, + "learning_rate": 1.3389573057185807e-05, + "loss": 0.9653, + "step": 5841 + }, + { + "epoch": 1.2249947578108618, + "grad_norm": 6.499346902308113, + "learning_rate": 1.338744310373917e-05, + "loss": 0.7606, + "step": 5842 + }, + { + "epoch": 1.2252044453763893, + "grad_norm": 5.899680961661284, + "learning_rate": 1.33853129766802e-05, + "loss": 0.5521, + "step": 5843 + }, + { + "epoch": 1.2254141329419166, + "grad_norm": 7.695289028320727, + "learning_rate": 1.3383182676118067e-05, + "loss": 1.0382, + "step": 5844 + }, + { + "epoch": 1.225623820507444, + "grad_norm": 5.960646240841862, + "learning_rate": 1.3381052202161957e-05, + "loss": 0.6716, + "step": 5845 + }, + { + "epoch": 1.2258335080729712, + "grad_norm": 6.363195936557256, + "learning_rate": 1.3378921554921056e-05, + "loss": 0.7683, + "step": 5846 + }, + { + "epoch": 1.2260431956384987, + "grad_norm": 6.415364351995851, + "learning_rate": 1.3376790734504564e-05, + "loss": 0.8462, + "step": 5847 + }, + { + "epoch": 1.226252883204026, + "grad_norm": 7.603025485521094, + "learning_rate": 1.3374659741021692e-05, + "loss": 0.9919, + "step": 5848 + }, + { + "epoch": 1.2264625707695533, + "grad_norm": 6.292475174734174, + "learning_rate": 1.3372528574581653e-05, + "loss": 0.7473, + "step": 5849 + }, + { + "epoch": 1.2266722583350806, + "grad_norm": 7.320104911572224, + "learning_rate": 1.3370397235293676e-05, + "loss": 0.8022, + "step": 5850 + }, + { + "epoch": 1.2268819459006082, + "grad_norm": 7.197282582784523, + "learning_rate": 1.3368265723266995e-05, + "loss": 0.9807, + "step": 5851 + }, + { + "epoch": 1.2270916334661355, + "grad_norm": 7.576549698116118, + "learning_rate": 1.3366134038610855e-05, + "loss": 0.9105, + "step": 5852 + }, + { + "epoch": 1.2273013210316628, + "grad_norm": 7.039019074210788, + "learning_rate": 1.3364002181434503e-05, + "loss": 0.9779, + "step": 5853 + }, + { + "epoch": 1.22751100859719, + "grad_norm": 6.181529393634165, + "learning_rate": 1.3361870151847211e-05, + "loss": 0.657, + "step": 5854 + }, + { + "epoch": 1.2277206961627176, + "grad_norm": 6.8707855577299, + "learning_rate": 1.335973794995824e-05, + "loss": 0.7589, + "step": 5855 + }, + { + "epoch": 1.227930383728245, + "grad_norm": 6.804275218537813, + "learning_rate": 1.3357605575876872e-05, + "loss": 0.821, + "step": 5856 + }, + { + "epoch": 1.2281400712937722, + "grad_norm": 8.743401802296475, + "learning_rate": 1.3355473029712393e-05, + "loss": 1.0944, + "step": 5857 + }, + { + "epoch": 1.2283497588592995, + "grad_norm": 8.962245158971008, + "learning_rate": 1.3353340311574101e-05, + "loss": 0.8815, + "step": 5858 + }, + { + "epoch": 1.228559446424827, + "grad_norm": 8.356869150861097, + "learning_rate": 1.3351207421571304e-05, + "loss": 0.9533, + "step": 5859 + }, + { + "epoch": 1.2287691339903544, + "grad_norm": 7.54195111888495, + "learning_rate": 1.3349074359813317e-05, + "loss": 0.7816, + "step": 5860 + }, + { + "epoch": 1.2289788215558817, + "grad_norm": 6.902442437315407, + "learning_rate": 1.3346941126409454e-05, + "loss": 0.7994, + "step": 5861 + }, + { + "epoch": 1.2291885091214092, + "grad_norm": 9.026227841637313, + "learning_rate": 1.3344807721469058e-05, + "loss": 0.8414, + "step": 5862 + }, + { + "epoch": 1.2293981966869365, + "grad_norm": 7.49394277908484, + "learning_rate": 1.3342674145101464e-05, + "loss": 1.0682, + "step": 5863 + }, + { + "epoch": 1.2296078842524638, + "grad_norm": 8.065999280809644, + "learning_rate": 1.3340540397416022e-05, + "loss": 1.1124, + "step": 5864 + }, + { + "epoch": 1.2298175718179911, + "grad_norm": 6.282781222155777, + "learning_rate": 1.333840647852209e-05, + "loss": 0.6416, + "step": 5865 + }, + { + "epoch": 1.2300272593835186, + "grad_norm": 5.932990583435032, + "learning_rate": 1.3336272388529037e-05, + "loss": 0.6896, + "step": 5866 + }, + { + "epoch": 1.230236946949046, + "grad_norm": 7.978718459026874, + "learning_rate": 1.3334138127546237e-05, + "loss": 0.995, + "step": 5867 + }, + { + "epoch": 1.2304466345145733, + "grad_norm": 7.252933352731293, + "learning_rate": 1.3332003695683076e-05, + "loss": 0.9244, + "step": 5868 + }, + { + "epoch": 1.2306563220801006, + "grad_norm": 7.184662412193372, + "learning_rate": 1.3329869093048945e-05, + "loss": 0.8365, + "step": 5869 + }, + { + "epoch": 1.230866009645628, + "grad_norm": 7.230314584662014, + "learning_rate": 1.332773431975325e-05, + "loss": 0.9782, + "step": 5870 + }, + { + "epoch": 1.2310756972111554, + "grad_norm": 6.564339943227598, + "learning_rate": 1.3325599375905398e-05, + "loss": 0.7735, + "step": 5871 + }, + { + "epoch": 1.2312853847766827, + "grad_norm": 7.081147035610259, + "learning_rate": 1.3323464261614812e-05, + "loss": 0.9147, + "step": 5872 + }, + { + "epoch": 1.2314950723422102, + "grad_norm": 6.551852590609151, + "learning_rate": 1.3321328976990914e-05, + "loss": 0.9308, + "step": 5873 + }, + { + "epoch": 1.2317047599077375, + "grad_norm": 7.67515793737451, + "learning_rate": 1.3319193522143147e-05, + "loss": 0.9228, + "step": 5874 + }, + { + "epoch": 1.2319144474732648, + "grad_norm": 10.019820072331635, + "learning_rate": 1.3317057897180958e-05, + "loss": 1.0824, + "step": 5875 + }, + { + "epoch": 1.2321241350387921, + "grad_norm": 6.691562703091251, + "learning_rate": 1.3314922102213794e-05, + "loss": 0.8399, + "step": 5876 + }, + { + "epoch": 1.2323338226043195, + "grad_norm": 5.566115250426404, + "learning_rate": 1.3312786137351128e-05, + "loss": 0.7229, + "step": 5877 + }, + { + "epoch": 1.232543510169847, + "grad_norm": 8.668973974034632, + "learning_rate": 1.3310650002702424e-05, + "loss": 1.0549, + "step": 5878 + }, + { + "epoch": 1.2327531977353743, + "grad_norm": 8.331689781545807, + "learning_rate": 1.3308513698377166e-05, + "loss": 1.0334, + "step": 5879 + }, + { + "epoch": 1.2329628853009016, + "grad_norm": 5.538892876202873, + "learning_rate": 1.3306377224484841e-05, + "loss": 0.686, + "step": 5880 + }, + { + "epoch": 1.2331725728664291, + "grad_norm": 8.221833789811694, + "learning_rate": 1.3304240581134953e-05, + "loss": 0.8611, + "step": 5881 + }, + { + "epoch": 1.2333822604319564, + "grad_norm": 10.305839729022868, + "learning_rate": 1.3302103768436998e-05, + "loss": 1.1235, + "step": 5882 + }, + { + "epoch": 1.2335919479974837, + "grad_norm": 8.713198883172174, + "learning_rate": 1.3299966786500502e-05, + "loss": 0.8551, + "step": 5883 + }, + { + "epoch": 1.233801635563011, + "grad_norm": 7.012397694723294, + "learning_rate": 1.3297829635434982e-05, + "loss": 1.0189, + "step": 5884 + }, + { + "epoch": 1.2340113231285386, + "grad_norm": 7.369799091393783, + "learning_rate": 1.3295692315349977e-05, + "loss": 0.9282, + "step": 5885 + }, + { + "epoch": 1.2342210106940659, + "grad_norm": 9.457098879881555, + "learning_rate": 1.3293554826355021e-05, + "loss": 0.7594, + "step": 5886 + }, + { + "epoch": 1.2344306982595932, + "grad_norm": 6.554775324120577, + "learning_rate": 1.3291417168559671e-05, + "loss": 0.7042, + "step": 5887 + }, + { + "epoch": 1.2346403858251205, + "grad_norm": 8.81949058365872, + "learning_rate": 1.3289279342073477e-05, + "loss": 0.8508, + "step": 5888 + }, + { + "epoch": 1.234850073390648, + "grad_norm": 6.594094679849617, + "learning_rate": 1.3287141347006018e-05, + "loss": 0.8406, + "step": 5889 + }, + { + "epoch": 1.2350597609561753, + "grad_norm": 8.230547223398219, + "learning_rate": 1.3285003183466861e-05, + "loss": 0.9463, + "step": 5890 + }, + { + "epoch": 1.2352694485217026, + "grad_norm": 7.241667764598087, + "learning_rate": 1.3282864851565593e-05, + "loss": 0.6958, + "step": 5891 + }, + { + "epoch": 1.2354791360872301, + "grad_norm": 6.277505374648453, + "learning_rate": 1.3280726351411807e-05, + "loss": 0.6448, + "step": 5892 + }, + { + "epoch": 1.2356888236527575, + "grad_norm": 7.868498131415277, + "learning_rate": 1.3278587683115109e-05, + "loss": 0.9432, + "step": 5893 + }, + { + "epoch": 1.2358985112182848, + "grad_norm": 7.717147087834003, + "learning_rate": 1.3276448846785102e-05, + "loss": 0.8075, + "step": 5894 + }, + { + "epoch": 1.236108198783812, + "grad_norm": 7.610037105363087, + "learning_rate": 1.327430984253141e-05, + "loss": 1.0663, + "step": 5895 + }, + { + "epoch": 1.2363178863493394, + "grad_norm": 6.854136103788508, + "learning_rate": 1.327217067046366e-05, + "loss": 0.8529, + "step": 5896 + }, + { + "epoch": 1.236527573914867, + "grad_norm": 7.23677996899156, + "learning_rate": 1.3270031330691486e-05, + "loss": 0.8035, + "step": 5897 + }, + { + "epoch": 1.2367372614803942, + "grad_norm": 7.319395921905429, + "learning_rate": 1.3267891823324538e-05, + "loss": 0.9607, + "step": 5898 + }, + { + "epoch": 1.2369469490459215, + "grad_norm": 8.266866387025116, + "learning_rate": 1.3265752148472465e-05, + "loss": 0.8838, + "step": 5899 + }, + { + "epoch": 1.237156636611449, + "grad_norm": 6.529974215078343, + "learning_rate": 1.3263612306244929e-05, + "loss": 0.7246, + "step": 5900 + }, + { + "epoch": 1.2373663241769763, + "grad_norm": 7.331703540721676, + "learning_rate": 1.3261472296751603e-05, + "loss": 0.7812, + "step": 5901 + }, + { + "epoch": 1.2375760117425036, + "grad_norm": 7.8433271606143204, + "learning_rate": 1.3259332120102165e-05, + "loss": 0.9865, + "step": 5902 + }, + { + "epoch": 1.237785699308031, + "grad_norm": 7.222709676453146, + "learning_rate": 1.3257191776406301e-05, + "loss": 0.8065, + "step": 5903 + }, + { + "epoch": 1.2379953868735585, + "grad_norm": 7.5281333583430685, + "learning_rate": 1.3255051265773713e-05, + "loss": 0.9838, + "step": 5904 + }, + { + "epoch": 1.2382050744390858, + "grad_norm": 7.797138188646394, + "learning_rate": 1.3252910588314096e-05, + "loss": 0.9126, + "step": 5905 + }, + { + "epoch": 1.238414762004613, + "grad_norm": 8.284317230859948, + "learning_rate": 1.3250769744137176e-05, + "loss": 0.8831, + "step": 5906 + }, + { + "epoch": 1.2386244495701404, + "grad_norm": 7.942609078798129, + "learning_rate": 1.3248628733352664e-05, + "loss": 1.069, + "step": 5907 + }, + { + "epoch": 1.238834137135668, + "grad_norm": 8.081077799230988, + "learning_rate": 1.3246487556070298e-05, + "loss": 0.8513, + "step": 5908 + }, + { + "epoch": 1.2390438247011952, + "grad_norm": 7.556997169509274, + "learning_rate": 1.324434621239981e-05, + "loss": 0.8617, + "step": 5909 + }, + { + "epoch": 1.2392535122667225, + "grad_norm": 7.170677953907514, + "learning_rate": 1.3242204702450956e-05, + "loss": 0.9087, + "step": 5910 + }, + { + "epoch": 1.23946319983225, + "grad_norm": 8.97477581899254, + "learning_rate": 1.3240063026333485e-05, + "loss": 1.0813, + "step": 5911 + }, + { + "epoch": 1.2396728873977774, + "grad_norm": 8.183028029277942, + "learning_rate": 1.3237921184157165e-05, + "loss": 1.0831, + "step": 5912 + }, + { + "epoch": 1.2398825749633047, + "grad_norm": 7.7608822384758644, + "learning_rate": 1.3235779176031768e-05, + "loss": 0.9606, + "step": 5913 + }, + { + "epoch": 1.240092262528832, + "grad_norm": 9.086904608769242, + "learning_rate": 1.3233637002067077e-05, + "loss": 0.9414, + "step": 5914 + }, + { + "epoch": 1.2403019500943593, + "grad_norm": 6.580209896617262, + "learning_rate": 1.3231494662372879e-05, + "loss": 0.8229, + "step": 5915 + }, + { + "epoch": 1.2405116376598868, + "grad_norm": 6.783376017936515, + "learning_rate": 1.3229352157058976e-05, + "loss": 0.8793, + "step": 5916 + }, + { + "epoch": 1.2407213252254141, + "grad_norm": 7.159058781754046, + "learning_rate": 1.3227209486235175e-05, + "loss": 0.7659, + "step": 5917 + }, + { + "epoch": 1.2409310127909414, + "grad_norm": 7.184444666981426, + "learning_rate": 1.3225066650011292e-05, + "loss": 0.7275, + "step": 5918 + }, + { + "epoch": 1.241140700356469, + "grad_norm": 5.427450237819178, + "learning_rate": 1.3222923648497145e-05, + "loss": 0.725, + "step": 5919 + }, + { + "epoch": 1.2413503879219963, + "grad_norm": 8.222908120250661, + "learning_rate": 1.3220780481802575e-05, + "loss": 0.8327, + "step": 5920 + }, + { + "epoch": 1.2415600754875236, + "grad_norm": 7.200754136202807, + "learning_rate": 1.321863715003742e-05, + "loss": 0.8968, + "step": 5921 + }, + { + "epoch": 1.2417697630530509, + "grad_norm": 6.4969008449583505, + "learning_rate": 1.3216493653311524e-05, + "loss": 0.7659, + "step": 5922 + }, + { + "epoch": 1.2419794506185784, + "grad_norm": 7.062470127460296, + "learning_rate": 1.3214349991734752e-05, + "loss": 0.7546, + "step": 5923 + }, + { + "epoch": 1.2421891381841057, + "grad_norm": 7.564135408119068, + "learning_rate": 1.3212206165416968e-05, + "loss": 0.8186, + "step": 5924 + }, + { + "epoch": 1.242398825749633, + "grad_norm": 7.040137155336007, + "learning_rate": 1.3210062174468045e-05, + "loss": 0.9118, + "step": 5925 + }, + { + "epoch": 1.2426085133151603, + "grad_norm": 5.983356165001147, + "learning_rate": 1.320791801899787e-05, + "loss": 0.6286, + "step": 5926 + }, + { + "epoch": 1.2428182008806878, + "grad_norm": 7.993299373583643, + "learning_rate": 1.3205773699116335e-05, + "loss": 1.0812, + "step": 5927 + }, + { + "epoch": 1.2430278884462151, + "grad_norm": 6.637840888719781, + "learning_rate": 1.320362921493333e-05, + "loss": 0.7251, + "step": 5928 + }, + { + "epoch": 1.2432375760117425, + "grad_norm": 6.563449298794639, + "learning_rate": 1.320148456655878e-05, + "loss": 0.8709, + "step": 5929 + }, + { + "epoch": 1.24344726357727, + "grad_norm": 7.581467297966803, + "learning_rate": 1.3199339754102588e-05, + "loss": 0.7732, + "step": 5930 + }, + { + "epoch": 1.2436569511427973, + "grad_norm": 6.8058803671494665, + "learning_rate": 1.3197194777674686e-05, + "loss": 0.9513, + "step": 5931 + }, + { + "epoch": 1.2438666387083246, + "grad_norm": 6.32267997053899, + "learning_rate": 1.3195049637385009e-05, + "loss": 0.8289, + "step": 5932 + }, + { + "epoch": 1.244076326273852, + "grad_norm": 6.875654871593721, + "learning_rate": 1.3192904333343495e-05, + "loss": 0.8299, + "step": 5933 + }, + { + "epoch": 1.2442860138393792, + "grad_norm": 7.086009104465749, + "learning_rate": 1.3190758865660094e-05, + "loss": 0.8272, + "step": 5934 + }, + { + "epoch": 1.2444957014049067, + "grad_norm": 7.251641598268852, + "learning_rate": 1.3188613234444769e-05, + "loss": 0.7386, + "step": 5935 + }, + { + "epoch": 1.244705388970434, + "grad_norm": 6.987856155535251, + "learning_rate": 1.3186467439807485e-05, + "loss": 0.8563, + "step": 5936 + }, + { + "epoch": 1.2449150765359613, + "grad_norm": 6.508274965666494, + "learning_rate": 1.3184321481858215e-05, + "loss": 0.7127, + "step": 5937 + }, + { + "epoch": 1.2451247641014889, + "grad_norm": 7.31527588142438, + "learning_rate": 1.318217536070695e-05, + "loss": 0.8898, + "step": 5938 + }, + { + "epoch": 1.2453344516670162, + "grad_norm": 6.915994760615957, + "learning_rate": 1.3180029076463679e-05, + "loss": 0.8771, + "step": 5939 + }, + { + "epoch": 1.2455441392325435, + "grad_norm": 6.828163396740463, + "learning_rate": 1.3177882629238398e-05, + "loss": 0.9628, + "step": 5940 + }, + { + "epoch": 1.2457538267980708, + "grad_norm": 6.068641268380271, + "learning_rate": 1.3175736019141125e-05, + "loss": 0.7807, + "step": 5941 + }, + { + "epoch": 1.2459635143635983, + "grad_norm": 7.6292257742566525, + "learning_rate": 1.3173589246281872e-05, + "loss": 1.144, + "step": 5942 + }, + { + "epoch": 1.2461732019291256, + "grad_norm": 7.133807228112248, + "learning_rate": 1.3171442310770666e-05, + "loss": 0.8635, + "step": 5943 + }, + { + "epoch": 1.246382889494653, + "grad_norm": 6.643684556682252, + "learning_rate": 1.316929521271754e-05, + "loss": 0.8486, + "step": 5944 + }, + { + "epoch": 1.2465925770601802, + "grad_norm": 6.992939035955345, + "learning_rate": 1.3167147952232537e-05, + "loss": 0.7983, + "step": 5945 + }, + { + "epoch": 1.2468022646257078, + "grad_norm": 6.585236598907351, + "learning_rate": 1.316500052942571e-05, + "loss": 0.7214, + "step": 5946 + }, + { + "epoch": 1.247011952191235, + "grad_norm": 8.448339676903569, + "learning_rate": 1.3162852944407112e-05, + "loss": 0.7745, + "step": 5947 + }, + { + "epoch": 1.2472216397567624, + "grad_norm": 7.327930690348754, + "learning_rate": 1.3160705197286819e-05, + "loss": 0.868, + "step": 5948 + }, + { + "epoch": 1.24743132732229, + "grad_norm": 7.1969647483099495, + "learning_rate": 1.3158557288174905e-05, + "loss": 0.6885, + "step": 5949 + }, + { + "epoch": 1.2476410148878172, + "grad_norm": 6.961021058922199, + "learning_rate": 1.3156409217181448e-05, + "loss": 0.8385, + "step": 5950 + }, + { + "epoch": 1.2478507024533445, + "grad_norm": 10.56644835049462, + "learning_rate": 1.3154260984416542e-05, + "loss": 1.0422, + "step": 5951 + }, + { + "epoch": 1.2480603900188718, + "grad_norm": 6.223421505031867, + "learning_rate": 1.3152112589990293e-05, + "loss": 0.7384, + "step": 5952 + }, + { + "epoch": 1.2482700775843993, + "grad_norm": 6.9770299030327685, + "learning_rate": 1.3149964034012807e-05, + "loss": 0.8129, + "step": 5953 + }, + { + "epoch": 1.2484797651499266, + "grad_norm": 7.711382813914579, + "learning_rate": 1.31478153165942e-05, + "loss": 0.7714, + "step": 5954 + }, + { + "epoch": 1.248689452715454, + "grad_norm": 8.345489197240461, + "learning_rate": 1.3145666437844597e-05, + "loss": 0.8987, + "step": 5955 + }, + { + "epoch": 1.2488991402809813, + "grad_norm": 7.0084718188874655, + "learning_rate": 1.3143517397874137e-05, + "loss": 0.7883, + "step": 5956 + }, + { + "epoch": 1.2491088278465088, + "grad_norm": 8.071885252970802, + "learning_rate": 1.3141368196792954e-05, + "loss": 0.8244, + "step": 5957 + }, + { + "epoch": 1.249318515412036, + "grad_norm": 7.179253261128992, + "learning_rate": 1.3139218834711205e-05, + "loss": 0.778, + "step": 5958 + }, + { + "epoch": 1.2495282029775634, + "grad_norm": 8.735841735426847, + "learning_rate": 1.3137069311739047e-05, + "loss": 1.0267, + "step": 5959 + }, + { + "epoch": 1.2497378905430907, + "grad_norm": 8.12252458657569, + "learning_rate": 1.3134919627986645e-05, + "loss": 0.8295, + "step": 5960 + }, + { + "epoch": 1.2499475781086182, + "grad_norm": 6.837977156436172, + "learning_rate": 1.3132769783564171e-05, + "loss": 0.7049, + "step": 5961 + }, + { + "epoch": 1.2501572656741455, + "grad_norm": 9.144628938970335, + "learning_rate": 1.3130619778581819e-05, + "loss": 0.9484, + "step": 5962 + }, + { + "epoch": 1.2503669532396728, + "grad_norm": 7.0659514310338505, + "learning_rate": 1.3128469613149766e-05, + "loss": 0.7887, + "step": 5963 + }, + { + "epoch": 1.2505766408052001, + "grad_norm": 6.622287998410311, + "learning_rate": 1.3126319287378225e-05, + "loss": 0.7791, + "step": 5964 + }, + { + "epoch": 1.2507863283707277, + "grad_norm": 6.773965691230501, + "learning_rate": 1.3124168801377395e-05, + "loss": 0.7283, + "step": 5965 + }, + { + "epoch": 1.250996015936255, + "grad_norm": 7.525730484773722, + "learning_rate": 1.3122018155257497e-05, + "loss": 0.8708, + "step": 5966 + }, + { + "epoch": 1.2512057035017823, + "grad_norm": 5.8133519404578164, + "learning_rate": 1.3119867349128752e-05, + "loss": 0.6654, + "step": 5967 + }, + { + "epoch": 1.2514153910673098, + "grad_norm": 8.545979295970767, + "learning_rate": 1.3117716383101398e-05, + "loss": 0.967, + "step": 5968 + }, + { + "epoch": 1.2516250786328371, + "grad_norm": 7.940506567516569, + "learning_rate": 1.3115565257285668e-05, + "loss": 0.9418, + "step": 5969 + }, + { + "epoch": 1.2518347661983644, + "grad_norm": 7.981152185695404, + "learning_rate": 1.3113413971791818e-05, + "loss": 0.8263, + "step": 5970 + }, + { + "epoch": 1.2520444537638917, + "grad_norm": 7.808213978015147, + "learning_rate": 1.31112625267301e-05, + "loss": 0.746, + "step": 5971 + }, + { + "epoch": 1.252254141329419, + "grad_norm": 9.568949866504491, + "learning_rate": 1.310911092221078e-05, + "loss": 1.1228, + "step": 5972 + }, + { + "epoch": 1.2524638288949466, + "grad_norm": 7.515001352412217, + "learning_rate": 1.3106959158344135e-05, + "loss": 0.7434, + "step": 5973 + }, + { + "epoch": 1.2526735164604739, + "grad_norm": 6.8584554543040515, + "learning_rate": 1.3104807235240444e-05, + "loss": 0.6662, + "step": 5974 + }, + { + "epoch": 1.2528832040260012, + "grad_norm": 6.656636325948288, + "learning_rate": 1.3102655153009996e-05, + "loss": 0.7594, + "step": 5975 + }, + { + "epoch": 1.2530928915915287, + "grad_norm": 6.278901699083973, + "learning_rate": 1.3100502911763091e-05, + "loss": 0.7159, + "step": 5976 + }, + { + "epoch": 1.253302579157056, + "grad_norm": 6.719974560368373, + "learning_rate": 1.3098350511610036e-05, + "loss": 0.6479, + "step": 5977 + }, + { + "epoch": 1.2535122667225833, + "grad_norm": 5.770458756407946, + "learning_rate": 1.3096197952661137e-05, + "loss": 0.6464, + "step": 5978 + }, + { + "epoch": 1.2537219542881108, + "grad_norm": 6.626169737696493, + "learning_rate": 1.3094045235026732e-05, + "loss": 0.7278, + "step": 5979 + }, + { + "epoch": 1.2539316418536381, + "grad_norm": 6.7624388321809015, + "learning_rate": 1.3091892358817138e-05, + "loss": 0.9228, + "step": 5980 + }, + { + "epoch": 1.2541413294191655, + "grad_norm": 6.652753800238416, + "learning_rate": 1.3089739324142696e-05, + "loss": 0.751, + "step": 5981 + }, + { + "epoch": 1.2543510169846928, + "grad_norm": 7.163766340043406, + "learning_rate": 1.3087586131113756e-05, + "loss": 0.9468, + "step": 5982 + }, + { + "epoch": 1.25456070455022, + "grad_norm": 6.314057693819456, + "learning_rate": 1.3085432779840672e-05, + "loss": 0.7903, + "step": 5983 + }, + { + "epoch": 1.2547703921157476, + "grad_norm": 7.479376297304569, + "learning_rate": 1.3083279270433806e-05, + "loss": 0.604, + "step": 5984 + }, + { + "epoch": 1.254980079681275, + "grad_norm": 6.846040216219421, + "learning_rate": 1.3081125603003532e-05, + "loss": 0.7936, + "step": 5985 + }, + { + "epoch": 1.2551897672468022, + "grad_norm": 6.713330089794746, + "learning_rate": 1.3078971777660226e-05, + "loss": 0.7736, + "step": 5986 + }, + { + "epoch": 1.2553994548123297, + "grad_norm": 6.142056844816593, + "learning_rate": 1.3076817794514277e-05, + "loss": 0.7146, + "step": 5987 + }, + { + "epoch": 1.255609142377857, + "grad_norm": 7.448684058330817, + "learning_rate": 1.3074663653676079e-05, + "loss": 0.9735, + "step": 5988 + }, + { + "epoch": 1.2558188299433843, + "grad_norm": 6.453839316937063, + "learning_rate": 1.3072509355256033e-05, + "loss": 0.8268, + "step": 5989 + }, + { + "epoch": 1.2560285175089116, + "grad_norm": 5.436625043722283, + "learning_rate": 1.3070354899364557e-05, + "loss": 0.664, + "step": 5990 + }, + { + "epoch": 1.256238205074439, + "grad_norm": 7.913564847351125, + "learning_rate": 1.3068200286112068e-05, + "loss": 0.8649, + "step": 5991 + }, + { + "epoch": 1.2564478926399665, + "grad_norm": 8.355716642123957, + "learning_rate": 1.3066045515608987e-05, + "loss": 1.018, + "step": 5992 + }, + { + "epoch": 1.2566575802054938, + "grad_norm": 8.504598837953978, + "learning_rate": 1.3063890587965762e-05, + "loss": 0.9613, + "step": 5993 + }, + { + "epoch": 1.256867267771021, + "grad_norm": 8.082977945167267, + "learning_rate": 1.3061735503292826e-05, + "loss": 1.0331, + "step": 5994 + }, + { + "epoch": 1.2570769553365486, + "grad_norm": 7.656673168146104, + "learning_rate": 1.3059580261700637e-05, + "loss": 0.811, + "step": 5995 + }, + { + "epoch": 1.257286642902076, + "grad_norm": 7.68150717223882, + "learning_rate": 1.3057424863299653e-05, + "loss": 0.7938, + "step": 5996 + }, + { + "epoch": 1.2574963304676032, + "grad_norm": 7.483904362855158, + "learning_rate": 1.3055269308200345e-05, + "loss": 0.8793, + "step": 5997 + }, + { + "epoch": 1.2577060180331308, + "grad_norm": 6.496794259275839, + "learning_rate": 1.3053113596513178e-05, + "loss": 0.6599, + "step": 5998 + }, + { + "epoch": 1.257915705598658, + "grad_norm": 7.199334344955251, + "learning_rate": 1.3050957728348652e-05, + "loss": 0.9995, + "step": 5999 + }, + { + "epoch": 1.2581253931641854, + "grad_norm": 7.875658972261774, + "learning_rate": 1.3048801703817247e-05, + "loss": 0.8635, + "step": 6000 + }, + { + "epoch": 1.2583350807297127, + "grad_norm": 6.627044726569807, + "learning_rate": 1.3046645523029466e-05, + "loss": 0.8329, + "step": 6001 + }, + { + "epoch": 1.25854476829524, + "grad_norm": 7.9505949122620745, + "learning_rate": 1.3044489186095819e-05, + "loss": 0.6114, + "step": 6002 + }, + { + "epoch": 1.2587544558607675, + "grad_norm": 6.8941852458443815, + "learning_rate": 1.304233269312682e-05, + "loss": 0.8679, + "step": 6003 + }, + { + "epoch": 1.2589641434262948, + "grad_norm": 6.804248132656482, + "learning_rate": 1.3040176044232994e-05, + "loss": 0.862, + "step": 6004 + }, + { + "epoch": 1.2591738309918221, + "grad_norm": 6.849484075270946, + "learning_rate": 1.3038019239524873e-05, + "loss": 0.817, + "step": 6005 + }, + { + "epoch": 1.2593835185573496, + "grad_norm": 7.10564018806051, + "learning_rate": 1.3035862279112998e-05, + "loss": 0.6012, + "step": 6006 + }, + { + "epoch": 1.259593206122877, + "grad_norm": 7.75684757656822, + "learning_rate": 1.3033705163107912e-05, + "loss": 0.7785, + "step": 6007 + }, + { + "epoch": 1.2598028936884043, + "grad_norm": 6.913700558192165, + "learning_rate": 1.303154789162018e-05, + "loss": 0.8866, + "step": 6008 + }, + { + "epoch": 1.2600125812539316, + "grad_norm": 6.236958463205739, + "learning_rate": 1.3029390464760356e-05, + "loss": 0.6957, + "step": 6009 + }, + { + "epoch": 1.2602222688194589, + "grad_norm": 8.121305291611979, + "learning_rate": 1.3027232882639021e-05, + "loss": 0.8579, + "step": 6010 + }, + { + "epoch": 1.2604319563849864, + "grad_norm": 6.815267139775197, + "learning_rate": 1.3025075145366747e-05, + "loss": 0.8185, + "step": 6011 + }, + { + "epoch": 1.2606416439505137, + "grad_norm": 7.530695160731996, + "learning_rate": 1.3022917253054125e-05, + "loss": 0.9568, + "step": 6012 + }, + { + "epoch": 1.260851331516041, + "grad_norm": 6.589340432408899, + "learning_rate": 1.3020759205811752e-05, + "loss": 0.8335, + "step": 6013 + }, + { + "epoch": 1.2610610190815685, + "grad_norm": 6.765269403660454, + "learning_rate": 1.3018601003750234e-05, + "loss": 0.7176, + "step": 6014 + }, + { + "epoch": 1.2612707066470958, + "grad_norm": 6.8406173248452715, + "learning_rate": 1.3016442646980172e-05, + "loss": 0.628, + "step": 6015 + }, + { + "epoch": 1.2614803942126231, + "grad_norm": 6.969548983193765, + "learning_rate": 1.3014284135612197e-05, + "loss": 0.7574, + "step": 6016 + }, + { + "epoch": 1.2616900817781507, + "grad_norm": 5.361427532379264, + "learning_rate": 1.3012125469756931e-05, + "loss": 0.5628, + "step": 6017 + }, + { + "epoch": 1.261899769343678, + "grad_norm": 8.301283503428039, + "learning_rate": 1.300996664952501e-05, + "loss": 1.138, + "step": 6018 + }, + { + "epoch": 1.2621094569092053, + "grad_norm": 6.659859284718956, + "learning_rate": 1.3007807675027078e-05, + "loss": 0.5891, + "step": 6019 + }, + { + "epoch": 1.2623191444747326, + "grad_norm": 6.643783342648083, + "learning_rate": 1.3005648546373787e-05, + "loss": 0.7563, + "step": 6020 + }, + { + "epoch": 1.26252883204026, + "grad_norm": 7.046259218656, + "learning_rate": 1.3003489263675793e-05, + "loss": 0.8667, + "step": 6021 + }, + { + "epoch": 1.2627385196057874, + "grad_norm": 9.26337847646165, + "learning_rate": 1.3001329827043767e-05, + "loss": 1.0952, + "step": 6022 + }, + { + "epoch": 1.2629482071713147, + "grad_norm": 6.705901126956028, + "learning_rate": 1.2999170236588378e-05, + "loss": 0.7866, + "step": 6023 + }, + { + "epoch": 1.263157894736842, + "grad_norm": 5.90191789854271, + "learning_rate": 1.2997010492420317e-05, + "loss": 0.7158, + "step": 6024 + }, + { + "epoch": 1.2633675823023696, + "grad_norm": 7.767689446794684, + "learning_rate": 1.2994850594650268e-05, + "loss": 0.8881, + "step": 6025 + }, + { + "epoch": 1.2635772698678969, + "grad_norm": 6.807524862006748, + "learning_rate": 1.2992690543388933e-05, + "loss": 0.828, + "step": 6026 + }, + { + "epoch": 1.2637869574334242, + "grad_norm": 7.918793122539964, + "learning_rate": 1.2990530338747015e-05, + "loss": 1.0704, + "step": 6027 + }, + { + "epoch": 1.2639966449989515, + "grad_norm": 8.641864479256418, + "learning_rate": 1.2988369980835233e-05, + "loss": 0.8381, + "step": 6028 + }, + { + "epoch": 1.264206332564479, + "grad_norm": 6.939859245375967, + "learning_rate": 1.2986209469764304e-05, + "loss": 0.9765, + "step": 6029 + }, + { + "epoch": 1.2644160201300063, + "grad_norm": 8.234343725937697, + "learning_rate": 1.2984048805644962e-05, + "loss": 0.8397, + "step": 6030 + }, + { + "epoch": 1.2646257076955336, + "grad_norm": 6.5880710535819125, + "learning_rate": 1.2981887988587943e-05, + "loss": 0.6833, + "step": 6031 + }, + { + "epoch": 1.264835395261061, + "grad_norm": 7.240868543477055, + "learning_rate": 1.2979727018703991e-05, + "loss": 0.7853, + "step": 6032 + }, + { + "epoch": 1.2650450828265885, + "grad_norm": 7.148033268239787, + "learning_rate": 1.2977565896103861e-05, + "loss": 0.7809, + "step": 6033 + }, + { + "epoch": 1.2652547703921158, + "grad_norm": 6.451611374971096, + "learning_rate": 1.2975404620898315e-05, + "loss": 0.5605, + "step": 6034 + }, + { + "epoch": 1.265464457957643, + "grad_norm": 6.937914034263406, + "learning_rate": 1.2973243193198126e-05, + "loss": 0.792, + "step": 6035 + }, + { + "epoch": 1.2656741455231706, + "grad_norm": 6.747569081053799, + "learning_rate": 1.297108161311406e-05, + "loss": 0.7364, + "step": 6036 + }, + { + "epoch": 1.265883833088698, + "grad_norm": 6.656940307998814, + "learning_rate": 1.2968919880756914e-05, + "loss": 1.0242, + "step": 6037 + }, + { + "epoch": 1.2660935206542252, + "grad_norm": 6.191860676734484, + "learning_rate": 1.296675799623747e-05, + "loss": 0.8907, + "step": 6038 + }, + { + "epoch": 1.2663032082197525, + "grad_norm": 6.493201876888931, + "learning_rate": 1.2964595959666538e-05, + "loss": 0.7854, + "step": 6039 + }, + { + "epoch": 1.2665128957852798, + "grad_norm": 6.0704996155040085, + "learning_rate": 1.2962433771154916e-05, + "loss": 0.7155, + "step": 6040 + }, + { + "epoch": 1.2667225833508073, + "grad_norm": 6.917858259674632, + "learning_rate": 1.296027143081343e-05, + "loss": 0.7123, + "step": 6041 + }, + { + "epoch": 1.2669322709163346, + "grad_norm": 7.460387487135502, + "learning_rate": 1.2958108938752894e-05, + "loss": 0.8623, + "step": 6042 + }, + { + "epoch": 1.267141958481862, + "grad_norm": 7.634917059123814, + "learning_rate": 1.2955946295084147e-05, + "loss": 0.8757, + "step": 6043 + }, + { + "epoch": 1.2673516460473895, + "grad_norm": 6.328522590226969, + "learning_rate": 1.2953783499918025e-05, + "loss": 0.8166, + "step": 6044 + }, + { + "epoch": 1.2675613336129168, + "grad_norm": 6.159955982999825, + "learning_rate": 1.2951620553365375e-05, + "loss": 0.7034, + "step": 6045 + }, + { + "epoch": 1.267771021178444, + "grad_norm": 6.750301776951597, + "learning_rate": 1.2949457455537056e-05, + "loss": 0.6254, + "step": 6046 + }, + { + "epoch": 1.2679807087439716, + "grad_norm": 7.668024026960067, + "learning_rate": 1.2947294206543923e-05, + "loss": 0.9864, + "step": 6047 + }, + { + "epoch": 1.268190396309499, + "grad_norm": 6.512050588598844, + "learning_rate": 1.294513080649685e-05, + "loss": 0.656, + "step": 6048 + }, + { + "epoch": 1.2684000838750262, + "grad_norm": 7.130803554925573, + "learning_rate": 1.2942967255506717e-05, + "loss": 0.9287, + "step": 6049 + }, + { + "epoch": 1.2686097714405535, + "grad_norm": 7.82416021284202, + "learning_rate": 1.2940803553684407e-05, + "loss": 0.7829, + "step": 6050 + }, + { + "epoch": 1.2688194590060808, + "grad_norm": 6.243479781429856, + "learning_rate": 1.2938639701140816e-05, + "loss": 0.632, + "step": 6051 + }, + { + "epoch": 1.2690291465716084, + "grad_norm": 6.792160871812018, + "learning_rate": 1.2936475697986842e-05, + "loss": 0.6227, + "step": 6052 + }, + { + "epoch": 1.2692388341371357, + "grad_norm": 7.777499392545686, + "learning_rate": 1.2934311544333393e-05, + "loss": 0.943, + "step": 6053 + }, + { + "epoch": 1.269448521702663, + "grad_norm": 7.6502452832396814, + "learning_rate": 1.293214724029139e-05, + "loss": 0.958, + "step": 6054 + }, + { + "epoch": 1.2696582092681905, + "grad_norm": 6.195129361157676, + "learning_rate": 1.2929982785971756e-05, + "loss": 0.7819, + "step": 6055 + }, + { + "epoch": 1.2698678968337178, + "grad_norm": 6.4037675961670235, + "learning_rate": 1.2927818181485421e-05, + "loss": 0.691, + "step": 6056 + }, + { + "epoch": 1.2700775843992451, + "grad_norm": 7.9612141730511174, + "learning_rate": 1.2925653426943327e-05, + "loss": 0.84, + "step": 6057 + }, + { + "epoch": 1.2702872719647724, + "grad_norm": 8.572304176909645, + "learning_rate": 1.2923488522456425e-05, + "loss": 0.9259, + "step": 6058 + }, + { + "epoch": 1.2704969595302997, + "grad_norm": 7.159929219365069, + "learning_rate": 1.2921323468135661e-05, + "loss": 0.7909, + "step": 6059 + }, + { + "epoch": 1.2707066470958273, + "grad_norm": 8.513035438239559, + "learning_rate": 1.2919158264092001e-05, + "loss": 0.885, + "step": 6060 + }, + { + "epoch": 1.2709163346613546, + "grad_norm": 7.623441285484064, + "learning_rate": 1.291699291043642e-05, + "loss": 0.9277, + "step": 6061 + }, + { + "epoch": 1.2711260222268819, + "grad_norm": 7.498560927209463, + "learning_rate": 1.2914827407279891e-05, + "loss": 0.9728, + "step": 6062 + }, + { + "epoch": 1.2713357097924094, + "grad_norm": 7.974983747964251, + "learning_rate": 1.2912661754733403e-05, + "loss": 0.9296, + "step": 6063 + }, + { + "epoch": 1.2715453973579367, + "grad_norm": 7.416491805394539, + "learning_rate": 1.2910495952907952e-05, + "loss": 0.838, + "step": 6064 + }, + { + "epoch": 1.271755084923464, + "grad_norm": 8.001422166957868, + "learning_rate": 1.2908330001914527e-05, + "loss": 0.6973, + "step": 6065 + }, + { + "epoch": 1.2719647724889915, + "grad_norm": 6.360282028382773, + "learning_rate": 1.2906163901864152e-05, + "loss": 0.6531, + "step": 6066 + }, + { + "epoch": 1.2721744600545188, + "grad_norm": 7.452211990230422, + "learning_rate": 1.2903997652867833e-05, + "loss": 0.9653, + "step": 6067 + }, + { + "epoch": 1.2723841476200461, + "grad_norm": 7.855459341293004, + "learning_rate": 1.2901831255036598e-05, + "loss": 1.06, + "step": 6068 + }, + { + "epoch": 1.2725938351855735, + "grad_norm": 8.16281631007094, + "learning_rate": 1.289966470848148e-05, + "loss": 1.1338, + "step": 6069 + }, + { + "epoch": 1.2728035227511008, + "grad_norm": 7.150712832277037, + "learning_rate": 1.2897498013313516e-05, + "loss": 0.8942, + "step": 6070 + }, + { + "epoch": 1.2730132103166283, + "grad_norm": 6.554744247212476, + "learning_rate": 1.2895331169643752e-05, + "loss": 0.7974, + "step": 6071 + }, + { + "epoch": 1.2732228978821556, + "grad_norm": 6.495659659133183, + "learning_rate": 1.2893164177583246e-05, + "loss": 0.8689, + "step": 6072 + }, + { + "epoch": 1.273432585447683, + "grad_norm": 7.7328977585341665, + "learning_rate": 1.2890997037243054e-05, + "loss": 0.856, + "step": 6073 + }, + { + "epoch": 1.2736422730132104, + "grad_norm": 6.800678746219489, + "learning_rate": 1.2888829748734256e-05, + "loss": 0.8856, + "step": 6074 + }, + { + "epoch": 1.2738519605787377, + "grad_norm": 7.112425812142012, + "learning_rate": 1.2886662312167917e-05, + "loss": 0.7343, + "step": 6075 + }, + { + "epoch": 1.274061648144265, + "grad_norm": 7.3156525836739865, + "learning_rate": 1.2884494727655131e-05, + "loss": 0.8694, + "step": 6076 + }, + { + "epoch": 1.2742713357097923, + "grad_norm": 6.900085487281027, + "learning_rate": 1.2882326995306986e-05, + "loss": 0.7334, + "step": 6077 + }, + { + "epoch": 1.2744810232753196, + "grad_norm": 7.349181343720241, + "learning_rate": 1.2880159115234587e-05, + "loss": 0.988, + "step": 6078 + }, + { + "epoch": 1.2746907108408472, + "grad_norm": 6.673340383981672, + "learning_rate": 1.287799108754903e-05, + "loss": 0.9803, + "step": 6079 + }, + { + "epoch": 1.2749003984063745, + "grad_norm": 6.362015148609689, + "learning_rate": 1.2875822912361446e-05, + "loss": 0.7034, + "step": 6080 + }, + { + "epoch": 1.2751100859719018, + "grad_norm": 6.877849565391634, + "learning_rate": 1.2873654589782947e-05, + "loss": 0.8221, + "step": 6081 + }, + { + "epoch": 1.2753197735374293, + "grad_norm": 6.735173092507871, + "learning_rate": 1.2871486119924665e-05, + "loss": 0.8304, + "step": 6082 + }, + { + "epoch": 1.2755294611029566, + "grad_norm": 8.650058965209823, + "learning_rate": 1.286931750289774e-05, + "loss": 0.9, + "step": 6083 + }, + { + "epoch": 1.275739148668484, + "grad_norm": 8.100644098662862, + "learning_rate": 1.2867148738813318e-05, + "loss": 0.9384, + "step": 6084 + }, + { + "epoch": 1.2759488362340115, + "grad_norm": 6.930392269531282, + "learning_rate": 1.2864979827782547e-05, + "loss": 0.8729, + "step": 6085 + }, + { + "epoch": 1.2761585237995388, + "grad_norm": 7.1686501662433315, + "learning_rate": 1.2862810769916592e-05, + "loss": 0.8384, + "step": 6086 + }, + { + "epoch": 1.276368211365066, + "grad_norm": 7.729322065311478, + "learning_rate": 1.2860641565326622e-05, + "loss": 0.9509, + "step": 6087 + }, + { + "epoch": 1.2765778989305934, + "grad_norm": 7.613119550576094, + "learning_rate": 1.2858472214123806e-05, + "loss": 0.9419, + "step": 6088 + }, + { + "epoch": 1.2767875864961207, + "grad_norm": 7.36653954638884, + "learning_rate": 1.2856302716419335e-05, + "loss": 0.7743, + "step": 6089 + }, + { + "epoch": 1.2769972740616482, + "grad_norm": 6.390528093042797, + "learning_rate": 1.2854133072324396e-05, + "loss": 0.7955, + "step": 6090 + }, + { + "epoch": 1.2772069616271755, + "grad_norm": 5.628202163015353, + "learning_rate": 1.2851963281950184e-05, + "loss": 0.7212, + "step": 6091 + }, + { + "epoch": 1.2774166491927028, + "grad_norm": 7.256642709037587, + "learning_rate": 1.2849793345407907e-05, + "loss": 0.9678, + "step": 6092 + }, + { + "epoch": 1.2776263367582303, + "grad_norm": 8.203907299432196, + "learning_rate": 1.2847623262808784e-05, + "loss": 0.9656, + "step": 6093 + }, + { + "epoch": 1.2778360243237576, + "grad_norm": 7.653669620694143, + "learning_rate": 1.2845453034264023e-05, + "loss": 0.9931, + "step": 6094 + }, + { + "epoch": 1.278045711889285, + "grad_norm": 5.2018417891389594, + "learning_rate": 1.2843282659884863e-05, + "loss": 0.6141, + "step": 6095 + }, + { + "epoch": 1.2782553994548123, + "grad_norm": 6.491677574933637, + "learning_rate": 1.2841112139782535e-05, + "loss": 0.6638, + "step": 6096 + }, + { + "epoch": 1.2784650870203396, + "grad_norm": 7.8962487830287476, + "learning_rate": 1.2838941474068278e-05, + "loss": 1.0162, + "step": 6097 + }, + { + "epoch": 1.278674774585867, + "grad_norm": 6.257543263376296, + "learning_rate": 1.2836770662853351e-05, + "loss": 0.7475, + "step": 6098 + }, + { + "epoch": 1.2788844621513944, + "grad_norm": 7.291020476653607, + "learning_rate": 1.2834599706249007e-05, + "loss": 0.8212, + "step": 6099 + }, + { + "epoch": 1.2790941497169217, + "grad_norm": 8.860685239849312, + "learning_rate": 1.283242860436651e-05, + "loss": 0.9843, + "step": 6100 + }, + { + "epoch": 1.2793038372824492, + "grad_norm": 6.262127315021293, + "learning_rate": 1.2830257357317138e-05, + "loss": 0.8299, + "step": 6101 + }, + { + "epoch": 1.2795135248479765, + "grad_norm": 6.7897875001443015, + "learning_rate": 1.282808596521216e-05, + "loss": 0.7245, + "step": 6102 + }, + { + "epoch": 1.2797232124135038, + "grad_norm": 7.236661491488742, + "learning_rate": 1.2825914428162878e-05, + "loss": 0.8534, + "step": 6103 + }, + { + "epoch": 1.2799328999790314, + "grad_norm": 7.727077872182112, + "learning_rate": 1.282374274628058e-05, + "loss": 0.8479, + "step": 6104 + }, + { + "epoch": 1.2801425875445587, + "grad_norm": 7.554489709844892, + "learning_rate": 1.2821570919676565e-05, + "loss": 0.9154, + "step": 6105 + }, + { + "epoch": 1.280352275110086, + "grad_norm": 7.912321894382789, + "learning_rate": 1.2819398948462147e-05, + "loss": 0.9591, + "step": 6106 + }, + { + "epoch": 1.2805619626756133, + "grad_norm": 6.882910961027078, + "learning_rate": 1.2817226832748646e-05, + "loss": 0.7885, + "step": 6107 + }, + { + "epoch": 1.2807716502411406, + "grad_norm": 6.7016453850866515, + "learning_rate": 1.281505457264738e-05, + "loss": 0.8226, + "step": 6108 + }, + { + "epoch": 1.2809813378066681, + "grad_norm": 7.422305512992405, + "learning_rate": 1.281288216826969e-05, + "loss": 0.8435, + "step": 6109 + }, + { + "epoch": 1.2811910253721954, + "grad_norm": 6.648104880251532, + "learning_rate": 1.2810709619726906e-05, + "loss": 0.8304, + "step": 6110 + }, + { + "epoch": 1.2814007129377227, + "grad_norm": 5.778646604397303, + "learning_rate": 1.280853692713038e-05, + "loss": 0.6677, + "step": 6111 + }, + { + "epoch": 1.2816104005032503, + "grad_norm": 6.710588260429533, + "learning_rate": 1.2806364090591466e-05, + "loss": 0.6831, + "step": 6112 + }, + { + "epoch": 1.2818200880687776, + "grad_norm": 7.879129650124248, + "learning_rate": 1.2804191110221521e-05, + "loss": 0.8493, + "step": 6113 + }, + { + "epoch": 1.2820297756343049, + "grad_norm": 6.883871276067967, + "learning_rate": 1.2802017986131922e-05, + "loss": 1.0, + "step": 6114 + }, + { + "epoch": 1.2822394631998322, + "grad_norm": 6.494667860298186, + "learning_rate": 1.279984471843404e-05, + "loss": 0.7792, + "step": 6115 + }, + { + "epoch": 1.2824491507653595, + "grad_norm": 6.35644516407903, + "learning_rate": 1.2797671307239264e-05, + "loss": 0.593, + "step": 6116 + }, + { + "epoch": 1.282658838330887, + "grad_norm": 7.156173154624485, + "learning_rate": 1.2795497752658974e-05, + "loss": 0.8694, + "step": 6117 + }, + { + "epoch": 1.2828685258964143, + "grad_norm": 6.284506067279038, + "learning_rate": 1.2793324054804584e-05, + "loss": 0.6494, + "step": 6118 + }, + { + "epoch": 1.2830782134619416, + "grad_norm": 6.769965149791385, + "learning_rate": 1.2791150213787484e-05, + "loss": 0.7165, + "step": 6119 + }, + { + "epoch": 1.2832879010274691, + "grad_norm": 5.818174607031319, + "learning_rate": 1.27889762297191e-05, + "loss": 0.7371, + "step": 6120 + }, + { + "epoch": 1.2834975885929965, + "grad_norm": 6.349746504024223, + "learning_rate": 1.2786802102710842e-05, + "loss": 0.7428, + "step": 6121 + }, + { + "epoch": 1.2837072761585238, + "grad_norm": 7.247691257801813, + "learning_rate": 1.2784627832874148e-05, + "loss": 0.9457, + "step": 6122 + }, + { + "epoch": 1.2839169637240513, + "grad_norm": 7.162199129832679, + "learning_rate": 1.2782453420320442e-05, + "loss": 0.7769, + "step": 6123 + }, + { + "epoch": 1.2841266512895786, + "grad_norm": 9.86637734570439, + "learning_rate": 1.2780278865161177e-05, + "loss": 1.0192, + "step": 6124 + }, + { + "epoch": 1.284336338855106, + "grad_norm": 6.877951487785994, + "learning_rate": 1.2778104167507797e-05, + "loss": 0.7039, + "step": 6125 + }, + { + "epoch": 1.2845460264206332, + "grad_norm": 7.543852311829568, + "learning_rate": 1.2775929327471759e-05, + "loss": 0.7649, + "step": 6126 + }, + { + "epoch": 1.2847557139861605, + "grad_norm": 7.855847615017012, + "learning_rate": 1.2773754345164529e-05, + "loss": 0.7525, + "step": 6127 + }, + { + "epoch": 1.284965401551688, + "grad_norm": 6.954972784567074, + "learning_rate": 1.277157922069758e-05, + "loss": 0.87, + "step": 6128 + }, + { + "epoch": 1.2851750891172153, + "grad_norm": 6.775680557506945, + "learning_rate": 1.2769403954182383e-05, + "loss": 0.7465, + "step": 6129 + }, + { + "epoch": 1.2853847766827426, + "grad_norm": 8.964950437471938, + "learning_rate": 1.2767228545730435e-05, + "loss": 0.947, + "step": 6130 + }, + { + "epoch": 1.2855944642482702, + "grad_norm": 8.494369027695468, + "learning_rate": 1.2765052995453223e-05, + "loss": 0.8859, + "step": 6131 + }, + { + "epoch": 1.2858041518137975, + "grad_norm": 8.734689756328109, + "learning_rate": 1.2762877303462248e-05, + "loss": 0.9103, + "step": 6132 + }, + { + "epoch": 1.2860138393793248, + "grad_norm": 7.247235156077307, + "learning_rate": 1.2760701469869021e-05, + "loss": 0.8347, + "step": 6133 + }, + { + "epoch": 1.286223526944852, + "grad_norm": 5.8549710878830385, + "learning_rate": 1.2758525494785055e-05, + "loss": 0.7539, + "step": 6134 + }, + { + "epoch": 1.2864332145103794, + "grad_norm": 6.86861172810041, + "learning_rate": 1.2756349378321873e-05, + "loss": 0.6946, + "step": 6135 + }, + { + "epoch": 1.286642902075907, + "grad_norm": 6.9489395083889285, + "learning_rate": 1.2754173120591005e-05, + "loss": 0.8046, + "step": 6136 + }, + { + "epoch": 1.2868525896414342, + "grad_norm": 7.692398615364792, + "learning_rate": 1.2751996721703987e-05, + "loss": 1.0384, + "step": 6137 + }, + { + "epoch": 1.2870622772069615, + "grad_norm": 6.780551861828205, + "learning_rate": 1.2749820181772365e-05, + "loss": 0.7138, + "step": 6138 + }, + { + "epoch": 1.287271964772489, + "grad_norm": 5.9624238111068735, + "learning_rate": 1.274764350090769e-05, + "loss": 0.732, + "step": 6139 + }, + { + "epoch": 1.2874816523380164, + "grad_norm": 7.050752486754102, + "learning_rate": 1.2745466679221518e-05, + "loss": 0.8291, + "step": 6140 + }, + { + "epoch": 1.2876913399035437, + "grad_norm": 6.030339450243202, + "learning_rate": 1.2743289716825418e-05, + "loss": 0.5257, + "step": 6141 + }, + { + "epoch": 1.2879010274690712, + "grad_norm": 5.989484226177406, + "learning_rate": 1.2741112613830962e-05, + "loss": 0.6679, + "step": 6142 + }, + { + "epoch": 1.2881107150345985, + "grad_norm": 7.534290183116791, + "learning_rate": 1.273893537034973e-05, + "loss": 0.7587, + "step": 6143 + }, + { + "epoch": 1.2883204026001258, + "grad_norm": 7.687929943174706, + "learning_rate": 1.2736757986493309e-05, + "loss": 0.7274, + "step": 6144 + }, + { + "epoch": 1.2885300901656531, + "grad_norm": 7.798303100421033, + "learning_rate": 1.2734580462373299e-05, + "loss": 0.9392, + "step": 6145 + }, + { + "epoch": 1.2887397777311804, + "grad_norm": 7.489515855238279, + "learning_rate": 1.273240279810129e-05, + "loss": 0.7213, + "step": 6146 + }, + { + "epoch": 1.288949465296708, + "grad_norm": 6.583220140184706, + "learning_rate": 1.2730224993788904e-05, + "loss": 0.8241, + "step": 6147 + }, + { + "epoch": 1.2891591528622353, + "grad_norm": 8.243555067153068, + "learning_rate": 1.2728047049547749e-05, + "loss": 1.1078, + "step": 6148 + }, + { + "epoch": 1.2893688404277626, + "grad_norm": 8.54262841754231, + "learning_rate": 1.2725868965489452e-05, + "loss": 0.8575, + "step": 6149 + }, + { + "epoch": 1.28957852799329, + "grad_norm": 7.626448356444029, + "learning_rate": 1.2723690741725643e-05, + "loss": 1.0221, + "step": 6150 + }, + { + "epoch": 1.2897882155588174, + "grad_norm": 7.467976606899438, + "learning_rate": 1.2721512378367962e-05, + "loss": 0.8748, + "step": 6151 + }, + { + "epoch": 1.2899979031243447, + "grad_norm": 6.404718249216459, + "learning_rate": 1.2719333875528046e-05, + "loss": 0.7069, + "step": 6152 + }, + { + "epoch": 1.290207590689872, + "grad_norm": 7.3896180902530695, + "learning_rate": 1.2717155233317557e-05, + "loss": 0.9844, + "step": 6153 + }, + { + "epoch": 1.2904172782553995, + "grad_norm": 6.653349139353941, + "learning_rate": 1.2714976451848147e-05, + "loss": 0.7254, + "step": 6154 + }, + { + "epoch": 1.2906269658209268, + "grad_norm": 7.098454842492728, + "learning_rate": 1.2712797531231487e-05, + "loss": 0.8049, + "step": 6155 + }, + { + "epoch": 1.2908366533864541, + "grad_norm": 6.965573618403277, + "learning_rate": 1.2710618471579247e-05, + "loss": 0.7915, + "step": 6156 + }, + { + "epoch": 1.2910463409519815, + "grad_norm": 8.366233220298378, + "learning_rate": 1.270843927300311e-05, + "loss": 0.7872, + "step": 6157 + }, + { + "epoch": 1.291256028517509, + "grad_norm": 8.97733366528058, + "learning_rate": 1.270625993561476e-05, + "loss": 0.9534, + "step": 6158 + }, + { + "epoch": 1.2914657160830363, + "grad_norm": 7.951734529966697, + "learning_rate": 1.2704080459525899e-05, + "loss": 0.893, + "step": 6159 + }, + { + "epoch": 1.2916754036485636, + "grad_norm": 7.530115672316366, + "learning_rate": 1.270190084484822e-05, + "loss": 0.7286, + "step": 6160 + }, + { + "epoch": 1.2918850912140911, + "grad_norm": 8.3191676028206, + "learning_rate": 1.269972109169344e-05, + "loss": 0.8012, + "step": 6161 + }, + { + "epoch": 1.2920947787796184, + "grad_norm": 5.866122689971722, + "learning_rate": 1.269754120017327e-05, + "loss": 0.6695, + "step": 6162 + }, + { + "epoch": 1.2923044663451457, + "grad_norm": 9.438116207239645, + "learning_rate": 1.2695361170399434e-05, + "loss": 0.862, + "step": 6163 + }, + { + "epoch": 1.292514153910673, + "grad_norm": 7.538726566420118, + "learning_rate": 1.2693181002483664e-05, + "loss": 0.7418, + "step": 6164 + }, + { + "epoch": 1.2927238414762003, + "grad_norm": 7.098749012874384, + "learning_rate": 1.2691000696537696e-05, + "loss": 0.7402, + "step": 6165 + }, + { + "epoch": 1.2929335290417279, + "grad_norm": 6.69861885754718, + "learning_rate": 1.2688820252673275e-05, + "loss": 0.9062, + "step": 6166 + }, + { + "epoch": 1.2931432166072552, + "grad_norm": 6.7637693582362015, + "learning_rate": 1.2686639671002152e-05, + "loss": 0.8883, + "step": 6167 + }, + { + "epoch": 1.2933529041727825, + "grad_norm": 7.192430401791591, + "learning_rate": 1.2684458951636089e-05, + "loss": 0.9579, + "step": 6168 + }, + { + "epoch": 1.29356259173831, + "grad_norm": 8.0021445281955, + "learning_rate": 1.2682278094686849e-05, + "loss": 0.9076, + "step": 6169 + }, + { + "epoch": 1.2937722793038373, + "grad_norm": 6.995822728720604, + "learning_rate": 1.2680097100266201e-05, + "loss": 0.9147, + "step": 6170 + }, + { + "epoch": 1.2939819668693646, + "grad_norm": 6.614901004069066, + "learning_rate": 1.267791596848593e-05, + "loss": 0.8785, + "step": 6171 + }, + { + "epoch": 1.2941916544348921, + "grad_norm": 6.191145878951838, + "learning_rate": 1.267573469945782e-05, + "loss": 0.5883, + "step": 6172 + }, + { + "epoch": 1.2944013420004195, + "grad_norm": 5.692107098234349, + "learning_rate": 1.2673553293293667e-05, + "loss": 0.5437, + "step": 6173 + }, + { + "epoch": 1.2946110295659468, + "grad_norm": 6.263618631972242, + "learning_rate": 1.2671371750105272e-05, + "loss": 0.7545, + "step": 6174 + }, + { + "epoch": 1.294820717131474, + "grad_norm": 8.84086188893648, + "learning_rate": 1.2669190070004437e-05, + "loss": 0.8851, + "step": 6175 + }, + { + "epoch": 1.2950304046970014, + "grad_norm": 5.77569231474779, + "learning_rate": 1.2667008253102987e-05, + "loss": 0.4937, + "step": 6176 + }, + { + "epoch": 1.295240092262529, + "grad_norm": 5.27394289420634, + "learning_rate": 1.2664826299512733e-05, + "loss": 0.6302, + "step": 6177 + }, + { + "epoch": 1.2954497798280562, + "grad_norm": 6.966634249306382, + "learning_rate": 1.2662644209345514e-05, + "loss": 0.6364, + "step": 6178 + }, + { + "epoch": 1.2956594673935835, + "grad_norm": 6.656312641528478, + "learning_rate": 1.2660461982713157e-05, + "loss": 0.8662, + "step": 6179 + }, + { + "epoch": 1.295869154959111, + "grad_norm": 6.660210323673568, + "learning_rate": 1.2658279619727513e-05, + "loss": 0.6697, + "step": 6180 + }, + { + "epoch": 1.2960788425246383, + "grad_norm": 6.992780659599443, + "learning_rate": 1.2656097120500424e-05, + "loss": 0.638, + "step": 6181 + }, + { + "epoch": 1.2962885300901656, + "grad_norm": 7.253678146980893, + "learning_rate": 1.2653914485143754e-05, + "loss": 0.8442, + "step": 6182 + }, + { + "epoch": 1.296498217655693, + "grad_norm": 6.780408377707726, + "learning_rate": 1.2651731713769361e-05, + "loss": 0.8438, + "step": 6183 + }, + { + "epoch": 1.2967079052212203, + "grad_norm": 8.44902355711107, + "learning_rate": 1.264954880648912e-05, + "loss": 1.0467, + "step": 6184 + }, + { + "epoch": 1.2969175927867478, + "grad_norm": 7.298844215461077, + "learning_rate": 1.2647365763414908e-05, + "loss": 0.8692, + "step": 6185 + }, + { + "epoch": 1.297127280352275, + "grad_norm": 7.996619592315084, + "learning_rate": 1.2645182584658611e-05, + "loss": 0.9413, + "step": 6186 + }, + { + "epoch": 1.2973369679178024, + "grad_norm": 8.22172170864594, + "learning_rate": 1.2642999270332113e-05, + "loss": 0.9664, + "step": 6187 + }, + { + "epoch": 1.29754665548333, + "grad_norm": 6.6840205404660775, + "learning_rate": 1.2640815820547323e-05, + "loss": 0.6972, + "step": 6188 + }, + { + "epoch": 1.2977563430488572, + "grad_norm": 6.021940935070364, + "learning_rate": 1.263863223541614e-05, + "loss": 0.6601, + "step": 6189 + }, + { + "epoch": 1.2979660306143845, + "grad_norm": 7.9848398702695, + "learning_rate": 1.2636448515050477e-05, + "loss": 0.9068, + "step": 6190 + }, + { + "epoch": 1.298175718179912, + "grad_norm": 7.849313248200001, + "learning_rate": 1.2634264659562257e-05, + "loss": 0.9521, + "step": 6191 + }, + { + "epoch": 1.2983854057454394, + "grad_norm": 7.042980585495604, + "learning_rate": 1.2632080669063403e-05, + "loss": 0.7827, + "step": 6192 + }, + { + "epoch": 1.2985950933109667, + "grad_norm": 5.8306174754017945, + "learning_rate": 1.2629896543665852e-05, + "loss": 0.6933, + "step": 6193 + }, + { + "epoch": 1.298804780876494, + "grad_norm": 6.567095279000707, + "learning_rate": 1.262771228348154e-05, + "loss": 0.8111, + "step": 6194 + }, + { + "epoch": 1.2990144684420213, + "grad_norm": 8.034754786432387, + "learning_rate": 1.2625527888622416e-05, + "loss": 0.6704, + "step": 6195 + }, + { + "epoch": 1.2992241560075488, + "grad_norm": 8.564244070731304, + "learning_rate": 1.2623343359200433e-05, + "loss": 0.9161, + "step": 6196 + }, + { + "epoch": 1.2994338435730761, + "grad_norm": 8.158401103285374, + "learning_rate": 1.2621158695327556e-05, + "loss": 0.8502, + "step": 6197 + }, + { + "epoch": 1.2996435311386034, + "grad_norm": 6.6271004349862555, + "learning_rate": 1.2618973897115746e-05, + "loss": 0.7858, + "step": 6198 + }, + { + "epoch": 1.299853218704131, + "grad_norm": 8.191276910140028, + "learning_rate": 1.2616788964676982e-05, + "loss": 0.8722, + "step": 6199 + }, + { + "epoch": 1.3000629062696583, + "grad_norm": 7.467146373614771, + "learning_rate": 1.2614603898123248e-05, + "loss": 0.814, + "step": 6200 + }, + { + "epoch": 1.3002725938351856, + "grad_norm": 7.7613071979043005, + "learning_rate": 1.2612418697566527e-05, + "loss": 0.9195, + "step": 6201 + }, + { + "epoch": 1.3004822814007129, + "grad_norm": 7.133067589477989, + "learning_rate": 1.2610233363118815e-05, + "loss": 0.8636, + "step": 6202 + }, + { + "epoch": 1.3006919689662402, + "grad_norm": 6.937473894995899, + "learning_rate": 1.2608047894892122e-05, + "loss": 0.8048, + "step": 6203 + }, + { + "epoch": 1.3009016565317677, + "grad_norm": 7.002110125166155, + "learning_rate": 1.2605862292998443e-05, + "loss": 0.7577, + "step": 6204 + }, + { + "epoch": 1.301111344097295, + "grad_norm": 7.178718517851093, + "learning_rate": 1.2603676557549809e-05, + "loss": 0.8693, + "step": 6205 + }, + { + "epoch": 1.3013210316628223, + "grad_norm": 7.050765313582229, + "learning_rate": 1.260149068865823e-05, + "loss": 0.6457, + "step": 6206 + }, + { + "epoch": 1.3015307192283498, + "grad_norm": 6.737620128872615, + "learning_rate": 1.2599304686435745e-05, + "loss": 0.9038, + "step": 6207 + }, + { + "epoch": 1.3017404067938771, + "grad_norm": 6.801114865686833, + "learning_rate": 1.2597118550994384e-05, + "loss": 0.7783, + "step": 6208 + }, + { + "epoch": 1.3019500943594045, + "grad_norm": 6.88166367174607, + "learning_rate": 1.2594932282446197e-05, + "loss": 0.9352, + "step": 6209 + }, + { + "epoch": 1.302159781924932, + "grad_norm": 6.8814549051690586, + "learning_rate": 1.2592745880903224e-05, + "loss": 0.7024, + "step": 6210 + }, + { + "epoch": 1.3023694694904593, + "grad_norm": 5.9316108255074305, + "learning_rate": 1.2590559346477533e-05, + "loss": 0.6481, + "step": 6211 + }, + { + "epoch": 1.3025791570559866, + "grad_norm": 7.619384966572234, + "learning_rate": 1.258837267928118e-05, + "loss": 0.7787, + "step": 6212 + }, + { + "epoch": 1.302788844621514, + "grad_norm": 7.902218059628242, + "learning_rate": 1.2586185879426239e-05, + "loss": 0.9446, + "step": 6213 + }, + { + "epoch": 1.3029985321870412, + "grad_norm": 8.260981574887053, + "learning_rate": 1.2583998947024783e-05, + "loss": 0.939, + "step": 6214 + }, + { + "epoch": 1.3032082197525687, + "grad_norm": 8.948175006028508, + "learning_rate": 1.2581811882188901e-05, + "loss": 1.0613, + "step": 6215 + }, + { + "epoch": 1.303417907318096, + "grad_norm": 7.731358091462994, + "learning_rate": 1.2579624685030682e-05, + "loss": 0.8221, + "step": 6216 + }, + { + "epoch": 1.3036275948836233, + "grad_norm": 7.9562591511478225, + "learning_rate": 1.2577437355662225e-05, + "loss": 0.8381, + "step": 6217 + }, + { + "epoch": 1.3038372824491509, + "grad_norm": 7.931766029245781, + "learning_rate": 1.2575249894195634e-05, + "loss": 0.9988, + "step": 6218 + }, + { + "epoch": 1.3040469700146782, + "grad_norm": 6.916426259421991, + "learning_rate": 1.2573062300743018e-05, + "loss": 0.7585, + "step": 6219 + }, + { + "epoch": 1.3042566575802055, + "grad_norm": 7.897788449144269, + "learning_rate": 1.2570874575416494e-05, + "loss": 0.8848, + "step": 6220 + }, + { + "epoch": 1.3044663451457328, + "grad_norm": 5.580257956304535, + "learning_rate": 1.256868671832819e-05, + "loss": 0.5303, + "step": 6221 + }, + { + "epoch": 1.30467603271126, + "grad_norm": 6.703353118238242, + "learning_rate": 1.2566498729590237e-05, + "loss": 0.7187, + "step": 6222 + }, + { + "epoch": 1.3048857202767876, + "grad_norm": 5.478876745082544, + "learning_rate": 1.2564310609314774e-05, + "loss": 0.5515, + "step": 6223 + }, + { + "epoch": 1.305095407842315, + "grad_norm": 5.963225542703108, + "learning_rate": 1.2562122357613944e-05, + "loss": 0.6011, + "step": 6224 + }, + { + "epoch": 1.3053050954078422, + "grad_norm": 7.443856457198042, + "learning_rate": 1.2559933974599899e-05, + "loss": 0.8735, + "step": 6225 + }, + { + "epoch": 1.3055147829733698, + "grad_norm": 6.518360011590578, + "learning_rate": 1.25577454603848e-05, + "loss": 0.7934, + "step": 6226 + }, + { + "epoch": 1.305724470538897, + "grad_norm": 6.859765635274767, + "learning_rate": 1.2555556815080805e-05, + "loss": 0.7529, + "step": 6227 + }, + { + "epoch": 1.3059341581044244, + "grad_norm": 7.174672001320122, + "learning_rate": 1.2553368038800096e-05, + "loss": 0.8242, + "step": 6228 + }, + { + "epoch": 1.306143845669952, + "grad_norm": 7.334078835896211, + "learning_rate": 1.2551179131654845e-05, + "loss": 0.925, + "step": 6229 + }, + { + "epoch": 1.3063535332354792, + "grad_norm": 8.84104936623884, + "learning_rate": 1.2548990093757239e-05, + "loss": 1.0665, + "step": 6230 + }, + { + "epoch": 1.3065632208010065, + "grad_norm": 6.683764237453939, + "learning_rate": 1.2546800925219467e-05, + "loss": 0.8067, + "step": 6231 + }, + { + "epoch": 1.3067729083665338, + "grad_norm": 7.561359685548278, + "learning_rate": 1.2544611626153736e-05, + "loss": 0.8613, + "step": 6232 + }, + { + "epoch": 1.3069825959320611, + "grad_norm": 6.3516183897025185, + "learning_rate": 1.2542422196672237e-05, + "loss": 0.6975, + "step": 6233 + }, + { + "epoch": 1.3071922834975886, + "grad_norm": 8.256583682673003, + "learning_rate": 1.25402326368872e-05, + "loss": 1.1666, + "step": 6234 + }, + { + "epoch": 1.307401971063116, + "grad_norm": 7.320410852219375, + "learning_rate": 1.253804294691083e-05, + "loss": 0.8327, + "step": 6235 + }, + { + "epoch": 1.3076116586286433, + "grad_norm": 6.919852334485194, + "learning_rate": 1.2535853126855355e-05, + "loss": 0.8599, + "step": 6236 + }, + { + "epoch": 1.3078213461941708, + "grad_norm": 6.233577757217376, + "learning_rate": 1.2533663176833012e-05, + "loss": 0.6749, + "step": 6237 + }, + { + "epoch": 1.308031033759698, + "grad_norm": 6.912195929497092, + "learning_rate": 1.2531473096956038e-05, + "loss": 0.7492, + "step": 6238 + }, + { + "epoch": 1.3082407213252254, + "grad_norm": 6.7561583098560325, + "learning_rate": 1.2529282887336673e-05, + "loss": 0.815, + "step": 6239 + }, + { + "epoch": 1.3084504088907527, + "grad_norm": 6.528387055380406, + "learning_rate": 1.2527092548087177e-05, + "loss": 0.649, + "step": 6240 + }, + { + "epoch": 1.30866009645628, + "grad_norm": 7.171033651995904, + "learning_rate": 1.2524902079319802e-05, + "loss": 0.9021, + "step": 6241 + }, + { + "epoch": 1.3088697840218075, + "grad_norm": 9.72475951462515, + "learning_rate": 1.2522711481146818e-05, + "loss": 1.1956, + "step": 6242 + }, + { + "epoch": 1.3090794715873348, + "grad_norm": 7.2600096596647505, + "learning_rate": 1.2520520753680497e-05, + "loss": 0.8849, + "step": 6243 + }, + { + "epoch": 1.3092891591528621, + "grad_norm": 5.913500284944334, + "learning_rate": 1.2518329897033112e-05, + "loss": 0.6277, + "step": 6244 + }, + { + "epoch": 1.3094988467183897, + "grad_norm": 8.174090455288136, + "learning_rate": 1.2516138911316954e-05, + "loss": 0.8386, + "step": 6245 + }, + { + "epoch": 1.309708534283917, + "grad_norm": 6.4104434727183826, + "learning_rate": 1.251394779664431e-05, + "loss": 0.7419, + "step": 6246 + }, + { + "epoch": 1.3099182218494443, + "grad_norm": 6.827930737515556, + "learning_rate": 1.2511756553127486e-05, + "loss": 0.8364, + "step": 6247 + }, + { + "epoch": 1.3101279094149718, + "grad_norm": 6.023726820027417, + "learning_rate": 1.250956518087878e-05, + "loss": 0.8679, + "step": 6248 + }, + { + "epoch": 1.3103375969804991, + "grad_norm": 7.290356646120866, + "learning_rate": 1.2507373680010506e-05, + "loss": 0.8436, + "step": 6249 + }, + { + "epoch": 1.3105472845460264, + "grad_norm": 7.045952828052618, + "learning_rate": 1.250518205063498e-05, + "loss": 0.8245, + "step": 6250 + }, + { + "epoch": 1.3107569721115537, + "grad_norm": 7.929375944688133, + "learning_rate": 1.250299029286453e-05, + "loss": 0.875, + "step": 6251 + }, + { + "epoch": 1.310966659677081, + "grad_norm": 6.2361467197973655, + "learning_rate": 1.2500798406811487e-05, + "loss": 0.6945, + "step": 6252 + }, + { + "epoch": 1.3111763472426086, + "grad_norm": 7.88603689770897, + "learning_rate": 1.2498606392588188e-05, + "loss": 0.9239, + "step": 6253 + }, + { + "epoch": 1.3113860348081359, + "grad_norm": 6.805703368438138, + "learning_rate": 1.2496414250306978e-05, + "loss": 0.8467, + "step": 6254 + }, + { + "epoch": 1.3115957223736632, + "grad_norm": 7.2557971101967, + "learning_rate": 1.2494221980080208e-05, + "loss": 0.7178, + "step": 6255 + }, + { + "epoch": 1.3118054099391907, + "grad_norm": 6.126031817438328, + "learning_rate": 1.2492029582020234e-05, + "loss": 0.5402, + "step": 6256 + }, + { + "epoch": 1.312015097504718, + "grad_norm": 9.042391314927801, + "learning_rate": 1.2489837056239425e-05, + "loss": 0.9687, + "step": 6257 + }, + { + "epoch": 1.3122247850702453, + "grad_norm": 7.70388123918556, + "learning_rate": 1.2487644402850145e-05, + "loss": 0.755, + "step": 6258 + }, + { + "epoch": 1.3124344726357726, + "grad_norm": 7.957026053713945, + "learning_rate": 1.2485451621964777e-05, + "loss": 0.8996, + "step": 6259 + }, + { + "epoch": 1.3126441602013, + "grad_norm": 5.407750929804625, + "learning_rate": 1.24832587136957e-05, + "loss": 0.6771, + "step": 6260 + }, + { + "epoch": 1.3128538477668275, + "grad_norm": 8.659960773380002, + "learning_rate": 1.248106567815531e-05, + "loss": 0.807, + "step": 6261 + }, + { + "epoch": 1.3130635353323548, + "grad_norm": 10.173007790676033, + "learning_rate": 1.2478872515455998e-05, + "loss": 1.0889, + "step": 6262 + }, + { + "epoch": 1.313273222897882, + "grad_norm": 9.373569317172365, + "learning_rate": 1.2476679225710173e-05, + "loss": 1.1735, + "step": 6263 + }, + { + "epoch": 1.3134829104634096, + "grad_norm": 8.332991228407897, + "learning_rate": 1.2474485809030241e-05, + "loss": 1.0026, + "step": 6264 + }, + { + "epoch": 1.313692598028937, + "grad_norm": 7.650601959912915, + "learning_rate": 1.2472292265528619e-05, + "loss": 0.8604, + "step": 6265 + }, + { + "epoch": 1.3139022855944642, + "grad_norm": 7.591348670661808, + "learning_rate": 1.2470098595317731e-05, + "loss": 0.872, + "step": 6266 + }, + { + "epoch": 1.3141119731599917, + "grad_norm": 7.620015234230331, + "learning_rate": 1.2467904798510007e-05, + "loss": 0.9644, + "step": 6267 + }, + { + "epoch": 1.314321660725519, + "grad_norm": 6.799257880734605, + "learning_rate": 1.2465710875217878e-05, + "loss": 0.8346, + "step": 6268 + }, + { + "epoch": 1.3145313482910463, + "grad_norm": 6.465629735286245, + "learning_rate": 1.2463516825553795e-05, + "loss": 0.7088, + "step": 6269 + }, + { + "epoch": 1.3147410358565736, + "grad_norm": 8.147248487687248, + "learning_rate": 1.2461322649630198e-05, + "loss": 0.8596, + "step": 6270 + }, + { + "epoch": 1.314950723422101, + "grad_norm": 6.740167444057646, + "learning_rate": 1.2459128347559545e-05, + "loss": 0.7263, + "step": 6271 + }, + { + "epoch": 1.3151604109876285, + "grad_norm": 7.608792824182061, + "learning_rate": 1.2456933919454304e-05, + "loss": 0.9543, + "step": 6272 + }, + { + "epoch": 1.3153700985531558, + "grad_norm": 9.134980162498588, + "learning_rate": 1.2454739365426933e-05, + "loss": 1.2386, + "step": 6273 + }, + { + "epoch": 1.315579786118683, + "grad_norm": 8.115495131305103, + "learning_rate": 1.2452544685589916e-05, + "loss": 0.9329, + "step": 6274 + }, + { + "epoch": 1.3157894736842106, + "grad_norm": 7.003016527845287, + "learning_rate": 1.245034988005573e-05, + "loss": 0.8459, + "step": 6275 + }, + { + "epoch": 1.315999161249738, + "grad_norm": 6.16954346933709, + "learning_rate": 1.2448154948936862e-05, + "loss": 0.8558, + "step": 6276 + }, + { + "epoch": 1.3162088488152652, + "grad_norm": 6.982659948598512, + "learning_rate": 1.2445959892345805e-05, + "loss": 0.8158, + "step": 6277 + }, + { + "epoch": 1.3164185363807925, + "grad_norm": 6.710069696713367, + "learning_rate": 1.2443764710395063e-05, + "loss": 0.8759, + "step": 6278 + }, + { + "epoch": 1.31662822394632, + "grad_norm": 7.61940215903079, + "learning_rate": 1.244156940319714e-05, + "loss": 0.7999, + "step": 6279 + }, + { + "epoch": 1.3168379115118474, + "grad_norm": 7.783529227160778, + "learning_rate": 1.2439373970864552e-05, + "loss": 1.0536, + "step": 6280 + }, + { + "epoch": 1.3170475990773747, + "grad_norm": 7.7734975778951405, + "learning_rate": 1.2437178413509816e-05, + "loss": 0.8884, + "step": 6281 + }, + { + "epoch": 1.317257286642902, + "grad_norm": 5.945537235183072, + "learning_rate": 1.2434982731245462e-05, + "loss": 0.6851, + "step": 6282 + }, + { + "epoch": 1.3174669742084295, + "grad_norm": 6.721230408887563, + "learning_rate": 1.2432786924184016e-05, + "loss": 0.8461, + "step": 6283 + }, + { + "epoch": 1.3176766617739568, + "grad_norm": 6.90454831103914, + "learning_rate": 1.2430590992438023e-05, + "loss": 0.7176, + "step": 6284 + }, + { + "epoch": 1.3178863493394841, + "grad_norm": 7.320784860157427, + "learning_rate": 1.242839493612002e-05, + "loss": 0.7848, + "step": 6285 + }, + { + "epoch": 1.3180960369050116, + "grad_norm": 8.33224399177018, + "learning_rate": 1.2426198755342574e-05, + "loss": 0.9776, + "step": 6286 + }, + { + "epoch": 1.318305724470539, + "grad_norm": 6.085017141732459, + "learning_rate": 1.2424002450218229e-05, + "loss": 0.8089, + "step": 6287 + }, + { + "epoch": 1.3185154120360663, + "grad_norm": 6.730832934143617, + "learning_rate": 1.2421806020859552e-05, + "loss": 0.8583, + "step": 6288 + }, + { + "epoch": 1.3187250996015936, + "grad_norm": 8.112140020359531, + "learning_rate": 1.241960946737912e-05, + "loss": 0.7246, + "step": 6289 + }, + { + "epoch": 1.3189347871671209, + "grad_norm": 7.18437059551874, + "learning_rate": 1.2417412789889506e-05, + "loss": 0.9671, + "step": 6290 + }, + { + "epoch": 1.3191444747326484, + "grad_norm": 9.315549163660037, + "learning_rate": 1.2415215988503291e-05, + "loss": 1.2177, + "step": 6291 + }, + { + "epoch": 1.3193541622981757, + "grad_norm": 9.26430064011248, + "learning_rate": 1.241301906333307e-05, + "loss": 1.2192, + "step": 6292 + }, + { + "epoch": 1.319563849863703, + "grad_norm": 5.868347478708639, + "learning_rate": 1.2410822014491438e-05, + "loss": 0.6913, + "step": 6293 + }, + { + "epoch": 1.3197735374292305, + "grad_norm": 6.40220282808585, + "learning_rate": 1.2408624842090994e-05, + "loss": 0.5633, + "step": 6294 + }, + { + "epoch": 1.3199832249947578, + "grad_norm": 7.273805561853192, + "learning_rate": 1.2406427546244349e-05, + "loss": 0.9107, + "step": 6295 + }, + { + "epoch": 1.3201929125602851, + "grad_norm": 7.1618487126270525, + "learning_rate": 1.2404230127064118e-05, + "loss": 0.9079, + "step": 6296 + }, + { + "epoch": 1.3204026001258127, + "grad_norm": 5.863546169625546, + "learning_rate": 1.2402032584662925e-05, + "loss": 0.7765, + "step": 6297 + }, + { + "epoch": 1.32061228769134, + "grad_norm": 7.93214834814956, + "learning_rate": 1.2399834919153396e-05, + "loss": 0.7991, + "step": 6298 + }, + { + "epoch": 1.3208219752568673, + "grad_norm": 5.172872739625734, + "learning_rate": 1.2397637130648163e-05, + "loss": 0.5715, + "step": 6299 + }, + { + "epoch": 1.3210316628223946, + "grad_norm": 6.96435165217951, + "learning_rate": 1.239543921925987e-05, + "loss": 0.9448, + "step": 6300 + }, + { + "epoch": 1.321241350387922, + "grad_norm": 7.346502662725581, + "learning_rate": 1.2393241185101162e-05, + "loss": 0.9498, + "step": 6301 + }, + { + "epoch": 1.3214510379534494, + "grad_norm": 7.082736286632278, + "learning_rate": 1.2391043028284693e-05, + "loss": 0.8144, + "step": 6302 + }, + { + "epoch": 1.3216607255189767, + "grad_norm": 7.377703071815039, + "learning_rate": 1.238884474892312e-05, + "loss": 1.0032, + "step": 6303 + }, + { + "epoch": 1.321870413084504, + "grad_norm": 8.853614857485917, + "learning_rate": 1.2386646347129111e-05, + "loss": 1.0391, + "step": 6304 + }, + { + "epoch": 1.3220801006500316, + "grad_norm": 8.22072462014297, + "learning_rate": 1.238444782301534e-05, + "loss": 0.8437, + "step": 6305 + }, + { + "epoch": 1.3222897882155589, + "grad_norm": 6.871814652013581, + "learning_rate": 1.2382249176694478e-05, + "loss": 0.9364, + "step": 6306 + }, + { + "epoch": 1.3224994757810862, + "grad_norm": 5.9843673399423025, + "learning_rate": 1.2380050408279218e-05, + "loss": 0.7063, + "step": 6307 + }, + { + "epoch": 1.3227091633466135, + "grad_norm": 7.544942755081471, + "learning_rate": 1.2377851517882244e-05, + "loss": 1.0531, + "step": 6308 + }, + { + "epoch": 1.3229188509121408, + "grad_norm": 6.516591888122146, + "learning_rate": 1.2375652505616258e-05, + "loss": 0.7674, + "step": 6309 + }, + { + "epoch": 1.3231285384776683, + "grad_norm": 5.882246292901434, + "learning_rate": 1.2373453371593958e-05, + "loss": 0.5633, + "step": 6310 + }, + { + "epoch": 1.3233382260431956, + "grad_norm": 8.486226793480947, + "learning_rate": 1.237125411592806e-05, + "loss": 0.9849, + "step": 6311 + }, + { + "epoch": 1.323547913608723, + "grad_norm": 6.134628721869582, + "learning_rate": 1.236905473873127e-05, + "loss": 0.7074, + "step": 6312 + }, + { + "epoch": 1.3237576011742505, + "grad_norm": 7.628740572680764, + "learning_rate": 1.236685524011632e-05, + "loss": 0.8129, + "step": 6313 + }, + { + "epoch": 1.3239672887397778, + "grad_norm": 7.077167196447652, + "learning_rate": 1.236465562019593e-05, + "loss": 0.7938, + "step": 6314 + }, + { + "epoch": 1.324176976305305, + "grad_norm": 6.952295180426609, + "learning_rate": 1.2362455879082844e-05, + "loss": 0.888, + "step": 6315 + }, + { + "epoch": 1.3243866638708326, + "grad_norm": 6.521558790019405, + "learning_rate": 1.2360256016889793e-05, + "loss": 0.7781, + "step": 6316 + }, + { + "epoch": 1.32459635143636, + "grad_norm": 7.056200840472986, + "learning_rate": 1.2358056033729527e-05, + "loss": 0.7345, + "step": 6317 + }, + { + "epoch": 1.3248060390018872, + "grad_norm": 6.851196852108395, + "learning_rate": 1.23558559297148e-05, + "loss": 0.6554, + "step": 6318 + }, + { + "epoch": 1.3250157265674145, + "grad_norm": 9.267435237156828, + "learning_rate": 1.2353655704958373e-05, + "loss": 1.1985, + "step": 6319 + }, + { + "epoch": 1.3252254141329418, + "grad_norm": 6.879265880065429, + "learning_rate": 1.2351455359573007e-05, + "loss": 0.6786, + "step": 6320 + }, + { + "epoch": 1.3254351016984693, + "grad_norm": 7.7026664890493866, + "learning_rate": 1.2349254893671475e-05, + "loss": 0.9577, + "step": 6321 + }, + { + "epoch": 1.3256447892639966, + "grad_norm": 7.602072685651666, + "learning_rate": 1.2347054307366555e-05, + "loss": 0.863, + "step": 6322 + }, + { + "epoch": 1.325854476829524, + "grad_norm": 7.014685597640618, + "learning_rate": 1.2344853600771032e-05, + "loss": 0.661, + "step": 6323 + }, + { + "epoch": 1.3260641643950515, + "grad_norm": 5.680576666929746, + "learning_rate": 1.2342652773997694e-05, + "loss": 0.6455, + "step": 6324 + }, + { + "epoch": 1.3262738519605788, + "grad_norm": 6.1321473145104965, + "learning_rate": 1.234045182715934e-05, + "loss": 0.6388, + "step": 6325 + }, + { + "epoch": 1.326483539526106, + "grad_norm": 7.224141368985771, + "learning_rate": 1.2338250760368768e-05, + "loss": 0.8501, + "step": 6326 + }, + { + "epoch": 1.3266932270916334, + "grad_norm": 8.731481236114174, + "learning_rate": 1.2336049573738791e-05, + "loss": 1.0735, + "step": 6327 + }, + { + "epoch": 1.3269029146571607, + "grad_norm": 7.569071014910265, + "learning_rate": 1.2333848267382223e-05, + "loss": 0.8619, + "step": 6328 + }, + { + "epoch": 1.3271126022226882, + "grad_norm": 8.22828417293366, + "learning_rate": 1.2331646841411879e-05, + "loss": 0.9784, + "step": 6329 + }, + { + "epoch": 1.3273222897882155, + "grad_norm": 8.502076770287367, + "learning_rate": 1.2329445295940593e-05, + "loss": 0.6743, + "step": 6330 + }, + { + "epoch": 1.3275319773537428, + "grad_norm": 6.770356632730843, + "learning_rate": 1.2327243631081196e-05, + "loss": 0.7733, + "step": 6331 + }, + { + "epoch": 1.3277416649192704, + "grad_norm": 8.349799986331119, + "learning_rate": 1.2325041846946523e-05, + "loss": 1.0666, + "step": 6332 + }, + { + "epoch": 1.3279513524847977, + "grad_norm": 7.585850026640918, + "learning_rate": 1.2322839943649427e-05, + "loss": 0.8339, + "step": 6333 + }, + { + "epoch": 1.328161040050325, + "grad_norm": 7.225764883975388, + "learning_rate": 1.2320637921302756e-05, + "loss": 0.701, + "step": 6334 + }, + { + "epoch": 1.3283707276158525, + "grad_norm": 8.291193251846932, + "learning_rate": 1.2318435780019363e-05, + "loss": 1.0607, + "step": 6335 + }, + { + "epoch": 1.3285804151813798, + "grad_norm": 6.879350407442858, + "learning_rate": 1.2316233519912117e-05, + "loss": 0.7429, + "step": 6336 + }, + { + "epoch": 1.3287901027469071, + "grad_norm": 7.037351227203288, + "learning_rate": 1.2314031141093886e-05, + "loss": 0.7068, + "step": 6337 + }, + { + "epoch": 1.3289997903124344, + "grad_norm": 8.644403595893284, + "learning_rate": 1.2311828643677543e-05, + "loss": 1.0035, + "step": 6338 + }, + { + "epoch": 1.3292094778779617, + "grad_norm": 5.734301025286374, + "learning_rate": 1.2309626027775973e-05, + "loss": 0.5971, + "step": 6339 + }, + { + "epoch": 1.3294191654434893, + "grad_norm": 8.881082823094001, + "learning_rate": 1.2307423293502066e-05, + "loss": 0.972, + "step": 6340 + }, + { + "epoch": 1.3296288530090166, + "grad_norm": 7.080786453067711, + "learning_rate": 1.2305220440968708e-05, + "loss": 0.83, + "step": 6341 + }, + { + "epoch": 1.3298385405745439, + "grad_norm": 7.059360822935252, + "learning_rate": 1.2303017470288812e-05, + "loss": 0.7991, + "step": 6342 + }, + { + "epoch": 1.3300482281400714, + "grad_norm": 6.174910090915003, + "learning_rate": 1.230081438157527e-05, + "loss": 0.6918, + "step": 6343 + }, + { + "epoch": 1.3302579157055987, + "grad_norm": 6.684230746079871, + "learning_rate": 1.2298611174941002e-05, + "loss": 0.7191, + "step": 6344 + }, + { + "epoch": 1.330467603271126, + "grad_norm": 8.175134555093182, + "learning_rate": 1.2296407850498925e-05, + "loss": 0.938, + "step": 6345 + }, + { + "epoch": 1.3306772908366533, + "grad_norm": 6.587376230369217, + "learning_rate": 1.2294204408361962e-05, + "loss": 0.8823, + "step": 6346 + }, + { + "epoch": 1.3308869784021806, + "grad_norm": 6.933278242642745, + "learning_rate": 1.2292000848643044e-05, + "loss": 0.9722, + "step": 6347 + }, + { + "epoch": 1.3310966659677081, + "grad_norm": 7.88074255954808, + "learning_rate": 1.2289797171455109e-05, + "loss": 0.8897, + "step": 6348 + }, + { + "epoch": 1.3313063535332355, + "grad_norm": 6.633974776496631, + "learning_rate": 1.2287593376911096e-05, + "loss": 0.8409, + "step": 6349 + }, + { + "epoch": 1.3315160410987628, + "grad_norm": 7.255480027582813, + "learning_rate": 1.2285389465123956e-05, + "loss": 0.9073, + "step": 6350 + }, + { + "epoch": 1.3317257286642903, + "grad_norm": 7.101523480423554, + "learning_rate": 1.2283185436206642e-05, + "loss": 0.5874, + "step": 6351 + }, + { + "epoch": 1.3319354162298176, + "grad_norm": 7.549834765619955, + "learning_rate": 1.2280981290272112e-05, + "loss": 1.0618, + "step": 6352 + }, + { + "epoch": 1.332145103795345, + "grad_norm": 7.029536159181028, + "learning_rate": 1.2278777027433334e-05, + "loss": 0.8477, + "step": 6353 + }, + { + "epoch": 1.3323547913608724, + "grad_norm": 7.517146697267649, + "learning_rate": 1.2276572647803284e-05, + "loss": 0.7355, + "step": 6354 + }, + { + "epoch": 1.3325644789263997, + "grad_norm": 6.562714521557585, + "learning_rate": 1.2274368151494937e-05, + "loss": 0.8249, + "step": 6355 + }, + { + "epoch": 1.332774166491927, + "grad_norm": 7.206542570712602, + "learning_rate": 1.2272163538621278e-05, + "loss": 0.8267, + "step": 6356 + }, + { + "epoch": 1.3329838540574543, + "grad_norm": 6.2312876764065255, + "learning_rate": 1.2269958809295296e-05, + "loss": 0.6921, + "step": 6357 + }, + { + "epoch": 1.3331935416229816, + "grad_norm": 6.137754752903873, + "learning_rate": 1.226775396362999e-05, + "loss": 0.798, + "step": 6358 + }, + { + "epoch": 1.3334032291885092, + "grad_norm": 7.2518022117251135, + "learning_rate": 1.2265549001738358e-05, + "loss": 0.812, + "step": 6359 + }, + { + "epoch": 1.3336129167540365, + "grad_norm": 5.488058748011243, + "learning_rate": 1.2263343923733408e-05, + "loss": 0.5067, + "step": 6360 + }, + { + "epoch": 1.3338226043195638, + "grad_norm": 7.215295401337767, + "learning_rate": 1.226113872972816e-05, + "loss": 0.7912, + "step": 6361 + }, + { + "epoch": 1.3340322918850913, + "grad_norm": 6.296936005428166, + "learning_rate": 1.2258933419835628e-05, + "loss": 0.6197, + "step": 6362 + }, + { + "epoch": 1.3342419794506186, + "grad_norm": 7.106974057993462, + "learning_rate": 1.2256727994168843e-05, + "loss": 0.7379, + "step": 6363 + }, + { + "epoch": 1.334451667016146, + "grad_norm": 5.747317717355062, + "learning_rate": 1.2254522452840828e-05, + "loss": 0.6364, + "step": 6364 + }, + { + "epoch": 1.3346613545816732, + "grad_norm": 8.347439591305442, + "learning_rate": 1.2252316795964636e-05, + "loss": 0.8809, + "step": 6365 + }, + { + "epoch": 1.3348710421472005, + "grad_norm": 7.409370748224249, + "learning_rate": 1.2250111023653295e-05, + "loss": 0.6623, + "step": 6366 + }, + { + "epoch": 1.335080729712728, + "grad_norm": 7.391833095726885, + "learning_rate": 1.2247905136019865e-05, + "loss": 0.7512, + "step": 6367 + }, + { + "epoch": 1.3352904172782554, + "grad_norm": 6.626935521070651, + "learning_rate": 1.2245699133177395e-05, + "loss": 0.6602, + "step": 6368 + }, + { + "epoch": 1.3355001048437827, + "grad_norm": 7.3282605346253025, + "learning_rate": 1.2243493015238954e-05, + "loss": 0.8294, + "step": 6369 + }, + { + "epoch": 1.3357097924093102, + "grad_norm": 6.415595360205275, + "learning_rate": 1.2241286782317601e-05, + "loss": 0.6436, + "step": 6370 + }, + { + "epoch": 1.3359194799748375, + "grad_norm": 7.417513561947137, + "learning_rate": 1.2239080434526414e-05, + "loss": 0.7973, + "step": 6371 + }, + { + "epoch": 1.3361291675403648, + "grad_norm": 6.453143399047254, + "learning_rate": 1.2236873971978472e-05, + "loss": 0.7547, + "step": 6372 + }, + { + "epoch": 1.3363388551058923, + "grad_norm": 8.321888873040013, + "learning_rate": 1.2234667394786859e-05, + "loss": 0.7975, + "step": 6373 + }, + { + "epoch": 1.3365485426714196, + "grad_norm": 6.716772455735117, + "learning_rate": 1.2232460703064664e-05, + "loss": 0.9022, + "step": 6374 + }, + { + "epoch": 1.336758230236947, + "grad_norm": 6.724140798235005, + "learning_rate": 1.223025389692499e-05, + "loss": 0.7211, + "step": 6375 + }, + { + "epoch": 1.3369679178024743, + "grad_norm": 7.869258431892303, + "learning_rate": 1.2228046976480932e-05, + "loss": 0.9175, + "step": 6376 + }, + { + "epoch": 1.3371776053680016, + "grad_norm": 7.924518384950294, + "learning_rate": 1.2225839941845603e-05, + "loss": 0.7832, + "step": 6377 + }, + { + "epoch": 1.337387292933529, + "grad_norm": 8.510591336006046, + "learning_rate": 1.2223632793132116e-05, + "loss": 0.7884, + "step": 6378 + }, + { + "epoch": 1.3375969804990564, + "grad_norm": 9.622231681010305, + "learning_rate": 1.2221425530453595e-05, + "loss": 1.0785, + "step": 6379 + }, + { + "epoch": 1.3378066680645837, + "grad_norm": 7.931903944174604, + "learning_rate": 1.221921815392316e-05, + "loss": 0.7796, + "step": 6380 + }, + { + "epoch": 1.3380163556301112, + "grad_norm": 6.775823015907663, + "learning_rate": 1.2217010663653945e-05, + "loss": 0.6253, + "step": 6381 + }, + { + "epoch": 1.3382260431956385, + "grad_norm": 7.0558142876389605, + "learning_rate": 1.2214803059759088e-05, + "loss": 0.918, + "step": 6382 + }, + { + "epoch": 1.3384357307611658, + "grad_norm": 7.522407203709215, + "learning_rate": 1.2212595342351734e-05, + "loss": 0.9256, + "step": 6383 + }, + { + "epoch": 1.3386454183266931, + "grad_norm": 7.931084352932363, + "learning_rate": 1.2210387511545032e-05, + "loss": 0.8428, + "step": 6384 + }, + { + "epoch": 1.3388551058922205, + "grad_norm": 6.371398574382828, + "learning_rate": 1.2208179567452134e-05, + "loss": 0.6863, + "step": 6385 + }, + { + "epoch": 1.339064793457748, + "grad_norm": 9.453617554040981, + "learning_rate": 1.2205971510186208e-05, + "loss": 1.0095, + "step": 6386 + }, + { + "epoch": 1.3392744810232753, + "grad_norm": 7.132838109713118, + "learning_rate": 1.2203763339860408e-05, + "loss": 0.8155, + "step": 6387 + }, + { + "epoch": 1.3394841685888026, + "grad_norm": 6.8325518687272275, + "learning_rate": 1.2201555056587922e-05, + "loss": 0.7387, + "step": 6388 + }, + { + "epoch": 1.3396938561543301, + "grad_norm": 6.710857920457937, + "learning_rate": 1.2199346660481919e-05, + "loss": 0.7284, + "step": 6389 + }, + { + "epoch": 1.3399035437198574, + "grad_norm": 7.065039756417791, + "learning_rate": 1.2197138151655584e-05, + "loss": 0.7537, + "step": 6390 + }, + { + "epoch": 1.3401132312853847, + "grad_norm": 7.1288391755488965, + "learning_rate": 1.2194929530222107e-05, + "loss": 0.791, + "step": 6391 + }, + { + "epoch": 1.3403229188509123, + "grad_norm": 6.374731983833911, + "learning_rate": 1.2192720796294688e-05, + "loss": 0.6228, + "step": 6392 + }, + { + "epoch": 1.3405326064164396, + "grad_norm": 8.434942651550807, + "learning_rate": 1.219051194998652e-05, + "loss": 0.808, + "step": 6393 + }, + { + "epoch": 1.3407422939819669, + "grad_norm": 5.737027653113215, + "learning_rate": 1.2188302991410821e-05, + "loss": 0.5594, + "step": 6394 + }, + { + "epoch": 1.3409519815474942, + "grad_norm": 6.47801744183265, + "learning_rate": 1.2186093920680794e-05, + "loss": 0.5943, + "step": 6395 + }, + { + "epoch": 1.3411616691130215, + "grad_norm": 6.630583631122121, + "learning_rate": 1.2183884737909664e-05, + "loss": 0.689, + "step": 6396 + }, + { + "epoch": 1.341371356678549, + "grad_norm": 7.937045736014091, + "learning_rate": 1.2181675443210654e-05, + "loss": 0.768, + "step": 6397 + }, + { + "epoch": 1.3415810442440763, + "grad_norm": 9.476028501241416, + "learning_rate": 1.2179466036696994e-05, + "loss": 1.0816, + "step": 6398 + }, + { + "epoch": 1.3417907318096036, + "grad_norm": 8.727766426721434, + "learning_rate": 1.2177256518481915e-05, + "loss": 1.1069, + "step": 6399 + }, + { + "epoch": 1.3420004193751311, + "grad_norm": 9.04448713877718, + "learning_rate": 1.2175046888678668e-05, + "loss": 0.9713, + "step": 6400 + }, + { + "epoch": 1.3422101069406585, + "grad_norm": 6.3406632649384305, + "learning_rate": 1.2172837147400493e-05, + "loss": 0.8137, + "step": 6401 + }, + { + "epoch": 1.3424197945061858, + "grad_norm": 6.503425996822555, + "learning_rate": 1.217062729476065e-05, + "loss": 0.5634, + "step": 6402 + }, + { + "epoch": 1.342629482071713, + "grad_norm": 7.485520643998392, + "learning_rate": 1.2168417330872386e-05, + "loss": 0.6971, + "step": 6403 + }, + { + "epoch": 1.3428391696372406, + "grad_norm": 6.900948390990102, + "learning_rate": 1.2166207255848978e-05, + "loss": 0.8474, + "step": 6404 + }, + { + "epoch": 1.343048857202768, + "grad_norm": 8.168064706358221, + "learning_rate": 1.216399706980369e-05, + "loss": 0.9946, + "step": 6405 + }, + { + "epoch": 1.3432585447682952, + "grad_norm": 6.172145807535987, + "learning_rate": 1.21617867728498e-05, + "loss": 0.6357, + "step": 6406 + }, + { + "epoch": 1.3434682323338225, + "grad_norm": 6.728455847487092, + "learning_rate": 1.2159576365100584e-05, + "loss": 0.6908, + "step": 6407 + }, + { + "epoch": 1.34367791989935, + "grad_norm": 6.9947057828290164, + "learning_rate": 1.2157365846669337e-05, + "loss": 0.8733, + "step": 6408 + }, + { + "epoch": 1.3438876074648773, + "grad_norm": 8.27785496575513, + "learning_rate": 1.215515521766935e-05, + "loss": 0.9759, + "step": 6409 + }, + { + "epoch": 1.3440972950304046, + "grad_norm": 6.430511938133592, + "learning_rate": 1.215294447821392e-05, + "loss": 0.6646, + "step": 6410 + }, + { + "epoch": 1.3443069825959322, + "grad_norm": 5.14303280616504, + "learning_rate": 1.215073362841635e-05, + "loss": 0.5087, + "step": 6411 + }, + { + "epoch": 1.3445166701614595, + "grad_norm": 6.680342091727118, + "learning_rate": 1.214852266838995e-05, + "loss": 0.7957, + "step": 6412 + }, + { + "epoch": 1.3447263577269868, + "grad_norm": 8.794716128425481, + "learning_rate": 1.2146311598248037e-05, + "loss": 0.8562, + "step": 6413 + }, + { + "epoch": 1.344936045292514, + "grad_norm": 6.686944028909398, + "learning_rate": 1.2144100418103933e-05, + "loss": 0.6927, + "step": 6414 + }, + { + "epoch": 1.3451457328580414, + "grad_norm": 7.705282594170068, + "learning_rate": 1.2141889128070965e-05, + "loss": 0.9205, + "step": 6415 + }, + { + "epoch": 1.345355420423569, + "grad_norm": 7.9855756472785, + "learning_rate": 1.2139677728262462e-05, + "loss": 0.8751, + "step": 6416 + }, + { + "epoch": 1.3455651079890962, + "grad_norm": 6.169298783562788, + "learning_rate": 1.2137466218791767e-05, + "loss": 0.6533, + "step": 6417 + }, + { + "epoch": 1.3457747955546235, + "grad_norm": 7.866182801275056, + "learning_rate": 1.2135254599772219e-05, + "loss": 0.9997, + "step": 6418 + }, + { + "epoch": 1.345984483120151, + "grad_norm": 7.112278664963097, + "learning_rate": 1.213304287131717e-05, + "loss": 0.7735, + "step": 6419 + }, + { + "epoch": 1.3461941706856784, + "grad_norm": 8.535440582934076, + "learning_rate": 1.2130831033539974e-05, + "loss": 0.8853, + "step": 6420 + }, + { + "epoch": 1.3464038582512057, + "grad_norm": 6.795340600781241, + "learning_rate": 1.2128619086553994e-05, + "loss": 0.804, + "step": 6421 + }, + { + "epoch": 1.3466135458167332, + "grad_norm": 8.144412468197274, + "learning_rate": 1.2126407030472588e-05, + "loss": 0.9034, + "step": 6422 + }, + { + "epoch": 1.3468232333822605, + "grad_norm": 6.560874991517755, + "learning_rate": 1.2124194865409141e-05, + "loss": 0.8112, + "step": 6423 + }, + { + "epoch": 1.3470329209477878, + "grad_norm": 7.879052556797238, + "learning_rate": 1.212198259147702e-05, + "loss": 0.8889, + "step": 6424 + }, + { + "epoch": 1.3472426085133151, + "grad_norm": 7.184043146537791, + "learning_rate": 1.211977020878961e-05, + "loss": 0.7173, + "step": 6425 + }, + { + "epoch": 1.3474522960788424, + "grad_norm": 6.831071393072331, + "learning_rate": 1.21175577174603e-05, + "loss": 0.7138, + "step": 6426 + }, + { + "epoch": 1.34766198364437, + "grad_norm": 7.102510532461891, + "learning_rate": 1.2115345117602487e-05, + "loss": 0.8524, + "step": 6427 + }, + { + "epoch": 1.3478716712098973, + "grad_norm": 7.455383826442386, + "learning_rate": 1.2113132409329561e-05, + "loss": 0.941, + "step": 6428 + }, + { + "epoch": 1.3480813587754246, + "grad_norm": 8.64800744206449, + "learning_rate": 1.2110919592754942e-05, + "loss": 0.9091, + "step": 6429 + }, + { + "epoch": 1.348291046340952, + "grad_norm": 7.821744815761876, + "learning_rate": 1.2108706667992027e-05, + "loss": 0.8368, + "step": 6430 + }, + { + "epoch": 1.3485007339064794, + "grad_norm": 7.7635782432908, + "learning_rate": 1.2106493635154241e-05, + "loss": 1.0088, + "step": 6431 + }, + { + "epoch": 1.3487104214720067, + "grad_norm": 6.625058904731215, + "learning_rate": 1.2104280494355e-05, + "loss": 0.6071, + "step": 6432 + }, + { + "epoch": 1.348920109037534, + "grad_norm": 5.698745091138195, + "learning_rate": 1.2102067245707736e-05, + "loss": 0.6146, + "step": 6433 + }, + { + "epoch": 1.3491297966030613, + "grad_norm": 7.124762109969999, + "learning_rate": 1.209985388932588e-05, + "loss": 0.838, + "step": 6434 + }, + { + "epoch": 1.3493394841685888, + "grad_norm": 7.005309830326782, + "learning_rate": 1.209764042532287e-05, + "loss": 0.7887, + "step": 6435 + }, + { + "epoch": 1.3495491717341161, + "grad_norm": 6.906327197292887, + "learning_rate": 1.209542685381215e-05, + "loss": 0.6673, + "step": 6436 + }, + { + "epoch": 1.3497588592996435, + "grad_norm": 6.542585245536598, + "learning_rate": 1.209321317490717e-05, + "loss": 0.7678, + "step": 6437 + }, + { + "epoch": 1.349968546865171, + "grad_norm": 5.235220682094636, + "learning_rate": 1.2090999388721382e-05, + "loss": 0.4319, + "step": 6438 + }, + { + "epoch": 1.3501782344306983, + "grad_norm": 6.115632969242415, + "learning_rate": 1.2088785495368252e-05, + "loss": 0.7086, + "step": 6439 + }, + { + "epoch": 1.3503879219962256, + "grad_norm": 6.768467139053652, + "learning_rate": 1.2086571494961239e-05, + "loss": 0.796, + "step": 6440 + }, + { + "epoch": 1.3505976095617531, + "grad_norm": 6.0479325799244545, + "learning_rate": 1.2084357387613821e-05, + "loss": 0.6299, + "step": 6441 + }, + { + "epoch": 1.3508072971272804, + "grad_norm": 7.589600298761827, + "learning_rate": 1.208214317343947e-05, + "loss": 0.7612, + "step": 6442 + }, + { + "epoch": 1.3510169846928077, + "grad_norm": 7.324217942047075, + "learning_rate": 1.207992885255167e-05, + "loss": 0.8527, + "step": 6443 + }, + { + "epoch": 1.351226672258335, + "grad_norm": 8.168365694404505, + "learning_rate": 1.2077714425063911e-05, + "loss": 0.9359, + "step": 6444 + }, + { + "epoch": 1.3514363598238623, + "grad_norm": 8.216643646648595, + "learning_rate": 1.207549989108968e-05, + "loss": 0.8615, + "step": 6445 + }, + { + "epoch": 1.3516460473893899, + "grad_norm": 7.387445932479297, + "learning_rate": 1.2073285250742483e-05, + "loss": 0.7637, + "step": 6446 + }, + { + "epoch": 1.3518557349549172, + "grad_norm": 6.752606635320653, + "learning_rate": 1.207107050413582e-05, + "loss": 0.7572, + "step": 6447 + }, + { + "epoch": 1.3520654225204445, + "grad_norm": 6.9863794957832575, + "learning_rate": 1.2068855651383203e-05, + "loss": 0.6041, + "step": 6448 + }, + { + "epoch": 1.352275110085972, + "grad_norm": 7.802703222849968, + "learning_rate": 1.2066640692598145e-05, + "loss": 0.7171, + "step": 6449 + }, + { + "epoch": 1.3524847976514993, + "grad_norm": 8.625470177129106, + "learning_rate": 1.2064425627894164e-05, + "loss": 1.0682, + "step": 6450 + }, + { + "epoch": 1.3526944852170266, + "grad_norm": 7.300274909959484, + "learning_rate": 1.2062210457384791e-05, + "loss": 0.7515, + "step": 6451 + }, + { + "epoch": 1.352904172782554, + "grad_norm": 9.046372541127317, + "learning_rate": 1.2059995181183557e-05, + "loss": 1.0834, + "step": 6452 + }, + { + "epoch": 1.3531138603480812, + "grad_norm": 9.38341953118875, + "learning_rate": 1.2057779799403998e-05, + "loss": 0.9231, + "step": 6453 + }, + { + "epoch": 1.3533235479136088, + "grad_norm": 6.331139161768647, + "learning_rate": 1.2055564312159651e-05, + "loss": 0.5789, + "step": 6454 + }, + { + "epoch": 1.353533235479136, + "grad_norm": 9.529222757466018, + "learning_rate": 1.205334871956407e-05, + "loss": 0.821, + "step": 6455 + }, + { + "epoch": 1.3537429230446634, + "grad_norm": 7.744715875135254, + "learning_rate": 1.2051133021730807e-05, + "loss": 0.6338, + "step": 6456 + }, + { + "epoch": 1.353952610610191, + "grad_norm": 8.629793055923283, + "learning_rate": 1.2048917218773415e-05, + "loss": 0.9461, + "step": 6457 + }, + { + "epoch": 1.3541622981757182, + "grad_norm": 7.105329218055937, + "learning_rate": 1.2046701310805466e-05, + "loss": 0.7343, + "step": 6458 + }, + { + "epoch": 1.3543719857412455, + "grad_norm": 7.18579049201621, + "learning_rate": 1.2044485297940524e-05, + "loss": 0.7233, + "step": 6459 + }, + { + "epoch": 1.354581673306773, + "grad_norm": 7.313754544047871, + "learning_rate": 1.2042269180292166e-05, + "loss": 0.7028, + "step": 6460 + }, + { + "epoch": 1.3547913608723003, + "grad_norm": 7.754111303601662, + "learning_rate": 1.2040052957973967e-05, + "loss": 0.848, + "step": 6461 + }, + { + "epoch": 1.3550010484378276, + "grad_norm": 8.795921450170088, + "learning_rate": 1.2037836631099518e-05, + "loss": 0.865, + "step": 6462 + }, + { + "epoch": 1.355210736003355, + "grad_norm": 8.845688286286869, + "learning_rate": 1.2035620199782406e-05, + "loss": 1.0648, + "step": 6463 + }, + { + "epoch": 1.3554204235688823, + "grad_norm": 6.461648798388557, + "learning_rate": 1.2033403664136227e-05, + "loss": 0.7458, + "step": 6464 + }, + { + "epoch": 1.3556301111344098, + "grad_norm": 6.961432123646212, + "learning_rate": 1.2031187024274586e-05, + "loss": 0.9058, + "step": 6465 + }, + { + "epoch": 1.355839798699937, + "grad_norm": 7.396042813381256, + "learning_rate": 1.2028970280311086e-05, + "loss": 0.9224, + "step": 6466 + }, + { + "epoch": 1.3560494862654644, + "grad_norm": 7.580249542149914, + "learning_rate": 1.2026753432359341e-05, + "loss": 0.8751, + "step": 6467 + }, + { + "epoch": 1.356259173830992, + "grad_norm": 6.734508260976132, + "learning_rate": 1.2024536480532965e-05, + "loss": 0.8256, + "step": 6468 + }, + { + "epoch": 1.3564688613965192, + "grad_norm": 6.4974724746756705, + "learning_rate": 1.2022319424945585e-05, + "loss": 0.6532, + "step": 6469 + }, + { + "epoch": 1.3566785489620465, + "grad_norm": 6.162735406962014, + "learning_rate": 1.2020102265710826e-05, + "loss": 0.7621, + "step": 6470 + }, + { + "epoch": 1.3568882365275738, + "grad_norm": 6.902084218873949, + "learning_rate": 1.2017885002942323e-05, + "loss": 0.8191, + "step": 6471 + }, + { + "epoch": 1.3570979240931011, + "grad_norm": 7.53779500821685, + "learning_rate": 1.2015667636753714e-05, + "loss": 0.9587, + "step": 6472 + }, + { + "epoch": 1.3573076116586287, + "grad_norm": 8.31024045278443, + "learning_rate": 1.2013450167258642e-05, + "loss": 1.0757, + "step": 6473 + }, + { + "epoch": 1.357517299224156, + "grad_norm": 7.124762292842106, + "learning_rate": 1.2011232594570754e-05, + "loss": 0.9512, + "step": 6474 + }, + { + "epoch": 1.3577269867896833, + "grad_norm": 7.44272313042543, + "learning_rate": 1.2009014918803712e-05, + "loss": 0.7056, + "step": 6475 + }, + { + "epoch": 1.3579366743552108, + "grad_norm": 7.672889390881073, + "learning_rate": 1.2006797140071169e-05, + "loss": 0.8154, + "step": 6476 + }, + { + "epoch": 1.3581463619207381, + "grad_norm": 7.028042465650182, + "learning_rate": 1.2004579258486793e-05, + "loss": 0.7657, + "step": 6477 + }, + { + "epoch": 1.3583560494862654, + "grad_norm": 6.489262918135886, + "learning_rate": 1.2002361274164252e-05, + "loss": 0.8441, + "step": 6478 + }, + { + "epoch": 1.358565737051793, + "grad_norm": 6.340888648852408, + "learning_rate": 1.2000143187217224e-05, + "loss": 0.6964, + "step": 6479 + }, + { + "epoch": 1.3587754246173203, + "grad_norm": 5.752937650051768, + "learning_rate": 1.1997924997759385e-05, + "loss": 0.707, + "step": 6480 + }, + { + "epoch": 1.3589851121828476, + "grad_norm": 7.239678776006695, + "learning_rate": 1.1995706705904428e-05, + "loss": 0.906, + "step": 6481 + }, + { + "epoch": 1.3591947997483749, + "grad_norm": 7.64144017228882, + "learning_rate": 1.1993488311766041e-05, + "loss": 0.813, + "step": 6482 + }, + { + "epoch": 1.3594044873139022, + "grad_norm": 6.54028863617211, + "learning_rate": 1.199126981545792e-05, + "loss": 0.7876, + "step": 6483 + }, + { + "epoch": 1.3596141748794297, + "grad_norm": 7.012584194285227, + "learning_rate": 1.1989051217093766e-05, + "loss": 0.97, + "step": 6484 + }, + { + "epoch": 1.359823862444957, + "grad_norm": 8.063412388983604, + "learning_rate": 1.198683251678729e-05, + "loss": 0.9218, + "step": 6485 + }, + { + "epoch": 1.3600335500104843, + "grad_norm": 8.908621547150124, + "learning_rate": 1.1984613714652198e-05, + "loss": 1.1506, + "step": 6486 + }, + { + "epoch": 1.3602432375760118, + "grad_norm": 6.41822774933903, + "learning_rate": 1.1982394810802213e-05, + "loss": 0.8226, + "step": 6487 + }, + { + "epoch": 1.3604529251415391, + "grad_norm": 6.989330817839758, + "learning_rate": 1.1980175805351055e-05, + "loss": 0.8343, + "step": 6488 + }, + { + "epoch": 1.3606626127070665, + "grad_norm": 6.872273175290653, + "learning_rate": 1.1977956698412454e-05, + "loss": 0.8181, + "step": 6489 + }, + { + "epoch": 1.3608723002725938, + "grad_norm": 6.302555376329823, + "learning_rate": 1.1975737490100139e-05, + "loss": 0.5962, + "step": 6490 + }, + { + "epoch": 1.361081987838121, + "grad_norm": 9.38762671515917, + "learning_rate": 1.1973518180527852e-05, + "loss": 1.0459, + "step": 6491 + }, + { + "epoch": 1.3612916754036486, + "grad_norm": 9.107687670358956, + "learning_rate": 1.1971298769809336e-05, + "loss": 0.9862, + "step": 6492 + }, + { + "epoch": 1.361501362969176, + "grad_norm": 7.1577075495267355, + "learning_rate": 1.1969079258058338e-05, + "loss": 0.795, + "step": 6493 + }, + { + "epoch": 1.3617110505347032, + "grad_norm": 10.223138145907203, + "learning_rate": 1.1966859645388616e-05, + "loss": 0.9621, + "step": 6494 + }, + { + "epoch": 1.3619207381002307, + "grad_norm": 6.145630463673914, + "learning_rate": 1.1964639931913921e-05, + "loss": 0.6194, + "step": 6495 + }, + { + "epoch": 1.362130425665758, + "grad_norm": 6.453689442702025, + "learning_rate": 1.1962420117748026e-05, + "loss": 0.8461, + "step": 6496 + }, + { + "epoch": 1.3623401132312853, + "grad_norm": 7.745999489324701, + "learning_rate": 1.1960200203004695e-05, + "loss": 0.9723, + "step": 6497 + }, + { + "epoch": 1.3625498007968129, + "grad_norm": 8.2748131648448, + "learning_rate": 1.1957980187797703e-05, + "loss": 0.9234, + "step": 6498 + }, + { + "epoch": 1.3627594883623402, + "grad_norm": 7.071999670757471, + "learning_rate": 1.195576007224083e-05, + "loss": 0.8486, + "step": 6499 + }, + { + "epoch": 1.3629691759278675, + "grad_norm": 7.867219262641423, + "learning_rate": 1.1953539856447862e-05, + "loss": 0.9015, + "step": 6500 + }, + { + "epoch": 1.3631788634933948, + "grad_norm": 7.391838717048641, + "learning_rate": 1.1951319540532588e-05, + "loss": 0.9211, + "step": 6501 + }, + { + "epoch": 1.363388551058922, + "grad_norm": 7.052136177553956, + "learning_rate": 1.1949099124608805e-05, + "loss": 0.7448, + "step": 6502 + }, + { + "epoch": 1.3635982386244496, + "grad_norm": 6.855026564778334, + "learning_rate": 1.1946878608790308e-05, + "loss": 0.7957, + "step": 6503 + }, + { + "epoch": 1.363807926189977, + "grad_norm": 7.975799689868046, + "learning_rate": 1.1944657993190908e-05, + "loss": 0.9342, + "step": 6504 + }, + { + "epoch": 1.3640176137555042, + "grad_norm": 6.45224396444523, + "learning_rate": 1.1942437277924413e-05, + "loss": 0.7104, + "step": 6505 + }, + { + "epoch": 1.3642273013210318, + "grad_norm": 7.690934506271937, + "learning_rate": 1.1940216463104637e-05, + "loss": 0.8596, + "step": 6506 + }, + { + "epoch": 1.364436988886559, + "grad_norm": 7.584933779773345, + "learning_rate": 1.1937995548845403e-05, + "loss": 1.0767, + "step": 6507 + }, + { + "epoch": 1.3646466764520864, + "grad_norm": 8.272794502206157, + "learning_rate": 1.1935774535260539e-05, + "loss": 0.9301, + "step": 6508 + }, + { + "epoch": 1.3648563640176137, + "grad_norm": 9.093163977335657, + "learning_rate": 1.1933553422463866e-05, + "loss": 1.0307, + "step": 6509 + }, + { + "epoch": 1.365066051583141, + "grad_norm": 8.218454500695204, + "learning_rate": 1.1931332210569231e-05, + "loss": 0.8428, + "step": 6510 + }, + { + "epoch": 1.3652757391486685, + "grad_norm": 7.800110874587736, + "learning_rate": 1.192911089969047e-05, + "loss": 0.7242, + "step": 6511 + }, + { + "epoch": 1.3654854267141958, + "grad_norm": 7.149805351880659, + "learning_rate": 1.1926889489941428e-05, + "loss": 0.8924, + "step": 6512 + }, + { + "epoch": 1.3656951142797231, + "grad_norm": 6.185942571839799, + "learning_rate": 1.192466798143596e-05, + "loss": 0.6214, + "step": 6513 + }, + { + "epoch": 1.3659048018452506, + "grad_norm": 7.893766220554455, + "learning_rate": 1.1922446374287917e-05, + "loss": 0.8213, + "step": 6514 + }, + { + "epoch": 1.366114489410778, + "grad_norm": 7.622345164180235, + "learning_rate": 1.1920224668611165e-05, + "loss": 0.8677, + "step": 6515 + }, + { + "epoch": 1.3663241769763053, + "grad_norm": 7.4316140844846155, + "learning_rate": 1.1918002864519567e-05, + "loss": 0.8886, + "step": 6516 + }, + { + "epoch": 1.3665338645418328, + "grad_norm": 6.8587455534025485, + "learning_rate": 1.1915780962126997e-05, + "loss": 0.7774, + "step": 6517 + }, + { + "epoch": 1.36674355210736, + "grad_norm": 6.172199307142109, + "learning_rate": 1.1913558961547328e-05, + "loss": 0.641, + "step": 6518 + }, + { + "epoch": 1.3669532396728874, + "grad_norm": 8.750493772232634, + "learning_rate": 1.1911336862894446e-05, + "loss": 1.1288, + "step": 6519 + }, + { + "epoch": 1.3671629272384147, + "grad_norm": 6.659723599216442, + "learning_rate": 1.1909114666282233e-05, + "loss": 0.6658, + "step": 6520 + }, + { + "epoch": 1.367372614803942, + "grad_norm": 8.520531440855127, + "learning_rate": 1.1906892371824583e-05, + "loss": 0.9822, + "step": 6521 + }, + { + "epoch": 1.3675823023694695, + "grad_norm": 6.943030750636971, + "learning_rate": 1.1904669979635391e-05, + "loss": 0.6321, + "step": 6522 + }, + { + "epoch": 1.3677919899349968, + "grad_norm": 8.088101730410482, + "learning_rate": 1.1902447489828559e-05, + "loss": 0.7461, + "step": 6523 + }, + { + "epoch": 1.3680016775005241, + "grad_norm": 6.13136213200197, + "learning_rate": 1.1900224902517992e-05, + "loss": 0.695, + "step": 6524 + }, + { + "epoch": 1.3682113650660517, + "grad_norm": 7.553332933092758, + "learning_rate": 1.1898002217817606e-05, + "loss": 0.8562, + "step": 6525 + }, + { + "epoch": 1.368421052631579, + "grad_norm": 7.833618812952744, + "learning_rate": 1.1895779435841313e-05, + "loss": 0.8995, + "step": 6526 + }, + { + "epoch": 1.3686307401971063, + "grad_norm": 6.599145703370401, + "learning_rate": 1.1893556556703037e-05, + "loss": 0.7193, + "step": 6527 + }, + { + "epoch": 1.3688404277626336, + "grad_norm": 8.287542650753487, + "learning_rate": 1.1891333580516701e-05, + "loss": 0.9613, + "step": 6528 + }, + { + "epoch": 1.3690501153281611, + "grad_norm": 8.436060212836612, + "learning_rate": 1.188911050739624e-05, + "loss": 1.0079, + "step": 6529 + }, + { + "epoch": 1.3692598028936884, + "grad_norm": 7.204025088454156, + "learning_rate": 1.1886887337455587e-05, + "loss": 0.8264, + "step": 6530 + }, + { + "epoch": 1.3694694904592157, + "grad_norm": 7.766508692077376, + "learning_rate": 1.188466407080869e-05, + "loss": 0.769, + "step": 6531 + }, + { + "epoch": 1.369679178024743, + "grad_norm": 6.453662159813938, + "learning_rate": 1.1882440707569485e-05, + "loss": 0.6975, + "step": 6532 + }, + { + "epoch": 1.3698888655902706, + "grad_norm": 9.257651786173767, + "learning_rate": 1.1880217247851934e-05, + "loss": 0.8093, + "step": 6533 + }, + { + "epoch": 1.3700985531557979, + "grad_norm": 8.260206614196193, + "learning_rate": 1.1877993691769985e-05, + "loss": 0.8786, + "step": 6534 + }, + { + "epoch": 1.3703082407213252, + "grad_norm": 7.913741648930507, + "learning_rate": 1.1875770039437602e-05, + "loss": 0.854, + "step": 6535 + }, + { + "epoch": 1.3705179282868527, + "grad_norm": 8.100140023906857, + "learning_rate": 1.1873546290968751e-05, + "loss": 0.9358, + "step": 6536 + }, + { + "epoch": 1.37072761585238, + "grad_norm": 7.1798431205280195, + "learning_rate": 1.1871322446477405e-05, + "loss": 0.7839, + "step": 6537 + }, + { + "epoch": 1.3709373034179073, + "grad_norm": 8.432020634333737, + "learning_rate": 1.1869098506077535e-05, + "loss": 1.1451, + "step": 6538 + }, + { + "epoch": 1.3711469909834346, + "grad_norm": 6.22945718774862, + "learning_rate": 1.1866874469883127e-05, + "loss": 0.7857, + "step": 6539 + }, + { + "epoch": 1.371356678548962, + "grad_norm": 6.21727889179072, + "learning_rate": 1.1864650338008166e-05, + "loss": 0.6953, + "step": 6540 + }, + { + "epoch": 1.3715663661144895, + "grad_norm": 8.71230058440118, + "learning_rate": 1.1862426110566639e-05, + "loss": 1.1317, + "step": 6541 + }, + { + "epoch": 1.3717760536800168, + "grad_norm": 6.476294272681391, + "learning_rate": 1.1860201787672543e-05, + "loss": 0.8774, + "step": 6542 + }, + { + "epoch": 1.371985741245544, + "grad_norm": 7.773833533123625, + "learning_rate": 1.1857977369439879e-05, + "loss": 0.7438, + "step": 6543 + }, + { + "epoch": 1.3721954288110716, + "grad_norm": 5.861100426941333, + "learning_rate": 1.1855752855982651e-05, + "loss": 0.6423, + "step": 6544 + }, + { + "epoch": 1.372405116376599, + "grad_norm": 6.574254840637303, + "learning_rate": 1.1853528247414874e-05, + "loss": 0.6343, + "step": 6545 + }, + { + "epoch": 1.3726148039421262, + "grad_norm": 6.893377801274218, + "learning_rate": 1.185130354385056e-05, + "loss": 0.7643, + "step": 6546 + }, + { + "epoch": 1.3728244915076537, + "grad_norm": 6.167877950507337, + "learning_rate": 1.1849078745403723e-05, + "loss": 0.7681, + "step": 6547 + }, + { + "epoch": 1.373034179073181, + "grad_norm": 6.292095878116126, + "learning_rate": 1.1846853852188398e-05, + "loss": 0.7757, + "step": 6548 + }, + { + "epoch": 1.3732438666387083, + "grad_norm": 7.708002096251597, + "learning_rate": 1.184462886431861e-05, + "loss": 0.7935, + "step": 6549 + }, + { + "epoch": 1.3734535542042356, + "grad_norm": 10.22684780710782, + "learning_rate": 1.184240378190839e-05, + "loss": 0.9295, + "step": 6550 + }, + { + "epoch": 1.373663241769763, + "grad_norm": 7.7082793971969155, + "learning_rate": 1.1840178605071784e-05, + "loss": 1.0751, + "step": 6551 + }, + { + "epoch": 1.3738729293352905, + "grad_norm": 7.399360127156178, + "learning_rate": 1.1837953333922833e-05, + "loss": 0.8041, + "step": 6552 + }, + { + "epoch": 1.3740826169008178, + "grad_norm": 6.41490753045952, + "learning_rate": 1.1835727968575582e-05, + "loss": 0.717, + "step": 6553 + }, + { + "epoch": 1.374292304466345, + "grad_norm": 6.447438474924967, + "learning_rate": 1.1833502509144093e-05, + "loss": 0.7296, + "step": 6554 + }, + { + "epoch": 1.3745019920318726, + "grad_norm": 7.69345772460767, + "learning_rate": 1.1831276955742417e-05, + "loss": 1.0048, + "step": 6555 + }, + { + "epoch": 1.3747116795974, + "grad_norm": 8.525720443769169, + "learning_rate": 1.1829051308484623e-05, + "loss": 0.9567, + "step": 6556 + }, + { + "epoch": 1.3749213671629272, + "grad_norm": 7.873091609319561, + "learning_rate": 1.1826825567484777e-05, + "loss": 1.0782, + "step": 6557 + }, + { + "epoch": 1.3751310547284545, + "grad_norm": 7.660062956156774, + "learning_rate": 1.1824599732856952e-05, + "loss": 0.7907, + "step": 6558 + }, + { + "epoch": 1.3753407422939818, + "grad_norm": 8.336168050259152, + "learning_rate": 1.1822373804715224e-05, + "loss": 0.9899, + "step": 6559 + }, + { + "epoch": 1.3755504298595094, + "grad_norm": 6.66942467141876, + "learning_rate": 1.182014778317368e-05, + "loss": 0.7631, + "step": 6560 + }, + { + "epoch": 1.3757601174250367, + "grad_norm": 6.967669774536097, + "learning_rate": 1.1817921668346404e-05, + "loss": 1.0413, + "step": 6561 + }, + { + "epoch": 1.375969804990564, + "grad_norm": 6.984940057443896, + "learning_rate": 1.1815695460347491e-05, + "loss": 0.6704, + "step": 6562 + }, + { + "epoch": 1.3761794925560915, + "grad_norm": 6.300850976332309, + "learning_rate": 1.1813469159291036e-05, + "loss": 0.7338, + "step": 6563 + }, + { + "epoch": 1.3763891801216188, + "grad_norm": 9.362977625730599, + "learning_rate": 1.181124276529114e-05, + "loss": 1.1666, + "step": 6564 + }, + { + "epoch": 1.3765988676871461, + "grad_norm": 7.339314243966552, + "learning_rate": 1.180901627846191e-05, + "loss": 0.9496, + "step": 6565 + }, + { + "epoch": 1.3768085552526736, + "grad_norm": 6.486303707330697, + "learning_rate": 1.1806789698917464e-05, + "loss": 0.7866, + "step": 6566 + }, + { + "epoch": 1.377018242818201, + "grad_norm": 7.359170399846788, + "learning_rate": 1.1804563026771905e-05, + "loss": 0.8162, + "step": 6567 + }, + { + "epoch": 1.3772279303837283, + "grad_norm": 9.306617448383625, + "learning_rate": 1.1802336262139368e-05, + "loss": 0.8148, + "step": 6568 + }, + { + "epoch": 1.3774376179492556, + "grad_norm": 6.352273582171359, + "learning_rate": 1.1800109405133967e-05, + "loss": 0.5752, + "step": 6569 + }, + { + "epoch": 1.3776473055147829, + "grad_norm": 8.336345222891403, + "learning_rate": 1.1797882455869836e-05, + "loss": 0.8743, + "step": 6570 + }, + { + "epoch": 1.3778569930803104, + "grad_norm": 8.12712759602131, + "learning_rate": 1.1795655414461112e-05, + "loss": 1.1835, + "step": 6571 + }, + { + "epoch": 1.3780666806458377, + "grad_norm": 5.809284087732241, + "learning_rate": 1.1793428281021935e-05, + "loss": 0.7845, + "step": 6572 + }, + { + "epoch": 1.378276368211365, + "grad_norm": 5.6606781214139685, + "learning_rate": 1.1791201055666446e-05, + "loss": 0.6728, + "step": 6573 + }, + { + "epoch": 1.3784860557768925, + "grad_norm": 6.066486406165189, + "learning_rate": 1.1788973738508797e-05, + "loss": 0.6765, + "step": 6574 + }, + { + "epoch": 1.3786957433424198, + "grad_norm": 5.912181407690022, + "learning_rate": 1.1786746329663144e-05, + "loss": 0.8034, + "step": 6575 + }, + { + "epoch": 1.3789054309079471, + "grad_norm": 7.044504600913205, + "learning_rate": 1.1784518829243635e-05, + "loss": 0.7407, + "step": 6576 + }, + { + "epoch": 1.3791151184734745, + "grad_norm": 7.5961482495600245, + "learning_rate": 1.1782291237364449e-05, + "loss": 0.945, + "step": 6577 + }, + { + "epoch": 1.3793248060390018, + "grad_norm": 7.352758266258367, + "learning_rate": 1.1780063554139742e-05, + "loss": 0.8457, + "step": 6578 + }, + { + "epoch": 1.3795344936045293, + "grad_norm": 6.1562648419925, + "learning_rate": 1.177783577968369e-05, + "loss": 0.6212, + "step": 6579 + }, + { + "epoch": 1.3797441811700566, + "grad_norm": 7.851459440698408, + "learning_rate": 1.1775607914110474e-05, + "loss": 0.9907, + "step": 6580 + }, + { + "epoch": 1.379953868735584, + "grad_norm": 7.247724791776623, + "learning_rate": 1.1773379957534273e-05, + "loss": 0.7796, + "step": 6581 + }, + { + "epoch": 1.3801635563011114, + "grad_norm": 7.191071626854504, + "learning_rate": 1.1771151910069267e-05, + "loss": 0.8473, + "step": 6582 + }, + { + "epoch": 1.3803732438666387, + "grad_norm": 6.3841339014981155, + "learning_rate": 1.176892377182966e-05, + "loss": 0.8311, + "step": 6583 + }, + { + "epoch": 1.380582931432166, + "grad_norm": 6.818605530534439, + "learning_rate": 1.176669554292964e-05, + "loss": 0.9068, + "step": 6584 + }, + { + "epoch": 1.3807926189976936, + "grad_norm": 5.85568504907392, + "learning_rate": 1.176446722348341e-05, + "loss": 0.6317, + "step": 6585 + }, + { + "epoch": 1.3810023065632209, + "grad_norm": 6.8586862871137, + "learning_rate": 1.1762238813605171e-05, + "loss": 0.7116, + "step": 6586 + }, + { + "epoch": 1.3812119941287482, + "grad_norm": 6.846916365487637, + "learning_rate": 1.1760010313409143e-05, + "loss": 0.6638, + "step": 6587 + }, + { + "epoch": 1.3814216816942755, + "grad_norm": 7.163554757813141, + "learning_rate": 1.1757781723009525e-05, + "loss": 0.9004, + "step": 6588 + }, + { + "epoch": 1.3816313692598028, + "grad_norm": 6.731100352203539, + "learning_rate": 1.175555304252055e-05, + "loss": 0.7153, + "step": 6589 + }, + { + "epoch": 1.3818410568253303, + "grad_norm": 6.891538944043689, + "learning_rate": 1.1753324272056436e-05, + "loss": 0.8494, + "step": 6590 + }, + { + "epoch": 1.3820507443908576, + "grad_norm": 8.392547245159692, + "learning_rate": 1.1751095411731414e-05, + "loss": 0.8276, + "step": 6591 + }, + { + "epoch": 1.382260431956385, + "grad_norm": 8.85170138099449, + "learning_rate": 1.1748866461659713e-05, + "loss": 1.0985, + "step": 6592 + }, + { + "epoch": 1.3824701195219125, + "grad_norm": 9.415444169015652, + "learning_rate": 1.1746637421955573e-05, + "loss": 1.0191, + "step": 6593 + }, + { + "epoch": 1.3826798070874398, + "grad_norm": 8.414984492939231, + "learning_rate": 1.1744408292733236e-05, + "loss": 0.9688, + "step": 6594 + }, + { + "epoch": 1.382889494652967, + "grad_norm": 6.15159211441699, + "learning_rate": 1.1742179074106948e-05, + "loss": 0.8505, + "step": 6595 + }, + { + "epoch": 1.3830991822184944, + "grad_norm": 7.311576724479214, + "learning_rate": 1.1739949766190959e-05, + "loss": 0.7363, + "step": 6596 + }, + { + "epoch": 1.3833088697840217, + "grad_norm": 7.2252640343780365, + "learning_rate": 1.173772036909953e-05, + "loss": 0.7245, + "step": 6597 + }, + { + "epoch": 1.3835185573495492, + "grad_norm": 6.7671096570807805, + "learning_rate": 1.1735490882946918e-05, + "loss": 0.8324, + "step": 6598 + }, + { + "epoch": 1.3837282449150765, + "grad_norm": 6.038666799327115, + "learning_rate": 1.1733261307847386e-05, + "loss": 0.8066, + "step": 6599 + }, + { + "epoch": 1.3839379324806038, + "grad_norm": 6.93936333838456, + "learning_rate": 1.1731031643915207e-05, + "loss": 0.8516, + "step": 6600 + }, + { + "epoch": 1.3841476200461313, + "grad_norm": 9.151150395658048, + "learning_rate": 1.1728801891264653e-05, + "loss": 1.1863, + "step": 6601 + }, + { + "epoch": 1.3843573076116586, + "grad_norm": 6.433808575594778, + "learning_rate": 1.1726572050010003e-05, + "loss": 0.9425, + "step": 6602 + }, + { + "epoch": 1.384566995177186, + "grad_norm": 6.586625059049893, + "learning_rate": 1.1724342120265543e-05, + "loss": 0.7479, + "step": 6603 + }, + { + "epoch": 1.3847766827427135, + "grad_norm": 7.870109632890547, + "learning_rate": 1.172211210214556e-05, + "loss": 0.9992, + "step": 6604 + }, + { + "epoch": 1.3849863703082408, + "grad_norm": 6.727053483684335, + "learning_rate": 1.1719881995764339e-05, + "loss": 0.8071, + "step": 6605 + }, + { + "epoch": 1.385196057873768, + "grad_norm": 6.434914581165896, + "learning_rate": 1.1717651801236189e-05, + "loss": 0.7525, + "step": 6606 + }, + { + "epoch": 1.3854057454392954, + "grad_norm": 7.616129780227017, + "learning_rate": 1.1715421518675403e-05, + "loss": 0.9002, + "step": 6607 + }, + { + "epoch": 1.3856154330048227, + "grad_norm": 7.488054598928749, + "learning_rate": 1.1713191148196288e-05, + "loss": 0.8735, + "step": 6608 + }, + { + "epoch": 1.3858251205703502, + "grad_norm": 5.949504204393372, + "learning_rate": 1.1710960689913156e-05, + "loss": 0.5876, + "step": 6609 + }, + { + "epoch": 1.3860348081358775, + "grad_norm": 7.394034745091445, + "learning_rate": 1.1708730143940321e-05, + "loss": 0.8589, + "step": 6610 + }, + { + "epoch": 1.3862444957014048, + "grad_norm": 6.002640102669201, + "learning_rate": 1.1706499510392099e-05, + "loss": 0.6967, + "step": 6611 + }, + { + "epoch": 1.3864541832669324, + "grad_norm": 7.2500665976079945, + "learning_rate": 1.1704268789382824e-05, + "loss": 0.7438, + "step": 6612 + }, + { + "epoch": 1.3866638708324597, + "grad_norm": 5.8573168399815, + "learning_rate": 1.1702037981026814e-05, + "loss": 0.6853, + "step": 6613 + }, + { + "epoch": 1.386873558397987, + "grad_norm": 7.3844037753007425, + "learning_rate": 1.1699807085438406e-05, + "loss": 0.8578, + "step": 6614 + }, + { + "epoch": 1.3870832459635143, + "grad_norm": 6.815948282668088, + "learning_rate": 1.1697576102731936e-05, + "loss": 0.7754, + "step": 6615 + }, + { + "epoch": 1.3872929335290416, + "grad_norm": 6.014188177500772, + "learning_rate": 1.1695345033021748e-05, + "loss": 0.6792, + "step": 6616 + }, + { + "epoch": 1.3875026210945691, + "grad_norm": 6.750373722780624, + "learning_rate": 1.1693113876422184e-05, + "loss": 0.8654, + "step": 6617 + }, + { + "epoch": 1.3877123086600964, + "grad_norm": 7.003411895232289, + "learning_rate": 1.16908826330476e-05, + "loss": 0.8274, + "step": 6618 + }, + { + "epoch": 1.3879219962256237, + "grad_norm": 7.2807614294050405, + "learning_rate": 1.1688651303012347e-05, + "loss": 0.9513, + "step": 6619 + }, + { + "epoch": 1.3881316837911513, + "grad_norm": 8.022940085852035, + "learning_rate": 1.1686419886430787e-05, + "loss": 0.9033, + "step": 6620 + }, + { + "epoch": 1.3883413713566786, + "grad_norm": 8.701788482053715, + "learning_rate": 1.1684188383417281e-05, + "loss": 0.8921, + "step": 6621 + }, + { + "epoch": 1.3885510589222059, + "grad_norm": 5.80839781135677, + "learning_rate": 1.16819567940862e-05, + "loss": 0.7583, + "step": 6622 + }, + { + "epoch": 1.3887607464877334, + "grad_norm": 6.441235085451961, + "learning_rate": 1.1679725118551915e-05, + "loss": 0.6999, + "step": 6623 + }, + { + "epoch": 1.3889704340532607, + "grad_norm": 8.155314715242946, + "learning_rate": 1.1677493356928803e-05, + "loss": 1.0398, + "step": 6624 + }, + { + "epoch": 1.389180121618788, + "grad_norm": 7.350547983014046, + "learning_rate": 1.1675261509331253e-05, + "loss": 0.9179, + "step": 6625 + }, + { + "epoch": 1.3893898091843153, + "grad_norm": 7.870354504337803, + "learning_rate": 1.167302957587364e-05, + "loss": 0.98, + "step": 6626 + }, + { + "epoch": 1.3895994967498426, + "grad_norm": 8.192615353854334, + "learning_rate": 1.1670797556670363e-05, + "loss": 0.8264, + "step": 6627 + }, + { + "epoch": 1.3898091843153701, + "grad_norm": 8.517072470381054, + "learning_rate": 1.1668565451835812e-05, + "loss": 1.1305, + "step": 6628 + }, + { + "epoch": 1.3900188718808975, + "grad_norm": 6.9074091103139255, + "learning_rate": 1.1666333261484386e-05, + "loss": 0.8471, + "step": 6629 + }, + { + "epoch": 1.3902285594464248, + "grad_norm": 7.474998169210219, + "learning_rate": 1.1664100985730492e-05, + "loss": 0.7854, + "step": 6630 + }, + { + "epoch": 1.3904382470119523, + "grad_norm": 7.572328663667642, + "learning_rate": 1.1661868624688535e-05, + "loss": 0.8249, + "step": 6631 + }, + { + "epoch": 1.3906479345774796, + "grad_norm": 7.057146820761422, + "learning_rate": 1.1659636178472928e-05, + "loss": 0.7033, + "step": 6632 + }, + { + "epoch": 1.390857622143007, + "grad_norm": 6.370976195399844, + "learning_rate": 1.1657403647198091e-05, + "loss": 0.6608, + "step": 6633 + }, + { + "epoch": 1.3910673097085342, + "grad_norm": 6.567614297864201, + "learning_rate": 1.165517103097844e-05, + "loss": 0.7177, + "step": 6634 + }, + { + "epoch": 1.3912769972740615, + "grad_norm": 7.143718740544215, + "learning_rate": 1.1652938329928405e-05, + "loss": 0.8184, + "step": 6635 + }, + { + "epoch": 1.391486684839589, + "grad_norm": 7.242472116809556, + "learning_rate": 1.1650705544162413e-05, + "loss": 0.9029, + "step": 6636 + }, + { + "epoch": 1.3916963724051163, + "grad_norm": 6.456009190251499, + "learning_rate": 1.1648472673794898e-05, + "loss": 0.8493, + "step": 6637 + }, + { + "epoch": 1.3919060599706436, + "grad_norm": 7.517263882621033, + "learning_rate": 1.16462397189403e-05, + "loss": 0.8861, + "step": 6638 + }, + { + "epoch": 1.3921157475361712, + "grad_norm": 8.784274984651768, + "learning_rate": 1.1644006679713063e-05, + "loss": 1.1381, + "step": 6639 + }, + { + "epoch": 1.3923254351016985, + "grad_norm": 6.796901845056859, + "learning_rate": 1.1641773556227626e-05, + "loss": 0.7094, + "step": 6640 + }, + { + "epoch": 1.3925351226672258, + "grad_norm": 5.872266529179356, + "learning_rate": 1.1639540348598455e-05, + "loss": 0.448, + "step": 6641 + }, + { + "epoch": 1.3927448102327533, + "grad_norm": 5.757750829161933, + "learning_rate": 1.1637307056939993e-05, + "loss": 0.7383, + "step": 6642 + }, + { + "epoch": 1.3929544977982806, + "grad_norm": 6.881902975592492, + "learning_rate": 1.1635073681366704e-05, + "loss": 0.7825, + "step": 6643 + }, + { + "epoch": 1.393164185363808, + "grad_norm": 7.445904927085417, + "learning_rate": 1.1632840221993054e-05, + "loss": 0.898, + "step": 6644 + }, + { + "epoch": 1.3933738729293352, + "grad_norm": 6.971941975998979, + "learning_rate": 1.1630606678933514e-05, + "loss": 0.8775, + "step": 6645 + }, + { + "epoch": 1.3935835604948625, + "grad_norm": 5.5757962412670725, + "learning_rate": 1.1628373052302546e-05, + "loss": 0.5841, + "step": 6646 + }, + { + "epoch": 1.39379324806039, + "grad_norm": 10.42929692519301, + "learning_rate": 1.1626139342214642e-05, + "loss": 1.3279, + "step": 6647 + }, + { + "epoch": 1.3940029356259174, + "grad_norm": 7.391261008883606, + "learning_rate": 1.1623905548784274e-05, + "loss": 0.7357, + "step": 6648 + }, + { + "epoch": 1.3942126231914447, + "grad_norm": 7.126826664677415, + "learning_rate": 1.1621671672125927e-05, + "loss": 0.7001, + "step": 6649 + }, + { + "epoch": 1.3944223107569722, + "grad_norm": 7.708544415835774, + "learning_rate": 1.1619437712354098e-05, + "loss": 0.7852, + "step": 6650 + }, + { + "epoch": 1.3946319983224995, + "grad_norm": 6.520850634050474, + "learning_rate": 1.1617203669583275e-05, + "loss": 0.6675, + "step": 6651 + }, + { + "epoch": 1.3948416858880268, + "grad_norm": 6.448882571330285, + "learning_rate": 1.1614969543927958e-05, + "loss": 0.5892, + "step": 6652 + }, + { + "epoch": 1.3950513734535541, + "grad_norm": 8.012185991687247, + "learning_rate": 1.1612735335502655e-05, + "loss": 1.0787, + "step": 6653 + }, + { + "epoch": 1.3952610610190816, + "grad_norm": 8.95374544308986, + "learning_rate": 1.1610501044421865e-05, + "loss": 0.7078, + "step": 6654 + }, + { + "epoch": 1.395470748584609, + "grad_norm": 6.720733172879704, + "learning_rate": 1.1608266670800106e-05, + "loss": 0.7862, + "step": 6655 + }, + { + "epoch": 1.3956804361501363, + "grad_norm": 6.60468007911462, + "learning_rate": 1.160603221475189e-05, + "loss": 0.7212, + "step": 6656 + }, + { + "epoch": 1.3958901237156636, + "grad_norm": 7.410094855264841, + "learning_rate": 1.1603797676391735e-05, + "loss": 0.8174, + "step": 6657 + }, + { + "epoch": 1.396099811281191, + "grad_norm": 7.267933478916165, + "learning_rate": 1.1601563055834174e-05, + "loss": 0.6736, + "step": 6658 + }, + { + "epoch": 1.3963094988467184, + "grad_norm": 6.56079821418887, + "learning_rate": 1.1599328353193722e-05, + "loss": 0.6477, + "step": 6659 + }, + { + "epoch": 1.3965191864122457, + "grad_norm": 8.05796146824699, + "learning_rate": 1.159709356858492e-05, + "loss": 0.8721, + "step": 6660 + }, + { + "epoch": 1.3967288739777732, + "grad_norm": 8.210484700802608, + "learning_rate": 1.1594858702122303e-05, + "loss": 0.8318, + "step": 6661 + }, + { + "epoch": 1.3969385615433005, + "grad_norm": 6.626723191904108, + "learning_rate": 1.1592623753920414e-05, + "loss": 0.768, + "step": 6662 + }, + { + "epoch": 1.3971482491088278, + "grad_norm": 5.648183128524098, + "learning_rate": 1.1590388724093788e-05, + "loss": 0.5109, + "step": 6663 + }, + { + "epoch": 1.3973579366743551, + "grad_norm": 6.6720359335313155, + "learning_rate": 1.158815361275699e-05, + "loss": 0.7033, + "step": 6664 + }, + { + "epoch": 1.3975676242398825, + "grad_norm": 6.644300058167578, + "learning_rate": 1.158591842002456e-05, + "loss": 0.791, + "step": 6665 + }, + { + "epoch": 1.39777731180541, + "grad_norm": 6.99716113977253, + "learning_rate": 1.1583683146011062e-05, + "loss": 0.7895, + "step": 6666 + }, + { + "epoch": 1.3979869993709373, + "grad_norm": 8.822283805176465, + "learning_rate": 1.1581447790831055e-05, + "loss": 1.0284, + "step": 6667 + }, + { + "epoch": 1.3981966869364646, + "grad_norm": 7.471159556788238, + "learning_rate": 1.1579212354599107e-05, + "loss": 0.7165, + "step": 6668 + }, + { + "epoch": 1.3984063745019921, + "grad_norm": 5.996239090689492, + "learning_rate": 1.1576976837429783e-05, + "loss": 0.7642, + "step": 6669 + }, + { + "epoch": 1.3986160620675194, + "grad_norm": 7.690131599098741, + "learning_rate": 1.1574741239437666e-05, + "loss": 0.8927, + "step": 6670 + }, + { + "epoch": 1.3988257496330467, + "grad_norm": 7.542280628270704, + "learning_rate": 1.1572505560737329e-05, + "loss": 0.8225, + "step": 6671 + }, + { + "epoch": 1.3990354371985743, + "grad_norm": 7.425715383220487, + "learning_rate": 1.1570269801443352e-05, + "loss": 0.8586, + "step": 6672 + }, + { + "epoch": 1.3992451247641016, + "grad_norm": 7.665122626708507, + "learning_rate": 1.1568033961670324e-05, + "loss": 0.7262, + "step": 6673 + }, + { + "epoch": 1.3994548123296289, + "grad_norm": 7.027987050218133, + "learning_rate": 1.1565798041532836e-05, + "loss": 0.9316, + "step": 6674 + }, + { + "epoch": 1.3996644998951562, + "grad_norm": 6.23455783090704, + "learning_rate": 1.1563562041145484e-05, + "loss": 0.5578, + "step": 6675 + }, + { + "epoch": 1.3998741874606835, + "grad_norm": 5.651723177620124, + "learning_rate": 1.1561325960622866e-05, + "loss": 0.6555, + "step": 6676 + }, + { + "epoch": 1.400083875026211, + "grad_norm": 7.631585497905252, + "learning_rate": 1.1559089800079584e-05, + "loss": 0.9897, + "step": 6677 + }, + { + "epoch": 1.4002935625917383, + "grad_norm": 6.232764541741989, + "learning_rate": 1.1556853559630246e-05, + "loss": 0.7395, + "step": 6678 + }, + { + "epoch": 1.4005032501572656, + "grad_norm": 6.22801725460777, + "learning_rate": 1.155461723938946e-05, + "loss": 0.6478, + "step": 6679 + }, + { + "epoch": 1.4007129377227931, + "grad_norm": 5.925271108309575, + "learning_rate": 1.1552380839471847e-05, + "loss": 0.5511, + "step": 6680 + }, + { + "epoch": 1.4009226252883205, + "grad_norm": 6.257955492364735, + "learning_rate": 1.1550144359992022e-05, + "loss": 0.4483, + "step": 6681 + }, + { + "epoch": 1.4011323128538478, + "grad_norm": 6.343931791855393, + "learning_rate": 1.154790780106461e-05, + "loss": 0.6076, + "step": 6682 + }, + { + "epoch": 1.401342000419375, + "grad_norm": 8.031149173356003, + "learning_rate": 1.1545671162804238e-05, + "loss": 0.9635, + "step": 6683 + }, + { + "epoch": 1.4015516879849024, + "grad_norm": 8.257850143433433, + "learning_rate": 1.1543434445325539e-05, + "loss": 0.8289, + "step": 6684 + }, + { + "epoch": 1.40176137555043, + "grad_norm": 6.436673506554482, + "learning_rate": 1.154119764874315e-05, + "loss": 0.7955, + "step": 6685 + }, + { + "epoch": 1.4019710631159572, + "grad_norm": 7.388303940153177, + "learning_rate": 1.1538960773171703e-05, + "loss": 0.731, + "step": 6686 + }, + { + "epoch": 1.4021807506814845, + "grad_norm": 7.4479717084489145, + "learning_rate": 1.1536723818725855e-05, + "loss": 0.7709, + "step": 6687 + }, + { + "epoch": 1.402390438247012, + "grad_norm": 7.986727123830757, + "learning_rate": 1.153448678552024e-05, + "loss": 0.7147, + "step": 6688 + }, + { + "epoch": 1.4026001258125393, + "grad_norm": 7.98665383696985, + "learning_rate": 1.1532249673669518e-05, + "loss": 1.04, + "step": 6689 + }, + { + "epoch": 1.4028098133780667, + "grad_norm": 7.9263985683917015, + "learning_rate": 1.153001248328834e-05, + "loss": 0.8171, + "step": 6690 + }, + { + "epoch": 1.4030195009435942, + "grad_norm": 8.234110955107315, + "learning_rate": 1.1527775214491373e-05, + "loss": 1.0076, + "step": 6691 + }, + { + "epoch": 1.4032291885091215, + "grad_norm": 5.728570372407337, + "learning_rate": 1.1525537867393271e-05, + "loss": 0.7229, + "step": 6692 + }, + { + "epoch": 1.4034388760746488, + "grad_norm": 7.338062605341393, + "learning_rate": 1.1523300442108712e-05, + "loss": 0.8714, + "step": 6693 + }, + { + "epoch": 1.403648563640176, + "grad_norm": 7.772863859547496, + "learning_rate": 1.1521062938752361e-05, + "loss": 0.814, + "step": 6694 + }, + { + "epoch": 1.4038582512057034, + "grad_norm": 9.347894945695758, + "learning_rate": 1.1518825357438899e-05, + "loss": 0.8665, + "step": 6695 + }, + { + "epoch": 1.404067938771231, + "grad_norm": 7.686234776394881, + "learning_rate": 1.1516587698283e-05, + "loss": 0.7991, + "step": 6696 + }, + { + "epoch": 1.4042776263367582, + "grad_norm": 7.671059535136048, + "learning_rate": 1.1514349961399354e-05, + "loss": 1.0598, + "step": 6697 + }, + { + "epoch": 1.4044873139022855, + "grad_norm": 7.836100909154196, + "learning_rate": 1.151211214690264e-05, + "loss": 0.8187, + "step": 6698 + }, + { + "epoch": 1.404697001467813, + "grad_norm": 7.632462960110242, + "learning_rate": 1.1509874254907562e-05, + "loss": 0.8339, + "step": 6699 + }, + { + "epoch": 1.4049066890333404, + "grad_norm": 8.871719374825206, + "learning_rate": 1.150763628552881e-05, + "loss": 1.1294, + "step": 6700 + }, + { + "epoch": 1.4051163765988677, + "grad_norm": 6.559798980788915, + "learning_rate": 1.1505398238881078e-05, + "loss": 0.6144, + "step": 6701 + }, + { + "epoch": 1.405326064164395, + "grad_norm": 8.17831443941465, + "learning_rate": 1.150316011507908e-05, + "loss": 0.8922, + "step": 6702 + }, + { + "epoch": 1.4055357517299223, + "grad_norm": 8.11354201627378, + "learning_rate": 1.1500921914237516e-05, + "loss": 0.9922, + "step": 6703 + }, + { + "epoch": 1.4057454392954498, + "grad_norm": 7.116864222188996, + "learning_rate": 1.1498683636471101e-05, + "loss": 0.656, + "step": 6704 + }, + { + "epoch": 1.4059551268609771, + "grad_norm": 7.542254159001361, + "learning_rate": 1.1496445281894552e-05, + "loss": 0.916, + "step": 6705 + }, + { + "epoch": 1.4061648144265044, + "grad_norm": 4.891630131140577, + "learning_rate": 1.1494206850622589e-05, + "loss": 0.5368, + "step": 6706 + }, + { + "epoch": 1.406374501992032, + "grad_norm": 6.93797773384164, + "learning_rate": 1.1491968342769931e-05, + "loss": 0.7061, + "step": 6707 + }, + { + "epoch": 1.4065841895575593, + "grad_norm": 7.017046514112816, + "learning_rate": 1.1489729758451306e-05, + "loss": 0.7606, + "step": 6708 + }, + { + "epoch": 1.4067938771230866, + "grad_norm": 7.278615592127141, + "learning_rate": 1.148749109778145e-05, + "loss": 0.8107, + "step": 6709 + }, + { + "epoch": 1.407003564688614, + "grad_norm": 8.159660732897715, + "learning_rate": 1.1485252360875093e-05, + "loss": 0.8696, + "step": 6710 + }, + { + "epoch": 1.4072132522541414, + "grad_norm": 7.977671974336682, + "learning_rate": 1.1483013547846975e-05, + "loss": 1.0339, + "step": 6711 + }, + { + "epoch": 1.4074229398196687, + "grad_norm": 6.6734354443972315, + "learning_rate": 1.1480774658811844e-05, + "loss": 0.6853, + "step": 6712 + }, + { + "epoch": 1.407632627385196, + "grad_norm": 7.180669595183709, + "learning_rate": 1.1478535693884441e-05, + "loss": 0.6942, + "step": 6713 + }, + { + "epoch": 1.4078423149507233, + "grad_norm": 5.820651066616864, + "learning_rate": 1.1476296653179521e-05, + "loss": 0.7162, + "step": 6714 + }, + { + "epoch": 1.4080520025162508, + "grad_norm": 6.75856456320359, + "learning_rate": 1.1474057536811836e-05, + "loss": 0.7588, + "step": 6715 + }, + { + "epoch": 1.4082616900817782, + "grad_norm": 6.8938669947813, + "learning_rate": 1.1471818344896148e-05, + "loss": 0.7462, + "step": 6716 + }, + { + "epoch": 1.4084713776473055, + "grad_norm": 6.5945010223640415, + "learning_rate": 1.1469579077547215e-05, + "loss": 0.735, + "step": 6717 + }, + { + "epoch": 1.408681065212833, + "grad_norm": 6.18346029394847, + "learning_rate": 1.1467339734879806e-05, + "loss": 0.6428, + "step": 6718 + }, + { + "epoch": 1.4088907527783603, + "grad_norm": 6.683122716945392, + "learning_rate": 1.1465100317008689e-05, + "loss": 0.7731, + "step": 6719 + }, + { + "epoch": 1.4091004403438876, + "grad_norm": 6.039894996750581, + "learning_rate": 1.1462860824048643e-05, + "loss": 0.5133, + "step": 6720 + }, + { + "epoch": 1.409310127909415, + "grad_norm": 5.965861207019783, + "learning_rate": 1.1460621256114438e-05, + "loss": 0.5742, + "step": 6721 + }, + { + "epoch": 1.4095198154749422, + "grad_norm": 7.801798550247179, + "learning_rate": 1.1458381613320863e-05, + "loss": 0.8277, + "step": 6722 + }, + { + "epoch": 1.4097295030404697, + "grad_norm": 6.304478333269873, + "learning_rate": 1.1456141895782701e-05, + "loss": 0.6967, + "step": 6723 + }, + { + "epoch": 1.409939190605997, + "grad_norm": 6.188005309802785, + "learning_rate": 1.145390210361474e-05, + "loss": 0.6774, + "step": 6724 + }, + { + "epoch": 1.4101488781715243, + "grad_norm": 6.169355337765976, + "learning_rate": 1.1451662236931775e-05, + "loss": 0.7409, + "step": 6725 + }, + { + "epoch": 1.4103585657370519, + "grad_norm": 7.982775636463912, + "learning_rate": 1.1449422295848606e-05, + "loss": 1.0103, + "step": 6726 + }, + { + "epoch": 1.4105682533025792, + "grad_norm": 8.892824784948115, + "learning_rate": 1.1447182280480024e-05, + "loss": 1.2514, + "step": 6727 + }, + { + "epoch": 1.4107779408681065, + "grad_norm": 6.370409539353546, + "learning_rate": 1.1444942190940846e-05, + "loss": 0.7717, + "step": 6728 + }, + { + "epoch": 1.410987628433634, + "grad_norm": 6.8691096710535895, + "learning_rate": 1.1442702027345873e-05, + "loss": 0.8624, + "step": 6729 + }, + { + "epoch": 1.4111973159991613, + "grad_norm": 6.516750276258667, + "learning_rate": 1.1440461789809919e-05, + "loss": 0.8874, + "step": 6730 + }, + { + "epoch": 1.4114070035646886, + "grad_norm": 6.928663245017479, + "learning_rate": 1.1438221478447798e-05, + "loss": 0.6907, + "step": 6731 + }, + { + "epoch": 1.411616691130216, + "grad_norm": 8.727437240819333, + "learning_rate": 1.143598109337433e-05, + "loss": 0.958, + "step": 6732 + }, + { + "epoch": 1.4118263786957432, + "grad_norm": 7.3211381000363716, + "learning_rate": 1.1433740634704345e-05, + "loss": 0.928, + "step": 6733 + }, + { + "epoch": 1.4120360662612708, + "grad_norm": 7.592545888747638, + "learning_rate": 1.1431500102552665e-05, + "loss": 0.9485, + "step": 6734 + }, + { + "epoch": 1.412245753826798, + "grad_norm": 7.8112810049339805, + "learning_rate": 1.142925949703412e-05, + "loss": 0.867, + "step": 6735 + }, + { + "epoch": 1.4124554413923254, + "grad_norm": 8.289506472042069, + "learning_rate": 1.1427018818263546e-05, + "loss": 0.923, + "step": 6736 + }, + { + "epoch": 1.412665128957853, + "grad_norm": 8.916790717111908, + "learning_rate": 1.1424778066355786e-05, + "loss": 1.0108, + "step": 6737 + }, + { + "epoch": 1.4128748165233802, + "grad_norm": 6.348259163613662, + "learning_rate": 1.1422537241425677e-05, + "loss": 0.7022, + "step": 6738 + }, + { + "epoch": 1.4130845040889075, + "grad_norm": 6.536542319791054, + "learning_rate": 1.1420296343588067e-05, + "loss": 0.7582, + "step": 6739 + }, + { + "epoch": 1.4132941916544348, + "grad_norm": 6.943847595090378, + "learning_rate": 1.1418055372957806e-05, + "loss": 0.6813, + "step": 6740 + }, + { + "epoch": 1.4135038792199621, + "grad_norm": 5.963084966633167, + "learning_rate": 1.1415814329649751e-05, + "loss": 0.6869, + "step": 6741 + }, + { + "epoch": 1.4137135667854897, + "grad_norm": 6.319514219174101, + "learning_rate": 1.141357321377875e-05, + "loss": 0.7139, + "step": 6742 + }, + { + "epoch": 1.413923254351017, + "grad_norm": 7.326589281094628, + "learning_rate": 1.1411332025459674e-05, + "loss": 0.9211, + "step": 6743 + }, + { + "epoch": 1.4141329419165443, + "grad_norm": 5.734591328297965, + "learning_rate": 1.140909076480738e-05, + "loss": 0.6822, + "step": 6744 + }, + { + "epoch": 1.4143426294820718, + "grad_norm": 8.314877954884183, + "learning_rate": 1.1406849431936746e-05, + "loss": 1.0942, + "step": 6745 + }, + { + "epoch": 1.414552317047599, + "grad_norm": 7.305383448928347, + "learning_rate": 1.1404608026962635e-05, + "loss": 0.7597, + "step": 6746 + }, + { + "epoch": 1.4147620046131264, + "grad_norm": 8.686699770359398, + "learning_rate": 1.1402366549999927e-05, + "loss": 1.0398, + "step": 6747 + }, + { + "epoch": 1.414971692178654, + "grad_norm": 6.371654515055473, + "learning_rate": 1.1400125001163503e-05, + "loss": 0.9606, + "step": 6748 + }, + { + "epoch": 1.4151813797441812, + "grad_norm": 6.075498973301181, + "learning_rate": 1.1397883380568242e-05, + "loss": 0.6941, + "step": 6749 + }, + { + "epoch": 1.4153910673097085, + "grad_norm": 5.84278018462633, + "learning_rate": 1.1395641688329031e-05, + "loss": 0.6959, + "step": 6750 + }, + { + "epoch": 1.4156007548752358, + "grad_norm": 7.199695647428923, + "learning_rate": 1.1393399924560769e-05, + "loss": 0.8448, + "step": 6751 + }, + { + "epoch": 1.4158104424407632, + "grad_norm": 6.503234197931621, + "learning_rate": 1.1391158089378341e-05, + "loss": 0.6662, + "step": 6752 + }, + { + "epoch": 1.4160201300062907, + "grad_norm": 7.327085431679781, + "learning_rate": 1.1388916182896645e-05, + "loss": 0.877, + "step": 6753 + }, + { + "epoch": 1.416229817571818, + "grad_norm": 6.245139488377247, + "learning_rate": 1.138667420523059e-05, + "loss": 0.704, + "step": 6754 + }, + { + "epoch": 1.4164395051373453, + "grad_norm": 6.934263139837212, + "learning_rate": 1.1384432156495078e-05, + "loss": 0.9103, + "step": 6755 + }, + { + "epoch": 1.4166491927028728, + "grad_norm": 7.046466192923788, + "learning_rate": 1.1382190036805012e-05, + "loss": 0.6756, + "step": 6756 + }, + { + "epoch": 1.4168588802684001, + "grad_norm": 7.032428613400029, + "learning_rate": 1.1379947846275311e-05, + "loss": 0.7064, + "step": 6757 + }, + { + "epoch": 1.4170685678339274, + "grad_norm": 5.801448440714787, + "learning_rate": 1.1377705585020891e-05, + "loss": 0.7222, + "step": 6758 + }, + { + "epoch": 1.4172782553994547, + "grad_norm": 7.167286371211719, + "learning_rate": 1.1375463253156669e-05, + "loss": 0.7781, + "step": 6759 + }, + { + "epoch": 1.417487942964982, + "grad_norm": 6.907400720677501, + "learning_rate": 1.1373220850797571e-05, + "loss": 0.955, + "step": 6760 + }, + { + "epoch": 1.4176976305305096, + "grad_norm": 6.467355626817842, + "learning_rate": 1.137097837805852e-05, + "loss": 0.7319, + "step": 6761 + }, + { + "epoch": 1.4179073180960369, + "grad_norm": 7.167739619716819, + "learning_rate": 1.1368735835054452e-05, + "loss": 0.7917, + "step": 6762 + }, + { + "epoch": 1.4181170056615642, + "grad_norm": 6.551186382007419, + "learning_rate": 1.1366493221900297e-05, + "loss": 0.7372, + "step": 6763 + }, + { + "epoch": 1.4183266932270917, + "grad_norm": 6.579719688252546, + "learning_rate": 1.1364250538710997e-05, + "loss": 0.7559, + "step": 6764 + }, + { + "epoch": 1.418536380792619, + "grad_norm": 6.222482984031967, + "learning_rate": 1.1362007785601486e-05, + "loss": 0.7952, + "step": 6765 + }, + { + "epoch": 1.4187460683581463, + "grad_norm": 7.59299136814988, + "learning_rate": 1.1359764962686717e-05, + "loss": 0.8762, + "step": 6766 + }, + { + "epoch": 1.4189557559236738, + "grad_norm": 6.6228444780873605, + "learning_rate": 1.1357522070081635e-05, + "loss": 0.7169, + "step": 6767 + }, + { + "epoch": 1.4191654434892012, + "grad_norm": 7.751878120518925, + "learning_rate": 1.1355279107901191e-05, + "loss": 0.7775, + "step": 6768 + }, + { + "epoch": 1.4193751310547285, + "grad_norm": 6.496836818213145, + "learning_rate": 1.1353036076260343e-05, + "loss": 0.5969, + "step": 6769 + }, + { + "epoch": 1.4195848186202558, + "grad_norm": 7.605081227279532, + "learning_rate": 1.135079297527405e-05, + "loss": 0.7277, + "step": 6770 + }, + { + "epoch": 1.419794506185783, + "grad_norm": 8.290769996194726, + "learning_rate": 1.1348549805057271e-05, + "loss": 0.893, + "step": 6771 + }, + { + "epoch": 1.4200041937513106, + "grad_norm": 5.623628443157594, + "learning_rate": 1.134630656572498e-05, + "loss": 0.6528, + "step": 6772 + }, + { + "epoch": 1.420213881316838, + "grad_norm": 7.391427346614643, + "learning_rate": 1.1344063257392138e-05, + "loss": 0.6381, + "step": 6773 + }, + { + "epoch": 1.4204235688823652, + "grad_norm": 6.226915171518154, + "learning_rate": 1.1341819880173727e-05, + "loss": 0.7045, + "step": 6774 + }, + { + "epoch": 1.4206332564478927, + "grad_norm": 7.79484729066865, + "learning_rate": 1.1339576434184718e-05, + "loss": 0.8221, + "step": 6775 + }, + { + "epoch": 1.42084294401342, + "grad_norm": 7.988799461913488, + "learning_rate": 1.1337332919540094e-05, + "loss": 0.9877, + "step": 6776 + }, + { + "epoch": 1.4210526315789473, + "grad_norm": 7.338870329231204, + "learning_rate": 1.133508933635484e-05, + "loss": 0.7249, + "step": 6777 + }, + { + "epoch": 1.4212623191444747, + "grad_norm": 7.466283641881016, + "learning_rate": 1.133284568474394e-05, + "loss": 0.837, + "step": 6778 + }, + { + "epoch": 1.4214720067100022, + "grad_norm": 7.317575295886132, + "learning_rate": 1.1330601964822385e-05, + "loss": 0.8228, + "step": 6779 + }, + { + "epoch": 1.4216816942755295, + "grad_norm": 6.930856716815753, + "learning_rate": 1.1328358176705174e-05, + "loss": 0.7348, + "step": 6780 + }, + { + "epoch": 1.4218913818410568, + "grad_norm": 8.411142307383761, + "learning_rate": 1.1326114320507303e-05, + "loss": 0.863, + "step": 6781 + }, + { + "epoch": 1.422101069406584, + "grad_norm": 6.016921555142951, + "learning_rate": 1.132387039634377e-05, + "loss": 0.6091, + "step": 6782 + }, + { + "epoch": 1.4223107569721116, + "grad_norm": 6.512092902098799, + "learning_rate": 1.1321626404329585e-05, + "loss": 0.7196, + "step": 6783 + }, + { + "epoch": 1.422520444537639, + "grad_norm": 7.065472327619744, + "learning_rate": 1.1319382344579755e-05, + "loss": 0.7822, + "step": 6784 + }, + { + "epoch": 1.4227301321031662, + "grad_norm": 8.793812604725721, + "learning_rate": 1.1317138217209291e-05, + "loss": 0.8931, + "step": 6785 + }, + { + "epoch": 1.4229398196686938, + "grad_norm": 6.323260672700587, + "learning_rate": 1.1314894022333212e-05, + "loss": 0.6437, + "step": 6786 + }, + { + "epoch": 1.423149507234221, + "grad_norm": 7.090874068478431, + "learning_rate": 1.131264976006653e-05, + "loss": 0.728, + "step": 6787 + }, + { + "epoch": 1.4233591947997484, + "grad_norm": 8.311724824048913, + "learning_rate": 1.1310405430524272e-05, + "loss": 0.9696, + "step": 6788 + }, + { + "epoch": 1.4235688823652757, + "grad_norm": 7.316499776186425, + "learning_rate": 1.1308161033821461e-05, + "loss": 0.7765, + "step": 6789 + }, + { + "epoch": 1.423778569930803, + "grad_norm": 6.93233136223138, + "learning_rate": 1.130591657007313e-05, + "loss": 0.7739, + "step": 6790 + }, + { + "epoch": 1.4239882574963305, + "grad_norm": 8.415228827715094, + "learning_rate": 1.130367203939431e-05, + "loss": 0.9769, + "step": 6791 + }, + { + "epoch": 1.4241979450618578, + "grad_norm": 7.746226707073446, + "learning_rate": 1.1301427441900035e-05, + "loss": 0.9966, + "step": 6792 + }, + { + "epoch": 1.4244076326273851, + "grad_norm": 7.013120406748664, + "learning_rate": 1.1299182777705351e-05, + "loss": 0.7732, + "step": 6793 + }, + { + "epoch": 1.4246173201929127, + "grad_norm": 6.547606610261444, + "learning_rate": 1.1296938046925289e-05, + "loss": 0.8093, + "step": 6794 + }, + { + "epoch": 1.42482700775844, + "grad_norm": 7.424236635872814, + "learning_rate": 1.129469324967491e-05, + "loss": 0.8441, + "step": 6795 + }, + { + "epoch": 1.4250366953239673, + "grad_norm": 7.354512909720858, + "learning_rate": 1.1292448386069252e-05, + "loss": 0.7541, + "step": 6796 + }, + { + "epoch": 1.4252463828894948, + "grad_norm": 6.068157592023968, + "learning_rate": 1.1290203456223374e-05, + "loss": 0.6598, + "step": 6797 + }, + { + "epoch": 1.425456070455022, + "grad_norm": 8.365753287769886, + "learning_rate": 1.1287958460252331e-05, + "loss": 0.8944, + "step": 6798 + }, + { + "epoch": 1.4256657580205494, + "grad_norm": 6.919343432815605, + "learning_rate": 1.1285713398271185e-05, + "loss": 0.6958, + "step": 6799 + }, + { + "epoch": 1.4258754455860767, + "grad_norm": 8.196646092298492, + "learning_rate": 1.1283468270394993e-05, + "loss": 1.1285, + "step": 6800 + }, + { + "epoch": 1.426085133151604, + "grad_norm": 7.218896919051745, + "learning_rate": 1.128122307673883e-05, + "loss": 0.8807, + "step": 6801 + }, + { + "epoch": 1.4262948207171315, + "grad_norm": 7.785695421465841, + "learning_rate": 1.1278977817417762e-05, + "loss": 0.8562, + "step": 6802 + }, + { + "epoch": 1.4265045082826588, + "grad_norm": 6.778392624836859, + "learning_rate": 1.127673249254686e-05, + "loss": 0.7605, + "step": 6803 + }, + { + "epoch": 1.4267141958481862, + "grad_norm": 7.101900565039615, + "learning_rate": 1.1274487102241205e-05, + "loss": 0.6591, + "step": 6804 + }, + { + "epoch": 1.4269238834137137, + "grad_norm": 8.561583302191645, + "learning_rate": 1.1272241646615874e-05, + "loss": 1.1022, + "step": 6805 + }, + { + "epoch": 1.427133570979241, + "grad_norm": 6.15805347737487, + "learning_rate": 1.1269996125785954e-05, + "loss": 0.7484, + "step": 6806 + }, + { + "epoch": 1.4273432585447683, + "grad_norm": 5.717574163656914, + "learning_rate": 1.1267750539866531e-05, + "loss": 0.5766, + "step": 6807 + }, + { + "epoch": 1.4275529461102956, + "grad_norm": 8.552316564846208, + "learning_rate": 1.126550488897269e-05, + "loss": 0.6417, + "step": 6808 + }, + { + "epoch": 1.427762633675823, + "grad_norm": 6.416222677272059, + "learning_rate": 1.1263259173219533e-05, + "loss": 0.9058, + "step": 6809 + }, + { + "epoch": 1.4279723212413504, + "grad_norm": 7.061573179795442, + "learning_rate": 1.1261013392722149e-05, + "loss": 0.9154, + "step": 6810 + }, + { + "epoch": 1.4281820088068777, + "grad_norm": 7.211097942811558, + "learning_rate": 1.1258767547595641e-05, + "loss": 0.9448, + "step": 6811 + }, + { + "epoch": 1.428391696372405, + "grad_norm": 7.471317820971958, + "learning_rate": 1.1256521637955116e-05, + "loss": 0.9165, + "step": 6812 + }, + { + "epoch": 1.4286013839379326, + "grad_norm": 6.838154164727039, + "learning_rate": 1.1254275663915674e-05, + "loss": 0.7439, + "step": 6813 + }, + { + "epoch": 1.4288110715034599, + "grad_norm": 6.269883511501781, + "learning_rate": 1.125202962559243e-05, + "loss": 0.857, + "step": 6814 + }, + { + "epoch": 1.4290207590689872, + "grad_norm": 7.609776467570679, + "learning_rate": 1.12497835231005e-05, + "loss": 0.8754, + "step": 6815 + }, + { + "epoch": 1.4292304466345147, + "grad_norm": 6.48014643829328, + "learning_rate": 1.1247537356554994e-05, + "loss": 0.8153, + "step": 6816 + }, + { + "epoch": 1.429440134200042, + "grad_norm": 7.295291520523818, + "learning_rate": 1.1245291126071035e-05, + "loss": 0.8505, + "step": 6817 + }, + { + "epoch": 1.4296498217655693, + "grad_norm": 6.889204933847784, + "learning_rate": 1.1243044831763744e-05, + "loss": 0.7721, + "step": 6818 + }, + { + "epoch": 1.4298595093310966, + "grad_norm": 8.469373231386127, + "learning_rate": 1.1240798473748252e-05, + "loss": 0.999, + "step": 6819 + }, + { + "epoch": 1.430069196896624, + "grad_norm": 6.509745627852387, + "learning_rate": 1.1238552052139684e-05, + "loss": 0.745, + "step": 6820 + }, + { + "epoch": 1.4302788844621515, + "grad_norm": 8.08729757098183, + "learning_rate": 1.1236305567053176e-05, + "loss": 0.8381, + "step": 6821 + }, + { + "epoch": 1.4304885720276788, + "grad_norm": 6.463844679219682, + "learning_rate": 1.1234059018603867e-05, + "loss": 0.7346, + "step": 6822 + }, + { + "epoch": 1.430698259593206, + "grad_norm": 6.506432636843981, + "learning_rate": 1.1231812406906885e-05, + "loss": 0.7153, + "step": 6823 + }, + { + "epoch": 1.4309079471587336, + "grad_norm": 8.201294357528045, + "learning_rate": 1.122956573207739e-05, + "loss": 0.9634, + "step": 6824 + }, + { + "epoch": 1.431117634724261, + "grad_norm": 7.680738299593677, + "learning_rate": 1.1227318994230512e-05, + "loss": 0.9054, + "step": 6825 + }, + { + "epoch": 1.4313273222897882, + "grad_norm": 5.994903141195533, + "learning_rate": 1.122507219348141e-05, + "loss": 0.7382, + "step": 6826 + }, + { + "epoch": 1.4315370098553155, + "grad_norm": 5.707624272431289, + "learning_rate": 1.122282532994523e-05, + "loss": 0.7145, + "step": 6827 + }, + { + "epoch": 1.4317466974208428, + "grad_norm": 7.897009353019369, + "learning_rate": 1.1220578403737137e-05, + "loss": 0.89, + "step": 6828 + }, + { + "epoch": 1.4319563849863703, + "grad_norm": 6.319534739876409, + "learning_rate": 1.1218331414972277e-05, + "loss": 0.7071, + "step": 6829 + }, + { + "epoch": 1.4321660725518977, + "grad_norm": 5.15359936519419, + "learning_rate": 1.1216084363765825e-05, + "loss": 0.6287, + "step": 6830 + }, + { + "epoch": 1.432375760117425, + "grad_norm": 6.737012297382711, + "learning_rate": 1.1213837250232935e-05, + "loss": 0.7068, + "step": 6831 + }, + { + "epoch": 1.4325854476829525, + "grad_norm": 7.5346853377034675, + "learning_rate": 1.1211590074488783e-05, + "loss": 0.8438, + "step": 6832 + }, + { + "epoch": 1.4327951352484798, + "grad_norm": 7.816487233258562, + "learning_rate": 1.1209342836648538e-05, + "loss": 1.0707, + "step": 6833 + }, + { + "epoch": 1.433004822814007, + "grad_norm": 6.861830249273758, + "learning_rate": 1.1207095536827376e-05, + "loss": 0.9528, + "step": 6834 + }, + { + "epoch": 1.4332145103795346, + "grad_norm": 8.068642141290168, + "learning_rate": 1.1204848175140472e-05, + "loss": 0.8905, + "step": 6835 + }, + { + "epoch": 1.433424197945062, + "grad_norm": 5.713705973028899, + "learning_rate": 1.1202600751703013e-05, + "loss": 0.5972, + "step": 6836 + }, + { + "epoch": 1.4336338855105892, + "grad_norm": 7.042698191155072, + "learning_rate": 1.1200353266630173e-05, + "loss": 0.887, + "step": 6837 + }, + { + "epoch": 1.4338435730761165, + "grad_norm": 8.732268642872002, + "learning_rate": 1.1198105720037154e-05, + "loss": 1.1082, + "step": 6838 + }, + { + "epoch": 1.4340532606416438, + "grad_norm": 6.380987528333441, + "learning_rate": 1.1195858112039136e-05, + "loss": 0.7097, + "step": 6839 + }, + { + "epoch": 1.4342629482071714, + "grad_norm": 8.448333331418867, + "learning_rate": 1.1193610442751313e-05, + "loss": 0.833, + "step": 6840 + }, + { + "epoch": 1.4344726357726987, + "grad_norm": 6.381589242359744, + "learning_rate": 1.1191362712288886e-05, + "loss": 0.7075, + "step": 6841 + }, + { + "epoch": 1.434682323338226, + "grad_norm": 7.89764556914232, + "learning_rate": 1.1189114920767054e-05, + "loss": 0.7255, + "step": 6842 + }, + { + "epoch": 1.4348920109037535, + "grad_norm": 8.372742998193994, + "learning_rate": 1.118686706830102e-05, + "loss": 0.8649, + "step": 6843 + }, + { + "epoch": 1.4351016984692808, + "grad_norm": 5.711536870467814, + "learning_rate": 1.1184619155005989e-05, + "loss": 0.694, + "step": 6844 + }, + { + "epoch": 1.4353113860348081, + "grad_norm": 6.1391252319342575, + "learning_rate": 1.1182371180997175e-05, + "loss": 0.8052, + "step": 6845 + }, + { + "epoch": 1.4355210736003354, + "grad_norm": 8.167742332903346, + "learning_rate": 1.1180123146389783e-05, + "loss": 0.7993, + "step": 6846 + }, + { + "epoch": 1.4357307611658627, + "grad_norm": 9.865808095658805, + "learning_rate": 1.117787505129904e-05, + "loss": 1.1074, + "step": 6847 + }, + { + "epoch": 1.4359404487313903, + "grad_norm": 7.46567715244991, + "learning_rate": 1.117562689584015e-05, + "loss": 0.6687, + "step": 6848 + }, + { + "epoch": 1.4361501362969176, + "grad_norm": 5.684198534797, + "learning_rate": 1.1173378680128346e-05, + "loss": 0.5612, + "step": 6849 + }, + { + "epoch": 1.4363598238624449, + "grad_norm": 8.10394183334663, + "learning_rate": 1.1171130404278849e-05, + "loss": 0.8482, + "step": 6850 + }, + { + "epoch": 1.4365695114279724, + "grad_norm": 6.615306366319538, + "learning_rate": 1.1168882068406887e-05, + "loss": 0.5142, + "step": 6851 + }, + { + "epoch": 1.4367791989934997, + "grad_norm": 7.571027684801783, + "learning_rate": 1.1166633672627692e-05, + "loss": 0.9545, + "step": 6852 + }, + { + "epoch": 1.436988886559027, + "grad_norm": 7.590467612895191, + "learning_rate": 1.1164385217056499e-05, + "loss": 0.9034, + "step": 6853 + }, + { + "epoch": 1.4371985741245545, + "grad_norm": 6.51941949486874, + "learning_rate": 1.1162136701808543e-05, + "loss": 0.8776, + "step": 6854 + }, + { + "epoch": 1.4374082616900818, + "grad_norm": 6.992731292453129, + "learning_rate": 1.1159888126999065e-05, + "loss": 0.6193, + "step": 6855 + }, + { + "epoch": 1.4376179492556092, + "grad_norm": 6.27191166475011, + "learning_rate": 1.115763949274331e-05, + "loss": 0.8363, + "step": 6856 + }, + { + "epoch": 1.4378276368211365, + "grad_norm": 5.535020701068475, + "learning_rate": 1.1155390799156525e-05, + "loss": 0.6222, + "step": 6857 + }, + { + "epoch": 1.4380373243866638, + "grad_norm": 7.695522514372399, + "learning_rate": 1.115314204635395e-05, + "loss": 0.9623, + "step": 6858 + }, + { + "epoch": 1.4382470119521913, + "grad_norm": 7.296892077667993, + "learning_rate": 1.1150893234450855e-05, + "loss": 0.7106, + "step": 6859 + }, + { + "epoch": 1.4384566995177186, + "grad_norm": 7.673352068248968, + "learning_rate": 1.114864436356248e-05, + "loss": 0.8278, + "step": 6860 + }, + { + "epoch": 1.438666387083246, + "grad_norm": 7.040973132530379, + "learning_rate": 1.1146395433804089e-05, + "loss": 0.8479, + "step": 6861 + }, + { + "epoch": 1.4388760746487734, + "grad_norm": 6.607635933500828, + "learning_rate": 1.1144146445290944e-05, + "loss": 0.7321, + "step": 6862 + }, + { + "epoch": 1.4390857622143007, + "grad_norm": 5.495097877891542, + "learning_rate": 1.114189739813831e-05, + "loss": 0.638, + "step": 6863 + }, + { + "epoch": 1.439295449779828, + "grad_norm": 7.301819846091403, + "learning_rate": 1.1139648292461453e-05, + "loss": 0.8304, + "step": 6864 + }, + { + "epoch": 1.4395051373453553, + "grad_norm": 7.131639686568112, + "learning_rate": 1.1137399128375648e-05, + "loss": 0.6566, + "step": 6865 + }, + { + "epoch": 1.4397148249108827, + "grad_norm": 7.524496340441047, + "learning_rate": 1.113514990599616e-05, + "loss": 0.8597, + "step": 6866 + }, + { + "epoch": 1.4399245124764102, + "grad_norm": 6.956022342830008, + "learning_rate": 1.1132900625438274e-05, + "loss": 0.6969, + "step": 6867 + }, + { + "epoch": 1.4401342000419375, + "grad_norm": 7.546059746077672, + "learning_rate": 1.1130651286817265e-05, + "loss": 0.7141, + "step": 6868 + }, + { + "epoch": 1.4403438876074648, + "grad_norm": 9.220226344601553, + "learning_rate": 1.1128401890248415e-05, + "loss": 1.0678, + "step": 6869 + }, + { + "epoch": 1.4405535751729923, + "grad_norm": 8.926742237162193, + "learning_rate": 1.1126152435847012e-05, + "loss": 0.8513, + "step": 6870 + }, + { + "epoch": 1.4407632627385196, + "grad_norm": 6.901832700893549, + "learning_rate": 1.1123902923728342e-05, + "loss": 0.7339, + "step": 6871 + }, + { + "epoch": 1.440972950304047, + "grad_norm": 6.371586708776338, + "learning_rate": 1.11216533540077e-05, + "loss": 0.7701, + "step": 6872 + }, + { + "epoch": 1.4411826378695745, + "grad_norm": 6.795254688220955, + "learning_rate": 1.1119403726800377e-05, + "loss": 0.9154, + "step": 6873 + }, + { + "epoch": 1.4413923254351018, + "grad_norm": 5.90342045155571, + "learning_rate": 1.1117154042221677e-05, + "loss": 0.5962, + "step": 6874 + }, + { + "epoch": 1.441602013000629, + "grad_norm": 7.298245507032124, + "learning_rate": 1.1114904300386886e-05, + "loss": 0.8072, + "step": 6875 + }, + { + "epoch": 1.4418117005661564, + "grad_norm": 6.481776073143847, + "learning_rate": 1.1112654501411321e-05, + "loss": 0.9831, + "step": 6876 + }, + { + "epoch": 1.4420213881316837, + "grad_norm": 5.918697646547978, + "learning_rate": 1.1110404645410282e-05, + "loss": 0.4852, + "step": 6877 + }, + { + "epoch": 1.4422310756972112, + "grad_norm": 6.6511943933593365, + "learning_rate": 1.1108154732499079e-05, + "loss": 0.7606, + "step": 6878 + }, + { + "epoch": 1.4424407632627385, + "grad_norm": 7.504540261000289, + "learning_rate": 1.1105904762793023e-05, + "loss": 0.875, + "step": 6879 + }, + { + "epoch": 1.4426504508282658, + "grad_norm": 7.7672005691832755, + "learning_rate": 1.1103654736407432e-05, + "loss": 0.9111, + "step": 6880 + }, + { + "epoch": 1.4428601383937933, + "grad_norm": 6.606686948993776, + "learning_rate": 1.1101404653457619e-05, + "loss": 0.7731, + "step": 6881 + }, + { + "epoch": 1.4430698259593207, + "grad_norm": 7.963783671621207, + "learning_rate": 1.109915451405891e-05, + "loss": 0.8656, + "step": 6882 + }, + { + "epoch": 1.443279513524848, + "grad_norm": 8.344906365838602, + "learning_rate": 1.1096904318326623e-05, + "loss": 0.8629, + "step": 6883 + }, + { + "epoch": 1.4434892010903753, + "grad_norm": 8.368263836560889, + "learning_rate": 1.109465406637609e-05, + "loss": 0.9468, + "step": 6884 + }, + { + "epoch": 1.4436988886559026, + "grad_norm": 8.323988382216479, + "learning_rate": 1.1092403758322633e-05, + "loss": 0.7567, + "step": 6885 + }, + { + "epoch": 1.44390857622143, + "grad_norm": 7.880848924046126, + "learning_rate": 1.1090153394281595e-05, + "loss": 0.8679, + "step": 6886 + }, + { + "epoch": 1.4441182637869574, + "grad_norm": 6.838945078947538, + "learning_rate": 1.10879029743683e-05, + "loss": 0.9316, + "step": 6887 + }, + { + "epoch": 1.4443279513524847, + "grad_norm": 7.1032783562023365, + "learning_rate": 1.1085652498698091e-05, + "loss": 0.8852, + "step": 6888 + }, + { + "epoch": 1.4445376389180122, + "grad_norm": 7.021332674682388, + "learning_rate": 1.1083401967386311e-05, + "loss": 0.6722, + "step": 6889 + }, + { + "epoch": 1.4447473264835395, + "grad_norm": 8.515355877604629, + "learning_rate": 1.1081151380548299e-05, + "loss": 0.9394, + "step": 6890 + }, + { + "epoch": 1.4449570140490668, + "grad_norm": 6.193472587828282, + "learning_rate": 1.1078900738299402e-05, + "loss": 0.6843, + "step": 6891 + }, + { + "epoch": 1.4451667016145944, + "grad_norm": 5.782952204333246, + "learning_rate": 1.1076650040754975e-05, + "loss": 0.6531, + "step": 6892 + }, + { + "epoch": 1.4453763891801217, + "grad_norm": 8.732258653960399, + "learning_rate": 1.1074399288030365e-05, + "loss": 0.7034, + "step": 6893 + }, + { + "epoch": 1.445586076745649, + "grad_norm": 5.386831830278344, + "learning_rate": 1.1072148480240925e-05, + "loss": 0.6337, + "step": 6894 + }, + { + "epoch": 1.4457957643111763, + "grad_norm": 5.908290954283815, + "learning_rate": 1.1069897617502018e-05, + "loss": 0.678, + "step": 6895 + }, + { + "epoch": 1.4460054518767036, + "grad_norm": 7.897323336675538, + "learning_rate": 1.1067646699929004e-05, + "loss": 0.9813, + "step": 6896 + }, + { + "epoch": 1.4462151394422311, + "grad_norm": 7.993555765918302, + "learning_rate": 1.1065395727637245e-05, + "loss": 0.9201, + "step": 6897 + }, + { + "epoch": 1.4464248270077584, + "grad_norm": 6.7154006922599905, + "learning_rate": 1.1063144700742103e-05, + "loss": 0.7165, + "step": 6898 + }, + { + "epoch": 1.4466345145732857, + "grad_norm": 6.641355129101737, + "learning_rate": 1.1060893619358955e-05, + "loss": 0.7253, + "step": 6899 + }, + { + "epoch": 1.4468442021388133, + "grad_norm": 7.115315108583384, + "learning_rate": 1.1058642483603167e-05, + "loss": 0.7449, + "step": 6900 + }, + { + "epoch": 1.4470538897043406, + "grad_norm": 7.191787097083664, + "learning_rate": 1.1056391293590117e-05, + "loss": 0.8129, + "step": 6901 + }, + { + "epoch": 1.4472635772698679, + "grad_norm": 6.969260975628303, + "learning_rate": 1.1054140049435178e-05, + "loss": 0.8854, + "step": 6902 + }, + { + "epoch": 1.4474732648353952, + "grad_norm": 7.015662062043676, + "learning_rate": 1.1051888751253738e-05, + "loss": 0.6776, + "step": 6903 + }, + { + "epoch": 1.4476829524009227, + "grad_norm": 6.568530606508727, + "learning_rate": 1.104963739916117e-05, + "loss": 0.7591, + "step": 6904 + }, + { + "epoch": 1.44789263996645, + "grad_norm": 6.801360167705214, + "learning_rate": 1.104738599327287e-05, + "loss": 0.7625, + "step": 6905 + }, + { + "epoch": 1.4481023275319773, + "grad_norm": 6.697832381429832, + "learning_rate": 1.1045134533704218e-05, + "loss": 0.609, + "step": 6906 + }, + { + "epoch": 1.4483120150975046, + "grad_norm": 7.044685071598635, + "learning_rate": 1.1042883020570608e-05, + "loss": 0.7233, + "step": 6907 + }, + { + "epoch": 1.4485217026630322, + "grad_norm": 9.776300223920158, + "learning_rate": 1.1040631453987436e-05, + "loss": 0.8672, + "step": 6908 + }, + { + "epoch": 1.4487313902285595, + "grad_norm": 8.534666058673096, + "learning_rate": 1.1038379834070099e-05, + "loss": 0.9572, + "step": 6909 + }, + { + "epoch": 1.4489410777940868, + "grad_norm": 7.29111684840825, + "learning_rate": 1.1036128160933987e-05, + "loss": 0.7511, + "step": 6910 + }, + { + "epoch": 1.4491507653596143, + "grad_norm": 6.587983089197357, + "learning_rate": 1.1033876434694517e-05, + "loss": 0.7783, + "step": 6911 + }, + { + "epoch": 1.4493604529251416, + "grad_norm": 6.774896749683167, + "learning_rate": 1.1031624655467085e-05, + "loss": 0.9562, + "step": 6912 + }, + { + "epoch": 1.449570140490669, + "grad_norm": 7.952914255325333, + "learning_rate": 1.1029372823367099e-05, + "loss": 0.8752, + "step": 6913 + }, + { + "epoch": 1.4497798280561962, + "grad_norm": 8.1949826088685, + "learning_rate": 1.1027120938509969e-05, + "loss": 1.0126, + "step": 6914 + }, + { + "epoch": 1.4499895156217235, + "grad_norm": 7.138411227267545, + "learning_rate": 1.1024869001011114e-05, + "loss": 0.9027, + "step": 6915 + }, + { + "epoch": 1.450199203187251, + "grad_norm": 7.18447107420487, + "learning_rate": 1.1022617010985941e-05, + "loss": 0.6824, + "step": 6916 + }, + { + "epoch": 1.4504088907527783, + "grad_norm": 7.850307653510362, + "learning_rate": 1.1020364968549875e-05, + "loss": 0.9505, + "step": 6917 + }, + { + "epoch": 1.4506185783183057, + "grad_norm": 7.386083095944167, + "learning_rate": 1.1018112873818335e-05, + "loss": 0.6007, + "step": 6918 + }, + { + "epoch": 1.4508282658838332, + "grad_norm": 6.882912068940383, + "learning_rate": 1.1015860726906743e-05, + "loss": 0.7787, + "step": 6919 + }, + { + "epoch": 1.4510379534493605, + "grad_norm": 8.149664522439808, + "learning_rate": 1.1013608527930526e-05, + "loss": 0.839, + "step": 6920 + }, + { + "epoch": 1.4512476410148878, + "grad_norm": 7.844721802081185, + "learning_rate": 1.1011356277005114e-05, + "loss": 0.9131, + "step": 6921 + }, + { + "epoch": 1.4514573285804153, + "grad_norm": 7.842989979261236, + "learning_rate": 1.1009103974245938e-05, + "loss": 0.7796, + "step": 6922 + }, + { + "epoch": 1.4516670161459426, + "grad_norm": 6.5184647608922655, + "learning_rate": 1.1006851619768435e-05, + "loss": 0.8229, + "step": 6923 + }, + { + "epoch": 1.45187670371147, + "grad_norm": 8.890786859810532, + "learning_rate": 1.100459921368804e-05, + "loss": 0.9806, + "step": 6924 + }, + { + "epoch": 1.4520863912769972, + "grad_norm": 7.1443871621232855, + "learning_rate": 1.1002346756120192e-05, + "loss": 0.7822, + "step": 6925 + }, + { + "epoch": 1.4522960788425245, + "grad_norm": 7.322218271571962, + "learning_rate": 1.1000094247180337e-05, + "loss": 0.7353, + "step": 6926 + }, + { + "epoch": 1.452505766408052, + "grad_norm": 8.857157899483214, + "learning_rate": 1.0997841686983914e-05, + "loss": 1.0475, + "step": 6927 + }, + { + "epoch": 1.4527154539735794, + "grad_norm": 6.739530633147822, + "learning_rate": 1.0995589075646376e-05, + "loss": 0.7762, + "step": 6928 + }, + { + "epoch": 1.4529251415391067, + "grad_norm": 6.875342787384439, + "learning_rate": 1.099333641328317e-05, + "loss": 0.847, + "step": 6929 + }, + { + "epoch": 1.4531348291046342, + "grad_norm": 5.843777896819617, + "learning_rate": 1.099108370000975e-05, + "loss": 0.7353, + "step": 6930 + }, + { + "epoch": 1.4533445166701615, + "grad_norm": 6.41658678749503, + "learning_rate": 1.098883093594157e-05, + "loss": 0.7813, + "step": 6931 + }, + { + "epoch": 1.4535542042356888, + "grad_norm": 6.463762765442355, + "learning_rate": 1.0986578121194093e-05, + "loss": 0.5983, + "step": 6932 + }, + { + "epoch": 1.4537638918012161, + "grad_norm": 7.7043754225778205, + "learning_rate": 1.0984325255882772e-05, + "loss": 0.9623, + "step": 6933 + }, + { + "epoch": 1.4539735793667434, + "grad_norm": 7.018612668854298, + "learning_rate": 1.0982072340123078e-05, + "loss": 0.7825, + "step": 6934 + }, + { + "epoch": 1.454183266932271, + "grad_norm": 7.79225597229119, + "learning_rate": 1.097981937403047e-05, + "loss": 0.9175, + "step": 6935 + }, + { + "epoch": 1.4543929544977983, + "grad_norm": 7.797663450271207, + "learning_rate": 1.0977566357720423e-05, + "loss": 0.8233, + "step": 6936 + }, + { + "epoch": 1.4546026420633256, + "grad_norm": 8.021999765422642, + "learning_rate": 1.0975313291308402e-05, + "loss": 0.8661, + "step": 6937 + }, + { + "epoch": 1.454812329628853, + "grad_norm": 8.005624943576937, + "learning_rate": 1.0973060174909888e-05, + "loss": 0.7796, + "step": 6938 + }, + { + "epoch": 1.4550220171943804, + "grad_norm": 7.25699804286934, + "learning_rate": 1.0970807008640345e-05, + "loss": 0.9146, + "step": 6939 + }, + { + "epoch": 1.4552317047599077, + "grad_norm": 7.110779914008801, + "learning_rate": 1.0968553792615267e-05, + "loss": 0.8783, + "step": 6940 + }, + { + "epoch": 1.4554413923254352, + "grad_norm": 4.574432018591386, + "learning_rate": 1.0966300526950122e-05, + "loss": 0.4127, + "step": 6941 + }, + { + "epoch": 1.4556510798909625, + "grad_norm": 10.85671678433408, + "learning_rate": 1.09640472117604e-05, + "loss": 0.9038, + "step": 6942 + }, + { + "epoch": 1.4558607674564898, + "grad_norm": 7.174455694701378, + "learning_rate": 1.0961793847161586e-05, + "loss": 0.7581, + "step": 6943 + }, + { + "epoch": 1.4560704550220172, + "grad_norm": 7.772469407331134, + "learning_rate": 1.095954043326917e-05, + "loss": 0.8159, + "step": 6944 + }, + { + "epoch": 1.4562801425875445, + "grad_norm": 8.198968373274605, + "learning_rate": 1.0957286970198642e-05, + "loss": 0.929, + "step": 6945 + }, + { + "epoch": 1.456489830153072, + "grad_norm": 5.512110921599867, + "learning_rate": 1.0955033458065499e-05, + "loss": 0.5957, + "step": 6946 + }, + { + "epoch": 1.4566995177185993, + "grad_norm": 7.141574330632033, + "learning_rate": 1.095277989698523e-05, + "loss": 0.5706, + "step": 6947 + }, + { + "epoch": 1.4569092052841266, + "grad_norm": 8.475745144124659, + "learning_rate": 1.0950526287073343e-05, + "loss": 0.7797, + "step": 6948 + }, + { + "epoch": 1.4571188928496541, + "grad_norm": 8.053372807005887, + "learning_rate": 1.0948272628445331e-05, + "loss": 1.1298, + "step": 6949 + }, + { + "epoch": 1.4573285804151814, + "grad_norm": 7.3863802655281265, + "learning_rate": 1.0946018921216705e-05, + "loss": 0.7573, + "step": 6950 + }, + { + "epoch": 1.4575382679807087, + "grad_norm": 6.318406022325628, + "learning_rate": 1.0943765165502968e-05, + "loss": 0.6765, + "step": 6951 + }, + { + "epoch": 1.457747955546236, + "grad_norm": 8.882307898432142, + "learning_rate": 1.094151136141963e-05, + "loss": 1.0538, + "step": 6952 + }, + { + "epoch": 1.4579576431117633, + "grad_norm": 5.350578988501194, + "learning_rate": 1.0939257509082202e-05, + "loss": 0.6303, + "step": 6953 + }, + { + "epoch": 1.4581673306772909, + "grad_norm": 7.422614753146623, + "learning_rate": 1.0937003608606194e-05, + "loss": 0.8199, + "step": 6954 + }, + { + "epoch": 1.4583770182428182, + "grad_norm": 6.7216625049908485, + "learning_rate": 1.0934749660107131e-05, + "loss": 0.7224, + "step": 6955 + }, + { + "epoch": 1.4585867058083455, + "grad_norm": 7.7932868476710055, + "learning_rate": 1.0932495663700522e-05, + "loss": 0.7766, + "step": 6956 + }, + { + "epoch": 1.458796393373873, + "grad_norm": 6.493093401288736, + "learning_rate": 1.0930241619501898e-05, + "loss": 0.6023, + "step": 6957 + }, + { + "epoch": 1.4590060809394003, + "grad_norm": 9.06989515130504, + "learning_rate": 1.0927987527626773e-05, + "loss": 1.0163, + "step": 6958 + }, + { + "epoch": 1.4592157685049276, + "grad_norm": 6.748799009956378, + "learning_rate": 1.092573338819068e-05, + "loss": 0.813, + "step": 6959 + }, + { + "epoch": 1.4594254560704552, + "grad_norm": 7.319475751817692, + "learning_rate": 1.0923479201309144e-05, + "loss": 0.8586, + "step": 6960 + }, + { + "epoch": 1.4596351436359825, + "grad_norm": 7.424438084687504, + "learning_rate": 1.0921224967097701e-05, + "loss": 0.8178, + "step": 6961 + }, + { + "epoch": 1.4598448312015098, + "grad_norm": 7.589390352914218, + "learning_rate": 1.0918970685671874e-05, + "loss": 0.9653, + "step": 6962 + }, + { + "epoch": 1.460054518767037, + "grad_norm": 7.186533920004667, + "learning_rate": 1.0916716357147213e-05, + "loss": 0.7361, + "step": 6963 + }, + { + "epoch": 1.4602642063325644, + "grad_norm": 7.992026114395873, + "learning_rate": 1.0914461981639244e-05, + "loss": 0.8174, + "step": 6964 + }, + { + "epoch": 1.460473893898092, + "grad_norm": 7.6054046786861385, + "learning_rate": 1.0912207559263513e-05, + "loss": 0.8008, + "step": 6965 + }, + { + "epoch": 1.4606835814636192, + "grad_norm": 7.1621197819276565, + "learning_rate": 1.0909953090135561e-05, + "loss": 0.8023, + "step": 6966 + }, + { + "epoch": 1.4608932690291465, + "grad_norm": 6.5964627963707265, + "learning_rate": 1.090769857437094e-05, + "loss": 0.7509, + "step": 6967 + }, + { + "epoch": 1.461102956594674, + "grad_norm": 6.768847624565883, + "learning_rate": 1.0905444012085188e-05, + "loss": 0.7515, + "step": 6968 + }, + { + "epoch": 1.4613126441602013, + "grad_norm": 7.105126024005378, + "learning_rate": 1.0903189403393862e-05, + "loss": 0.7554, + "step": 6969 + }, + { + "epoch": 1.4615223317257287, + "grad_norm": 6.47221993332316, + "learning_rate": 1.090093474841251e-05, + "loss": 0.7285, + "step": 6970 + }, + { + "epoch": 1.461732019291256, + "grad_norm": 6.021824291062265, + "learning_rate": 1.0898680047256693e-05, + "loss": 0.646, + "step": 6971 + }, + { + "epoch": 1.4619417068567833, + "grad_norm": 8.91745955274642, + "learning_rate": 1.0896425300041961e-05, + "loss": 0.9178, + "step": 6972 + }, + { + "epoch": 1.4621513944223108, + "grad_norm": 8.02855279616266, + "learning_rate": 1.0894170506883882e-05, + "loss": 0.9764, + "step": 6973 + }, + { + "epoch": 1.462361081987838, + "grad_norm": 7.5319468766420705, + "learning_rate": 1.0891915667898009e-05, + "loss": 0.869, + "step": 6974 + }, + { + "epoch": 1.4625707695533654, + "grad_norm": 6.851904643187208, + "learning_rate": 1.0889660783199918e-05, + "loss": 0.7802, + "step": 6975 + }, + { + "epoch": 1.462780457118893, + "grad_norm": 7.895049987192538, + "learning_rate": 1.0887405852905165e-05, + "loss": 1.0078, + "step": 6976 + }, + { + "epoch": 1.4629901446844202, + "grad_norm": 6.174017721352524, + "learning_rate": 1.088515087712932e-05, + "loss": 0.7581, + "step": 6977 + }, + { + "epoch": 1.4631998322499475, + "grad_norm": 6.175165017766711, + "learning_rate": 1.0882895855987961e-05, + "loss": 0.6452, + "step": 6978 + }, + { + "epoch": 1.463409519815475, + "grad_norm": 5.647606342514414, + "learning_rate": 1.0880640789596654e-05, + "loss": 0.6178, + "step": 6979 + }, + { + "epoch": 1.4636192073810024, + "grad_norm": 6.195371793785506, + "learning_rate": 1.0878385678070984e-05, + "loss": 0.7581, + "step": 6980 + }, + { + "epoch": 1.4638288949465297, + "grad_norm": 7.686737086656785, + "learning_rate": 1.0876130521526522e-05, + "loss": 0.7766, + "step": 6981 + }, + { + "epoch": 1.464038582512057, + "grad_norm": 6.5541007330019445, + "learning_rate": 1.0873875320078852e-05, + "loss": 0.7089, + "step": 6982 + }, + { + "epoch": 1.4642482700775843, + "grad_norm": 9.367968565476625, + "learning_rate": 1.0871620073843554e-05, + "loss": 0.9481, + "step": 6983 + }, + { + "epoch": 1.4644579576431118, + "grad_norm": 7.6243250261817765, + "learning_rate": 1.086936478293622e-05, + "loss": 0.7618, + "step": 6984 + }, + { + "epoch": 1.4646676452086391, + "grad_norm": 6.676643628936052, + "learning_rate": 1.086710944747243e-05, + "loss": 0.6947, + "step": 6985 + }, + { + "epoch": 1.4648773327741664, + "grad_norm": 7.294701536903166, + "learning_rate": 1.0864854067567777e-05, + "loss": 0.864, + "step": 6986 + }, + { + "epoch": 1.465087020339694, + "grad_norm": 5.057533797854403, + "learning_rate": 1.0862598643337854e-05, + "loss": 0.4011, + "step": 6987 + }, + { + "epoch": 1.4652967079052213, + "grad_norm": 7.207787166278459, + "learning_rate": 1.086034317489825e-05, + "loss": 0.7775, + "step": 6988 + }, + { + "epoch": 1.4655063954707486, + "grad_norm": 6.6724294883215345, + "learning_rate": 1.0858087662364572e-05, + "loss": 0.767, + "step": 6989 + }, + { + "epoch": 1.4657160830362759, + "grad_norm": 7.472615112773731, + "learning_rate": 1.0855832105852413e-05, + "loss": 0.9126, + "step": 6990 + }, + { + "epoch": 1.4659257706018032, + "grad_norm": 7.340186983267856, + "learning_rate": 1.085357650547737e-05, + "loss": 0.6979, + "step": 6991 + }, + { + "epoch": 1.4661354581673307, + "grad_norm": 6.690484240622299, + "learning_rate": 1.0851320861355055e-05, + "loss": 0.7461, + "step": 6992 + }, + { + "epoch": 1.466345145732858, + "grad_norm": 6.980456969283783, + "learning_rate": 1.0849065173601066e-05, + "loss": 0.821, + "step": 6993 + }, + { + "epoch": 1.4665548332983853, + "grad_norm": 5.824463113043737, + "learning_rate": 1.0846809442331014e-05, + "loss": 0.5662, + "step": 6994 + }, + { + "epoch": 1.4667645208639128, + "grad_norm": 9.593140131129912, + "learning_rate": 1.0844553667660511e-05, + "loss": 0.9659, + "step": 6995 + }, + { + "epoch": 1.4669742084294402, + "grad_norm": 6.736554412298324, + "learning_rate": 1.084229784970517e-05, + "loss": 0.7312, + "step": 6996 + }, + { + "epoch": 1.4671838959949675, + "grad_norm": 6.496612408555153, + "learning_rate": 1.0840041988580596e-05, + "loss": 0.5665, + "step": 6997 + }, + { + "epoch": 1.467393583560495, + "grad_norm": 6.989413718189927, + "learning_rate": 1.0837786084402422e-05, + "loss": 0.8089, + "step": 6998 + }, + { + "epoch": 1.4676032711260223, + "grad_norm": 7.831308445194834, + "learning_rate": 1.0835530137286252e-05, + "loss": 0.8865, + "step": 6999 + }, + { + "epoch": 1.4678129586915496, + "grad_norm": 8.014899865426337, + "learning_rate": 1.0833274147347716e-05, + "loss": 1.039, + "step": 7000 + }, + { + "epoch": 1.468022646257077, + "grad_norm": 7.435002702350244, + "learning_rate": 1.0831018114702435e-05, + "loss": 0.7409, + "step": 7001 + }, + { + "epoch": 1.4682323338226042, + "grad_norm": 6.257718963912202, + "learning_rate": 1.082876203946603e-05, + "loss": 0.6339, + "step": 7002 + }, + { + "epoch": 1.4684420213881317, + "grad_norm": 6.452257626006419, + "learning_rate": 1.0826505921754137e-05, + "loss": 0.7642, + "step": 7003 + }, + { + "epoch": 1.468651708953659, + "grad_norm": 6.50086412279567, + "learning_rate": 1.0824249761682382e-05, + "loss": 0.7113, + "step": 7004 + }, + { + "epoch": 1.4688613965191863, + "grad_norm": 6.584702201126405, + "learning_rate": 1.0821993559366398e-05, + "loss": 0.7537, + "step": 7005 + }, + { + "epoch": 1.4690710840847139, + "grad_norm": 8.099909513306217, + "learning_rate": 1.0819737314921817e-05, + "loss": 0.945, + "step": 7006 + }, + { + "epoch": 1.4692807716502412, + "grad_norm": 7.5890492483824055, + "learning_rate": 1.0817481028464274e-05, + "loss": 0.8279, + "step": 7007 + }, + { + "epoch": 1.4694904592157685, + "grad_norm": 8.40989966317815, + "learning_rate": 1.0815224700109414e-05, + "loss": 0.7376, + "step": 7008 + }, + { + "epoch": 1.4697001467812958, + "grad_norm": 6.415005363112198, + "learning_rate": 1.0812968329972873e-05, + "loss": 0.7743, + "step": 7009 + }, + { + "epoch": 1.469909834346823, + "grad_norm": 7.024819439683643, + "learning_rate": 1.0810711918170296e-05, + "loss": 0.7637, + "step": 7010 + }, + { + "epoch": 1.4701195219123506, + "grad_norm": 6.746138328466806, + "learning_rate": 1.0808455464817326e-05, + "loss": 0.7718, + "step": 7011 + }, + { + "epoch": 1.470329209477878, + "grad_norm": 7.058141736254412, + "learning_rate": 1.0806198970029612e-05, + "loss": 0.6958, + "step": 7012 + }, + { + "epoch": 1.4705388970434052, + "grad_norm": 7.815646219691848, + "learning_rate": 1.0803942433922804e-05, + "loss": 0.8925, + "step": 7013 + }, + { + "epoch": 1.4707485846089328, + "grad_norm": 8.719273444241164, + "learning_rate": 1.0801685856612547e-05, + "loss": 0.9824, + "step": 7014 + }, + { + "epoch": 1.47095827217446, + "grad_norm": 8.69743326489451, + "learning_rate": 1.0799429238214502e-05, + "loss": 0.8838, + "step": 7015 + }, + { + "epoch": 1.4711679597399874, + "grad_norm": 8.689867937918082, + "learning_rate": 1.0797172578844324e-05, + "loss": 0.9967, + "step": 7016 + }, + { + "epoch": 1.471377647305515, + "grad_norm": 7.708944331930854, + "learning_rate": 1.0794915878617667e-05, + "loss": 0.7831, + "step": 7017 + }, + { + "epoch": 1.4715873348710422, + "grad_norm": 7.684149756432796, + "learning_rate": 1.0792659137650191e-05, + "loss": 0.785, + "step": 7018 + }, + { + "epoch": 1.4717970224365695, + "grad_norm": 7.3221774509317505, + "learning_rate": 1.0790402356057563e-05, + "loss": 0.8536, + "step": 7019 + }, + { + "epoch": 1.4720067100020968, + "grad_norm": 7.175964656047589, + "learning_rate": 1.0788145533955439e-05, + "loss": 0.7382, + "step": 7020 + }, + { + "epoch": 1.4722163975676241, + "grad_norm": 7.614592141334634, + "learning_rate": 1.0785888671459494e-05, + "loss": 0.8799, + "step": 7021 + }, + { + "epoch": 1.4724260851331517, + "grad_norm": 6.016435102544791, + "learning_rate": 1.0783631768685389e-05, + "loss": 0.6672, + "step": 7022 + }, + { + "epoch": 1.472635772698679, + "grad_norm": 6.485757201309465, + "learning_rate": 1.0781374825748795e-05, + "loss": 0.6662, + "step": 7023 + }, + { + "epoch": 1.4728454602642063, + "grad_norm": 6.632678340312162, + "learning_rate": 1.077911784276539e-05, + "loss": 0.7688, + "step": 7024 + }, + { + "epoch": 1.4730551478297338, + "grad_norm": 5.984807281779909, + "learning_rate": 1.0776860819850846e-05, + "loss": 0.5357, + "step": 7025 + }, + { + "epoch": 1.473264835395261, + "grad_norm": 6.582498979221595, + "learning_rate": 1.077460375712083e-05, + "loss": 0.7566, + "step": 7026 + }, + { + "epoch": 1.4734745229607884, + "grad_norm": 7.997955168965671, + "learning_rate": 1.0772346654691033e-05, + "loss": 0.8477, + "step": 7027 + }, + { + "epoch": 1.4736842105263157, + "grad_norm": 8.261956867547118, + "learning_rate": 1.0770089512677129e-05, + "loss": 0.7434, + "step": 7028 + }, + { + "epoch": 1.4738938980918432, + "grad_norm": 8.138387642820739, + "learning_rate": 1.07678323311948e-05, + "loss": 0.8062, + "step": 7029 + }, + { + "epoch": 1.4741035856573705, + "grad_norm": 6.918887172985046, + "learning_rate": 1.0765575110359735e-05, + "loss": 0.6932, + "step": 7030 + }, + { + "epoch": 1.4743132732228978, + "grad_norm": 9.969833676299642, + "learning_rate": 1.0763317850287618e-05, + "loss": 1.1169, + "step": 7031 + }, + { + "epoch": 1.4745229607884252, + "grad_norm": 8.112814820289168, + "learning_rate": 1.0761060551094135e-05, + "loss": 0.8434, + "step": 7032 + }, + { + "epoch": 1.4747326483539527, + "grad_norm": 6.706794037787037, + "learning_rate": 1.0758803212894981e-05, + "loss": 0.8027, + "step": 7033 + }, + { + "epoch": 1.47494233591948, + "grad_norm": 6.307318762916073, + "learning_rate": 1.0756545835805847e-05, + "loss": 0.6003, + "step": 7034 + }, + { + "epoch": 1.4751520234850073, + "grad_norm": 6.697717672924268, + "learning_rate": 1.0754288419942422e-05, + "loss": 0.8563, + "step": 7035 + }, + { + "epoch": 1.4753617110505348, + "grad_norm": 6.226644419426775, + "learning_rate": 1.0752030965420413e-05, + "loss": 0.6525, + "step": 7036 + }, + { + "epoch": 1.4755713986160621, + "grad_norm": 7.208126792628617, + "learning_rate": 1.0749773472355509e-05, + "loss": 0.6862, + "step": 7037 + }, + { + "epoch": 1.4757810861815894, + "grad_norm": 8.257804750603077, + "learning_rate": 1.0747515940863416e-05, + "loss": 1.0113, + "step": 7038 + }, + { + "epoch": 1.4759907737471167, + "grad_norm": 8.731089854922134, + "learning_rate": 1.0745258371059832e-05, + "loss": 0.88, + "step": 7039 + }, + { + "epoch": 1.476200461312644, + "grad_norm": 6.58570148768025, + "learning_rate": 1.074300076306047e-05, + "loss": 0.5896, + "step": 7040 + }, + { + "epoch": 1.4764101488781716, + "grad_norm": 6.398412668904247, + "learning_rate": 1.0740743116981023e-05, + "loss": 0.5519, + "step": 7041 + }, + { + "epoch": 1.4766198364436989, + "grad_norm": 6.41983844347742, + "learning_rate": 1.0738485432937211e-05, + "loss": 0.8377, + "step": 7042 + }, + { + "epoch": 1.4768295240092262, + "grad_norm": 7.482973538452351, + "learning_rate": 1.0736227711044738e-05, + "loss": 0.779, + "step": 7043 + }, + { + "epoch": 1.4770392115747537, + "grad_norm": 7.751294241402007, + "learning_rate": 1.0733969951419317e-05, + "loss": 0.9654, + "step": 7044 + }, + { + "epoch": 1.477248899140281, + "grad_norm": 7.00281367606814, + "learning_rate": 1.0731712154176663e-05, + "loss": 0.6882, + "step": 7045 + }, + { + "epoch": 1.4774585867058083, + "grad_norm": 8.84644468361504, + "learning_rate": 1.0729454319432495e-05, + "loss": 0.8992, + "step": 7046 + }, + { + "epoch": 1.4776682742713358, + "grad_norm": 8.600213816495094, + "learning_rate": 1.0727196447302523e-05, + "loss": 0.8576, + "step": 7047 + }, + { + "epoch": 1.4778779618368632, + "grad_norm": 7.332713326080814, + "learning_rate": 1.0724938537902476e-05, + "loss": 0.9805, + "step": 7048 + }, + { + "epoch": 1.4780876494023905, + "grad_norm": 7.780696758266012, + "learning_rate": 1.072268059134807e-05, + "loss": 0.8381, + "step": 7049 + }, + { + "epoch": 1.4782973369679178, + "grad_norm": 6.007873542597066, + "learning_rate": 1.0720422607755032e-05, + "loss": 0.608, + "step": 7050 + }, + { + "epoch": 1.478507024533445, + "grad_norm": 7.278521117226383, + "learning_rate": 1.0718164587239084e-05, + "loss": 0.786, + "step": 7051 + }, + { + "epoch": 1.4787167120989726, + "grad_norm": 6.708216431656241, + "learning_rate": 1.0715906529915956e-05, + "loss": 0.8159, + "step": 7052 + }, + { + "epoch": 1.4789263996645, + "grad_norm": 9.06760666816895, + "learning_rate": 1.0713648435901376e-05, + "loss": 0.8638, + "step": 7053 + }, + { + "epoch": 1.4791360872300272, + "grad_norm": 7.160288261397438, + "learning_rate": 1.0711390305311077e-05, + "loss": 0.8218, + "step": 7054 + }, + { + "epoch": 1.4793457747955547, + "grad_norm": 9.012669355243682, + "learning_rate": 1.0709132138260788e-05, + "loss": 1.0622, + "step": 7055 + }, + { + "epoch": 1.479555462361082, + "grad_norm": 8.541998536022632, + "learning_rate": 1.0706873934866251e-05, + "loss": 0.9541, + "step": 7056 + }, + { + "epoch": 1.4797651499266093, + "grad_norm": 7.394470655159761, + "learning_rate": 1.0704615695243197e-05, + "loss": 0.7982, + "step": 7057 + }, + { + "epoch": 1.4799748374921367, + "grad_norm": 6.928453318865293, + "learning_rate": 1.0702357419507367e-05, + "loss": 0.6739, + "step": 7058 + }, + { + "epoch": 1.480184525057664, + "grad_norm": 7.729687714889983, + "learning_rate": 1.07000991077745e-05, + "loss": 0.9483, + "step": 7059 + }, + { + "epoch": 1.4803942126231915, + "grad_norm": 6.751826441820275, + "learning_rate": 1.069784076016034e-05, + "loss": 0.7243, + "step": 7060 + }, + { + "epoch": 1.4806039001887188, + "grad_norm": 7.674851410482444, + "learning_rate": 1.0695582376780632e-05, + "loss": 0.7664, + "step": 7061 + }, + { + "epoch": 1.480813587754246, + "grad_norm": 10.9091376673557, + "learning_rate": 1.0693323957751118e-05, + "loss": 1.0597, + "step": 7062 + }, + { + "epoch": 1.4810232753197736, + "grad_norm": 7.616185609511507, + "learning_rate": 1.0691065503187552e-05, + "loss": 0.8427, + "step": 7063 + }, + { + "epoch": 1.481232962885301, + "grad_norm": 7.184007538353708, + "learning_rate": 1.0688807013205675e-05, + "loss": 0.7318, + "step": 7064 + }, + { + "epoch": 1.4814426504508282, + "grad_norm": 7.07671838193026, + "learning_rate": 1.068654848792125e-05, + "loss": 0.8175, + "step": 7065 + }, + { + "epoch": 1.4816523380163558, + "grad_norm": 5.543034183164377, + "learning_rate": 1.0684289927450022e-05, + "loss": 0.4713, + "step": 7066 + }, + { + "epoch": 1.481862025581883, + "grad_norm": 7.13368237576877, + "learning_rate": 1.0682031331907748e-05, + "loss": 0.685, + "step": 7067 + }, + { + "epoch": 1.4820717131474104, + "grad_norm": 7.007583046402595, + "learning_rate": 1.0679772701410185e-05, + "loss": 0.6954, + "step": 7068 + }, + { + "epoch": 1.4822814007129377, + "grad_norm": 6.347053411761595, + "learning_rate": 1.0677514036073096e-05, + "loss": 0.6355, + "step": 7069 + }, + { + "epoch": 1.482491088278465, + "grad_norm": 5.882246874755733, + "learning_rate": 1.0675255336012232e-05, + "loss": 0.6091, + "step": 7070 + }, + { + "epoch": 1.4827007758439925, + "grad_norm": 8.70443673196323, + "learning_rate": 1.0672996601343363e-05, + "loss": 1.0314, + "step": 7071 + }, + { + "epoch": 1.4829104634095198, + "grad_norm": 7.295387126584218, + "learning_rate": 1.0670737832182251e-05, + "loss": 0.9358, + "step": 7072 + }, + { + "epoch": 1.4831201509750471, + "grad_norm": 8.552708804555555, + "learning_rate": 1.0668479028644661e-05, + "loss": 0.9032, + "step": 7073 + }, + { + "epoch": 1.4833298385405747, + "grad_norm": 6.593498558096608, + "learning_rate": 1.0666220190846361e-05, + "loss": 0.6793, + "step": 7074 + }, + { + "epoch": 1.483539526106102, + "grad_norm": 6.793962967857767, + "learning_rate": 1.0663961318903124e-05, + "loss": 0.683, + "step": 7075 + }, + { + "epoch": 1.4837492136716293, + "grad_norm": 6.9491709341116366, + "learning_rate": 1.0661702412930715e-05, + "loss": 0.8255, + "step": 7076 + }, + { + "epoch": 1.4839589012371566, + "grad_norm": 6.815572034374432, + "learning_rate": 1.065944347304491e-05, + "loss": 0.7957, + "step": 7077 + }, + { + "epoch": 1.4841685888026839, + "grad_norm": 7.4930338013630955, + "learning_rate": 1.0657184499361484e-05, + "loss": 0.683, + "step": 7078 + }, + { + "epoch": 1.4843782763682114, + "grad_norm": 5.597508605891956, + "learning_rate": 1.0654925491996211e-05, + "loss": 0.6202, + "step": 7079 + }, + { + "epoch": 1.4845879639337387, + "grad_norm": 6.641568660444248, + "learning_rate": 1.0652666451064873e-05, + "loss": 0.702, + "step": 7080 + }, + { + "epoch": 1.484797651499266, + "grad_norm": 6.330997846313807, + "learning_rate": 1.0650407376683246e-05, + "loss": 0.8758, + "step": 7081 + }, + { + "epoch": 1.4850073390647935, + "grad_norm": 7.884791234314769, + "learning_rate": 1.0648148268967112e-05, + "loss": 0.916, + "step": 7082 + }, + { + "epoch": 1.4852170266303208, + "grad_norm": 8.354607136848015, + "learning_rate": 1.0645889128032256e-05, + "loss": 0.8773, + "step": 7083 + }, + { + "epoch": 1.4854267141958482, + "grad_norm": 8.354384325179506, + "learning_rate": 1.064362995399446e-05, + "loss": 0.8201, + "step": 7084 + }, + { + "epoch": 1.4856364017613757, + "grad_norm": 5.928129725964623, + "learning_rate": 1.0641370746969517e-05, + "loss": 0.7918, + "step": 7085 + }, + { + "epoch": 1.485846089326903, + "grad_norm": 6.809599512678502, + "learning_rate": 1.0639111507073207e-05, + "loss": 0.9387, + "step": 7086 + }, + { + "epoch": 1.4860557768924303, + "grad_norm": 6.8006304783626375, + "learning_rate": 1.0636852234421327e-05, + "loss": 0.7452, + "step": 7087 + }, + { + "epoch": 1.4862654644579576, + "grad_norm": 5.909975032555732, + "learning_rate": 1.0634592929129664e-05, + "loss": 0.6667, + "step": 7088 + }, + { + "epoch": 1.486475152023485, + "grad_norm": 6.788903863916933, + "learning_rate": 1.063233359131401e-05, + "loss": 0.7743, + "step": 7089 + }, + { + "epoch": 1.4866848395890124, + "grad_norm": 7.900361320384673, + "learning_rate": 1.0630074221090165e-05, + "loss": 0.7609, + "step": 7090 + }, + { + "epoch": 1.4868945271545397, + "grad_norm": 6.951496769696583, + "learning_rate": 1.0627814818573923e-05, + "loss": 0.8681, + "step": 7091 + }, + { + "epoch": 1.487104214720067, + "grad_norm": 6.666027230728512, + "learning_rate": 1.0625555383881085e-05, + "loss": 0.7307, + "step": 7092 + }, + { + "epoch": 1.4873139022855946, + "grad_norm": 7.0428805657809725, + "learning_rate": 1.0623295917127444e-05, + "loss": 0.7903, + "step": 7093 + }, + { + "epoch": 1.4875235898511219, + "grad_norm": 7.692608990527666, + "learning_rate": 1.0621036418428811e-05, + "loss": 0.7342, + "step": 7094 + }, + { + "epoch": 1.4877332774166492, + "grad_norm": 8.810065710230523, + "learning_rate": 1.0618776887900981e-05, + "loss": 0.995, + "step": 7095 + }, + { + "epoch": 1.4879429649821765, + "grad_norm": 7.943922376038015, + "learning_rate": 1.0616517325659762e-05, + "loss": 0.9918, + "step": 7096 + }, + { + "epoch": 1.4881526525477038, + "grad_norm": 5.885855247850953, + "learning_rate": 1.0614257731820966e-05, + "loss": 0.6516, + "step": 7097 + }, + { + "epoch": 1.4883623401132313, + "grad_norm": 7.575614536473457, + "learning_rate": 1.0611998106500392e-05, + "loss": 0.8066, + "step": 7098 + }, + { + "epoch": 1.4885720276787586, + "grad_norm": 5.799156292745133, + "learning_rate": 1.0609738449813852e-05, + "loss": 0.7026, + "step": 7099 + }, + { + "epoch": 1.488781715244286, + "grad_norm": 7.73046256773992, + "learning_rate": 1.0607478761877162e-05, + "loss": 0.9519, + "step": 7100 + }, + { + "epoch": 1.4889914028098135, + "grad_norm": 6.728291396791191, + "learning_rate": 1.060521904280613e-05, + "loss": 0.7349, + "step": 7101 + }, + { + "epoch": 1.4892010903753408, + "grad_norm": 6.670827173872375, + "learning_rate": 1.0602959292716573e-05, + "loss": 0.7416, + "step": 7102 + }, + { + "epoch": 1.489410777940868, + "grad_norm": 7.850341996181215, + "learning_rate": 1.0600699511724307e-05, + "loss": 0.926, + "step": 7103 + }, + { + "epoch": 1.4896204655063956, + "grad_norm": 6.211664527651134, + "learning_rate": 1.059843969994515e-05, + "loss": 0.711, + "step": 7104 + }, + { + "epoch": 1.489830153071923, + "grad_norm": 6.784092200667709, + "learning_rate": 1.0596179857494915e-05, + "loss": 0.648, + "step": 7105 + }, + { + "epoch": 1.4900398406374502, + "grad_norm": 6.247031838771317, + "learning_rate": 1.0593919984489434e-05, + "loss": 0.6548, + "step": 7106 + }, + { + "epoch": 1.4902495282029775, + "grad_norm": 6.670129063461565, + "learning_rate": 1.0591660081044522e-05, + "loss": 0.6964, + "step": 7107 + }, + { + "epoch": 1.4904592157685048, + "grad_norm": 9.076212510423709, + "learning_rate": 1.0589400147276003e-05, + "loss": 0.9677, + "step": 7108 + }, + { + "epoch": 1.4906689033340323, + "grad_norm": 7.004491665347137, + "learning_rate": 1.0587140183299707e-05, + "loss": 0.8185, + "step": 7109 + }, + { + "epoch": 1.4908785908995597, + "grad_norm": 5.926748577927218, + "learning_rate": 1.0584880189231456e-05, + "loss": 0.5383, + "step": 7110 + }, + { + "epoch": 1.491088278465087, + "grad_norm": 8.009512627556292, + "learning_rate": 1.0582620165187084e-05, + "loss": 0.7928, + "step": 7111 + }, + { + "epoch": 1.4912979660306145, + "grad_norm": 8.319617757536445, + "learning_rate": 1.0580360111282413e-05, + "loss": 0.817, + "step": 7112 + }, + { + "epoch": 1.4915076535961418, + "grad_norm": 6.521398451776895, + "learning_rate": 1.0578100027633282e-05, + "loss": 0.8551, + "step": 7113 + }, + { + "epoch": 1.491717341161669, + "grad_norm": 6.501170337580323, + "learning_rate": 1.0575839914355524e-05, + "loss": 0.7031, + "step": 7114 + }, + { + "epoch": 1.4919270287271964, + "grad_norm": 7.70242112258516, + "learning_rate": 1.057357977156497e-05, + "loss": 0.849, + "step": 7115 + }, + { + "epoch": 1.4921367162927237, + "grad_norm": 8.273358334747023, + "learning_rate": 1.057131959937746e-05, + "loss": 0.8167, + "step": 7116 + }, + { + "epoch": 1.4923464038582512, + "grad_norm": 6.55221399949956, + "learning_rate": 1.0569059397908828e-05, + "loss": 0.8744, + "step": 7117 + }, + { + "epoch": 1.4925560914237785, + "grad_norm": 6.602488510373889, + "learning_rate": 1.0566799167274913e-05, + "loss": 0.8125, + "step": 7118 + }, + { + "epoch": 1.4927657789893058, + "grad_norm": 7.278827156180899, + "learning_rate": 1.056453890759156e-05, + "loss": 1.0564, + "step": 7119 + }, + { + "epoch": 1.4929754665548334, + "grad_norm": 6.533391902459943, + "learning_rate": 1.0562278618974607e-05, + "loss": 0.7626, + "step": 7120 + }, + { + "epoch": 1.4931851541203607, + "grad_norm": 7.360576633789399, + "learning_rate": 1.0560018301539902e-05, + "loss": 0.8486, + "step": 7121 + }, + { + "epoch": 1.493394841685888, + "grad_norm": 6.063777729639732, + "learning_rate": 1.0557757955403283e-05, + "loss": 0.629, + "step": 7122 + }, + { + "epoch": 1.4936045292514155, + "grad_norm": 7.860633878440131, + "learning_rate": 1.0555497580680607e-05, + "loss": 0.9613, + "step": 7123 + }, + { + "epoch": 1.4938142168169428, + "grad_norm": 6.607117353261462, + "learning_rate": 1.0553237177487714e-05, + "loss": 0.5875, + "step": 7124 + }, + { + "epoch": 1.4940239043824701, + "grad_norm": 6.18348446597019, + "learning_rate": 1.0550976745940455e-05, + "loss": 0.6631, + "step": 7125 + }, + { + "epoch": 1.4942335919479974, + "grad_norm": 6.493629319562998, + "learning_rate": 1.0548716286154685e-05, + "loss": 0.7772, + "step": 7126 + }, + { + "epoch": 1.4944432795135247, + "grad_norm": 7.247846480353934, + "learning_rate": 1.0546455798246253e-05, + "loss": 0.9058, + "step": 7127 + }, + { + "epoch": 1.4946529670790523, + "grad_norm": 6.462098426719643, + "learning_rate": 1.054419528233101e-05, + "loss": 0.757, + "step": 7128 + }, + { + "epoch": 1.4948626546445796, + "grad_norm": 7.536226537452106, + "learning_rate": 1.054193473852482e-05, + "loss": 0.753, + "step": 7129 + }, + { + "epoch": 1.4950723422101069, + "grad_norm": 8.903094909867505, + "learning_rate": 1.0539674166943532e-05, + "loss": 1.1463, + "step": 7130 + }, + { + "epoch": 1.4952820297756344, + "grad_norm": 7.449940098681308, + "learning_rate": 1.0537413567703008e-05, + "loss": 0.8703, + "step": 7131 + }, + { + "epoch": 1.4954917173411617, + "grad_norm": 6.3937215255975355, + "learning_rate": 1.0535152940919105e-05, + "loss": 0.7058, + "step": 7132 + }, + { + "epoch": 1.495701404906689, + "grad_norm": 6.5330801012804045, + "learning_rate": 1.0532892286707688e-05, + "loss": 0.6088, + "step": 7133 + }, + { + "epoch": 1.4959110924722163, + "grad_norm": 6.741138782388606, + "learning_rate": 1.0530631605184614e-05, + "loss": 0.703, + "step": 7134 + }, + { + "epoch": 1.4961207800377436, + "grad_norm": 7.384332731443193, + "learning_rate": 1.0528370896465751e-05, + "loss": 0.6708, + "step": 7135 + }, + { + "epoch": 1.4963304676032712, + "grad_norm": 6.023507675827106, + "learning_rate": 1.0526110160666964e-05, + "loss": 0.6056, + "step": 7136 + }, + { + "epoch": 1.4965401551687985, + "grad_norm": 7.97545497278103, + "learning_rate": 1.0523849397904118e-05, + "loss": 0.7755, + "step": 7137 + }, + { + "epoch": 1.4967498427343258, + "grad_norm": 6.725913918968319, + "learning_rate": 1.0521588608293082e-05, + "loss": 0.6248, + "step": 7138 + }, + { + "epoch": 1.4969595302998533, + "grad_norm": 7.299922568348619, + "learning_rate": 1.0519327791949725e-05, + "loss": 0.9233, + "step": 7139 + }, + { + "epoch": 1.4971692178653806, + "grad_norm": 7.613481060741226, + "learning_rate": 1.0517066948989919e-05, + "loss": 0.7219, + "step": 7140 + }, + { + "epoch": 1.497378905430908, + "grad_norm": 7.706368618701546, + "learning_rate": 1.0514806079529534e-05, + "loss": 0.7726, + "step": 7141 + }, + { + "epoch": 1.4975885929964354, + "grad_norm": 7.132751188557857, + "learning_rate": 1.0512545183684444e-05, + "loss": 0.8031, + "step": 7142 + }, + { + "epoch": 1.4977982805619627, + "grad_norm": 9.094150284104597, + "learning_rate": 1.0510284261570525e-05, + "loss": 1.0268, + "step": 7143 + }, + { + "epoch": 1.49800796812749, + "grad_norm": 5.861597995112914, + "learning_rate": 1.0508023313303655e-05, + "loss": 0.5802, + "step": 7144 + }, + { + "epoch": 1.4982176556930173, + "grad_norm": 4.892347243375909, + "learning_rate": 1.0505762338999704e-05, + "loss": 0.4412, + "step": 7145 + }, + { + "epoch": 1.4984273432585447, + "grad_norm": 6.860161570568778, + "learning_rate": 1.050350133877456e-05, + "loss": 0.8146, + "step": 7146 + }, + { + "epoch": 1.4986370308240722, + "grad_norm": 5.771369608917613, + "learning_rate": 1.0501240312744098e-05, + "loss": 0.6055, + "step": 7147 + }, + { + "epoch": 1.4988467183895995, + "grad_norm": 5.493139122461287, + "learning_rate": 1.04989792610242e-05, + "loss": 0.5286, + "step": 7148 + }, + { + "epoch": 1.4990564059551268, + "grad_norm": 7.373294583637393, + "learning_rate": 1.049671818373075e-05, + "loss": 0.767, + "step": 7149 + }, + { + "epoch": 1.4992660935206543, + "grad_norm": 6.122362195600923, + "learning_rate": 1.0494457080979633e-05, + "loss": 0.6081, + "step": 7150 + }, + { + "epoch": 1.4994757810861816, + "grad_norm": 7.762090281486593, + "learning_rate": 1.0492195952886728e-05, + "loss": 0.762, + "step": 7151 + }, + { + "epoch": 1.499685468651709, + "grad_norm": 5.704882530260195, + "learning_rate": 1.0489934799567934e-05, + "loss": 0.5562, + "step": 7152 + }, + { + "epoch": 1.4998951562172362, + "grad_norm": 7.216254968003204, + "learning_rate": 1.0487673621139127e-05, + "loss": 0.8937, + "step": 7153 + }, + { + "epoch": 1.5001048437827635, + "grad_norm": 8.018995218808998, + "learning_rate": 1.0485412417716201e-05, + "loss": 0.898, + "step": 7154 + }, + { + "epoch": 1.500314531348291, + "grad_norm": 7.535309299083411, + "learning_rate": 1.0483151189415048e-05, + "loss": 0.823, + "step": 7155 + }, + { + "epoch": 1.5005242189138184, + "grad_norm": 6.6389659882095815, + "learning_rate": 1.048088993635156e-05, + "loss": 0.7675, + "step": 7156 + }, + { + "epoch": 1.5007339064793457, + "grad_norm": 6.137192533774715, + "learning_rate": 1.0478628658641625e-05, + "loss": 0.5862, + "step": 7157 + }, + { + "epoch": 1.5009435940448732, + "grad_norm": 9.724634951212895, + "learning_rate": 1.0476367356401145e-05, + "loss": 0.976, + "step": 7158 + }, + { + "epoch": 1.5011532816104005, + "grad_norm": 7.553107397457866, + "learning_rate": 1.047410602974601e-05, + "loss": 0.899, + "step": 7159 + }, + { + "epoch": 1.5013629691759278, + "grad_norm": 6.496685621029923, + "learning_rate": 1.0471844678792118e-05, + "loss": 0.635, + "step": 7160 + }, + { + "epoch": 1.5015726567414553, + "grad_norm": 8.60126898721266, + "learning_rate": 1.0469583303655368e-05, + "loss": 0.9175, + "step": 7161 + }, + { + "epoch": 1.5017823443069824, + "grad_norm": 6.102222299496603, + "learning_rate": 1.0467321904451659e-05, + "loss": 0.6187, + "step": 7162 + }, + { + "epoch": 1.50199203187251, + "grad_norm": 7.262707474690552, + "learning_rate": 1.0465060481296892e-05, + "loss": 0.8693, + "step": 7163 + }, + { + "epoch": 1.5022017194380375, + "grad_norm": 7.534613266914976, + "learning_rate": 1.0462799034306972e-05, + "loss": 0.9057, + "step": 7164 + }, + { + "epoch": 1.5024114070035646, + "grad_norm": 7.388669203299701, + "learning_rate": 1.0460537563597795e-05, + "loss": 0.7615, + "step": 7165 + }, + { + "epoch": 1.502621094569092, + "grad_norm": 6.353890458374306, + "learning_rate": 1.045827606928527e-05, + "loss": 0.7831, + "step": 7166 + }, + { + "epoch": 1.5028307821346194, + "grad_norm": 5.788662861917991, + "learning_rate": 1.0456014551485301e-05, + "loss": 0.5901, + "step": 7167 + }, + { + "epoch": 1.5030404697001467, + "grad_norm": 6.766691258491545, + "learning_rate": 1.0453753010313796e-05, + "loss": 0.7368, + "step": 7168 + }, + { + "epoch": 1.5032501572656742, + "grad_norm": 7.317303649848834, + "learning_rate": 1.0451491445886661e-05, + "loss": 0.6686, + "step": 7169 + }, + { + "epoch": 1.5034598448312015, + "grad_norm": 7.476714891516945, + "learning_rate": 1.0449229858319808e-05, + "loss": 0.6706, + "step": 7170 + }, + { + "epoch": 1.5036695323967288, + "grad_norm": 6.157120425574784, + "learning_rate": 1.0446968247729145e-05, + "loss": 0.7993, + "step": 7171 + }, + { + "epoch": 1.5038792199622564, + "grad_norm": 7.5198079174717085, + "learning_rate": 1.0444706614230585e-05, + "loss": 1.0423, + "step": 7172 + }, + { + "epoch": 1.5040889075277835, + "grad_norm": 8.825016520129456, + "learning_rate": 1.0442444957940042e-05, + "loss": 0.9199, + "step": 7173 + }, + { + "epoch": 1.504298595093311, + "grad_norm": 6.610466576115415, + "learning_rate": 1.0440183278973419e-05, + "loss": 0.6741, + "step": 7174 + }, + { + "epoch": 1.5045082826588383, + "grad_norm": 7.212215351430875, + "learning_rate": 1.0437921577446648e-05, + "loss": 0.8964, + "step": 7175 + }, + { + "epoch": 1.5047179702243656, + "grad_norm": 7.11337712268857, + "learning_rate": 1.0435659853475633e-05, + "loss": 0.7989, + "step": 7176 + }, + { + "epoch": 1.5049276577898931, + "grad_norm": 8.084673998427256, + "learning_rate": 1.0433398107176295e-05, + "loss": 0.8495, + "step": 7177 + }, + { + "epoch": 1.5051373453554204, + "grad_norm": 6.262565861287114, + "learning_rate": 1.0431136338664554e-05, + "loss": 0.7085, + "step": 7178 + }, + { + "epoch": 1.5053470329209477, + "grad_norm": 7.353710557246072, + "learning_rate": 1.0428874548056327e-05, + "loss": 0.7513, + "step": 7179 + }, + { + "epoch": 1.5055567204864753, + "grad_norm": 7.158545080632724, + "learning_rate": 1.0426612735467535e-05, + "loss": 0.7538, + "step": 7180 + }, + { + "epoch": 1.5057664080520026, + "grad_norm": 6.908563202222414, + "learning_rate": 1.0424350901014102e-05, + "loss": 0.6736, + "step": 7181 + }, + { + "epoch": 1.5059760956175299, + "grad_norm": 7.466766633010677, + "learning_rate": 1.0422089044811947e-05, + "loss": 0.7598, + "step": 7182 + }, + { + "epoch": 1.5061857831830574, + "grad_norm": 7.558929816372385, + "learning_rate": 1.0419827166976996e-05, + "loss": 0.8435, + "step": 7183 + }, + { + "epoch": 1.5063954707485845, + "grad_norm": 10.863052746407542, + "learning_rate": 1.0417565267625177e-05, + "loss": 1.2898, + "step": 7184 + }, + { + "epoch": 1.506605158314112, + "grad_norm": 6.529234545538859, + "learning_rate": 1.0415303346872414e-05, + "loss": 0.7826, + "step": 7185 + }, + { + "epoch": 1.5068148458796393, + "grad_norm": 6.765601062704781, + "learning_rate": 1.0413041404834632e-05, + "loss": 0.6716, + "step": 7186 + }, + { + "epoch": 1.5070245334451666, + "grad_norm": 7.202051853890094, + "learning_rate": 1.0410779441627762e-05, + "loss": 0.7864, + "step": 7187 + }, + { + "epoch": 1.5072342210106942, + "grad_norm": 6.3788151048314505, + "learning_rate": 1.0408517457367736e-05, + "loss": 0.7534, + "step": 7188 + }, + { + "epoch": 1.5074439085762215, + "grad_norm": 8.477930517598896, + "learning_rate": 1.0406255452170478e-05, + "loss": 0.9195, + "step": 7189 + }, + { + "epoch": 1.5076535961417488, + "grad_norm": 6.317121233635021, + "learning_rate": 1.0403993426151923e-05, + "loss": 0.579, + "step": 7190 + }, + { + "epoch": 1.5078632837072763, + "grad_norm": 7.333313896323223, + "learning_rate": 1.0401731379428006e-05, + "loss": 0.7964, + "step": 7191 + }, + { + "epoch": 1.5080729712728034, + "grad_norm": 7.065798215079973, + "learning_rate": 1.0399469312114661e-05, + "loss": 0.792, + "step": 7192 + }, + { + "epoch": 1.508282658838331, + "grad_norm": 7.497041454262129, + "learning_rate": 1.039720722432782e-05, + "loss": 0.9291, + "step": 7193 + }, + { + "epoch": 1.5084923464038582, + "grad_norm": 5.177489335698658, + "learning_rate": 1.0394945116183419e-05, + "loss": 0.5616, + "step": 7194 + }, + { + "epoch": 1.5087020339693855, + "grad_norm": 6.919790561306532, + "learning_rate": 1.0392682987797396e-05, + "loss": 0.8215, + "step": 7195 + }, + { + "epoch": 1.508911721534913, + "grad_norm": 7.1870891246285415, + "learning_rate": 1.0390420839285687e-05, + "loss": 0.944, + "step": 7196 + }, + { + "epoch": 1.5091214091004403, + "grad_norm": 7.943277737057516, + "learning_rate": 1.0388158670764234e-05, + "loss": 0.7801, + "step": 7197 + }, + { + "epoch": 1.5093310966659677, + "grad_norm": 6.202826949647392, + "learning_rate": 1.0385896482348976e-05, + "loss": 0.6971, + "step": 7198 + }, + { + "epoch": 1.5095407842314952, + "grad_norm": 8.699193719331987, + "learning_rate": 1.0383634274155854e-05, + "loss": 1.0111, + "step": 7199 + }, + { + "epoch": 1.5097504717970225, + "grad_norm": 7.73420748794142, + "learning_rate": 1.0381372046300813e-05, + "loss": 0.8511, + "step": 7200 + }, + { + "epoch": 1.5099601593625498, + "grad_norm": 6.6104197555678645, + "learning_rate": 1.0379109798899787e-05, + "loss": 0.8213, + "step": 7201 + }, + { + "epoch": 1.5101698469280773, + "grad_norm": 5.810383249912145, + "learning_rate": 1.0376847532068732e-05, + "loss": 0.8499, + "step": 7202 + }, + { + "epoch": 1.5103795344936044, + "grad_norm": 7.077879558422452, + "learning_rate": 1.0374585245923583e-05, + "loss": 0.6363, + "step": 7203 + }, + { + "epoch": 1.510589222059132, + "grad_norm": 6.0736232434991155, + "learning_rate": 1.0372322940580295e-05, + "loss": 0.6029, + "step": 7204 + }, + { + "epoch": 1.5107989096246592, + "grad_norm": 7.686288769454458, + "learning_rate": 1.0370060616154807e-05, + "loss": 0.6943, + "step": 7205 + }, + { + "epoch": 1.5110085971901865, + "grad_norm": 6.058421053168277, + "learning_rate": 1.0367798272763071e-05, + "loss": 0.693, + "step": 7206 + }, + { + "epoch": 1.511218284755714, + "grad_norm": 6.41309111946237, + "learning_rate": 1.0365535910521036e-05, + "loss": 0.6205, + "step": 7207 + }, + { + "epoch": 1.5114279723212414, + "grad_norm": 6.758311557258199, + "learning_rate": 1.0363273529544655e-05, + "loss": 0.7815, + "step": 7208 + }, + { + "epoch": 1.5116376598867687, + "grad_norm": 4.823608543630581, + "learning_rate": 1.036101112994987e-05, + "loss": 0.4516, + "step": 7209 + }, + { + "epoch": 1.5118473474522962, + "grad_norm": 6.382616307968191, + "learning_rate": 1.0358748711852643e-05, + "loss": 0.4889, + "step": 7210 + }, + { + "epoch": 1.5120570350178233, + "grad_norm": 8.05645790725201, + "learning_rate": 1.035648627536892e-05, + "loss": 0.8713, + "step": 7211 + }, + { + "epoch": 1.5122667225833508, + "grad_norm": 7.039288261309405, + "learning_rate": 1.0354223820614655e-05, + "loss": 0.6426, + "step": 7212 + }, + { + "epoch": 1.5124764101488781, + "grad_norm": 6.969950464849892, + "learning_rate": 1.0351961347705809e-05, + "loss": 0.6918, + "step": 7213 + }, + { + "epoch": 1.5126860977144054, + "grad_norm": 6.863030918953516, + "learning_rate": 1.0349698856758332e-05, + "loss": 0.9141, + "step": 7214 + }, + { + "epoch": 1.512895785279933, + "grad_norm": 7.293898322341821, + "learning_rate": 1.034743634788818e-05, + "loss": 0.9232, + "step": 7215 + }, + { + "epoch": 1.5131054728454603, + "grad_norm": 7.76004929003657, + "learning_rate": 1.0345173821211316e-05, + "loss": 1.0377, + "step": 7216 + }, + { + "epoch": 1.5133151604109876, + "grad_norm": 7.919419896437171, + "learning_rate": 1.0342911276843692e-05, + "loss": 0.9963, + "step": 7217 + }, + { + "epoch": 1.513524847976515, + "grad_norm": 7.379983765498563, + "learning_rate": 1.0340648714901271e-05, + "loss": 0.8489, + "step": 7218 + }, + { + "epoch": 1.5137345355420424, + "grad_norm": 5.337259017427037, + "learning_rate": 1.0338386135500014e-05, + "loss": 0.5561, + "step": 7219 + }, + { + "epoch": 1.5139442231075697, + "grad_norm": 7.34990730614422, + "learning_rate": 1.0336123538755877e-05, + "loss": 0.8141, + "step": 7220 + }, + { + "epoch": 1.5141539106730972, + "grad_norm": 6.871520929411175, + "learning_rate": 1.0333860924784829e-05, + "loss": 0.6368, + "step": 7221 + }, + { + "epoch": 1.5143635982386243, + "grad_norm": 5.657223691434179, + "learning_rate": 1.0331598293702829e-05, + "loss": 0.6661, + "step": 7222 + }, + { + "epoch": 1.5145732858041518, + "grad_norm": 7.410537803410694, + "learning_rate": 1.0329335645625845e-05, + "loss": 0.8287, + "step": 7223 + }, + { + "epoch": 1.5147829733696792, + "grad_norm": 8.441122876873957, + "learning_rate": 1.032707298066983e-05, + "loss": 0.9426, + "step": 7224 + }, + { + "epoch": 1.5149926609352065, + "grad_norm": 9.247675299506014, + "learning_rate": 1.0324810298950765e-05, + "loss": 1.0272, + "step": 7225 + }, + { + "epoch": 1.515202348500734, + "grad_norm": 7.794645679477385, + "learning_rate": 1.0322547600584605e-05, + "loss": 0.9103, + "step": 7226 + }, + { + "epoch": 1.5154120360662613, + "grad_norm": 7.786649540905256, + "learning_rate": 1.032028488568732e-05, + "loss": 0.8813, + "step": 7227 + }, + { + "epoch": 1.5156217236317886, + "grad_norm": 6.655706695481443, + "learning_rate": 1.031802215437488e-05, + "loss": 0.7194, + "step": 7228 + }, + { + "epoch": 1.5158314111973161, + "grad_norm": 6.954621659693343, + "learning_rate": 1.0315759406763258e-05, + "loss": 0.7674, + "step": 7229 + }, + { + "epoch": 1.5160410987628432, + "grad_norm": 6.984153889403306, + "learning_rate": 1.031349664296841e-05, + "loss": 0.8589, + "step": 7230 + }, + { + "epoch": 1.5162507863283707, + "grad_norm": 7.249698906023001, + "learning_rate": 1.0311233863106323e-05, + "loss": 0.6932, + "step": 7231 + }, + { + "epoch": 1.516460473893898, + "grad_norm": 9.692383956716936, + "learning_rate": 1.0308971067292954e-05, + "loss": 0.899, + "step": 7232 + }, + { + "epoch": 1.5166701614594253, + "grad_norm": 9.53633659426326, + "learning_rate": 1.0306708255644287e-05, + "loss": 0.9927, + "step": 7233 + }, + { + "epoch": 1.5168798490249529, + "grad_norm": 6.123473730677284, + "learning_rate": 1.0304445428276288e-05, + "loss": 0.7277, + "step": 7234 + }, + { + "epoch": 1.5170895365904802, + "grad_norm": 8.102528268053565, + "learning_rate": 1.0302182585304932e-05, + "loss": 0.9082, + "step": 7235 + }, + { + "epoch": 1.5172992241560075, + "grad_norm": 7.334231919023435, + "learning_rate": 1.0299919726846195e-05, + "loss": 0.6974, + "step": 7236 + }, + { + "epoch": 1.517508911721535, + "grad_norm": 7.123208459904657, + "learning_rate": 1.0297656853016055e-05, + "loss": 0.5872, + "step": 7237 + }, + { + "epoch": 1.5177185992870623, + "grad_norm": 8.009548849173681, + "learning_rate": 1.0295393963930481e-05, + "loss": 0.9495, + "step": 7238 + }, + { + "epoch": 1.5179282868525896, + "grad_norm": 7.284894923337498, + "learning_rate": 1.0293131059705456e-05, + "loss": 0.844, + "step": 7239 + }, + { + "epoch": 1.5181379744181172, + "grad_norm": 7.016237756778589, + "learning_rate": 1.0290868140456956e-05, + "loss": 0.9339, + "step": 7240 + }, + { + "epoch": 1.5183476619836442, + "grad_norm": 6.28625131915815, + "learning_rate": 1.0288605206300956e-05, + "loss": 0.7448, + "step": 7241 + }, + { + "epoch": 1.5185573495491718, + "grad_norm": 6.534574981172653, + "learning_rate": 1.0286342257353441e-05, + "loss": 0.7578, + "step": 7242 + }, + { + "epoch": 1.518767037114699, + "grad_norm": 5.611130303800543, + "learning_rate": 1.0284079293730392e-05, + "loss": 0.5366, + "step": 7243 + }, + { + "epoch": 1.5189767246802264, + "grad_norm": 4.5642509274877865, + "learning_rate": 1.0281816315547778e-05, + "loss": 0.4823, + "step": 7244 + }, + { + "epoch": 1.519186412245754, + "grad_norm": 6.401566719099642, + "learning_rate": 1.0279553322921598e-05, + "loss": 0.6233, + "step": 7245 + }, + { + "epoch": 1.5193960998112812, + "grad_norm": 7.143747667768428, + "learning_rate": 1.0277290315967821e-05, + "loss": 0.8247, + "step": 7246 + }, + { + "epoch": 1.5196057873768085, + "grad_norm": 8.408142469765536, + "learning_rate": 1.0275027294802436e-05, + "loss": 1.02, + "step": 7247 + }, + { + "epoch": 1.519815474942336, + "grad_norm": 6.059746834660585, + "learning_rate": 1.0272764259541423e-05, + "loss": 0.73, + "step": 7248 + }, + { + "epoch": 1.5200251625078631, + "grad_norm": 7.778963544199173, + "learning_rate": 1.027050121030077e-05, + "loss": 0.6743, + "step": 7249 + }, + { + "epoch": 1.5202348500733907, + "grad_norm": 6.672043312599691, + "learning_rate": 1.0268238147196459e-05, + "loss": 0.7725, + "step": 7250 + }, + { + "epoch": 1.520444537638918, + "grad_norm": 7.11011299767315, + "learning_rate": 1.0265975070344478e-05, + "loss": 0.8617, + "step": 7251 + }, + { + "epoch": 1.5206542252044453, + "grad_norm": 6.894065383808769, + "learning_rate": 1.0263711979860817e-05, + "loss": 0.7816, + "step": 7252 + }, + { + "epoch": 1.5208639127699728, + "grad_norm": 7.042356650676313, + "learning_rate": 1.0261448875861455e-05, + "loss": 0.8459, + "step": 7253 + }, + { + "epoch": 1.5210736003355, + "grad_norm": 7.990217081096455, + "learning_rate": 1.0259185758462389e-05, + "loss": 0.8728, + "step": 7254 + }, + { + "epoch": 1.5212832879010274, + "grad_norm": 7.909815437368781, + "learning_rate": 1.0256922627779599e-05, + "loss": 0.875, + "step": 7255 + }, + { + "epoch": 1.521492975466555, + "grad_norm": 7.313489515598973, + "learning_rate": 1.025465948392908e-05, + "loss": 0.8089, + "step": 7256 + }, + { + "epoch": 1.5217026630320822, + "grad_norm": 6.47980345112776, + "learning_rate": 1.0252396327026823e-05, + "loss": 0.6915, + "step": 7257 + }, + { + "epoch": 1.5219123505976095, + "grad_norm": 6.86446174612249, + "learning_rate": 1.0250133157188817e-05, + "loss": 0.7712, + "step": 7258 + }, + { + "epoch": 1.522122038163137, + "grad_norm": 7.41267377064116, + "learning_rate": 1.0247869974531048e-05, + "loss": 0.6989, + "step": 7259 + }, + { + "epoch": 1.5223317257286642, + "grad_norm": 6.24550741734712, + "learning_rate": 1.0245606779169518e-05, + "loss": 0.6255, + "step": 7260 + }, + { + "epoch": 1.5225414132941917, + "grad_norm": 6.0988744283174405, + "learning_rate": 1.0243343571220212e-05, + "loss": 0.6784, + "step": 7261 + }, + { + "epoch": 1.522751100859719, + "grad_norm": 7.322015036871099, + "learning_rate": 1.0241080350799127e-05, + "loss": 0.7716, + "step": 7262 + }, + { + "epoch": 1.5229607884252463, + "grad_norm": 5.308205613697077, + "learning_rate": 1.0238817118022255e-05, + "loss": 0.6083, + "step": 7263 + }, + { + "epoch": 1.5231704759907738, + "grad_norm": 6.750314424286773, + "learning_rate": 1.0236553873005592e-05, + "loss": 0.8001, + "step": 7264 + }, + { + "epoch": 1.5233801635563011, + "grad_norm": 6.9471760582639845, + "learning_rate": 1.0234290615865132e-05, + "loss": 0.8286, + "step": 7265 + }, + { + "epoch": 1.5235898511218284, + "grad_norm": 8.595175343566673, + "learning_rate": 1.0232027346716872e-05, + "loss": 0.7378, + "step": 7266 + }, + { + "epoch": 1.523799538687356, + "grad_norm": 6.978178774724283, + "learning_rate": 1.0229764065676808e-05, + "loss": 0.767, + "step": 7267 + }, + { + "epoch": 1.524009226252883, + "grad_norm": 7.909728023217, + "learning_rate": 1.0227500772860938e-05, + "loss": 0.7439, + "step": 7268 + }, + { + "epoch": 1.5242189138184106, + "grad_norm": 6.7516061056682615, + "learning_rate": 1.0225237468385258e-05, + "loss": 0.8037, + "step": 7269 + }, + { + "epoch": 1.524428601383938, + "grad_norm": 6.825298250966481, + "learning_rate": 1.0222974152365768e-05, + "loss": 0.7308, + "step": 7270 + }, + { + "epoch": 1.5246382889494652, + "grad_norm": 5.93066276224902, + "learning_rate": 1.0220710824918465e-05, + "loss": 0.6715, + "step": 7271 + }, + { + "epoch": 1.5248479765149927, + "grad_norm": 8.634307663669126, + "learning_rate": 1.0218447486159348e-05, + "loss": 0.8895, + "step": 7272 + }, + { + "epoch": 1.52505766408052, + "grad_norm": 7.960698578537101, + "learning_rate": 1.021618413620442e-05, + "loss": 0.8666, + "step": 7273 + }, + { + "epoch": 1.5252673516460473, + "grad_norm": 6.555442657152822, + "learning_rate": 1.0213920775169681e-05, + "loss": 0.7264, + "step": 7274 + }, + { + "epoch": 1.5254770392115748, + "grad_norm": 7.076378488169698, + "learning_rate": 1.0211657403171132e-05, + "loss": 0.722, + "step": 7275 + }, + { + "epoch": 1.5256867267771022, + "grad_norm": 7.995447325329171, + "learning_rate": 1.020939402032477e-05, + "loss": 0.9173, + "step": 7276 + }, + { + "epoch": 1.5258964143426295, + "grad_norm": 7.672499086111196, + "learning_rate": 1.0207130626746604e-05, + "loss": 0.7231, + "step": 7277 + }, + { + "epoch": 1.526106101908157, + "grad_norm": 6.595953215911844, + "learning_rate": 1.0204867222552632e-05, + "loss": 0.6305, + "step": 7278 + }, + { + "epoch": 1.526315789473684, + "grad_norm": 8.422016427253935, + "learning_rate": 1.0202603807858862e-05, + "loss": 0.8106, + "step": 7279 + }, + { + "epoch": 1.5265254770392116, + "grad_norm": 7.201992731771172, + "learning_rate": 1.0200340382781294e-05, + "loss": 0.7639, + "step": 7280 + }, + { + "epoch": 1.526735164604739, + "grad_norm": 7.14176316732443, + "learning_rate": 1.0198076947435938e-05, + "loss": 0.7653, + "step": 7281 + }, + { + "epoch": 1.5269448521702662, + "grad_norm": 7.898189974244631, + "learning_rate": 1.0195813501938789e-05, + "loss": 0.9381, + "step": 7282 + }, + { + "epoch": 1.5271545397357937, + "grad_norm": 6.279219730579641, + "learning_rate": 1.0193550046405862e-05, + "loss": 0.6652, + "step": 7283 + }, + { + "epoch": 1.527364227301321, + "grad_norm": 6.380413107099824, + "learning_rate": 1.019128658095316e-05, + "loss": 0.6105, + "step": 7284 + }, + { + "epoch": 1.5275739148668483, + "grad_norm": 5.750222569963986, + "learning_rate": 1.0189023105696684e-05, + "loss": 0.5727, + "step": 7285 + }, + { + "epoch": 1.5277836024323759, + "grad_norm": 6.753279777162709, + "learning_rate": 1.018675962075245e-05, + "loss": 0.7523, + "step": 7286 + }, + { + "epoch": 1.527993289997903, + "grad_norm": 7.759657203738766, + "learning_rate": 1.0184496126236462e-05, + "loss": 0.941, + "step": 7287 + }, + { + "epoch": 1.5282029775634305, + "grad_norm": 4.688936286266487, + "learning_rate": 1.0182232622264725e-05, + "loss": 0.471, + "step": 7288 + }, + { + "epoch": 1.528412665128958, + "grad_norm": 7.366677720685867, + "learning_rate": 1.0179969108953254e-05, + "loss": 0.7325, + "step": 7289 + }, + { + "epoch": 1.528622352694485, + "grad_norm": 6.918557412332455, + "learning_rate": 1.017770558641805e-05, + "loss": 0.8737, + "step": 7290 + }, + { + "epoch": 1.5288320402600126, + "grad_norm": 6.703537761683248, + "learning_rate": 1.0175442054775126e-05, + "loss": 0.7144, + "step": 7291 + }, + { + "epoch": 1.52904172782554, + "grad_norm": 7.629493612637875, + "learning_rate": 1.0173178514140496e-05, + "loss": 0.7616, + "step": 7292 + }, + { + "epoch": 1.5292514153910672, + "grad_norm": 6.519719882508098, + "learning_rate": 1.0170914964630165e-05, + "loss": 0.7256, + "step": 7293 + }, + { + "epoch": 1.5294611029565948, + "grad_norm": 6.141777620174371, + "learning_rate": 1.0168651406360144e-05, + "loss": 0.6655, + "step": 7294 + }, + { + "epoch": 1.529670790522122, + "grad_norm": 7.033435258898474, + "learning_rate": 1.016638783944645e-05, + "loss": 0.7145, + "step": 7295 + }, + { + "epoch": 1.5298804780876494, + "grad_norm": 7.151885961000227, + "learning_rate": 1.0164124264005086e-05, + "loss": 0.807, + "step": 7296 + }, + { + "epoch": 1.530090165653177, + "grad_norm": 7.885985467659819, + "learning_rate": 1.0161860680152073e-05, + "loss": 0.8227, + "step": 7297 + }, + { + "epoch": 1.530299853218704, + "grad_norm": 7.425592980463889, + "learning_rate": 1.0159597088003417e-05, + "loss": 0.7464, + "step": 7298 + }, + { + "epoch": 1.5305095407842315, + "grad_norm": 8.65554673087192, + "learning_rate": 1.0157333487675134e-05, + "loss": 0.8471, + "step": 7299 + }, + { + "epoch": 1.5307192283497588, + "grad_norm": 7.028338666433179, + "learning_rate": 1.0155069879283238e-05, + "loss": 0.8137, + "step": 7300 + }, + { + "epoch": 1.5309289159152861, + "grad_norm": 6.794506875293018, + "learning_rate": 1.0152806262943737e-05, + "loss": 0.7932, + "step": 7301 + }, + { + "epoch": 1.5311386034808137, + "grad_norm": 7.977245355990587, + "learning_rate": 1.0150542638772654e-05, + "loss": 0.8832, + "step": 7302 + }, + { + "epoch": 1.531348291046341, + "grad_norm": 7.558654451742566, + "learning_rate": 1.0148279006886e-05, + "loss": 0.763, + "step": 7303 + }, + { + "epoch": 1.5315579786118683, + "grad_norm": 7.816927046299833, + "learning_rate": 1.014601536739979e-05, + "loss": 0.9116, + "step": 7304 + }, + { + "epoch": 1.5317676661773958, + "grad_norm": 6.539313831222138, + "learning_rate": 1.0143751720430036e-05, + "loss": 0.6733, + "step": 7305 + }, + { + "epoch": 1.531977353742923, + "grad_norm": 8.173099263599498, + "learning_rate": 1.014148806609276e-05, + "loss": 0.8424, + "step": 7306 + }, + { + "epoch": 1.5321870413084504, + "grad_norm": 8.270424255388624, + "learning_rate": 1.0139224404503974e-05, + "loss": 0.8122, + "step": 7307 + }, + { + "epoch": 1.532396728873978, + "grad_norm": 7.466165008970468, + "learning_rate": 1.0136960735779695e-05, + "loss": 0.9181, + "step": 7308 + }, + { + "epoch": 1.532606416439505, + "grad_norm": 7.204561997311021, + "learning_rate": 1.0134697060035943e-05, + "loss": 0.7996, + "step": 7309 + }, + { + "epoch": 1.5328161040050325, + "grad_norm": 6.789451362186487, + "learning_rate": 1.0132433377388733e-05, + "loss": 0.8782, + "step": 7310 + }, + { + "epoch": 1.5330257915705598, + "grad_norm": 7.758820342234318, + "learning_rate": 1.0130169687954078e-05, + "loss": 0.8946, + "step": 7311 + }, + { + "epoch": 1.5332354791360872, + "grad_norm": 8.588861096994258, + "learning_rate": 1.0127905991848009e-05, + "loss": 0.8825, + "step": 7312 + }, + { + "epoch": 1.5334451667016147, + "grad_norm": 5.799675147245732, + "learning_rate": 1.012564228918653e-05, + "loss": 0.6987, + "step": 7313 + }, + { + "epoch": 1.533654854267142, + "grad_norm": 6.232088840101841, + "learning_rate": 1.0123378580085667e-05, + "loss": 0.4624, + "step": 7314 + }, + { + "epoch": 1.5338645418326693, + "grad_norm": 8.451537630388748, + "learning_rate": 1.0121114864661436e-05, + "loss": 0.7845, + "step": 7315 + }, + { + "epoch": 1.5340742293981968, + "grad_norm": 8.013331779572823, + "learning_rate": 1.0118851143029865e-05, + "loss": 0.9236, + "step": 7316 + }, + { + "epoch": 1.534283916963724, + "grad_norm": 8.287058797066555, + "learning_rate": 1.0116587415306958e-05, + "loss": 0.9076, + "step": 7317 + }, + { + "epoch": 1.5344936045292514, + "grad_norm": 7.091103764952184, + "learning_rate": 1.0114323681608748e-05, + "loss": 0.7458, + "step": 7318 + }, + { + "epoch": 1.5347032920947787, + "grad_norm": 8.257460961578008, + "learning_rate": 1.011205994205125e-05, + "loss": 0.7686, + "step": 7319 + }, + { + "epoch": 1.534912979660306, + "grad_norm": 7.6694782967617705, + "learning_rate": 1.0109796196750487e-05, + "loss": 0.6978, + "step": 7320 + }, + { + "epoch": 1.5351226672258336, + "grad_norm": 8.286857548109896, + "learning_rate": 1.0107532445822477e-05, + "loss": 0.7187, + "step": 7321 + }, + { + "epoch": 1.5353323547913609, + "grad_norm": 7.054008020796386, + "learning_rate": 1.0105268689383245e-05, + "loss": 0.7754, + "step": 7322 + }, + { + "epoch": 1.5355420423568882, + "grad_norm": 8.845225316840892, + "learning_rate": 1.0103004927548808e-05, + "loss": 0.9422, + "step": 7323 + }, + { + "epoch": 1.5357517299224157, + "grad_norm": 7.347927872566806, + "learning_rate": 1.010074116043519e-05, + "loss": 0.7987, + "step": 7324 + }, + { + "epoch": 1.535961417487943, + "grad_norm": 6.345647155606473, + "learning_rate": 1.0098477388158412e-05, + "loss": 0.6966, + "step": 7325 + }, + { + "epoch": 1.5361711050534703, + "grad_norm": 5.77880545480286, + "learning_rate": 1.00962136108345e-05, + "loss": 0.6293, + "step": 7326 + }, + { + "epoch": 1.5363807926189978, + "grad_norm": 8.560415297830199, + "learning_rate": 1.009394982857947e-05, + "loss": 0.9623, + "step": 7327 + }, + { + "epoch": 1.536590480184525, + "grad_norm": 7.26537345789201, + "learning_rate": 1.009168604150935e-05, + "loss": 0.904, + "step": 7328 + }, + { + "epoch": 1.5368001677500525, + "grad_norm": 11.573828364253833, + "learning_rate": 1.0089422249740162e-05, + "loss": 1.1564, + "step": 7329 + }, + { + "epoch": 1.5370098553155798, + "grad_norm": 5.907906731351163, + "learning_rate": 1.0087158453387926e-05, + "loss": 0.7361, + "step": 7330 + }, + { + "epoch": 1.537219542881107, + "grad_norm": 7.54048879527474, + "learning_rate": 1.0084894652568671e-05, + "loss": 0.8141, + "step": 7331 + }, + { + "epoch": 1.5374292304466346, + "grad_norm": 6.969313078367815, + "learning_rate": 1.0082630847398416e-05, + "loss": 0.6609, + "step": 7332 + }, + { + "epoch": 1.537638918012162, + "grad_norm": 6.105329665419921, + "learning_rate": 1.0080367037993191e-05, + "loss": 0.7284, + "step": 7333 + }, + { + "epoch": 1.5378486055776892, + "grad_norm": 6.6100200195733425, + "learning_rate": 1.0078103224469009e-05, + "loss": 0.8241, + "step": 7334 + }, + { + "epoch": 1.5380582931432167, + "grad_norm": 8.642240882371405, + "learning_rate": 1.0075839406941907e-05, + "loss": 0.8906, + "step": 7335 + }, + { + "epoch": 1.5382679807087438, + "grad_norm": 9.21746784159912, + "learning_rate": 1.00735755855279e-05, + "loss": 0.8199, + "step": 7336 + }, + { + "epoch": 1.5384776682742713, + "grad_norm": 6.234086328470299, + "learning_rate": 1.007131176034302e-05, + "loss": 0.7233, + "step": 7337 + }, + { + "epoch": 1.5386873558397987, + "grad_norm": 6.571711999693177, + "learning_rate": 1.0069047931503286e-05, + "loss": 0.7688, + "step": 7338 + }, + { + "epoch": 1.538897043405326, + "grad_norm": 6.003520274056556, + "learning_rate": 1.006678409912473e-05, + "loss": 0.6299, + "step": 7339 + }, + { + "epoch": 1.5391067309708535, + "grad_norm": 5.83944702746518, + "learning_rate": 1.0064520263323364e-05, + "loss": 0.6076, + "step": 7340 + }, + { + "epoch": 1.5393164185363808, + "grad_norm": 5.649722688795755, + "learning_rate": 1.0062256424215233e-05, + "loss": 0.6365, + "step": 7341 + }, + { + "epoch": 1.539526106101908, + "grad_norm": 6.439930669469153, + "learning_rate": 1.0059992581916348e-05, + "loss": 0.96, + "step": 7342 + }, + { + "epoch": 1.5397357936674356, + "grad_norm": 9.902383883054467, + "learning_rate": 1.0057728736542738e-05, + "loss": 1.0087, + "step": 7343 + }, + { + "epoch": 1.539945481232963, + "grad_norm": 6.391528948494521, + "learning_rate": 1.005546488821043e-05, + "loss": 0.6244, + "step": 7344 + }, + { + "epoch": 1.5401551687984902, + "grad_norm": 6.383548870939659, + "learning_rate": 1.0053201037035453e-05, + "loss": 0.6891, + "step": 7345 + }, + { + "epoch": 1.5403648563640178, + "grad_norm": 7.598552313921071, + "learning_rate": 1.0050937183133826e-05, + "loss": 0.9693, + "step": 7346 + }, + { + "epoch": 1.5405745439295448, + "grad_norm": 6.547856236853699, + "learning_rate": 1.0048673326621585e-05, + "loss": 0.6411, + "step": 7347 + }, + { + "epoch": 1.5407842314950724, + "grad_norm": 5.427315215263923, + "learning_rate": 1.0046409467614748e-05, + "loss": 0.5787, + "step": 7348 + }, + { + "epoch": 1.5409939190605997, + "grad_norm": 6.551414962943264, + "learning_rate": 1.0044145606229347e-05, + "loss": 0.6684, + "step": 7349 + }, + { + "epoch": 1.541203606626127, + "grad_norm": 7.182514978300086, + "learning_rate": 1.0041881742581407e-05, + "loss": 0.8352, + "step": 7350 + }, + { + "epoch": 1.5414132941916545, + "grad_norm": 6.466233656104143, + "learning_rate": 1.0039617876786954e-05, + "loss": 0.8065, + "step": 7351 + }, + { + "epoch": 1.5416229817571818, + "grad_norm": 6.088522211625142, + "learning_rate": 1.0037354008962018e-05, + "loss": 0.5606, + "step": 7352 + }, + { + "epoch": 1.5418326693227091, + "grad_norm": 7.773234586128021, + "learning_rate": 1.0035090139222624e-05, + "loss": 1.0263, + "step": 7353 + }, + { + "epoch": 1.5420423568882367, + "grad_norm": 6.845897787158895, + "learning_rate": 1.0032826267684796e-05, + "loss": 0.7058, + "step": 7354 + }, + { + "epoch": 1.5422520444537637, + "grad_norm": 5.3232289991003725, + "learning_rate": 1.003056239446457e-05, + "loss": 0.5401, + "step": 7355 + }, + { + "epoch": 1.5424617320192913, + "grad_norm": 7.171126056712241, + "learning_rate": 1.0028298519677966e-05, + "loss": 0.8042, + "step": 7356 + }, + { + "epoch": 1.5426714195848186, + "grad_norm": 8.433038781632222, + "learning_rate": 1.0026034643441011e-05, + "loss": 0.9363, + "step": 7357 + }, + { + "epoch": 1.5428811071503459, + "grad_norm": 8.134643153062074, + "learning_rate": 1.0023770765869739e-05, + "loss": 0.9954, + "step": 7358 + }, + { + "epoch": 1.5430907947158734, + "grad_norm": 7.124723165032045, + "learning_rate": 1.0021506887080172e-05, + "loss": 0.7058, + "step": 7359 + }, + { + "epoch": 1.5433004822814007, + "grad_norm": 8.50121184521389, + "learning_rate": 1.0019243007188342e-05, + "loss": 0.8233, + "step": 7360 + }, + { + "epoch": 1.543510169846928, + "grad_norm": 7.651948067525507, + "learning_rate": 1.0016979126310271e-05, + "loss": 0.7866, + "step": 7361 + }, + { + "epoch": 1.5437198574124555, + "grad_norm": 8.082797857629302, + "learning_rate": 1.0014715244561995e-05, + "loss": 0.792, + "step": 7362 + }, + { + "epoch": 1.5439295449779828, + "grad_norm": 6.998884789419031, + "learning_rate": 1.0012451362059533e-05, + "loss": 0.8672, + "step": 7363 + }, + { + "epoch": 1.5441392325435102, + "grad_norm": 9.8663186863454, + "learning_rate": 1.001018747891892e-05, + "loss": 1.1342, + "step": 7364 + }, + { + "epoch": 1.5443489201090377, + "grad_norm": 7.299144624447819, + "learning_rate": 1.0007923595256179e-05, + "loss": 0.7506, + "step": 7365 + }, + { + "epoch": 1.5445586076745648, + "grad_norm": 7.520050021767935, + "learning_rate": 1.000565971118734e-05, + "loss": 0.8727, + "step": 7366 + }, + { + "epoch": 1.5447682952400923, + "grad_norm": 5.602260144967309, + "learning_rate": 1.0003395826828433e-05, + "loss": 0.5291, + "step": 7367 + }, + { + "epoch": 1.5449779828056196, + "grad_norm": 7.031603714470065, + "learning_rate": 1.0001131942295485e-05, + "loss": 0.7103, + "step": 7368 + }, + { + "epoch": 1.545187670371147, + "grad_norm": 6.76785456779502, + "learning_rate": 9.998868057704519e-06, + "loss": 0.8379, + "step": 7369 + }, + { + "epoch": 1.5453973579366744, + "grad_norm": 6.514405194543272, + "learning_rate": 9.996604173171567e-06, + "loss": 0.6836, + "step": 7370 + }, + { + "epoch": 1.5456070455022017, + "grad_norm": 6.390203505480608, + "learning_rate": 9.994340288812661e-06, + "loss": 0.5797, + "step": 7371 + }, + { + "epoch": 1.545816733067729, + "grad_norm": 6.114566608284368, + "learning_rate": 9.992076404743823e-06, + "loss": 0.7431, + "step": 7372 + }, + { + "epoch": 1.5460264206332566, + "grad_norm": 9.2910983968259, + "learning_rate": 9.989812521081081e-06, + "loss": 1.0739, + "step": 7373 + }, + { + "epoch": 1.5462361081987837, + "grad_norm": 7.383611383307065, + "learning_rate": 9.98754863794047e-06, + "loss": 0.9107, + "step": 7374 + }, + { + "epoch": 1.5464457957643112, + "grad_norm": 6.911824932750947, + "learning_rate": 9.985284755438006e-06, + "loss": 0.8516, + "step": 7375 + }, + { + "epoch": 1.5466554833298385, + "grad_norm": 6.499021184220128, + "learning_rate": 9.983020873689729e-06, + "loss": 0.6742, + "step": 7376 + }, + { + "epoch": 1.5468651708953658, + "grad_norm": 7.037560189645708, + "learning_rate": 9.980756992811662e-06, + "loss": 0.7315, + "step": 7377 + }, + { + "epoch": 1.5470748584608933, + "grad_norm": 7.150666859336036, + "learning_rate": 9.978493112919831e-06, + "loss": 0.87, + "step": 7378 + }, + { + "epoch": 1.5472845460264206, + "grad_norm": 8.515569495317697, + "learning_rate": 9.976229234130261e-06, + "loss": 1.0724, + "step": 7379 + }, + { + "epoch": 1.547494233591948, + "grad_norm": 6.84047306235906, + "learning_rate": 9.97396535655899e-06, + "loss": 0.7746, + "step": 7380 + }, + { + "epoch": 1.5477039211574755, + "grad_norm": 8.25125969698863, + "learning_rate": 9.971701480322038e-06, + "loss": 1.1777, + "step": 7381 + }, + { + "epoch": 1.5479136087230028, + "grad_norm": 7.422910125224376, + "learning_rate": 9.969437605535434e-06, + "loss": 0.8754, + "step": 7382 + }, + { + "epoch": 1.54812329628853, + "grad_norm": 6.36336117328942, + "learning_rate": 9.967173732315207e-06, + "loss": 0.6979, + "step": 7383 + }, + { + "epoch": 1.5483329838540576, + "grad_norm": 5.766541204693824, + "learning_rate": 9.964909860777378e-06, + "loss": 0.6505, + "step": 7384 + }, + { + "epoch": 1.5485426714195847, + "grad_norm": 8.4724776724924, + "learning_rate": 9.962645991037984e-06, + "loss": 0.9283, + "step": 7385 + }, + { + "epoch": 1.5487523589851122, + "grad_norm": 5.937185265173824, + "learning_rate": 9.960382123213049e-06, + "loss": 0.705, + "step": 7386 + }, + { + "epoch": 1.5489620465506395, + "grad_norm": 7.36011256077543, + "learning_rate": 9.958118257418596e-06, + "loss": 0.6625, + "step": 7387 + }, + { + "epoch": 1.5491717341161668, + "grad_norm": 6.244074231971001, + "learning_rate": 9.955854393770655e-06, + "loss": 0.6501, + "step": 7388 + }, + { + "epoch": 1.5493814216816943, + "grad_norm": 7.774768067067479, + "learning_rate": 9.953590532385256e-06, + "loss": 0.9274, + "step": 7389 + }, + { + "epoch": 1.5495911092472217, + "grad_norm": 7.553429689084709, + "learning_rate": 9.951326673378418e-06, + "loss": 0.7823, + "step": 7390 + }, + { + "epoch": 1.549800796812749, + "grad_norm": 6.5687810237587, + "learning_rate": 9.949062816866176e-06, + "loss": 0.6806, + "step": 7391 + }, + { + "epoch": 1.5500104843782765, + "grad_norm": 7.082892471930119, + "learning_rate": 9.946798962964552e-06, + "loss": 0.7919, + "step": 7392 + }, + { + "epoch": 1.5502201719438036, + "grad_norm": 9.797756382432269, + "learning_rate": 9.944535111789572e-06, + "loss": 0.9143, + "step": 7393 + }, + { + "epoch": 1.550429859509331, + "grad_norm": 5.60964219046708, + "learning_rate": 9.942271263457264e-06, + "loss": 0.5399, + "step": 7394 + }, + { + "epoch": 1.5506395470748586, + "grad_norm": 7.972707391497301, + "learning_rate": 9.940007418083657e-06, + "loss": 0.9266, + "step": 7395 + }, + { + "epoch": 1.5508492346403857, + "grad_norm": 8.286666657127727, + "learning_rate": 9.93774357578477e-06, + "loss": 0.9395, + "step": 7396 + }, + { + "epoch": 1.5510589222059132, + "grad_norm": 7.8268969623340094, + "learning_rate": 9.935479736676634e-06, + "loss": 0.95, + "step": 7397 + }, + { + "epoch": 1.5512686097714405, + "grad_norm": 7.044189881490119, + "learning_rate": 9.933215900875275e-06, + "loss": 0.6886, + "step": 7398 + }, + { + "epoch": 1.5514782973369678, + "grad_norm": 5.651967220808202, + "learning_rate": 9.930952068496716e-06, + "loss": 0.4918, + "step": 7399 + }, + { + "epoch": 1.5516879849024954, + "grad_norm": 5.061012068538018, + "learning_rate": 9.928688239656982e-06, + "loss": 0.5931, + "step": 7400 + }, + { + "epoch": 1.5518976724680227, + "grad_norm": 8.138325659370816, + "learning_rate": 9.926424414472104e-06, + "loss": 1.005, + "step": 7401 + }, + { + "epoch": 1.55210736003355, + "grad_norm": 6.886416599279407, + "learning_rate": 9.924160593058095e-06, + "loss": 0.7032, + "step": 7402 + }, + { + "epoch": 1.5523170475990775, + "grad_norm": 5.901054460204976, + "learning_rate": 9.921896775530991e-06, + "loss": 0.5227, + "step": 7403 + }, + { + "epoch": 1.5525267351646046, + "grad_norm": 6.940740870612526, + "learning_rate": 9.919632962006814e-06, + "loss": 0.7305, + "step": 7404 + }, + { + "epoch": 1.5527364227301321, + "grad_norm": 5.97440926309287, + "learning_rate": 9.917369152601586e-06, + "loss": 0.4689, + "step": 7405 + }, + { + "epoch": 1.5529461102956594, + "grad_norm": 7.1044651219747115, + "learning_rate": 9.91510534743133e-06, + "loss": 0.605, + "step": 7406 + }, + { + "epoch": 1.5531557978611867, + "grad_norm": 5.713273754640501, + "learning_rate": 9.912841546612075e-06, + "loss": 0.5209, + "step": 7407 + }, + { + "epoch": 1.5533654854267143, + "grad_norm": 6.536196489934422, + "learning_rate": 9.910577750259841e-06, + "loss": 0.5689, + "step": 7408 + }, + { + "epoch": 1.5535751729922416, + "grad_norm": 6.230477255932016, + "learning_rate": 9.908313958490652e-06, + "loss": 0.6734, + "step": 7409 + }, + { + "epoch": 1.5537848605577689, + "grad_norm": 7.439722408684949, + "learning_rate": 9.906050171420534e-06, + "loss": 0.7466, + "step": 7410 + }, + { + "epoch": 1.5539945481232964, + "grad_norm": 6.7371071189214815, + "learning_rate": 9.903786389165503e-06, + "loss": 0.7161, + "step": 7411 + }, + { + "epoch": 1.5542042356888235, + "grad_norm": 5.238648078897976, + "learning_rate": 9.901522611841588e-06, + "loss": 0.614, + "step": 7412 + }, + { + "epoch": 1.554413923254351, + "grad_norm": 7.648841098154688, + "learning_rate": 9.899258839564814e-06, + "loss": 0.8154, + "step": 7413 + }, + { + "epoch": 1.5546236108198785, + "grad_norm": 6.435799416317733, + "learning_rate": 9.896995072451197e-06, + "loss": 0.6055, + "step": 7414 + }, + { + "epoch": 1.5548332983854056, + "grad_norm": 6.061287014767684, + "learning_rate": 9.894731310616757e-06, + "loss": 0.5308, + "step": 7415 + }, + { + "epoch": 1.5550429859509332, + "grad_norm": 8.491850374920919, + "learning_rate": 9.892467554177527e-06, + "loss": 0.9534, + "step": 7416 + }, + { + "epoch": 1.5552526735164605, + "grad_norm": 7.269798493482319, + "learning_rate": 9.890203803249515e-06, + "loss": 0.6649, + "step": 7417 + }, + { + "epoch": 1.5554623610819878, + "grad_norm": 7.400409560532959, + "learning_rate": 9.887940057948751e-06, + "loss": 0.8717, + "step": 7418 + }, + { + "epoch": 1.5556720486475153, + "grad_norm": 4.9203649847954, + "learning_rate": 9.885676318391257e-06, + "loss": 0.439, + "step": 7419 + }, + { + "epoch": 1.5558817362130426, + "grad_norm": 7.194978466393346, + "learning_rate": 9.883412584693046e-06, + "loss": 0.6881, + "step": 7420 + }, + { + "epoch": 1.55609142377857, + "grad_norm": 5.993666010139874, + "learning_rate": 9.88114885697014e-06, + "loss": 0.5605, + "step": 7421 + }, + { + "epoch": 1.5563011113440974, + "grad_norm": 7.826221927628422, + "learning_rate": 9.878885135338567e-06, + "loss": 0.8106, + "step": 7422 + }, + { + "epoch": 1.5565107989096245, + "grad_norm": 8.556001663078536, + "learning_rate": 9.876621419914336e-06, + "loss": 0.825, + "step": 7423 + }, + { + "epoch": 1.556720486475152, + "grad_norm": 6.4766126111350175, + "learning_rate": 9.874357710813473e-06, + "loss": 0.6068, + "step": 7424 + }, + { + "epoch": 1.5569301740406793, + "grad_norm": 7.039615496126097, + "learning_rate": 9.872094008151998e-06, + "loss": 0.7329, + "step": 7425 + }, + { + "epoch": 1.5571398616062067, + "grad_norm": 7.808146805302258, + "learning_rate": 9.869830312045923e-06, + "loss": 0.7329, + "step": 7426 + }, + { + "epoch": 1.5573495491717342, + "grad_norm": 9.181438045786505, + "learning_rate": 9.86756662261127e-06, + "loss": 0.8702, + "step": 7427 + }, + { + "epoch": 1.5575592367372615, + "grad_norm": 7.221367404832524, + "learning_rate": 9.86530293996406e-06, + "loss": 0.8016, + "step": 7428 + }, + { + "epoch": 1.5577689243027888, + "grad_norm": 7.910725948399105, + "learning_rate": 9.863039264220307e-06, + "loss": 0.9236, + "step": 7429 + }, + { + "epoch": 1.5579786118683163, + "grad_norm": 7.6315821739079235, + "learning_rate": 9.860775595496028e-06, + "loss": 0.6841, + "step": 7430 + }, + { + "epoch": 1.5581882994338436, + "grad_norm": 9.748350604359869, + "learning_rate": 9.85851193390724e-06, + "loss": 1.0554, + "step": 7431 + }, + { + "epoch": 1.558397986999371, + "grad_norm": 9.683576405345406, + "learning_rate": 9.856248279569965e-06, + "loss": 0.961, + "step": 7432 + }, + { + "epoch": 1.5586076745648985, + "grad_norm": 6.301012582428548, + "learning_rate": 9.853984632600213e-06, + "loss": 0.6483, + "step": 7433 + }, + { + "epoch": 1.5588173621304255, + "grad_norm": 7.639637671200591, + "learning_rate": 9.851720993114002e-06, + "loss": 0.7748, + "step": 7434 + }, + { + "epoch": 1.559027049695953, + "grad_norm": 6.989799799631193, + "learning_rate": 9.849457361227348e-06, + "loss": 0.7327, + "step": 7435 + }, + { + "epoch": 1.5592367372614804, + "grad_norm": 6.6445969175248045, + "learning_rate": 9.847193737056265e-06, + "loss": 0.5949, + "step": 7436 + }, + { + "epoch": 1.5594464248270077, + "grad_norm": 8.196983726113416, + "learning_rate": 9.844930120716764e-06, + "loss": 0.9433, + "step": 7437 + }, + { + "epoch": 1.5596561123925352, + "grad_norm": 7.044213272565848, + "learning_rate": 9.84266651232487e-06, + "loss": 0.764, + "step": 7438 + }, + { + "epoch": 1.5598657999580625, + "grad_norm": 5.699596672260946, + "learning_rate": 9.840402911996587e-06, + "loss": 0.5077, + "step": 7439 + }, + { + "epoch": 1.5600754875235898, + "grad_norm": 7.175423056665053, + "learning_rate": 9.83813931984793e-06, + "loss": 0.7726, + "step": 7440 + }, + { + "epoch": 1.5602851750891173, + "grad_norm": 6.055914141470011, + "learning_rate": 9.835875735994917e-06, + "loss": 0.7445, + "step": 7441 + }, + { + "epoch": 1.5604948626546444, + "grad_norm": 6.866791139508365, + "learning_rate": 9.833612160553553e-06, + "loss": 0.8377, + "step": 7442 + }, + { + "epoch": 1.560704550220172, + "grad_norm": 5.682240899799926, + "learning_rate": 9.831348593639856e-06, + "loss": 0.6318, + "step": 7443 + }, + { + "epoch": 1.5609142377856993, + "grad_norm": 6.697821738660285, + "learning_rate": 9.829085035369839e-06, + "loss": 0.7465, + "step": 7444 + }, + { + "epoch": 1.5611239253512266, + "grad_norm": 8.4265197811356, + "learning_rate": 9.826821485859508e-06, + "loss": 0.84, + "step": 7445 + }, + { + "epoch": 1.561333612916754, + "grad_norm": 7.758443727004098, + "learning_rate": 9.824557945224872e-06, + "loss": 0.8567, + "step": 7446 + }, + { + "epoch": 1.5615433004822814, + "grad_norm": 6.32218140957602, + "learning_rate": 9.822294413581956e-06, + "loss": 0.6237, + "step": 7447 + }, + { + "epoch": 1.5617529880478087, + "grad_norm": 5.738486413228919, + "learning_rate": 9.82003089104675e-06, + "loss": 0.6485, + "step": 7448 + }, + { + "epoch": 1.5619626756133362, + "grad_norm": 6.243931410198658, + "learning_rate": 9.817767377735277e-06, + "loss": 0.7206, + "step": 7449 + }, + { + "epoch": 1.5621723631788635, + "grad_norm": 9.12705963059562, + "learning_rate": 9.815503873763543e-06, + "loss": 0.9281, + "step": 7450 + }, + { + "epoch": 1.5623820507443908, + "grad_norm": 6.853861817698881, + "learning_rate": 9.813240379247553e-06, + "loss": 0.7684, + "step": 7451 + }, + { + "epoch": 1.5625917383099184, + "grad_norm": 6.2619693168152954, + "learning_rate": 9.810976894303316e-06, + "loss": 0.7056, + "step": 7452 + }, + { + "epoch": 1.5628014258754455, + "grad_norm": 6.327822853310262, + "learning_rate": 9.808713419046847e-06, + "loss": 0.6128, + "step": 7453 + }, + { + "epoch": 1.563011113440973, + "grad_norm": 9.16285206304075, + "learning_rate": 9.806449953594141e-06, + "loss": 1.195, + "step": 7454 + }, + { + "epoch": 1.5632208010065003, + "grad_norm": 7.116947249577371, + "learning_rate": 9.804186498061213e-06, + "loss": 0.8245, + "step": 7455 + }, + { + "epoch": 1.5634304885720276, + "grad_norm": 7.116804055019194, + "learning_rate": 9.801923052564068e-06, + "loss": 0.6003, + "step": 7456 + }, + { + "epoch": 1.5636401761375551, + "grad_norm": 7.621269926462281, + "learning_rate": 9.799659617218707e-06, + "loss": 0.8642, + "step": 7457 + }, + { + "epoch": 1.5638498637030824, + "grad_norm": 7.742711612185244, + "learning_rate": 9.797396192141138e-06, + "loss": 0.9105, + "step": 7458 + }, + { + "epoch": 1.5640595512686097, + "grad_norm": 6.965607872390398, + "learning_rate": 9.79513277744737e-06, + "loss": 0.8853, + "step": 7459 + }, + { + "epoch": 1.5642692388341373, + "grad_norm": 6.756849768589031, + "learning_rate": 9.792869373253399e-06, + "loss": 0.751, + "step": 7460 + }, + { + "epoch": 1.5644789263996643, + "grad_norm": 6.6637820610235945, + "learning_rate": 9.790605979675231e-06, + "loss": 0.7562, + "step": 7461 + }, + { + "epoch": 1.5646886139651919, + "grad_norm": 7.003379069793895, + "learning_rate": 9.788342596828875e-06, + "loss": 0.6947, + "step": 7462 + }, + { + "epoch": 1.5648983015307192, + "grad_norm": 7.146558886184148, + "learning_rate": 9.78607922483032e-06, + "loss": 0.6611, + "step": 7463 + }, + { + "epoch": 1.5651079890962465, + "grad_norm": 10.160253778038685, + "learning_rate": 9.78381586379558e-06, + "loss": 1.1366, + "step": 7464 + }, + { + "epoch": 1.565317676661774, + "grad_norm": 7.119847253049567, + "learning_rate": 9.781552513840654e-06, + "loss": 0.656, + "step": 7465 + }, + { + "epoch": 1.5655273642273013, + "grad_norm": 8.890059129015276, + "learning_rate": 9.779289175081539e-06, + "loss": 0.9618, + "step": 7466 + }, + { + "epoch": 1.5657370517928286, + "grad_norm": 6.5823168964265175, + "learning_rate": 9.777025847634235e-06, + "loss": 0.8161, + "step": 7467 + }, + { + "epoch": 1.5659467393583562, + "grad_norm": 6.992804144728043, + "learning_rate": 9.774762531614747e-06, + "loss": 0.7692, + "step": 7468 + }, + { + "epoch": 1.5661564269238835, + "grad_norm": 8.453073228194512, + "learning_rate": 9.772499227139063e-06, + "loss": 1.0109, + "step": 7469 + }, + { + "epoch": 1.5663661144894108, + "grad_norm": 8.639264144664015, + "learning_rate": 9.770235934323193e-06, + "loss": 0.9519, + "step": 7470 + }, + { + "epoch": 1.5665758020549383, + "grad_norm": 7.291100729142389, + "learning_rate": 9.767972653283131e-06, + "loss": 0.8416, + "step": 7471 + }, + { + "epoch": 1.5667854896204654, + "grad_norm": 8.386398995828655, + "learning_rate": 9.765709384134871e-06, + "loss": 0.8636, + "step": 7472 + }, + { + "epoch": 1.566995177185993, + "grad_norm": 7.569159697271252, + "learning_rate": 9.76344612699441e-06, + "loss": 0.9601, + "step": 7473 + }, + { + "epoch": 1.5672048647515202, + "grad_norm": 7.123754178319306, + "learning_rate": 9.76118288197775e-06, + "loss": 0.8016, + "step": 7474 + }, + { + "epoch": 1.5674145523170475, + "grad_norm": 5.7134592128711015, + "learning_rate": 9.758919649200875e-06, + "loss": 0.6273, + "step": 7475 + }, + { + "epoch": 1.567624239882575, + "grad_norm": 6.365461833380269, + "learning_rate": 9.75665642877979e-06, + "loss": 0.7107, + "step": 7476 + }, + { + "epoch": 1.5678339274481023, + "grad_norm": 5.676818018675542, + "learning_rate": 9.754393220830487e-06, + "loss": 0.526, + "step": 7477 + }, + { + "epoch": 1.5680436150136297, + "grad_norm": 7.830523842657074, + "learning_rate": 9.752130025468954e-06, + "loss": 0.8589, + "step": 7478 + }, + { + "epoch": 1.5682533025791572, + "grad_norm": 7.396452662931585, + "learning_rate": 9.749866842811186e-06, + "loss": 0.8333, + "step": 7479 + }, + { + "epoch": 1.5684629901446843, + "grad_norm": 6.48636418906607, + "learning_rate": 9.74760367297318e-06, + "loss": 0.7618, + "step": 7480 + }, + { + "epoch": 1.5686726777102118, + "grad_norm": 7.248439210261221, + "learning_rate": 9.745340516070921e-06, + "loss": 0.7922, + "step": 7481 + }, + { + "epoch": 1.568882365275739, + "grad_norm": 7.6972765277571416, + "learning_rate": 9.743077372220403e-06, + "loss": 0.772, + "step": 7482 + }, + { + "epoch": 1.5690920528412664, + "grad_norm": 7.4256166688719665, + "learning_rate": 9.740814241537616e-06, + "loss": 0.7461, + "step": 7483 + }, + { + "epoch": 1.569301740406794, + "grad_norm": 6.6431578924629155, + "learning_rate": 9.738551124138548e-06, + "loss": 0.743, + "step": 7484 + }, + { + "epoch": 1.5695114279723212, + "grad_norm": 7.329326809923285, + "learning_rate": 9.736288020139185e-06, + "loss": 0.8285, + "step": 7485 + }, + { + "epoch": 1.5697211155378485, + "grad_norm": 7.1880629119512545, + "learning_rate": 9.734024929655525e-06, + "loss": 0.773, + "step": 7486 + }, + { + "epoch": 1.569930803103376, + "grad_norm": 5.582308808457554, + "learning_rate": 9.731761852803544e-06, + "loss": 0.6851, + "step": 7487 + }, + { + "epoch": 1.5701404906689034, + "grad_norm": 6.053800125006612, + "learning_rate": 9.729498789699234e-06, + "loss": 0.6111, + "step": 7488 + }, + { + "epoch": 1.5703501782344307, + "grad_norm": 7.219497942778008, + "learning_rate": 9.727235740458582e-06, + "loss": 0.6762, + "step": 7489 + }, + { + "epoch": 1.5705598657999582, + "grad_norm": 8.521334094887038, + "learning_rate": 9.72497270519757e-06, + "loss": 0.8987, + "step": 7490 + }, + { + "epoch": 1.5707695533654853, + "grad_norm": 7.009516597054003, + "learning_rate": 9.722709684032182e-06, + "loss": 0.7874, + "step": 7491 + }, + { + "epoch": 1.5709792409310128, + "grad_norm": 6.47722224926297, + "learning_rate": 9.720446677078406e-06, + "loss": 0.6186, + "step": 7492 + }, + { + "epoch": 1.5711889284965401, + "grad_norm": 8.663255608378202, + "learning_rate": 9.718183684452223e-06, + "loss": 0.8661, + "step": 7493 + }, + { + "epoch": 1.5713986160620674, + "grad_norm": 7.134489860552042, + "learning_rate": 9.715920706269611e-06, + "loss": 0.8573, + "step": 7494 + }, + { + "epoch": 1.571608303627595, + "grad_norm": 6.597442510139765, + "learning_rate": 9.713657742646559e-06, + "loss": 0.8199, + "step": 7495 + }, + { + "epoch": 1.5718179911931223, + "grad_norm": 7.461486140895039, + "learning_rate": 9.711394793699045e-06, + "loss": 0.7018, + "step": 7496 + }, + { + "epoch": 1.5720276787586496, + "grad_norm": 7.488250612217129, + "learning_rate": 9.709131859543048e-06, + "loss": 0.7776, + "step": 7497 + }, + { + "epoch": 1.572237366324177, + "grad_norm": 6.307917033650578, + "learning_rate": 9.706868940294547e-06, + "loss": 0.5386, + "step": 7498 + }, + { + "epoch": 1.5724470538897042, + "grad_norm": 7.499994318930143, + "learning_rate": 9.704606036069522e-06, + "loss": 0.9418, + "step": 7499 + }, + { + "epoch": 1.5726567414552317, + "grad_norm": 6.352712573145573, + "learning_rate": 9.702343146983948e-06, + "loss": 0.7557, + "step": 7500 + }, + { + "epoch": 1.572866429020759, + "grad_norm": 7.459755694325853, + "learning_rate": 9.700080273153805e-06, + "loss": 0.8752, + "step": 7501 + }, + { + "epoch": 1.5730761165862863, + "grad_norm": 8.010113058389537, + "learning_rate": 9.69781741469507e-06, + "loss": 0.888, + "step": 7502 + }, + { + "epoch": 1.5732858041518138, + "grad_norm": 7.173241699338803, + "learning_rate": 9.695554571723716e-06, + "loss": 0.8885, + "step": 7503 + }, + { + "epoch": 1.5734954917173412, + "grad_norm": 5.617137493552929, + "learning_rate": 9.693291744355715e-06, + "loss": 0.4666, + "step": 7504 + }, + { + "epoch": 1.5737051792828685, + "grad_norm": 8.085811460821102, + "learning_rate": 9.69102893270705e-06, + "loss": 1.0287, + "step": 7505 + }, + { + "epoch": 1.573914866848396, + "grad_norm": 6.538928147186817, + "learning_rate": 9.688766136893682e-06, + "loss": 0.6976, + "step": 7506 + }, + { + "epoch": 1.5741245544139233, + "grad_norm": 5.7084207093394985, + "learning_rate": 9.68650335703159e-06, + "loss": 0.52, + "step": 7507 + }, + { + "epoch": 1.5743342419794506, + "grad_norm": 7.8446855839585385, + "learning_rate": 9.684240593236749e-06, + "loss": 0.7537, + "step": 7508 + }, + { + "epoch": 1.5745439295449781, + "grad_norm": 9.25493089398353, + "learning_rate": 9.681977845625123e-06, + "loss": 1.0807, + "step": 7509 + }, + { + "epoch": 1.5747536171105052, + "grad_norm": 8.692973847412153, + "learning_rate": 9.67971511431268e-06, + "loss": 0.8971, + "step": 7510 + }, + { + "epoch": 1.5749633046760327, + "grad_norm": 6.976075381288707, + "learning_rate": 9.677452399415398e-06, + "loss": 0.7109, + "step": 7511 + }, + { + "epoch": 1.57517299224156, + "grad_norm": 8.253237559108165, + "learning_rate": 9.67518970104924e-06, + "loss": 0.955, + "step": 7512 + }, + { + "epoch": 1.5753826798070873, + "grad_norm": 7.758732530833251, + "learning_rate": 9.67292701933017e-06, + "loss": 0.8814, + "step": 7513 + }, + { + "epoch": 1.5755923673726149, + "grad_norm": 8.06314963419161, + "learning_rate": 9.670664354374162e-06, + "loss": 0.859, + "step": 7514 + }, + { + "epoch": 1.5758020549381422, + "grad_norm": 6.5901242393198105, + "learning_rate": 9.668401706297174e-06, + "loss": 0.6697, + "step": 7515 + }, + { + "epoch": 1.5760117425036695, + "grad_norm": 7.1728352680891145, + "learning_rate": 9.666139075215171e-06, + "loss": 0.7886, + "step": 7516 + }, + { + "epoch": 1.576221430069197, + "grad_norm": 8.118085997577953, + "learning_rate": 9.663876461244124e-06, + "loss": 0.8422, + "step": 7517 + }, + { + "epoch": 1.576431117634724, + "grad_norm": 5.829857308472773, + "learning_rate": 9.66161386449999e-06, + "loss": 0.7008, + "step": 7518 + }, + { + "epoch": 1.5766408052002516, + "grad_norm": 7.376646148600872, + "learning_rate": 9.65935128509873e-06, + "loss": 0.6357, + "step": 7519 + }, + { + "epoch": 1.5768504927657792, + "grad_norm": 7.480249333462964, + "learning_rate": 9.657088723156313e-06, + "loss": 1.098, + "step": 7520 + }, + { + "epoch": 1.5770601803313062, + "grad_norm": 7.377854134648045, + "learning_rate": 9.654826178788688e-06, + "loss": 1.0076, + "step": 7521 + }, + { + "epoch": 1.5772698678968338, + "grad_norm": 7.550957039288108, + "learning_rate": 9.652563652111821e-06, + "loss": 0.7002, + "step": 7522 + }, + { + "epoch": 1.577479555462361, + "grad_norm": 6.399844844795855, + "learning_rate": 9.650301143241673e-06, + "loss": 0.6634, + "step": 7523 + }, + { + "epoch": 1.5776892430278884, + "grad_norm": 6.54201463432692, + "learning_rate": 9.648038652294195e-06, + "loss": 0.9115, + "step": 7524 + }, + { + "epoch": 1.577898930593416, + "grad_norm": 7.176741895892891, + "learning_rate": 9.645776179385345e-06, + "loss": 0.7798, + "step": 7525 + }, + { + "epoch": 1.5781086181589432, + "grad_norm": 6.150466206177959, + "learning_rate": 9.643513724631086e-06, + "loss": 0.6437, + "step": 7526 + }, + { + "epoch": 1.5783183057244705, + "grad_norm": 7.060595301967234, + "learning_rate": 9.64125128814736e-06, + "loss": 0.7624, + "step": 7527 + }, + { + "epoch": 1.578527993289998, + "grad_norm": 6.420338450913425, + "learning_rate": 9.638988870050132e-06, + "loss": 0.5972, + "step": 7528 + }, + { + "epoch": 1.5787376808555251, + "grad_norm": 6.19310094822929, + "learning_rate": 9.636726470455351e-06, + "loss": 0.7269, + "step": 7529 + }, + { + "epoch": 1.5789473684210527, + "grad_norm": 6.931568484899664, + "learning_rate": 9.634464089478967e-06, + "loss": 0.7508, + "step": 7530 + }, + { + "epoch": 1.57915705598658, + "grad_norm": 8.96362334964029, + "learning_rate": 9.63220172723693e-06, + "loss": 0.809, + "step": 7531 + }, + { + "epoch": 1.5793667435521073, + "grad_norm": 8.500684861990925, + "learning_rate": 9.629939383845198e-06, + "loss": 0.9827, + "step": 7532 + }, + { + "epoch": 1.5795764311176348, + "grad_norm": 6.261521623516531, + "learning_rate": 9.627677059419708e-06, + "loss": 0.7134, + "step": 7533 + }, + { + "epoch": 1.579786118683162, + "grad_norm": 6.6534545299229135, + "learning_rate": 9.625414754076417e-06, + "loss": 0.7252, + "step": 7534 + }, + { + "epoch": 1.5799958062486894, + "grad_norm": 6.1417236288698955, + "learning_rate": 9.623152467931273e-06, + "loss": 0.6064, + "step": 7535 + }, + { + "epoch": 1.580205493814217, + "grad_norm": 6.385380118770043, + "learning_rate": 9.620890201100217e-06, + "loss": 0.7248, + "step": 7536 + }, + { + "epoch": 1.580415181379744, + "grad_norm": 7.402160771858727, + "learning_rate": 9.61862795369919e-06, + "loss": 0.9799, + "step": 7537 + }, + { + "epoch": 1.5806248689452715, + "grad_norm": 6.230326862722612, + "learning_rate": 9.61636572584415e-06, + "loss": 0.671, + "step": 7538 + }, + { + "epoch": 1.580834556510799, + "grad_norm": 7.575365523441947, + "learning_rate": 9.614103517651026e-06, + "loss": 0.8565, + "step": 7539 + }, + { + "epoch": 1.5810442440763262, + "grad_norm": 6.948627075638781, + "learning_rate": 9.611841329235768e-06, + "loss": 0.7565, + "step": 7540 + }, + { + "epoch": 1.5812539316418537, + "grad_norm": 7.363592153405708, + "learning_rate": 9.609579160714318e-06, + "loss": 0.818, + "step": 7541 + }, + { + "epoch": 1.581463619207381, + "grad_norm": 7.83431952455529, + "learning_rate": 9.607317012202608e-06, + "loss": 0.8215, + "step": 7542 + }, + { + "epoch": 1.5816733067729083, + "grad_norm": 7.175861132628268, + "learning_rate": 9.605054883816584e-06, + "loss": 0.8704, + "step": 7543 + }, + { + "epoch": 1.5818829943384358, + "grad_norm": 6.493760770143072, + "learning_rate": 9.602792775672185e-06, + "loss": 0.7365, + "step": 7544 + }, + { + "epoch": 1.5820926819039631, + "grad_norm": 6.198319330361964, + "learning_rate": 9.600530687885344e-06, + "loss": 0.6595, + "step": 7545 + }, + { + "epoch": 1.5823023694694904, + "grad_norm": 9.560676981398183, + "learning_rate": 9.598268620571996e-06, + "loss": 0.8659, + "step": 7546 + }, + { + "epoch": 1.582512057035018, + "grad_norm": 7.366657585410217, + "learning_rate": 9.59600657384808e-06, + "loss": 0.8464, + "step": 7547 + }, + { + "epoch": 1.582721744600545, + "grad_norm": 7.2257100285519575, + "learning_rate": 9.593744547829526e-06, + "loss": 0.975, + "step": 7548 + }, + { + "epoch": 1.5829314321660726, + "grad_norm": 6.397330127305164, + "learning_rate": 9.591482542632268e-06, + "loss": 0.7875, + "step": 7549 + }, + { + "epoch": 1.5831411197315999, + "grad_norm": 5.951921373719367, + "learning_rate": 9.589220558372241e-06, + "loss": 0.5713, + "step": 7550 + }, + { + "epoch": 1.5833508072971272, + "grad_norm": 6.814451252091627, + "learning_rate": 9.586958595165373e-06, + "loss": 0.7571, + "step": 7551 + }, + { + "epoch": 1.5835604948626547, + "grad_norm": 5.729014870120971, + "learning_rate": 9.584696653127589e-06, + "loss": 0.506, + "step": 7552 + }, + { + "epoch": 1.583770182428182, + "grad_norm": 6.015967608269447, + "learning_rate": 9.582434732374823e-06, + "loss": 0.6572, + "step": 7553 + }, + { + "epoch": 1.5839798699937093, + "grad_norm": 7.131110594928549, + "learning_rate": 9.580172833023006e-06, + "loss": 0.9791, + "step": 7554 + }, + { + "epoch": 1.5841895575592368, + "grad_norm": 6.396493916540819, + "learning_rate": 9.577910955188056e-06, + "loss": 0.6287, + "step": 7555 + }, + { + "epoch": 1.5843992451247642, + "grad_norm": 5.898334677538959, + "learning_rate": 9.5756490989859e-06, + "loss": 0.6404, + "step": 7556 + }, + { + "epoch": 1.5846089326902915, + "grad_norm": 7.431070292088967, + "learning_rate": 9.573387264532469e-06, + "loss": 0.7677, + "step": 7557 + }, + { + "epoch": 1.584818620255819, + "grad_norm": 7.214486151736406, + "learning_rate": 9.571125451943674e-06, + "loss": 0.8318, + "step": 7558 + }, + { + "epoch": 1.585028307821346, + "grad_norm": 7.134508097611619, + "learning_rate": 9.568863661335447e-06, + "loss": 0.7964, + "step": 7559 + }, + { + "epoch": 1.5852379953868736, + "grad_norm": 7.438777391646582, + "learning_rate": 9.566601892823706e-06, + "loss": 0.7829, + "step": 7560 + }, + { + "epoch": 1.585447682952401, + "grad_norm": 8.35222137174369, + "learning_rate": 9.564340146524369e-06, + "loss": 0.7847, + "step": 7561 + }, + { + "epoch": 1.5856573705179282, + "grad_norm": 7.154482566491909, + "learning_rate": 9.562078422553352e-06, + "loss": 0.8228, + "step": 7562 + }, + { + "epoch": 1.5858670580834557, + "grad_norm": 10.21973009542303, + "learning_rate": 9.559816721026583e-06, + "loss": 0.8651, + "step": 7563 + }, + { + "epoch": 1.586076745648983, + "grad_norm": 6.0944646423921585, + "learning_rate": 9.557555042059963e-06, + "loss": 0.7973, + "step": 7564 + }, + { + "epoch": 1.5862864332145103, + "grad_norm": 7.033348811610297, + "learning_rate": 9.555293385769417e-06, + "loss": 0.7986, + "step": 7565 + }, + { + "epoch": 1.5864961207800379, + "grad_norm": 6.064191070214426, + "learning_rate": 9.553031752270858e-06, + "loss": 0.6064, + "step": 7566 + }, + { + "epoch": 1.586705808345565, + "grad_norm": 7.4518086165095925, + "learning_rate": 9.550770141680194e-06, + "loss": 1.0722, + "step": 7567 + }, + { + "epoch": 1.5869154959110925, + "grad_norm": 8.205309077106456, + "learning_rate": 9.548508554113339e-06, + "loss": 0.9551, + "step": 7568 + }, + { + "epoch": 1.5871251834766198, + "grad_norm": 5.992382118360831, + "learning_rate": 9.546246989686207e-06, + "loss": 0.5532, + "step": 7569 + }, + { + "epoch": 1.587334871042147, + "grad_norm": 6.145985110612443, + "learning_rate": 9.543985448514702e-06, + "loss": 0.5055, + "step": 7570 + }, + { + "epoch": 1.5875445586076746, + "grad_norm": 7.159649154096195, + "learning_rate": 9.541723930714734e-06, + "loss": 0.8941, + "step": 7571 + }, + { + "epoch": 1.587754246173202, + "grad_norm": 5.753477237891142, + "learning_rate": 9.53946243640221e-06, + "loss": 0.6239, + "step": 7572 + }, + { + "epoch": 1.5879639337387292, + "grad_norm": 5.684399582245832, + "learning_rate": 9.537200965693031e-06, + "loss": 0.6259, + "step": 7573 + }, + { + "epoch": 1.5881736213042568, + "grad_norm": 9.627128063925703, + "learning_rate": 9.53493951870311e-06, + "loss": 1.315, + "step": 7574 + }, + { + "epoch": 1.588383308869784, + "grad_norm": 8.099195278428454, + "learning_rate": 9.532678095548345e-06, + "loss": 0.6974, + "step": 7575 + }, + { + "epoch": 1.5885929964353114, + "grad_norm": 7.342496324371607, + "learning_rate": 9.530416696344636e-06, + "loss": 0.8866, + "step": 7576 + }, + { + "epoch": 1.588802684000839, + "grad_norm": 6.560006933999104, + "learning_rate": 9.528155321207885e-06, + "loss": 0.7291, + "step": 7577 + }, + { + "epoch": 1.589012371566366, + "grad_norm": 7.096134236351607, + "learning_rate": 9.525893970253997e-06, + "loss": 0.8271, + "step": 7578 + }, + { + "epoch": 1.5892220591318935, + "grad_norm": 6.604334480487465, + "learning_rate": 9.523632643598858e-06, + "loss": 0.7784, + "step": 7579 + }, + { + "epoch": 1.5894317466974208, + "grad_norm": 7.9393429132673665, + "learning_rate": 9.521371341358376e-06, + "loss": 0.6618, + "step": 7580 + }, + { + "epoch": 1.5896414342629481, + "grad_norm": 7.3498354194543785, + "learning_rate": 9.519110063648445e-06, + "loss": 0.6907, + "step": 7581 + }, + { + "epoch": 1.5898511218284757, + "grad_norm": 7.035034089666707, + "learning_rate": 9.516848810584955e-06, + "loss": 0.7239, + "step": 7582 + }, + { + "epoch": 1.590060809394003, + "grad_norm": 7.767931199004609, + "learning_rate": 9.514587582283799e-06, + "loss": 0.9196, + "step": 7583 + }, + { + "epoch": 1.5902704969595303, + "grad_norm": 6.314748512294622, + "learning_rate": 9.51232637886088e-06, + "loss": 0.5313, + "step": 7584 + }, + { + "epoch": 1.5904801845250578, + "grad_norm": 6.789016078591631, + "learning_rate": 9.51006520043207e-06, + "loss": 0.558, + "step": 7585 + }, + { + "epoch": 1.5906898720905849, + "grad_norm": 7.350103278827284, + "learning_rate": 9.507804047113272e-06, + "loss": 0.6295, + "step": 7586 + }, + { + "epoch": 1.5908995596561124, + "grad_norm": 6.9470304294354985, + "learning_rate": 9.505542919020372e-06, + "loss": 0.7123, + "step": 7587 + }, + { + "epoch": 1.5911092472216397, + "grad_norm": 6.654879317505086, + "learning_rate": 9.503281816269253e-06, + "loss": 0.8382, + "step": 7588 + }, + { + "epoch": 1.591318934787167, + "grad_norm": 7.225590537665633, + "learning_rate": 9.501020738975801e-06, + "loss": 0.9082, + "step": 7589 + }, + { + "epoch": 1.5915286223526945, + "grad_norm": 6.352394236644447, + "learning_rate": 9.498759687255905e-06, + "loss": 0.792, + "step": 7590 + }, + { + "epoch": 1.5917383099182218, + "grad_norm": 7.242417276568485, + "learning_rate": 9.496498661225443e-06, + "loss": 0.7897, + "step": 7591 + }, + { + "epoch": 1.5919479974837492, + "grad_norm": 7.209606441983618, + "learning_rate": 9.494237661000298e-06, + "loss": 0.7947, + "step": 7592 + }, + { + "epoch": 1.5921576850492767, + "grad_norm": 8.484306295654152, + "learning_rate": 9.49197668669635e-06, + "loss": 1.0704, + "step": 7593 + }, + { + "epoch": 1.592367372614804, + "grad_norm": 5.79309084056881, + "learning_rate": 9.489715738429479e-06, + "loss": 0.728, + "step": 7594 + }, + { + "epoch": 1.5925770601803313, + "grad_norm": 6.221914843784227, + "learning_rate": 9.487454816315556e-06, + "loss": 0.6808, + "step": 7595 + }, + { + "epoch": 1.5927867477458588, + "grad_norm": 7.202143424080865, + "learning_rate": 9.48519392047047e-06, + "loss": 0.9228, + "step": 7596 + }, + { + "epoch": 1.592996435311386, + "grad_norm": 7.629816182196042, + "learning_rate": 9.482933051010085e-06, + "loss": 0.5871, + "step": 7597 + }, + { + "epoch": 1.5932061228769134, + "grad_norm": 7.743035071591913, + "learning_rate": 9.480672208050278e-06, + "loss": 0.8004, + "step": 7598 + }, + { + "epoch": 1.5934158104424407, + "grad_norm": 7.143327766381275, + "learning_rate": 9.478411391706923e-06, + "loss": 0.7195, + "step": 7599 + }, + { + "epoch": 1.593625498007968, + "grad_norm": 8.063154950357209, + "learning_rate": 9.476150602095883e-06, + "loss": 0.9223, + "step": 7600 + }, + { + "epoch": 1.5938351855734956, + "grad_norm": 6.998763510581478, + "learning_rate": 9.473889839333037e-06, + "loss": 0.6741, + "step": 7601 + }, + { + "epoch": 1.5940448731390229, + "grad_norm": 7.795093008316764, + "learning_rate": 9.471629103534252e-06, + "loss": 0.9485, + "step": 7602 + }, + { + "epoch": 1.5942545607045502, + "grad_norm": 8.258486459264457, + "learning_rate": 9.469368394815391e-06, + "loss": 1.032, + "step": 7603 + }, + { + "epoch": 1.5944642482700777, + "grad_norm": 6.863645188415032, + "learning_rate": 9.467107713292315e-06, + "loss": 0.8017, + "step": 7604 + }, + { + "epoch": 1.5946739358356048, + "grad_norm": 8.569780407977742, + "learning_rate": 9.4648470590809e-06, + "loss": 0.7837, + "step": 7605 + }, + { + "epoch": 1.5948836234011323, + "grad_norm": 7.916736903579657, + "learning_rate": 9.462586432296995e-06, + "loss": 0.797, + "step": 7606 + }, + { + "epoch": 1.5950933109666596, + "grad_norm": 5.618496525361615, + "learning_rate": 9.46032583305647e-06, + "loss": 0.5166, + "step": 7607 + }, + { + "epoch": 1.595302998532187, + "grad_norm": 6.86309679295889, + "learning_rate": 9.458065261475184e-06, + "loss": 0.5524, + "step": 7608 + }, + { + "epoch": 1.5955126860977145, + "grad_norm": 7.08751977868575, + "learning_rate": 9.455804717668993e-06, + "loss": 0.8004, + "step": 7609 + }, + { + "epoch": 1.5957223736632418, + "grad_norm": 7.941673010461113, + "learning_rate": 9.45354420175375e-06, + "loss": 0.7947, + "step": 7610 + }, + { + "epoch": 1.595932061228769, + "grad_norm": 7.416563145582192, + "learning_rate": 9.451283713845321e-06, + "loss": 0.8703, + "step": 7611 + }, + { + "epoch": 1.5961417487942966, + "grad_norm": 6.769307770647489, + "learning_rate": 9.449023254059547e-06, + "loss": 0.6543, + "step": 7612 + }, + { + "epoch": 1.596351436359824, + "grad_norm": 8.958574064433897, + "learning_rate": 9.446762822512289e-06, + "loss": 0.784, + "step": 7613 + }, + { + "epoch": 1.5965611239253512, + "grad_norm": 7.001065084174718, + "learning_rate": 9.444502419319393e-06, + "loss": 0.6407, + "step": 7614 + }, + { + "epoch": 1.5967708114908787, + "grad_norm": 6.811153271369026, + "learning_rate": 9.442242044596719e-06, + "loss": 0.7529, + "step": 7615 + }, + { + "epoch": 1.5969804990564058, + "grad_norm": 8.573629056599192, + "learning_rate": 9.439981698460101e-06, + "loss": 0.8493, + "step": 7616 + }, + { + "epoch": 1.5971901866219333, + "grad_norm": 7.82309361622577, + "learning_rate": 9.437721381025395e-06, + "loss": 0.9077, + "step": 7617 + }, + { + "epoch": 1.5973998741874607, + "grad_norm": 6.690752380500164, + "learning_rate": 9.435461092408445e-06, + "loss": 0.6738, + "step": 7618 + }, + { + "epoch": 1.597609561752988, + "grad_norm": 8.078854930003832, + "learning_rate": 9.43320083272509e-06, + "loss": 0.8209, + "step": 7619 + }, + { + "epoch": 1.5978192493185155, + "grad_norm": 8.591697372307808, + "learning_rate": 9.430940602091174e-06, + "loss": 1.1009, + "step": 7620 + }, + { + "epoch": 1.5980289368840428, + "grad_norm": 7.3563687001377165, + "learning_rate": 9.428680400622544e-06, + "loss": 0.8315, + "step": 7621 + }, + { + "epoch": 1.59823862444957, + "grad_norm": 7.989395837742091, + "learning_rate": 9.426420228435031e-06, + "loss": 0.9364, + "step": 7622 + }, + { + "epoch": 1.5984483120150976, + "grad_norm": 6.770592746617301, + "learning_rate": 9.424160085644476e-06, + "loss": 0.7764, + "step": 7623 + }, + { + "epoch": 1.5986579995806247, + "grad_norm": 6.020115962274509, + "learning_rate": 9.421899972366721e-06, + "loss": 0.7114, + "step": 7624 + }, + { + "epoch": 1.5988676871461522, + "grad_norm": 6.647356031656806, + "learning_rate": 9.419639888717589e-06, + "loss": 0.6068, + "step": 7625 + }, + { + "epoch": 1.5990773747116795, + "grad_norm": 5.4915670355416815, + "learning_rate": 9.417379834812918e-06, + "loss": 0.6227, + "step": 7626 + }, + { + "epoch": 1.5992870622772069, + "grad_norm": 6.509542323178818, + "learning_rate": 9.415119810768546e-06, + "loss": 0.7962, + "step": 7627 + }, + { + "epoch": 1.5994967498427344, + "grad_norm": 7.494419036308688, + "learning_rate": 9.412859816700296e-06, + "loss": 0.9367, + "step": 7628 + }, + { + "epoch": 1.5997064374082617, + "grad_norm": 6.163821032850612, + "learning_rate": 9.410599852723999e-06, + "loss": 0.6636, + "step": 7629 + }, + { + "epoch": 1.599916124973789, + "grad_norm": 6.201386977542586, + "learning_rate": 9.408339918955483e-06, + "loss": 0.5742, + "step": 7630 + }, + { + "epoch": 1.6001258125393165, + "grad_norm": 6.58385629566131, + "learning_rate": 9.406080015510567e-06, + "loss": 0.7012, + "step": 7631 + }, + { + "epoch": 1.6003355001048438, + "grad_norm": 8.09143711172415, + "learning_rate": 9.403820142505085e-06, + "loss": 0.9602, + "step": 7632 + }, + { + "epoch": 1.6005451876703711, + "grad_norm": 7.573456172268836, + "learning_rate": 9.401560300054856e-06, + "loss": 0.8303, + "step": 7633 + }, + { + "epoch": 1.6007548752358987, + "grad_norm": 5.644970830874413, + "learning_rate": 9.399300488275698e-06, + "loss": 0.6366, + "step": 7634 + }, + { + "epoch": 1.6009645628014257, + "grad_norm": 7.525339440261543, + "learning_rate": 9.397040707283429e-06, + "loss": 0.9075, + "step": 7635 + }, + { + "epoch": 1.6011742503669533, + "grad_norm": 6.877027376852609, + "learning_rate": 9.394780957193875e-06, + "loss": 0.8368, + "step": 7636 + }, + { + "epoch": 1.6013839379324806, + "grad_norm": 7.312198668623047, + "learning_rate": 9.392521238122841e-06, + "loss": 0.6836, + "step": 7637 + }, + { + "epoch": 1.6015936254980079, + "grad_norm": 7.409437819147826, + "learning_rate": 9.39026155018615e-06, + "loss": 0.9561, + "step": 7638 + }, + { + "epoch": 1.6018033130635354, + "grad_norm": 6.1472965616939845, + "learning_rate": 9.388001893499613e-06, + "loss": 0.7141, + "step": 7639 + }, + { + "epoch": 1.6020130006290627, + "grad_norm": 8.147380707981435, + "learning_rate": 9.38574226817904e-06, + "loss": 1.0179, + "step": 7640 + }, + { + "epoch": 1.60222268819459, + "grad_norm": 6.27219879140679, + "learning_rate": 9.383482674340236e-06, + "loss": 0.6746, + "step": 7641 + }, + { + "epoch": 1.6024323757601175, + "grad_norm": 6.576457475137217, + "learning_rate": 9.381223112099022e-06, + "loss": 0.6924, + "step": 7642 + }, + { + "epoch": 1.6026420633256446, + "grad_norm": 6.816325966272055, + "learning_rate": 9.378963581571192e-06, + "loss": 0.8075, + "step": 7643 + }, + { + "epoch": 1.6028517508911722, + "grad_norm": 6.061143228797713, + "learning_rate": 9.376704082872557e-06, + "loss": 0.5856, + "step": 7644 + }, + { + "epoch": 1.6030614384566997, + "grad_norm": 6.61247757902698, + "learning_rate": 9.37444461611892e-06, + "loss": 0.7368, + "step": 7645 + }, + { + "epoch": 1.6032711260222268, + "grad_norm": 7.695586660506961, + "learning_rate": 9.37218518142608e-06, + "loss": 0.9097, + "step": 7646 + }, + { + "epoch": 1.6034808135877543, + "grad_norm": 7.350334955326874, + "learning_rate": 9.369925778909834e-06, + "loss": 0.9832, + "step": 7647 + }, + { + "epoch": 1.6036905011532816, + "grad_norm": 7.966025178095759, + "learning_rate": 9.367666408685993e-06, + "loss": 0.9826, + "step": 7648 + }, + { + "epoch": 1.603900188718809, + "grad_norm": 9.168642622306578, + "learning_rate": 9.36540707087034e-06, + "loss": 0.9032, + "step": 7649 + }, + { + "epoch": 1.6041098762843364, + "grad_norm": 6.810058942069842, + "learning_rate": 9.363147765578677e-06, + "loss": 0.9252, + "step": 7650 + }, + { + "epoch": 1.6043195638498637, + "grad_norm": 7.16135585593105, + "learning_rate": 9.360888492926796e-06, + "loss": 0.6588, + "step": 7651 + }, + { + "epoch": 1.604529251415391, + "grad_norm": 6.741823606793392, + "learning_rate": 9.358629253030486e-06, + "loss": 0.8774, + "step": 7652 + }, + { + "epoch": 1.6047389389809186, + "grad_norm": 6.75213300538725, + "learning_rate": 9.35637004600554e-06, + "loss": 0.7712, + "step": 7653 + }, + { + "epoch": 1.6049486265464457, + "grad_norm": 6.249395764617356, + "learning_rate": 9.354110871967747e-06, + "loss": 0.6762, + "step": 7654 + }, + { + "epoch": 1.6051583141119732, + "grad_norm": 5.5447810821287655, + "learning_rate": 9.351851731032892e-06, + "loss": 0.6068, + "step": 7655 + }, + { + "epoch": 1.6053680016775005, + "grad_norm": 6.109286996690537, + "learning_rate": 9.349592623316758e-06, + "loss": 0.6901, + "step": 7656 + }, + { + "epoch": 1.6055776892430278, + "grad_norm": 8.614548385817152, + "learning_rate": 9.347333548935134e-06, + "loss": 0.8464, + "step": 7657 + }, + { + "epoch": 1.6057873768085553, + "grad_norm": 7.324001671689483, + "learning_rate": 9.34507450800379e-06, + "loss": 0.5875, + "step": 7658 + }, + { + "epoch": 1.6059970643740826, + "grad_norm": 8.369331799018477, + "learning_rate": 9.34281550063852e-06, + "loss": 0.8612, + "step": 7659 + }, + { + "epoch": 1.60620675193961, + "grad_norm": 11.079365352820734, + "learning_rate": 9.340556526955095e-06, + "loss": 1.053, + "step": 7660 + }, + { + "epoch": 1.6064164395051375, + "grad_norm": 6.492289935691345, + "learning_rate": 9.33829758706929e-06, + "loss": 0.8331, + "step": 7661 + }, + { + "epoch": 1.6066261270706645, + "grad_norm": 6.264903971811581, + "learning_rate": 9.336038681096878e-06, + "loss": 0.6872, + "step": 7662 + }, + { + "epoch": 1.606835814636192, + "grad_norm": 6.455794627783083, + "learning_rate": 9.333779809153644e-06, + "loss": 0.8051, + "step": 7663 + }, + { + "epoch": 1.6070455022017196, + "grad_norm": 7.462378459358794, + "learning_rate": 9.33152097135534e-06, + "loss": 0.871, + "step": 7664 + }, + { + "epoch": 1.6072551897672467, + "grad_norm": 7.138445784854093, + "learning_rate": 9.329262167817752e-06, + "loss": 0.8272, + "step": 7665 + }, + { + "epoch": 1.6074648773327742, + "grad_norm": 6.689626612675976, + "learning_rate": 9.327003398656642e-06, + "loss": 0.7404, + "step": 7666 + }, + { + "epoch": 1.6076745648983015, + "grad_norm": 7.39643293709871, + "learning_rate": 9.324744663987773e-06, + "loss": 0.943, + "step": 7667 + }, + { + "epoch": 1.6078842524638288, + "grad_norm": 7.011284641014348, + "learning_rate": 9.322485963926909e-06, + "loss": 0.7704, + "step": 7668 + }, + { + "epoch": 1.6080939400293564, + "grad_norm": 8.530566651447783, + "learning_rate": 9.320227298589819e-06, + "loss": 0.8044, + "step": 7669 + }, + { + "epoch": 1.6083036275948837, + "grad_norm": 5.936296176278467, + "learning_rate": 9.317968668092254e-06, + "loss": 0.6623, + "step": 7670 + }, + { + "epoch": 1.608513315160411, + "grad_norm": 5.9878794218950615, + "learning_rate": 9.31571007254998e-06, + "loss": 0.5952, + "step": 7671 + }, + { + "epoch": 1.6087230027259385, + "grad_norm": 6.8333450404266545, + "learning_rate": 9.313451512078755e-06, + "loss": 0.7791, + "step": 7672 + }, + { + "epoch": 1.6089326902914656, + "grad_norm": 8.583954802582612, + "learning_rate": 9.311192986794327e-06, + "loss": 0.9966, + "step": 7673 + }, + { + "epoch": 1.609142377856993, + "grad_norm": 7.44724023065807, + "learning_rate": 9.30893449681245e-06, + "loss": 1.1628, + "step": 7674 + }, + { + "epoch": 1.6093520654225204, + "grad_norm": 8.696459747858393, + "learning_rate": 9.306676042248883e-06, + "loss": 0.6441, + "step": 7675 + }, + { + "epoch": 1.6095617529880477, + "grad_norm": 6.42364527281743, + "learning_rate": 9.304417623219373e-06, + "loss": 0.7912, + "step": 7676 + }, + { + "epoch": 1.6097714405535752, + "grad_norm": 5.694052347505275, + "learning_rate": 9.302159239839663e-06, + "loss": 0.4556, + "step": 7677 + }, + { + "epoch": 1.6099811281191025, + "grad_norm": 6.234753324298042, + "learning_rate": 9.299900892225501e-06, + "loss": 0.7282, + "step": 7678 + }, + { + "epoch": 1.6101908156846299, + "grad_norm": 5.74387733448822, + "learning_rate": 9.297642580492636e-06, + "loss": 0.6228, + "step": 7679 + }, + { + "epoch": 1.6104005032501574, + "grad_norm": 8.161814630229511, + "learning_rate": 9.295384304756807e-06, + "loss": 0.856, + "step": 7680 + }, + { + "epoch": 1.6106101908156847, + "grad_norm": 7.547190616900892, + "learning_rate": 9.293126065133752e-06, + "loss": 0.7523, + "step": 7681 + }, + { + "epoch": 1.610819878381212, + "grad_norm": 7.194895647727988, + "learning_rate": 9.290867861739214e-06, + "loss": 0.7067, + "step": 7682 + }, + { + "epoch": 1.6110295659467395, + "grad_norm": 6.073590239822679, + "learning_rate": 9.288609694688926e-06, + "loss": 0.6457, + "step": 7683 + }, + { + "epoch": 1.6112392535122666, + "grad_norm": 5.904870026897326, + "learning_rate": 9.286351564098626e-06, + "loss": 0.6146, + "step": 7684 + }, + { + "epoch": 1.6114489410777941, + "grad_norm": 7.807972922699126, + "learning_rate": 9.284093470084049e-06, + "loss": 0.7418, + "step": 7685 + }, + { + "epoch": 1.6116586286433214, + "grad_norm": 6.635005892113704, + "learning_rate": 9.28183541276092e-06, + "loss": 0.6635, + "step": 7686 + }, + { + "epoch": 1.6118683162088487, + "grad_norm": 6.588851396234455, + "learning_rate": 9.279577392244971e-06, + "loss": 0.7607, + "step": 7687 + }, + { + "epoch": 1.6120780037743763, + "grad_norm": 6.254578303206539, + "learning_rate": 9.277319408651934e-06, + "loss": 0.708, + "step": 7688 + }, + { + "epoch": 1.6122876913399036, + "grad_norm": 7.459935497985122, + "learning_rate": 9.275061462097525e-06, + "loss": 1.0482, + "step": 7689 + }, + { + "epoch": 1.6124973789054309, + "grad_norm": 6.847227334735842, + "learning_rate": 9.272803552697477e-06, + "loss": 0.736, + "step": 7690 + }, + { + "epoch": 1.6127070664709584, + "grad_norm": 6.202760044137037, + "learning_rate": 9.27054568056751e-06, + "loss": 0.5946, + "step": 7691 + }, + { + "epoch": 1.6129167540364855, + "grad_norm": 6.186885385079247, + "learning_rate": 9.268287845823339e-06, + "loss": 0.7232, + "step": 7692 + }, + { + "epoch": 1.613126441602013, + "grad_norm": 6.712424454639081, + "learning_rate": 9.266030048580685e-06, + "loss": 0.821, + "step": 7693 + }, + { + "epoch": 1.6133361291675403, + "grad_norm": 8.081996553281089, + "learning_rate": 9.263772288955269e-06, + "loss": 1.1096, + "step": 7694 + }, + { + "epoch": 1.6135458167330676, + "grad_norm": 8.163160732650868, + "learning_rate": 9.261514567062792e-06, + "loss": 0.9729, + "step": 7695 + }, + { + "epoch": 1.6137555042985952, + "grad_norm": 6.819508470747861, + "learning_rate": 9.259256883018979e-06, + "loss": 0.7247, + "step": 7696 + }, + { + "epoch": 1.6139651918641225, + "grad_norm": 7.1199469147694625, + "learning_rate": 9.256999236939537e-06, + "loss": 0.862, + "step": 7697 + }, + { + "epoch": 1.6141748794296498, + "grad_norm": 7.3173584164444785, + "learning_rate": 9.25474162894017e-06, + "loss": 0.5913, + "step": 7698 + }, + { + "epoch": 1.6143845669951773, + "grad_norm": 6.337892406269405, + "learning_rate": 9.252484059136586e-06, + "loss": 0.5719, + "step": 7699 + }, + { + "epoch": 1.6145942545607046, + "grad_norm": 6.646045209862763, + "learning_rate": 9.250226527644493e-06, + "loss": 0.7902, + "step": 7700 + }, + { + "epoch": 1.614803942126232, + "grad_norm": 6.752093276979535, + "learning_rate": 9.24796903457959e-06, + "loss": 0.6803, + "step": 7701 + }, + { + "epoch": 1.6150136296917594, + "grad_norm": 7.643069592921631, + "learning_rate": 9.245711580057578e-06, + "loss": 0.7636, + "step": 7702 + }, + { + "epoch": 1.6152233172572865, + "grad_norm": 6.454717226082348, + "learning_rate": 9.243454164194158e-06, + "loss": 0.6316, + "step": 7703 + }, + { + "epoch": 1.615433004822814, + "grad_norm": 6.7170588879563615, + "learning_rate": 9.241196787105022e-06, + "loss": 0.7073, + "step": 7704 + }, + { + "epoch": 1.6156426923883414, + "grad_norm": 6.914804418408023, + "learning_rate": 9.238939448905865e-06, + "loss": 0.7462, + "step": 7705 + }, + { + "epoch": 1.6158523799538687, + "grad_norm": 6.18811509733816, + "learning_rate": 9.236682149712386e-06, + "loss": 0.6062, + "step": 7706 + }, + { + "epoch": 1.6160620675193962, + "grad_norm": 8.158439158376162, + "learning_rate": 9.234424889640266e-06, + "loss": 0.7235, + "step": 7707 + }, + { + "epoch": 1.6162717550849235, + "grad_norm": 6.435878491576857, + "learning_rate": 9.232167668805202e-06, + "loss": 0.7737, + "step": 7708 + }, + { + "epoch": 1.6164814426504508, + "grad_norm": 6.668656512325095, + "learning_rate": 9.229910487322876e-06, + "loss": 0.5984, + "step": 7709 + }, + { + "epoch": 1.6166911302159783, + "grad_norm": 6.845463158922423, + "learning_rate": 9.22765334530897e-06, + "loss": 0.6837, + "step": 7710 + }, + { + "epoch": 1.6169008177815054, + "grad_norm": 6.389256813563026, + "learning_rate": 9.225396242879171e-06, + "loss": 0.6051, + "step": 7711 + }, + { + "epoch": 1.617110505347033, + "grad_norm": 5.680118568209008, + "learning_rate": 9.223139180149161e-06, + "loss": 0.4734, + "step": 7712 + }, + { + "epoch": 1.6173201929125602, + "grad_norm": 9.2518072146989, + "learning_rate": 9.220882157234613e-06, + "loss": 1.0221, + "step": 7713 + }, + { + "epoch": 1.6175298804780875, + "grad_norm": 6.583159640906543, + "learning_rate": 9.218625174251203e-06, + "loss": 0.7512, + "step": 7714 + }, + { + "epoch": 1.617739568043615, + "grad_norm": 6.778190902839815, + "learning_rate": 9.216368231314616e-06, + "loss": 0.5849, + "step": 7715 + }, + { + "epoch": 1.6179492556091424, + "grad_norm": 6.1380748567697525, + "learning_rate": 9.21411132854051e-06, + "loss": 0.5401, + "step": 7716 + }, + { + "epoch": 1.6181589431746697, + "grad_norm": 7.266616727456601, + "learning_rate": 9.211854466044561e-06, + "loss": 0.6502, + "step": 7717 + }, + { + "epoch": 1.6183686307401972, + "grad_norm": 7.306867193447994, + "learning_rate": 9.209597643942442e-06, + "loss": 0.6686, + "step": 7718 + }, + { + "epoch": 1.6185783183057245, + "grad_norm": 8.986654667226633, + "learning_rate": 9.207340862349812e-06, + "loss": 0.5342, + "step": 7719 + }, + { + "epoch": 1.6187880058712518, + "grad_norm": 7.330866395700292, + "learning_rate": 9.205084121382334e-06, + "loss": 0.8546, + "step": 7720 + }, + { + "epoch": 1.6189976934367794, + "grad_norm": 8.72673438380192, + "learning_rate": 9.202827421155683e-06, + "loss": 0.8632, + "step": 7721 + }, + { + "epoch": 1.6192073810023064, + "grad_norm": 6.702695749450438, + "learning_rate": 9.200570761785501e-06, + "loss": 0.5613, + "step": 7722 + }, + { + "epoch": 1.619417068567834, + "grad_norm": 6.616371381299311, + "learning_rate": 9.198314143387455e-06, + "loss": 0.6279, + "step": 7723 + }, + { + "epoch": 1.6196267561333613, + "grad_norm": 6.369013747765007, + "learning_rate": 9.196057566077203e-06, + "loss": 0.7452, + "step": 7724 + }, + { + "epoch": 1.6198364436988886, + "grad_norm": 7.613138906783497, + "learning_rate": 9.193801029970393e-06, + "loss": 0.9167, + "step": 7725 + }, + { + "epoch": 1.620046131264416, + "grad_norm": 9.037988270623964, + "learning_rate": 9.191544535182676e-06, + "loss": 0.8354, + "step": 7726 + }, + { + "epoch": 1.6202558188299434, + "grad_norm": 8.455700302451687, + "learning_rate": 9.189288081829707e-06, + "loss": 0.9112, + "step": 7727 + }, + { + "epoch": 1.6204655063954707, + "grad_norm": 7.0709877123798766, + "learning_rate": 9.18703167002713e-06, + "loss": 0.82, + "step": 7728 + }, + { + "epoch": 1.6206751939609982, + "grad_norm": 5.833403692857144, + "learning_rate": 9.184775299890589e-06, + "loss": 0.6453, + "step": 7729 + }, + { + "epoch": 1.6208848815265253, + "grad_norm": 6.505698151080973, + "learning_rate": 9.18251897153573e-06, + "loss": 0.8114, + "step": 7730 + }, + { + "epoch": 1.6210945690920529, + "grad_norm": 7.211377931681865, + "learning_rate": 9.180262685078188e-06, + "loss": 0.7509, + "step": 7731 + }, + { + "epoch": 1.6213042566575802, + "grad_norm": 7.229646746529025, + "learning_rate": 9.178006440633606e-06, + "loss": 0.6695, + "step": 7732 + }, + { + "epoch": 1.6215139442231075, + "grad_norm": 8.54145256196707, + "learning_rate": 9.175750238317622e-06, + "loss": 1.0609, + "step": 7733 + }, + { + "epoch": 1.621723631788635, + "grad_norm": 7.923460238884291, + "learning_rate": 9.173494078245866e-06, + "loss": 0.918, + "step": 7734 + }, + { + "epoch": 1.6219333193541623, + "grad_norm": 6.130069673907095, + "learning_rate": 9.171237960533972e-06, + "loss": 0.5995, + "step": 7735 + }, + { + "epoch": 1.6221430069196896, + "grad_norm": 6.610556106485422, + "learning_rate": 9.168981885297567e-06, + "loss": 0.7412, + "step": 7736 + }, + { + "epoch": 1.6223526944852171, + "grad_norm": 6.537399381012326, + "learning_rate": 9.166725852652287e-06, + "loss": 0.6613, + "step": 7737 + }, + { + "epoch": 1.6225623820507444, + "grad_norm": 6.549146352386218, + "learning_rate": 9.16446986271375e-06, + "loss": 0.6692, + "step": 7738 + }, + { + "epoch": 1.6227720696162717, + "grad_norm": 7.297369259180637, + "learning_rate": 9.162213915597582e-06, + "loss": 0.8235, + "step": 7739 + }, + { + "epoch": 1.6229817571817993, + "grad_norm": 8.056141968987534, + "learning_rate": 9.159958011419406e-06, + "loss": 1.0264, + "step": 7740 + }, + { + "epoch": 1.6231914447473264, + "grad_norm": 9.394002331696816, + "learning_rate": 9.157702150294834e-06, + "loss": 1.0481, + "step": 7741 + }, + { + "epoch": 1.6234011323128539, + "grad_norm": 8.525512423961626, + "learning_rate": 9.15544633233949e-06, + "loss": 0.8341, + "step": 7742 + }, + { + "epoch": 1.6236108198783812, + "grad_norm": 7.818781009419329, + "learning_rate": 9.153190557668988e-06, + "loss": 0.8095, + "step": 7743 + }, + { + "epoch": 1.6238205074439085, + "grad_norm": 5.356602716697762, + "learning_rate": 9.150934826398937e-06, + "loss": 0.4621, + "step": 7744 + }, + { + "epoch": 1.624030195009436, + "grad_norm": 7.193194220431343, + "learning_rate": 9.148679138644947e-06, + "loss": 0.8949, + "step": 7745 + }, + { + "epoch": 1.6242398825749633, + "grad_norm": 7.872547340112933, + "learning_rate": 9.146423494522633e-06, + "loss": 0.9217, + "step": 7746 + }, + { + "epoch": 1.6244495701404906, + "grad_norm": 7.276877173815217, + "learning_rate": 9.144167894147589e-06, + "loss": 0.7924, + "step": 7747 + }, + { + "epoch": 1.6246592577060182, + "grad_norm": 6.776400272317756, + "learning_rate": 9.14191233763543e-06, + "loss": 0.6, + "step": 7748 + }, + { + "epoch": 1.6248689452715452, + "grad_norm": 7.008105977822801, + "learning_rate": 9.139656825101751e-06, + "loss": 0.8169, + "step": 7749 + }, + { + "epoch": 1.6250786328370728, + "grad_norm": 7.4436658429806295, + "learning_rate": 9.13740135666215e-06, + "loss": 0.8182, + "step": 7750 + }, + { + "epoch": 1.6252883204026, + "grad_norm": 6.69991438648829, + "learning_rate": 9.135145932432223e-06, + "loss": 0.5884, + "step": 7751 + }, + { + "epoch": 1.6254980079681274, + "grad_norm": 7.099049344858954, + "learning_rate": 9.132890552527575e-06, + "loss": 0.7609, + "step": 7752 + }, + { + "epoch": 1.625707695533655, + "grad_norm": 6.523643052980017, + "learning_rate": 9.130635217063783e-06, + "loss": 0.6455, + "step": 7753 + }, + { + "epoch": 1.6259173830991822, + "grad_norm": 9.535756779848574, + "learning_rate": 9.128379926156446e-06, + "loss": 1.0559, + "step": 7754 + }, + { + "epoch": 1.6261270706647095, + "grad_norm": 7.467107567296489, + "learning_rate": 9.126124679921153e-06, + "loss": 0.8032, + "step": 7755 + }, + { + "epoch": 1.626336758230237, + "grad_norm": 6.459242281077416, + "learning_rate": 9.123869478473482e-06, + "loss": 0.7315, + "step": 7756 + }, + { + "epoch": 1.6265464457957644, + "grad_norm": 8.208146292785123, + "learning_rate": 9.121614321929018e-06, + "loss": 0.8942, + "step": 7757 + }, + { + "epoch": 1.6267561333612917, + "grad_norm": 5.875277495235046, + "learning_rate": 9.119359210403347e-06, + "loss": 0.7302, + "step": 7758 + }, + { + "epoch": 1.6269658209268192, + "grad_norm": 6.658796999545085, + "learning_rate": 9.117104144012044e-06, + "loss": 0.5731, + "step": 7759 + }, + { + "epoch": 1.6271755084923463, + "grad_norm": 8.764316501335495, + "learning_rate": 9.114849122870683e-06, + "loss": 0.8873, + "step": 7760 + }, + { + "epoch": 1.6273851960578738, + "grad_norm": 5.514092865093092, + "learning_rate": 9.112594147094842e-06, + "loss": 0.6521, + "step": 7761 + }, + { + "epoch": 1.627594883623401, + "grad_norm": 7.532641461594116, + "learning_rate": 9.110339216800085e-06, + "loss": 0.9602, + "step": 7762 + }, + { + "epoch": 1.6278045711889284, + "grad_norm": 6.584313082856506, + "learning_rate": 9.108084332101989e-06, + "loss": 0.7568, + "step": 7763 + }, + { + "epoch": 1.628014258754456, + "grad_norm": 9.240860782331735, + "learning_rate": 9.105829493116122e-06, + "loss": 0.9935, + "step": 7764 + }, + { + "epoch": 1.6282239463199832, + "grad_norm": 7.217072751833694, + "learning_rate": 9.10357469995804e-06, + "loss": 0.7553, + "step": 7765 + }, + { + "epoch": 1.6284336338855105, + "grad_norm": 7.225607625019034, + "learning_rate": 9.10131995274331e-06, + "loss": 0.7728, + "step": 7766 + }, + { + "epoch": 1.628643321451038, + "grad_norm": 8.019962886801872, + "learning_rate": 9.099065251587493e-06, + "loss": 0.7569, + "step": 7767 + }, + { + "epoch": 1.6288530090165652, + "grad_norm": 6.579211521914319, + "learning_rate": 9.096810596606142e-06, + "loss": 0.6929, + "step": 7768 + }, + { + "epoch": 1.6290626965820927, + "grad_norm": 7.993381902048428, + "learning_rate": 9.094555987914814e-06, + "loss": 0.8417, + "step": 7769 + }, + { + "epoch": 1.6292723841476202, + "grad_norm": 6.976837203922693, + "learning_rate": 9.092301425629066e-06, + "loss": 0.8017, + "step": 7770 + }, + { + "epoch": 1.6294820717131473, + "grad_norm": 6.343367025367347, + "learning_rate": 9.09004690986444e-06, + "loss": 0.7216, + "step": 7771 + }, + { + "epoch": 1.6296917592786748, + "grad_norm": 7.341908328338591, + "learning_rate": 9.087792440736489e-06, + "loss": 0.6986, + "step": 7772 + }, + { + "epoch": 1.6299014468442021, + "grad_norm": 7.76041552743114, + "learning_rate": 9.085538018360762e-06, + "loss": 0.9403, + "step": 7773 + }, + { + "epoch": 1.6301111344097294, + "grad_norm": 7.1234580369252924, + "learning_rate": 9.08328364285279e-06, + "loss": 0.7101, + "step": 7774 + }, + { + "epoch": 1.630320821975257, + "grad_norm": 7.985365803806052, + "learning_rate": 9.081029314328126e-06, + "loss": 0.9688, + "step": 7775 + }, + { + "epoch": 1.6305305095407843, + "grad_norm": 6.228212972436947, + "learning_rate": 9.078775032902306e-06, + "loss": 0.6933, + "step": 7776 + }, + { + "epoch": 1.6307401971063116, + "grad_norm": 7.978440310520784, + "learning_rate": 9.07652079869086e-06, + "loss": 0.6719, + "step": 7777 + }, + { + "epoch": 1.630949884671839, + "grad_norm": 8.266884138862261, + "learning_rate": 9.07426661180932e-06, + "loss": 0.8062, + "step": 7778 + }, + { + "epoch": 1.6311595722373662, + "grad_norm": 8.028440388954458, + "learning_rate": 9.072012472373229e-06, + "loss": 0.7219, + "step": 7779 + }, + { + "epoch": 1.6313692598028937, + "grad_norm": 6.094528031504201, + "learning_rate": 9.069758380498106e-06, + "loss": 0.5652, + "step": 7780 + }, + { + "epoch": 1.631578947368421, + "grad_norm": 6.660577313405789, + "learning_rate": 9.067504336299478e-06, + "loss": 0.6145, + "step": 7781 + }, + { + "epoch": 1.6317886349339483, + "grad_norm": 7.12195376059287, + "learning_rate": 9.065250339892874e-06, + "loss": 0.842, + "step": 7782 + }, + { + "epoch": 1.6319983224994759, + "grad_norm": 7.566507908931353, + "learning_rate": 9.062996391393808e-06, + "loss": 0.9187, + "step": 7783 + }, + { + "epoch": 1.6322080100650032, + "grad_norm": 7.93613470835728, + "learning_rate": 9.060742490917802e-06, + "loss": 0.7541, + "step": 7784 + }, + { + "epoch": 1.6324176976305305, + "grad_norm": 8.833268826509867, + "learning_rate": 9.058488638580374e-06, + "loss": 0.9715, + "step": 7785 + }, + { + "epoch": 1.632627385196058, + "grad_norm": 5.663605550223333, + "learning_rate": 9.056234834497035e-06, + "loss": 0.5545, + "step": 7786 + }, + { + "epoch": 1.632837072761585, + "grad_norm": 5.908037337664748, + "learning_rate": 9.053981078783298e-06, + "loss": 0.5953, + "step": 7787 + }, + { + "epoch": 1.6330467603271126, + "grad_norm": 7.200108804549128, + "learning_rate": 9.051727371554672e-06, + "loss": 0.6641, + "step": 7788 + }, + { + "epoch": 1.6332564478926401, + "grad_norm": 6.320811345698257, + "learning_rate": 9.04947371292666e-06, + "loss": 0.6844, + "step": 7789 + }, + { + "epoch": 1.6334661354581672, + "grad_norm": 8.757750205533895, + "learning_rate": 9.047220103014772e-06, + "loss": 0.9923, + "step": 7790 + }, + { + "epoch": 1.6336758230236947, + "grad_norm": 9.12307464401014, + "learning_rate": 9.044966541934508e-06, + "loss": 0.947, + "step": 7791 + }, + { + "epoch": 1.633885510589222, + "grad_norm": 6.13456740024168, + "learning_rate": 9.042713029801361e-06, + "loss": 0.6018, + "step": 7792 + }, + { + "epoch": 1.6340951981547494, + "grad_norm": 8.043935077832542, + "learning_rate": 9.040459566730831e-06, + "loss": 0.991, + "step": 7793 + }, + { + "epoch": 1.6343048857202769, + "grad_norm": 6.425912823532984, + "learning_rate": 9.038206152838419e-06, + "loss": 0.5988, + "step": 7794 + }, + { + "epoch": 1.6345145732858042, + "grad_norm": 5.748848278214841, + "learning_rate": 9.035952788239604e-06, + "loss": 0.4612, + "step": 7795 + }, + { + "epoch": 1.6347242608513315, + "grad_norm": 7.991892382924009, + "learning_rate": 9.03369947304988e-06, + "loss": 0.8539, + "step": 7796 + }, + { + "epoch": 1.634933948416859, + "grad_norm": 6.419182170042177, + "learning_rate": 9.031446207384737e-06, + "loss": 0.6415, + "step": 7797 + }, + { + "epoch": 1.635143635982386, + "grad_norm": 6.95291576151036, + "learning_rate": 9.029192991359657e-06, + "loss": 0.671, + "step": 7798 + }, + { + "epoch": 1.6353533235479136, + "grad_norm": 8.695520854777458, + "learning_rate": 9.026939825090115e-06, + "loss": 0.905, + "step": 7799 + }, + { + "epoch": 1.635563011113441, + "grad_norm": 8.150198190082445, + "learning_rate": 9.024686708691596e-06, + "loss": 0.8433, + "step": 7800 + }, + { + "epoch": 1.6357726986789682, + "grad_norm": 8.242038531787703, + "learning_rate": 9.022433642279579e-06, + "loss": 0.8004, + "step": 7801 + }, + { + "epoch": 1.6359823862444958, + "grad_norm": 6.7422538807376, + "learning_rate": 9.02018062596953e-06, + "loss": 0.7693, + "step": 7802 + }, + { + "epoch": 1.636192073810023, + "grad_norm": 7.955059268544132, + "learning_rate": 9.017927659876922e-06, + "loss": 0.7599, + "step": 7803 + }, + { + "epoch": 1.6364017613755504, + "grad_norm": 9.334740973055053, + "learning_rate": 9.01567474411723e-06, + "loss": 0.9984, + "step": 7804 + }, + { + "epoch": 1.636611448941078, + "grad_norm": 6.549294965642516, + "learning_rate": 9.013421878805909e-06, + "loss": 0.7306, + "step": 7805 + }, + { + "epoch": 1.6368211365066052, + "grad_norm": 6.551145417504215, + "learning_rate": 9.01116906405843e-06, + "loss": 0.5998, + "step": 7806 + }, + { + "epoch": 1.6370308240721325, + "grad_norm": 9.172331068984027, + "learning_rate": 9.008916299990254e-06, + "loss": 0.68, + "step": 7807 + }, + { + "epoch": 1.63724051163766, + "grad_norm": 5.972490240531828, + "learning_rate": 9.006663586716832e-06, + "loss": 0.5711, + "step": 7808 + }, + { + "epoch": 1.6374501992031871, + "grad_norm": 7.842216548639763, + "learning_rate": 9.004410924353625e-06, + "loss": 0.7333, + "step": 7809 + }, + { + "epoch": 1.6376598867687147, + "grad_norm": 7.551651092716445, + "learning_rate": 9.002158313016088e-06, + "loss": 0.8457, + "step": 7810 + }, + { + "epoch": 1.637869574334242, + "grad_norm": 6.171420578132184, + "learning_rate": 8.999905752819667e-06, + "loss": 0.6926, + "step": 7811 + }, + { + "epoch": 1.6380792618997693, + "grad_norm": 6.638668594376376, + "learning_rate": 8.99765324387981e-06, + "loss": 0.6111, + "step": 7812 + }, + { + "epoch": 1.6382889494652968, + "grad_norm": 6.414472911385584, + "learning_rate": 8.995400786311963e-06, + "loss": 0.6862, + "step": 7813 + }, + { + "epoch": 1.638498637030824, + "grad_norm": 7.556447301478495, + "learning_rate": 8.993148380231567e-06, + "loss": 0.8061, + "step": 7814 + }, + { + "epoch": 1.6387083245963514, + "grad_norm": 6.4943448568081354, + "learning_rate": 8.990896025754062e-06, + "loss": 0.7759, + "step": 7815 + }, + { + "epoch": 1.638918012161879, + "grad_norm": 7.040814841287184, + "learning_rate": 8.988643722994889e-06, + "loss": 0.7806, + "step": 7816 + }, + { + "epoch": 1.639127699727406, + "grad_norm": 5.721088550447536, + "learning_rate": 8.986391472069477e-06, + "loss": 0.5134, + "step": 7817 + }, + { + "epoch": 1.6393373872929335, + "grad_norm": 6.803899185579097, + "learning_rate": 8.98413927309326e-06, + "loss": 0.5648, + "step": 7818 + }, + { + "epoch": 1.6395470748584609, + "grad_norm": 6.7845898536229186, + "learning_rate": 8.98188712618167e-06, + "loss": 0.5402, + "step": 7819 + }, + { + "epoch": 1.6397567624239882, + "grad_norm": 7.477447277540528, + "learning_rate": 8.979635031450127e-06, + "loss": 0.8508, + "step": 7820 + }, + { + "epoch": 1.6399664499895157, + "grad_norm": 6.224998123392957, + "learning_rate": 8.977382989014059e-06, + "loss": 0.5612, + "step": 7821 + }, + { + "epoch": 1.640176137555043, + "grad_norm": 7.121249784422646, + "learning_rate": 8.97513099898889e-06, + "loss": 0.8514, + "step": 7822 + }, + { + "epoch": 1.6403858251205703, + "grad_norm": 8.125016724587613, + "learning_rate": 8.972879061490033e-06, + "loss": 1.0115, + "step": 7823 + }, + { + "epoch": 1.6405955126860978, + "grad_norm": 6.954146328934536, + "learning_rate": 8.970627176632901e-06, + "loss": 0.6946, + "step": 7824 + }, + { + "epoch": 1.6408052002516251, + "grad_norm": 7.361509274333507, + "learning_rate": 8.96837534453292e-06, + "loss": 0.7038, + "step": 7825 + }, + { + "epoch": 1.6410148878171524, + "grad_norm": 5.9429767418405595, + "learning_rate": 8.966123565305484e-06, + "loss": 0.6317, + "step": 7826 + }, + { + "epoch": 1.64122457538268, + "grad_norm": 8.631053737205175, + "learning_rate": 8.963871839066013e-06, + "loss": 0.8529, + "step": 7827 + }, + { + "epoch": 1.641434262948207, + "grad_norm": 6.809005778731341, + "learning_rate": 8.961620165929908e-06, + "loss": 0.7271, + "step": 7828 + }, + { + "epoch": 1.6416439505137346, + "grad_norm": 6.470193564204594, + "learning_rate": 8.959368546012568e-06, + "loss": 0.5662, + "step": 7829 + }, + { + "epoch": 1.6418536380792619, + "grad_norm": 5.865134245963868, + "learning_rate": 8.957116979429392e-06, + "loss": 0.6353, + "step": 7830 + }, + { + "epoch": 1.6420633256447892, + "grad_norm": 8.82369511174877, + "learning_rate": 8.954865466295788e-06, + "loss": 0.8361, + "step": 7831 + }, + { + "epoch": 1.6422730132103167, + "grad_norm": 8.316480211840165, + "learning_rate": 8.952614006727134e-06, + "loss": 1.0116, + "step": 7832 + }, + { + "epoch": 1.642482700775844, + "grad_norm": 5.954328749446922, + "learning_rate": 8.95036260083883e-06, + "loss": 0.6046, + "step": 7833 + }, + { + "epoch": 1.6426923883413713, + "grad_norm": 6.9716871060910215, + "learning_rate": 8.948111248746267e-06, + "loss": 0.6077, + "step": 7834 + }, + { + "epoch": 1.6429020759068989, + "grad_norm": 7.462563305615036, + "learning_rate": 8.945859950564824e-06, + "loss": 0.7996, + "step": 7835 + }, + { + "epoch": 1.643111763472426, + "grad_norm": 7.119581058729402, + "learning_rate": 8.943608706409885e-06, + "loss": 0.7473, + "step": 7836 + }, + { + "epoch": 1.6433214510379535, + "grad_norm": 9.038190989241881, + "learning_rate": 8.941357516396836e-06, + "loss": 0.8691, + "step": 7837 + }, + { + "epoch": 1.6435311386034808, + "grad_norm": 7.31329374559849, + "learning_rate": 8.939106380641048e-06, + "loss": 0.8647, + "step": 7838 + }, + { + "epoch": 1.643740826169008, + "grad_norm": 7.319239433420918, + "learning_rate": 8.936855299257898e-06, + "loss": 0.8125, + "step": 7839 + }, + { + "epoch": 1.6439505137345356, + "grad_norm": 7.906977798819099, + "learning_rate": 8.934604272362762e-06, + "loss": 0.8577, + "step": 7840 + }, + { + "epoch": 1.644160201300063, + "grad_norm": 8.097594850024912, + "learning_rate": 8.932353300070999e-06, + "loss": 0.8985, + "step": 7841 + }, + { + "epoch": 1.6443698888655902, + "grad_norm": 7.569917875226165, + "learning_rate": 8.930102382497983e-06, + "loss": 0.8273, + "step": 7842 + }, + { + "epoch": 1.6445795764311177, + "grad_norm": 8.37187211832073, + "learning_rate": 8.927851519759078e-06, + "loss": 0.876, + "step": 7843 + }, + { + "epoch": 1.644789263996645, + "grad_norm": 7.03920992343352, + "learning_rate": 8.92560071196964e-06, + "loss": 0.8471, + "step": 7844 + }, + { + "epoch": 1.6449989515621724, + "grad_norm": 7.064214331672824, + "learning_rate": 8.923349959245028e-06, + "loss": 0.6916, + "step": 7845 + }, + { + "epoch": 1.6452086391276999, + "grad_norm": 5.956189127791883, + "learning_rate": 8.921099261700601e-06, + "loss": 0.8166, + "step": 7846 + }, + { + "epoch": 1.645418326693227, + "grad_norm": 7.408802102290926, + "learning_rate": 8.918848619451704e-06, + "loss": 0.7926, + "step": 7847 + }, + { + "epoch": 1.6456280142587545, + "grad_norm": 5.948108974255877, + "learning_rate": 8.916598032613692e-06, + "loss": 0.6077, + "step": 7848 + }, + { + "epoch": 1.6458377018242818, + "grad_norm": 6.374324558714478, + "learning_rate": 8.914347501301912e-06, + "loss": 0.6134, + "step": 7849 + }, + { + "epoch": 1.646047389389809, + "grad_norm": 7.76905898909541, + "learning_rate": 8.912097025631704e-06, + "loss": 0.8627, + "step": 7850 + }, + { + "epoch": 1.6462570769553366, + "grad_norm": 6.180386828779047, + "learning_rate": 8.909846605718409e-06, + "loss": 0.5631, + "step": 7851 + }, + { + "epoch": 1.646466764520864, + "grad_norm": 7.69337411880533, + "learning_rate": 8.907596241677372e-06, + "loss": 0.8367, + "step": 7852 + }, + { + "epoch": 1.6466764520863912, + "grad_norm": 7.013424175565773, + "learning_rate": 8.905345933623915e-06, + "loss": 0.7733, + "step": 7853 + }, + { + "epoch": 1.6468861396519188, + "grad_norm": 7.323874911566226, + "learning_rate": 8.903095681673379e-06, + "loss": 0.8333, + "step": 7854 + }, + { + "epoch": 1.6470958272174459, + "grad_norm": 7.066074133945021, + "learning_rate": 8.900845485941095e-06, + "loss": 0.833, + "step": 7855 + }, + { + "epoch": 1.6473055147829734, + "grad_norm": 6.9880044649366075, + "learning_rate": 8.898595346542384e-06, + "loss": 0.9081, + "step": 7856 + }, + { + "epoch": 1.6475152023485007, + "grad_norm": 6.7151457572736275, + "learning_rate": 8.89634526359257e-06, + "loss": 0.6958, + "step": 7857 + }, + { + "epoch": 1.647724889914028, + "grad_norm": 6.103046094916113, + "learning_rate": 8.894095237206977e-06, + "loss": 0.5631, + "step": 7858 + }, + { + "epoch": 1.6479345774795555, + "grad_norm": 6.198228014444002, + "learning_rate": 8.891845267500924e-06, + "loss": 0.5738, + "step": 7859 + }, + { + "epoch": 1.6481442650450828, + "grad_norm": 5.692642421466629, + "learning_rate": 8.88959535458972e-06, + "loss": 0.654, + "step": 7860 + }, + { + "epoch": 1.6483539526106101, + "grad_norm": 7.795451330499722, + "learning_rate": 8.88734549858868e-06, + "loss": 0.9011, + "step": 7861 + }, + { + "epoch": 1.6485636401761377, + "grad_norm": 6.438274814043322, + "learning_rate": 8.885095699613118e-06, + "loss": 0.6424, + "step": 7862 + }, + { + "epoch": 1.648773327741665, + "grad_norm": 6.772463335780769, + "learning_rate": 8.882845957778328e-06, + "loss": 0.8589, + "step": 7863 + }, + { + "epoch": 1.6489830153071923, + "grad_norm": 6.779361235484254, + "learning_rate": 8.880596273199622e-06, + "loss": 0.6972, + "step": 7864 + }, + { + "epoch": 1.6491927028727198, + "grad_norm": 5.502511403079243, + "learning_rate": 8.878346645992302e-06, + "loss": 0.5442, + "step": 7865 + }, + { + "epoch": 1.6494023904382469, + "grad_norm": 7.044055865927619, + "learning_rate": 8.876097076271659e-06, + "loss": 0.6622, + "step": 7866 + }, + { + "epoch": 1.6496120780037744, + "grad_norm": 6.417535907442338, + "learning_rate": 8.87384756415299e-06, + "loss": 0.6642, + "step": 7867 + }, + { + "epoch": 1.6498217655693017, + "grad_norm": 7.832673470524373, + "learning_rate": 8.871598109751588e-06, + "loss": 0.7935, + "step": 7868 + }, + { + "epoch": 1.650031453134829, + "grad_norm": 6.5190577643641365, + "learning_rate": 8.869348713182739e-06, + "loss": 0.7521, + "step": 7869 + }, + { + "epoch": 1.6502411407003565, + "grad_norm": 8.338501964790682, + "learning_rate": 8.86709937456173e-06, + "loss": 0.9217, + "step": 7870 + }, + { + "epoch": 1.6504508282658839, + "grad_norm": 7.140592566652326, + "learning_rate": 8.864850094003842e-06, + "loss": 0.8156, + "step": 7871 + }, + { + "epoch": 1.6506605158314112, + "grad_norm": 6.760609089181651, + "learning_rate": 8.862600871624356e-06, + "loss": 0.741, + "step": 7872 + }, + { + "epoch": 1.6508702033969387, + "grad_norm": 6.394354328295808, + "learning_rate": 8.860351707538547e-06, + "loss": 0.5869, + "step": 7873 + }, + { + "epoch": 1.6510798909624658, + "grad_norm": 6.4446379516652765, + "learning_rate": 8.858102601861693e-06, + "loss": 0.7013, + "step": 7874 + }, + { + "epoch": 1.6512895785279933, + "grad_norm": 6.310371926910119, + "learning_rate": 8.855853554709059e-06, + "loss": 0.7111, + "step": 7875 + }, + { + "epoch": 1.6514992660935206, + "grad_norm": 9.498116192142136, + "learning_rate": 8.853604566195913e-06, + "loss": 0.9019, + "step": 7876 + }, + { + "epoch": 1.651708953659048, + "grad_norm": 7.251788175737595, + "learning_rate": 8.851355636437526e-06, + "loss": 0.7464, + "step": 7877 + }, + { + "epoch": 1.6519186412245754, + "grad_norm": 7.409117399119542, + "learning_rate": 8.84910676554915e-06, + "loss": 0.9153, + "step": 7878 + }, + { + "epoch": 1.6521283287901027, + "grad_norm": 7.88757298620162, + "learning_rate": 8.84685795364605e-06, + "loss": 0.7746, + "step": 7879 + }, + { + "epoch": 1.65233801635563, + "grad_norm": 7.325593811656621, + "learning_rate": 8.844609200843482e-06, + "loss": 0.7974, + "step": 7880 + }, + { + "epoch": 1.6525477039211576, + "grad_norm": 8.29674548135122, + "learning_rate": 8.842360507256693e-06, + "loss": 0.8739, + "step": 7881 + }, + { + "epoch": 1.6527573914866849, + "grad_norm": 5.277429836228957, + "learning_rate": 8.840111873000935e-06, + "loss": 0.4822, + "step": 7882 + }, + { + "epoch": 1.6529670790522122, + "grad_norm": 6.434260100921237, + "learning_rate": 8.837863298191462e-06, + "loss": 0.6273, + "step": 7883 + }, + { + "epoch": 1.6531767666177397, + "grad_norm": 7.405569585497389, + "learning_rate": 8.835614782943504e-06, + "loss": 0.8478, + "step": 7884 + }, + { + "epoch": 1.6533864541832668, + "grad_norm": 6.809530668500229, + "learning_rate": 8.83336632737231e-06, + "loss": 0.6153, + "step": 7885 + }, + { + "epoch": 1.6535961417487943, + "grad_norm": 7.1272301049012174, + "learning_rate": 8.831117931593116e-06, + "loss": 0.7542, + "step": 7886 + }, + { + "epoch": 1.6538058293143216, + "grad_norm": 7.816309694049327, + "learning_rate": 8.828869595721155e-06, + "loss": 0.7419, + "step": 7887 + }, + { + "epoch": 1.654015516879849, + "grad_norm": 6.8610751167557416, + "learning_rate": 8.826621319871656e-06, + "loss": 0.7819, + "step": 7888 + }, + { + "epoch": 1.6542252044453765, + "grad_norm": 6.377532304688891, + "learning_rate": 8.824373104159852e-06, + "loss": 0.6748, + "step": 7889 + }, + { + "epoch": 1.6544348920109038, + "grad_norm": 6.063294558958377, + "learning_rate": 8.822124948700966e-06, + "loss": 0.5882, + "step": 7890 + }, + { + "epoch": 1.654644579576431, + "grad_norm": 5.341939592459933, + "learning_rate": 8.819876853610217e-06, + "loss": 0.5559, + "step": 7891 + }, + { + "epoch": 1.6548542671419586, + "grad_norm": 6.361732821406228, + "learning_rate": 8.81762881900283e-06, + "loss": 0.6792, + "step": 7892 + }, + { + "epoch": 1.6550639547074857, + "grad_norm": 5.8629369787132, + "learning_rate": 8.815380844994013e-06, + "loss": 0.5539, + "step": 7893 + }, + { + "epoch": 1.6552736422730132, + "grad_norm": 6.5251112158498685, + "learning_rate": 8.81313293169898e-06, + "loss": 0.8682, + "step": 7894 + }, + { + "epoch": 1.6554833298385407, + "grad_norm": 8.40042396561854, + "learning_rate": 8.810885079232949e-06, + "loss": 0.9727, + "step": 7895 + }, + { + "epoch": 1.6556930174040678, + "grad_norm": 7.516003102280768, + "learning_rate": 8.808637287711117e-06, + "loss": 0.8874, + "step": 7896 + }, + { + "epoch": 1.6559027049695954, + "grad_norm": 7.050681047044563, + "learning_rate": 8.80638955724869e-06, + "loss": 0.7266, + "step": 7897 + }, + { + "epoch": 1.6561123925351227, + "grad_norm": 7.833052915303682, + "learning_rate": 8.804141887960871e-06, + "loss": 0.9493, + "step": 7898 + }, + { + "epoch": 1.65632208010065, + "grad_norm": 7.195135434102008, + "learning_rate": 8.801894279962849e-06, + "loss": 0.7243, + "step": 7899 + }, + { + "epoch": 1.6565317676661775, + "grad_norm": 5.41166646842517, + "learning_rate": 8.799646733369827e-06, + "loss": 0.5467, + "step": 7900 + }, + { + "epoch": 1.6567414552317048, + "grad_norm": 5.9291125246206695, + "learning_rate": 8.797399248296992e-06, + "loss": 0.7297, + "step": 7901 + }, + { + "epoch": 1.656951142797232, + "grad_norm": 6.208039422140017, + "learning_rate": 8.795151824859532e-06, + "loss": 0.6777, + "step": 7902 + }, + { + "epoch": 1.6571608303627596, + "grad_norm": 7.577800940929109, + "learning_rate": 8.792904463172626e-06, + "loss": 0.9739, + "step": 7903 + }, + { + "epoch": 1.6573705179282867, + "grad_norm": 5.202438717515364, + "learning_rate": 8.790657163351467e-06, + "loss": 0.5284, + "step": 7904 + }, + { + "epoch": 1.6575802054938142, + "grad_norm": 6.611731298738552, + "learning_rate": 8.788409925511219e-06, + "loss": 0.6986, + "step": 7905 + }, + { + "epoch": 1.6577898930593415, + "grad_norm": 6.1960149729945195, + "learning_rate": 8.786162749767068e-06, + "loss": 0.6538, + "step": 7906 + }, + { + "epoch": 1.6579995806248689, + "grad_norm": 6.093209579388984, + "learning_rate": 8.783915636234182e-06, + "loss": 0.6627, + "step": 7907 + }, + { + "epoch": 1.6582092681903964, + "grad_norm": 6.654488490519788, + "learning_rate": 8.781668585027727e-06, + "loss": 0.6589, + "step": 7908 + }, + { + "epoch": 1.6584189557559237, + "grad_norm": 8.942024539762086, + "learning_rate": 8.779421596262867e-06, + "loss": 0.941, + "step": 7909 + }, + { + "epoch": 1.658628643321451, + "grad_norm": 6.604483984739474, + "learning_rate": 8.777174670054773e-06, + "loss": 0.7595, + "step": 7910 + }, + { + "epoch": 1.6588383308869785, + "grad_norm": 6.480259971865652, + "learning_rate": 8.774927806518594e-06, + "loss": 0.6982, + "step": 7911 + }, + { + "epoch": 1.6590480184525056, + "grad_norm": 7.291289020903068, + "learning_rate": 8.77268100576949e-06, + "loss": 0.8235, + "step": 7912 + }, + { + "epoch": 1.6592577060180331, + "grad_norm": 7.916026620836054, + "learning_rate": 8.770434267922616e-06, + "loss": 0.8941, + "step": 7913 + }, + { + "epoch": 1.6594673935835607, + "grad_norm": 6.96395612257874, + "learning_rate": 8.768187593093116e-06, + "loss": 0.5578, + "step": 7914 + }, + { + "epoch": 1.6596770811490877, + "grad_norm": 7.690320321652676, + "learning_rate": 8.765940981396137e-06, + "loss": 0.7181, + "step": 7915 + }, + { + "epoch": 1.6598867687146153, + "grad_norm": 6.978207996729425, + "learning_rate": 8.763694432946826e-06, + "loss": 0.7441, + "step": 7916 + }, + { + "epoch": 1.6600964562801426, + "grad_norm": 7.342379795369436, + "learning_rate": 8.761447947860318e-06, + "loss": 0.8486, + "step": 7917 + }, + { + "epoch": 1.6603061438456699, + "grad_norm": 8.185879480307776, + "learning_rate": 8.759201526251751e-06, + "loss": 1.0746, + "step": 7918 + }, + { + "epoch": 1.6605158314111974, + "grad_norm": 7.043380014140624, + "learning_rate": 8.756955168236257e-06, + "loss": 0.6542, + "step": 7919 + }, + { + "epoch": 1.6607255189767247, + "grad_norm": 8.837546697426431, + "learning_rate": 8.754708873928969e-06, + "loss": 0.8123, + "step": 7920 + }, + { + "epoch": 1.660935206542252, + "grad_norm": 6.402749627382423, + "learning_rate": 8.75246264344501e-06, + "loss": 0.6905, + "step": 7921 + }, + { + "epoch": 1.6611448941077795, + "grad_norm": 6.29802610229713, + "learning_rate": 8.750216476899502e-06, + "loss": 0.6896, + "step": 7922 + }, + { + "epoch": 1.6613545816733066, + "grad_norm": 7.59738095478557, + "learning_rate": 8.747970374407571e-06, + "loss": 0.8244, + "step": 7923 + }, + { + "epoch": 1.6615642692388342, + "grad_norm": 7.4601749187730855, + "learning_rate": 8.745724336084327e-06, + "loss": 0.8671, + "step": 7924 + }, + { + "epoch": 1.6617739568043615, + "grad_norm": 7.30553266498635, + "learning_rate": 8.743478362044885e-06, + "loss": 0.7412, + "step": 7925 + }, + { + "epoch": 1.6619836443698888, + "grad_norm": 7.285038067828108, + "learning_rate": 8.74123245240436e-06, + "loss": 0.5466, + "step": 7926 + }, + { + "epoch": 1.6621933319354163, + "grad_norm": 7.193093442804558, + "learning_rate": 8.738986607277855e-06, + "loss": 0.89, + "step": 7927 + }, + { + "epoch": 1.6624030195009436, + "grad_norm": 7.06242895499402, + "learning_rate": 8.73674082678047e-06, + "loss": 0.6801, + "step": 7928 + }, + { + "epoch": 1.662612707066471, + "grad_norm": 8.133991032846687, + "learning_rate": 8.734495111027314e-06, + "loss": 0.9388, + "step": 7929 + }, + { + "epoch": 1.6628223946319984, + "grad_norm": 6.684325985452376, + "learning_rate": 8.732249460133472e-06, + "loss": 0.5345, + "step": 7930 + }, + { + "epoch": 1.6630320821975257, + "grad_norm": 6.997096318411566, + "learning_rate": 8.730003874214047e-06, + "loss": 0.6719, + "step": 7931 + }, + { + "epoch": 1.663241769763053, + "grad_norm": 6.503134897674551, + "learning_rate": 8.727758353384127e-06, + "loss": 0.6121, + "step": 7932 + }, + { + "epoch": 1.6634514573285806, + "grad_norm": 6.640523560113847, + "learning_rate": 8.725512897758799e-06, + "loss": 0.6475, + "step": 7933 + }, + { + "epoch": 1.6636611448941077, + "grad_norm": 7.127556023431062, + "learning_rate": 8.72326750745314e-06, + "loss": 0.7375, + "step": 7934 + }, + { + "epoch": 1.6638708324596352, + "grad_norm": 8.384738109462614, + "learning_rate": 8.721022182582243e-06, + "loss": 0.9261, + "step": 7935 + }, + { + "epoch": 1.6640805200251625, + "grad_norm": 7.685506103557741, + "learning_rate": 8.718776923261172e-06, + "loss": 0.8266, + "step": 7936 + }, + { + "epoch": 1.6642902075906898, + "grad_norm": 7.400069930778782, + "learning_rate": 8.716531729605007e-06, + "loss": 0.8713, + "step": 7937 + }, + { + "epoch": 1.6644998951562173, + "grad_norm": 7.050206434113921, + "learning_rate": 8.71428660172882e-06, + "loss": 0.6644, + "step": 7938 + }, + { + "epoch": 1.6647095827217446, + "grad_norm": 6.728694507009341, + "learning_rate": 8.712041539747672e-06, + "loss": 0.7154, + "step": 7939 + }, + { + "epoch": 1.664919270287272, + "grad_norm": 5.800972377152287, + "learning_rate": 8.709796543776626e-06, + "loss": 0.6231, + "step": 7940 + }, + { + "epoch": 1.6651289578527995, + "grad_norm": 7.135230790133888, + "learning_rate": 8.707551613930753e-06, + "loss": 0.7436, + "step": 7941 + }, + { + "epoch": 1.6653386454183265, + "grad_norm": 7.279366435575045, + "learning_rate": 8.705306750325093e-06, + "loss": 0.8324, + "step": 7942 + }, + { + "epoch": 1.665548332983854, + "grad_norm": 6.87544505970825, + "learning_rate": 8.703061953074711e-06, + "loss": 0.8227, + "step": 7943 + }, + { + "epoch": 1.6657580205493814, + "grad_norm": 8.186348512899643, + "learning_rate": 8.700817222294655e-06, + "loss": 0.8456, + "step": 7944 + }, + { + "epoch": 1.6659677081149087, + "grad_norm": 8.495462602433276, + "learning_rate": 8.698572558099968e-06, + "loss": 0.7951, + "step": 7945 + }, + { + "epoch": 1.6661773956804362, + "grad_norm": 7.315423616878592, + "learning_rate": 8.696327960605692e-06, + "loss": 0.8457, + "step": 7946 + }, + { + "epoch": 1.6663870832459635, + "grad_norm": 6.937869589563069, + "learning_rate": 8.694083429926874e-06, + "loss": 0.8152, + "step": 7947 + }, + { + "epoch": 1.6665967708114908, + "grad_norm": 6.44761682817487, + "learning_rate": 8.691838966178542e-06, + "loss": 0.555, + "step": 7948 + }, + { + "epoch": 1.6668064583770184, + "grad_norm": 8.031872900507722, + "learning_rate": 8.689594569475732e-06, + "loss": 0.6499, + "step": 7949 + }, + { + "epoch": 1.6670161459425457, + "grad_norm": 8.077811337552681, + "learning_rate": 8.687350239933476e-06, + "loss": 0.8433, + "step": 7950 + }, + { + "epoch": 1.667225833508073, + "grad_norm": 6.540601450125432, + "learning_rate": 8.685105977666793e-06, + "loss": 0.6988, + "step": 7951 + }, + { + "epoch": 1.6674355210736005, + "grad_norm": 5.925902040233703, + "learning_rate": 8.68286178279071e-06, + "loss": 0.7093, + "step": 7952 + }, + { + "epoch": 1.6676452086391276, + "grad_norm": 7.539617421179522, + "learning_rate": 8.680617655420248e-06, + "loss": 0.7996, + "step": 7953 + }, + { + "epoch": 1.667854896204655, + "grad_norm": 9.000955619924227, + "learning_rate": 8.678373595670417e-06, + "loss": 1.0206, + "step": 7954 + }, + { + "epoch": 1.6680645837701824, + "grad_norm": 5.7631604169046495, + "learning_rate": 8.676129603656232e-06, + "loss": 0.5531, + "step": 7955 + }, + { + "epoch": 1.6682742713357097, + "grad_norm": 6.64776660053362, + "learning_rate": 8.673885679492702e-06, + "loss": 0.7323, + "step": 7956 + }, + { + "epoch": 1.6684839589012372, + "grad_norm": 6.363107573233188, + "learning_rate": 8.67164182329483e-06, + "loss": 0.7072, + "step": 7957 + }, + { + "epoch": 1.6686936464667645, + "grad_norm": 7.7207073421066275, + "learning_rate": 8.669398035177617e-06, + "loss": 0.9101, + "step": 7958 + }, + { + "epoch": 1.6689033340322919, + "grad_norm": 5.865041843001931, + "learning_rate": 8.667154315256066e-06, + "loss": 0.64, + "step": 7959 + }, + { + "epoch": 1.6691130215978194, + "grad_norm": 6.759500038707333, + "learning_rate": 8.664910663645166e-06, + "loss": 0.6285, + "step": 7960 + }, + { + "epoch": 1.6693227091633465, + "grad_norm": 8.33527194595454, + "learning_rate": 8.662667080459906e-06, + "loss": 1.1407, + "step": 7961 + }, + { + "epoch": 1.669532396728874, + "grad_norm": 6.72025731500222, + "learning_rate": 8.660423565815285e-06, + "loss": 0.821, + "step": 7962 + }, + { + "epoch": 1.6697420842944013, + "grad_norm": 6.974269472223837, + "learning_rate": 8.658180119826275e-06, + "loss": 0.6479, + "step": 7963 + }, + { + "epoch": 1.6699517718599286, + "grad_norm": 6.2427910447622015, + "learning_rate": 8.655936742607862e-06, + "loss": 0.7162, + "step": 7964 + }, + { + "epoch": 1.6701614594254561, + "grad_norm": 6.257374494710077, + "learning_rate": 8.653693434275024e-06, + "loss": 0.575, + "step": 7965 + }, + { + "epoch": 1.6703711469909834, + "grad_norm": 7.222025461798093, + "learning_rate": 8.651450194942732e-06, + "loss": 0.7048, + "step": 7966 + }, + { + "epoch": 1.6705808345565107, + "grad_norm": 6.783236509834643, + "learning_rate": 8.649207024725951e-06, + "loss": 0.6906, + "step": 7967 + }, + { + "epoch": 1.6707905221220383, + "grad_norm": 8.073130144813996, + "learning_rate": 8.64696392373966e-06, + "loss": 0.9674, + "step": 7968 + }, + { + "epoch": 1.6710002096875656, + "grad_norm": 7.533899201903426, + "learning_rate": 8.644720892098812e-06, + "loss": 0.7443, + "step": 7969 + }, + { + "epoch": 1.6712098972530929, + "grad_norm": 7.1771266238803735, + "learning_rate": 8.64247792991837e-06, + "loss": 0.7647, + "step": 7970 + }, + { + "epoch": 1.6714195848186204, + "grad_norm": 6.81172204387807, + "learning_rate": 8.640235037313289e-06, + "loss": 0.5971, + "step": 7971 + }, + { + "epoch": 1.6716292723841475, + "grad_norm": 8.73713069557914, + "learning_rate": 8.637992214398519e-06, + "loss": 0.9854, + "step": 7972 + }, + { + "epoch": 1.671838959949675, + "grad_norm": 5.940784803064217, + "learning_rate": 8.635749461289007e-06, + "loss": 0.5894, + "step": 7973 + }, + { + "epoch": 1.6720486475152023, + "grad_norm": 8.179758079792682, + "learning_rate": 8.633506778099708e-06, + "loss": 0.8216, + "step": 7974 + }, + { + "epoch": 1.6722583350807296, + "grad_norm": 8.146931451699574, + "learning_rate": 8.631264164945551e-06, + "loss": 0.8369, + "step": 7975 + }, + { + "epoch": 1.6724680226462572, + "grad_norm": 8.496235430840583, + "learning_rate": 8.629021621941482e-06, + "loss": 0.9198, + "step": 7976 + }, + { + "epoch": 1.6726777102117845, + "grad_norm": 6.521546106626131, + "learning_rate": 8.626779149202436e-06, + "loss": 0.7298, + "step": 7977 + }, + { + "epoch": 1.6728873977773118, + "grad_norm": 5.430366371788389, + "learning_rate": 8.624536746843334e-06, + "loss": 0.4953, + "step": 7978 + }, + { + "epoch": 1.6730970853428393, + "grad_norm": 9.3905806556264, + "learning_rate": 8.62229441497911e-06, + "loss": 0.8371, + "step": 7979 + }, + { + "epoch": 1.6733067729083664, + "grad_norm": 6.917407619280939, + "learning_rate": 8.62005215372469e-06, + "loss": 0.7659, + "step": 7980 + }, + { + "epoch": 1.673516460473894, + "grad_norm": 7.608628742628341, + "learning_rate": 8.617809963194992e-06, + "loss": 0.8316, + "step": 7981 + }, + { + "epoch": 1.6737261480394212, + "grad_norm": 6.708058395228349, + "learning_rate": 8.615567843504925e-06, + "loss": 0.7325, + "step": 7982 + }, + { + "epoch": 1.6739358356049485, + "grad_norm": 7.804047468555352, + "learning_rate": 8.613325794769411e-06, + "loss": 0.792, + "step": 7983 + }, + { + "epoch": 1.674145523170476, + "grad_norm": 6.997780450640977, + "learning_rate": 8.611083817103356e-06, + "loss": 0.724, + "step": 7984 + }, + { + "epoch": 1.6743552107360034, + "grad_norm": 6.9139058076484705, + "learning_rate": 8.608841910621662e-06, + "loss": 0.813, + "step": 7985 + }, + { + "epoch": 1.6745648983015307, + "grad_norm": 7.99163256020912, + "learning_rate": 8.606600075439234e-06, + "loss": 0.9029, + "step": 7986 + }, + { + "epoch": 1.6747745858670582, + "grad_norm": 8.114803515769998, + "learning_rate": 8.60435831167097e-06, + "loss": 0.779, + "step": 7987 + }, + { + "epoch": 1.6749842734325855, + "grad_norm": 8.015682362584512, + "learning_rate": 8.60211661943176e-06, + "loss": 0.8846, + "step": 7988 + }, + { + "epoch": 1.6751939609981128, + "grad_norm": 6.843713158650937, + "learning_rate": 8.599874998836499e-06, + "loss": 0.7526, + "step": 7989 + }, + { + "epoch": 1.6754036485636403, + "grad_norm": 8.072510770494533, + "learning_rate": 8.597633450000076e-06, + "loss": 0.9538, + "step": 7990 + }, + { + "epoch": 1.6756133361291674, + "grad_norm": 6.692364131144934, + "learning_rate": 8.595391973037368e-06, + "loss": 0.7711, + "step": 7991 + }, + { + "epoch": 1.675823023694695, + "grad_norm": 6.818287099161792, + "learning_rate": 8.593150568063255e-06, + "loss": 0.7255, + "step": 7992 + }, + { + "epoch": 1.6760327112602222, + "grad_norm": 5.995940372667456, + "learning_rate": 8.590909235192622e-06, + "loss": 0.5283, + "step": 7993 + }, + { + "epoch": 1.6762423988257495, + "grad_norm": 8.206040461476935, + "learning_rate": 8.58866797454033e-06, + "loss": 0.7554, + "step": 7994 + }, + { + "epoch": 1.676452086391277, + "grad_norm": 5.896634995020528, + "learning_rate": 8.586426786221251e-06, + "loss": 0.6032, + "step": 7995 + }, + { + "epoch": 1.6766617739568044, + "grad_norm": 5.49976187378121, + "learning_rate": 8.584185670350256e-06, + "loss": 0.6236, + "step": 7996 + }, + { + "epoch": 1.6768714615223317, + "grad_norm": 7.888853783740242, + "learning_rate": 8.581944627042198e-06, + "loss": 0.8424, + "step": 7997 + }, + { + "epoch": 1.6770811490878592, + "grad_norm": 5.6785441124012515, + "learning_rate": 8.579703656411933e-06, + "loss": 0.6689, + "step": 7998 + }, + { + "epoch": 1.6772908366533863, + "grad_norm": 6.4909018131165075, + "learning_rate": 8.577462758574327e-06, + "loss": 0.5982, + "step": 7999 + }, + { + "epoch": 1.6775005242189138, + "grad_norm": 6.80960107317987, + "learning_rate": 8.575221933644217e-06, + "loss": 0.7077, + "step": 8000 + }, + { + "epoch": 1.6777102117844411, + "grad_norm": 6.663749593843585, + "learning_rate": 8.572981181736455e-06, + "loss": 0.6556, + "step": 8001 + }, + { + "epoch": 1.6779198993499684, + "grad_norm": 6.371710797541373, + "learning_rate": 8.570740502965884e-06, + "loss": 0.5609, + "step": 8002 + }, + { + "epoch": 1.678129586915496, + "grad_norm": 5.893576674590227, + "learning_rate": 8.56849989744734e-06, + "loss": 0.6534, + "step": 8003 + }, + { + "epoch": 1.6783392744810233, + "grad_norm": 7.46259514234426, + "learning_rate": 8.566259365295656e-06, + "loss": 0.7833, + "step": 8004 + }, + { + "epoch": 1.6785489620465506, + "grad_norm": 5.492003604128599, + "learning_rate": 8.564018906625671e-06, + "loss": 0.5244, + "step": 8005 + }, + { + "epoch": 1.678758649612078, + "grad_norm": 5.807446988057185, + "learning_rate": 8.561778521552206e-06, + "loss": 0.6379, + "step": 8006 + }, + { + "epoch": 1.6789683371776054, + "grad_norm": 6.620500762795703, + "learning_rate": 8.559538210190086e-06, + "loss": 0.6097, + "step": 8007 + }, + { + "epoch": 1.6791780247431327, + "grad_norm": 7.835773203155217, + "learning_rate": 8.557297972654132e-06, + "loss": 0.8796, + "step": 8008 + }, + { + "epoch": 1.6793877123086602, + "grad_norm": 7.909440111544509, + "learning_rate": 8.555057809059157e-06, + "loss": 0.8656, + "step": 8009 + }, + { + "epoch": 1.6795973998741873, + "grad_norm": 6.48791835471679, + "learning_rate": 8.552817719519974e-06, + "loss": 0.7394, + "step": 8010 + }, + { + "epoch": 1.6798070874397149, + "grad_norm": 7.500786048208998, + "learning_rate": 8.550577704151399e-06, + "loss": 0.8444, + "step": 8011 + }, + { + "epoch": 1.6800167750052422, + "grad_norm": 5.251224581490754, + "learning_rate": 8.548337763068227e-06, + "loss": 0.5368, + "step": 8012 + }, + { + "epoch": 1.6802264625707695, + "grad_norm": 6.95714058500969, + "learning_rate": 8.546097896385259e-06, + "loss": 0.8292, + "step": 8013 + }, + { + "epoch": 1.680436150136297, + "grad_norm": 6.009253704792945, + "learning_rate": 8.543858104217304e-06, + "loss": 0.5852, + "step": 8014 + }, + { + "epoch": 1.6806458377018243, + "grad_norm": 8.533104719606097, + "learning_rate": 8.541618386679138e-06, + "loss": 0.9196, + "step": 8015 + }, + { + "epoch": 1.6808555252673516, + "grad_norm": 5.700445457075318, + "learning_rate": 8.539378743885564e-06, + "loss": 0.6881, + "step": 8016 + }, + { + "epoch": 1.6810652128328791, + "grad_norm": 5.359108099220697, + "learning_rate": 8.537139175951362e-06, + "loss": 0.5455, + "step": 8017 + }, + { + "epoch": 1.6812749003984062, + "grad_norm": 6.452091949045865, + "learning_rate": 8.534899682991315e-06, + "loss": 0.6682, + "step": 8018 + }, + { + "epoch": 1.6814845879639337, + "grad_norm": 5.767348710092066, + "learning_rate": 8.532660265120195e-06, + "loss": 0.6773, + "step": 8019 + }, + { + "epoch": 1.6816942755294613, + "grad_norm": 7.390001888198283, + "learning_rate": 8.530420922452792e-06, + "loss": 0.86, + "step": 8020 + }, + { + "epoch": 1.6819039630949884, + "grad_norm": 9.175739609780958, + "learning_rate": 8.528181655103856e-06, + "loss": 0.8896, + "step": 8021 + }, + { + "epoch": 1.6821136506605159, + "grad_norm": 7.130144749929262, + "learning_rate": 8.525942463188165e-06, + "loss": 0.8211, + "step": 8022 + }, + { + "epoch": 1.6823233382260432, + "grad_norm": 6.707265513893993, + "learning_rate": 8.523703346820482e-06, + "loss": 0.7068, + "step": 8023 + }, + { + "epoch": 1.6825330257915705, + "grad_norm": 8.058055461919542, + "learning_rate": 8.521464306115562e-06, + "loss": 0.993, + "step": 8024 + }, + { + "epoch": 1.682742713357098, + "grad_norm": 6.913546534591349, + "learning_rate": 8.519225341188157e-06, + "loss": 0.7853, + "step": 8025 + }, + { + "epoch": 1.6829524009226253, + "grad_norm": 7.280290277923408, + "learning_rate": 8.516986452153026e-06, + "loss": 0.6362, + "step": 8026 + }, + { + "epoch": 1.6831620884881526, + "grad_norm": 6.784915773492927, + "learning_rate": 8.514747639124912e-06, + "loss": 0.7801, + "step": 8027 + }, + { + "epoch": 1.6833717760536802, + "grad_norm": 5.865843341449276, + "learning_rate": 8.512508902218555e-06, + "loss": 0.619, + "step": 8028 + }, + { + "epoch": 1.6835814636192072, + "grad_norm": 7.540393017189529, + "learning_rate": 8.510270241548699e-06, + "loss": 0.8786, + "step": 8029 + }, + { + "epoch": 1.6837911511847348, + "grad_norm": 7.777346796871491, + "learning_rate": 8.508031657230072e-06, + "loss": 0.7917, + "step": 8030 + }, + { + "epoch": 1.684000838750262, + "grad_norm": 6.188350220507667, + "learning_rate": 8.505793149377415e-06, + "loss": 0.6743, + "step": 8031 + }, + { + "epoch": 1.6842105263157894, + "grad_norm": 5.722682114044694, + "learning_rate": 8.50355471810545e-06, + "loss": 0.7378, + "step": 8032 + }, + { + "epoch": 1.684420213881317, + "grad_norm": 6.69646675371456, + "learning_rate": 8.5013163635289e-06, + "loss": 0.6308, + "step": 8033 + }, + { + "epoch": 1.6846299014468442, + "grad_norm": 6.7043084921855955, + "learning_rate": 8.499078085762487e-06, + "loss": 0.8286, + "step": 8034 + }, + { + "epoch": 1.6848395890123715, + "grad_norm": 5.571066289421345, + "learning_rate": 8.496839884920926e-06, + "loss": 0.4789, + "step": 8035 + }, + { + "epoch": 1.685049276577899, + "grad_norm": 5.77492340856858, + "learning_rate": 8.494601761118923e-06, + "loss": 0.5743, + "step": 8036 + }, + { + "epoch": 1.6852589641434261, + "grad_norm": 6.59637512418809, + "learning_rate": 8.492363714471194e-06, + "loss": 0.628, + "step": 8037 + }, + { + "epoch": 1.6854686517089537, + "grad_norm": 6.301564052694155, + "learning_rate": 8.490125745092443e-06, + "loss": 0.7673, + "step": 8038 + }, + { + "epoch": 1.6856783392744812, + "grad_norm": 6.05285412749015, + "learning_rate": 8.487887853097363e-06, + "loss": 0.6635, + "step": 8039 + }, + { + "epoch": 1.6858880268400083, + "grad_norm": 6.835660195856901, + "learning_rate": 8.485650038600649e-06, + "loss": 0.8073, + "step": 8040 + }, + { + "epoch": 1.6860977144055358, + "grad_norm": 7.414508280602449, + "learning_rate": 8.483412301717002e-06, + "loss": 0.9519, + "step": 8041 + }, + { + "epoch": 1.686307401971063, + "grad_norm": 6.194657060471599, + "learning_rate": 8.481174642561105e-06, + "loss": 0.5978, + "step": 8042 + }, + { + "epoch": 1.6865170895365904, + "grad_norm": 5.684880897112377, + "learning_rate": 8.47893706124764e-06, + "loss": 0.6504, + "step": 8043 + }, + { + "epoch": 1.686726777102118, + "grad_norm": 8.617961456899883, + "learning_rate": 8.476699557891288e-06, + "loss": 0.9196, + "step": 8044 + }, + { + "epoch": 1.6869364646676452, + "grad_norm": 6.747033439753016, + "learning_rate": 8.47446213260673e-06, + "loss": 0.6528, + "step": 8045 + }, + { + "epoch": 1.6871461522331725, + "grad_norm": 7.674984968295738, + "learning_rate": 8.47222478550863e-06, + "loss": 0.765, + "step": 8046 + }, + { + "epoch": 1.6873558397987, + "grad_norm": 7.049999230413108, + "learning_rate": 8.469987516711661e-06, + "loss": 0.7092, + "step": 8047 + }, + { + "epoch": 1.6875655273642272, + "grad_norm": 7.657758874030204, + "learning_rate": 8.467750326330487e-06, + "loss": 0.6388, + "step": 8048 + }, + { + "epoch": 1.6877752149297547, + "grad_norm": 5.256210060462586, + "learning_rate": 8.465513214479763e-06, + "loss": 0.4683, + "step": 8049 + }, + { + "epoch": 1.687984902495282, + "grad_norm": 7.4660354131694735, + "learning_rate": 8.463276181274146e-06, + "loss": 0.7944, + "step": 8050 + }, + { + "epoch": 1.6881945900608093, + "grad_norm": 7.0439093485868325, + "learning_rate": 8.461039226828299e-06, + "loss": 0.6515, + "step": 8051 + }, + { + "epoch": 1.6884042776263368, + "grad_norm": 8.192802278177494, + "learning_rate": 8.458802351256853e-06, + "loss": 0.9139, + "step": 8052 + }, + { + "epoch": 1.6886139651918641, + "grad_norm": 6.061250995956041, + "learning_rate": 8.45656555467446e-06, + "loss": 0.4878, + "step": 8053 + }, + { + "epoch": 1.6888236527573914, + "grad_norm": 6.245641532759829, + "learning_rate": 8.454328837195764e-06, + "loss": 0.6864, + "step": 8054 + }, + { + "epoch": 1.689033340322919, + "grad_norm": 9.770464898491316, + "learning_rate": 8.452092198935393e-06, + "loss": 1.0368, + "step": 8055 + }, + { + "epoch": 1.6892430278884463, + "grad_norm": 7.477597614411314, + "learning_rate": 8.449855640007978e-06, + "loss": 0.6894, + "step": 8056 + }, + { + "epoch": 1.6894527154539736, + "grad_norm": 7.503843832476244, + "learning_rate": 8.447619160528157e-06, + "loss": 0.6995, + "step": 8057 + }, + { + "epoch": 1.689662403019501, + "grad_norm": 7.699534111686624, + "learning_rate": 8.445382760610543e-06, + "loss": 0.77, + "step": 8058 + }, + { + "epoch": 1.6898720905850282, + "grad_norm": 6.987653156604058, + "learning_rate": 8.443146440369759e-06, + "loss": 0.8355, + "step": 8059 + }, + { + "epoch": 1.6900817781505557, + "grad_norm": 6.038029520529165, + "learning_rate": 8.440910199920421e-06, + "loss": 0.6061, + "step": 8060 + }, + { + "epoch": 1.690291465716083, + "grad_norm": 5.481871602360362, + "learning_rate": 8.438674039377137e-06, + "loss": 0.596, + "step": 8061 + }, + { + "epoch": 1.6905011532816103, + "grad_norm": 6.977803674676321, + "learning_rate": 8.436437958854517e-06, + "loss": 0.7972, + "step": 8062 + }, + { + "epoch": 1.6907108408471379, + "grad_norm": 6.20600554704157, + "learning_rate": 8.434201958467166e-06, + "loss": 0.6285, + "step": 8063 + }, + { + "epoch": 1.6909205284126652, + "grad_norm": 4.880448201168151, + "learning_rate": 8.43196603832968e-06, + "loss": 0.5071, + "step": 8064 + }, + { + "epoch": 1.6911302159781925, + "grad_norm": 6.700645796645517, + "learning_rate": 8.429730198556651e-06, + "loss": 0.7009, + "step": 8065 + }, + { + "epoch": 1.69133990354372, + "grad_norm": 6.31705939585013, + "learning_rate": 8.427494439262678e-06, + "loss": 0.6697, + "step": 8066 + }, + { + "epoch": 1.691549591109247, + "grad_norm": 6.072743926745923, + "learning_rate": 8.425258760562336e-06, + "loss": 0.7793, + "step": 8067 + }, + { + "epoch": 1.6917592786747746, + "grad_norm": 6.5431877333021955, + "learning_rate": 8.423023162570217e-06, + "loss": 0.8613, + "step": 8068 + }, + { + "epoch": 1.691968966240302, + "grad_norm": 8.262423609252041, + "learning_rate": 8.420787645400898e-06, + "loss": 0.8091, + "step": 8069 + }, + { + "epoch": 1.6921786538058292, + "grad_norm": 6.918021527136952, + "learning_rate": 8.418552209168948e-06, + "loss": 0.9488, + "step": 8070 + }, + { + "epoch": 1.6923883413713567, + "grad_norm": 6.482965702566805, + "learning_rate": 8.41631685398894e-06, + "loss": 0.6523, + "step": 8071 + }, + { + "epoch": 1.692598028936884, + "grad_norm": 7.941002900960024, + "learning_rate": 8.414081579975446e-06, + "loss": 0.6936, + "step": 8072 + }, + { + "epoch": 1.6928077165024114, + "grad_norm": 6.19876960841199, + "learning_rate": 8.411846387243013e-06, + "loss": 0.6662, + "step": 8073 + }, + { + "epoch": 1.6930174040679389, + "grad_norm": 7.233879723289626, + "learning_rate": 8.40961127590621e-06, + "loss": 0.7804, + "step": 8074 + }, + { + "epoch": 1.6932270916334662, + "grad_norm": 6.347237426292081, + "learning_rate": 8.407376246079591e-06, + "loss": 0.6534, + "step": 8075 + }, + { + "epoch": 1.6934367791989935, + "grad_norm": 6.679985133310644, + "learning_rate": 8.405141297877698e-06, + "loss": 0.6696, + "step": 8076 + }, + { + "epoch": 1.693646466764521, + "grad_norm": 5.767368427162944, + "learning_rate": 8.40290643141508e-06, + "loss": 0.7408, + "step": 8077 + }, + { + "epoch": 1.693856154330048, + "grad_norm": 5.945581495864111, + "learning_rate": 8.40067164680628e-06, + "loss": 0.61, + "step": 8078 + }, + { + "epoch": 1.6940658418955756, + "grad_norm": 6.993805198967216, + "learning_rate": 8.398436944165831e-06, + "loss": 0.6648, + "step": 8079 + }, + { + "epoch": 1.694275529461103, + "grad_norm": 6.019221520052035, + "learning_rate": 8.396202323608265e-06, + "loss": 0.6475, + "step": 8080 + }, + { + "epoch": 1.6944852170266302, + "grad_norm": 10.342347916573448, + "learning_rate": 8.393967785248114e-06, + "loss": 1.0226, + "step": 8081 + }, + { + "epoch": 1.6946949045921578, + "grad_norm": 7.331218085012214, + "learning_rate": 8.391733329199897e-06, + "loss": 0.8291, + "step": 8082 + }, + { + "epoch": 1.694904592157685, + "grad_norm": 8.712958023710193, + "learning_rate": 8.389498955578135e-06, + "loss": 1.0205, + "step": 8083 + }, + { + "epoch": 1.6951142797232124, + "grad_norm": 8.40866670117006, + "learning_rate": 8.387264664497349e-06, + "loss": 0.9455, + "step": 8084 + }, + { + "epoch": 1.69532396728874, + "grad_norm": 6.971151785869194, + "learning_rate": 8.385030456072043e-06, + "loss": 0.7345, + "step": 8085 + }, + { + "epoch": 1.695533654854267, + "grad_norm": 6.580461537753359, + "learning_rate": 8.382796330416729e-06, + "loss": 0.6863, + "step": 8086 + }, + { + "epoch": 1.6957433424197945, + "grad_norm": 5.130742848978635, + "learning_rate": 8.380562287645907e-06, + "loss": 0.5778, + "step": 8087 + }, + { + "epoch": 1.6959530299853218, + "grad_norm": 5.535010613704116, + "learning_rate": 8.378328327874074e-06, + "loss": 0.4527, + "step": 8088 + }, + { + "epoch": 1.6961627175508491, + "grad_norm": 6.695417726617908, + "learning_rate": 8.376094451215731e-06, + "loss": 0.6171, + "step": 8089 + }, + { + "epoch": 1.6963724051163767, + "grad_norm": 6.692135483844437, + "learning_rate": 8.373860657785363e-06, + "loss": 0.7069, + "step": 8090 + }, + { + "epoch": 1.696582092681904, + "grad_norm": 6.143929887689125, + "learning_rate": 8.371626947697457e-06, + "loss": 0.5504, + "step": 8091 + }, + { + "epoch": 1.6967917802474313, + "grad_norm": 5.952562922628196, + "learning_rate": 8.369393321066491e-06, + "loss": 0.5548, + "step": 8092 + }, + { + "epoch": 1.6970014678129588, + "grad_norm": 9.527228428806245, + "learning_rate": 8.36715977800695e-06, + "loss": 0.9963, + "step": 8093 + }, + { + "epoch": 1.697211155378486, + "grad_norm": 5.945155094502981, + "learning_rate": 8.364926318633298e-06, + "loss": 0.6632, + "step": 8094 + }, + { + "epoch": 1.6974208429440134, + "grad_norm": 8.982557083600318, + "learning_rate": 8.36269294306001e-06, + "loss": 0.7392, + "step": 8095 + }, + { + "epoch": 1.697630530509541, + "grad_norm": 7.58690850666072, + "learning_rate": 8.36045965140155e-06, + "loss": 0.7448, + "step": 8096 + }, + { + "epoch": 1.697840218075068, + "grad_norm": 6.450839161702425, + "learning_rate": 8.358226443772376e-06, + "loss": 0.776, + "step": 8097 + }, + { + "epoch": 1.6980499056405955, + "grad_norm": 8.272157740066207, + "learning_rate": 8.35599332028694e-06, + "loss": 0.7227, + "step": 8098 + }, + { + "epoch": 1.6982595932061229, + "grad_norm": 8.34221366224279, + "learning_rate": 8.353760281059705e-06, + "loss": 0.8269, + "step": 8099 + }, + { + "epoch": 1.6984692807716502, + "grad_norm": 8.064196321986154, + "learning_rate": 8.351527326205104e-06, + "loss": 0.9194, + "step": 8100 + }, + { + "epoch": 1.6986789683371777, + "grad_norm": 7.122059137776797, + "learning_rate": 8.349294455837589e-06, + "loss": 0.6501, + "step": 8101 + }, + { + "epoch": 1.698888655902705, + "grad_norm": 6.2678046415889, + "learning_rate": 8.347061670071598e-06, + "loss": 0.5761, + "step": 8102 + }, + { + "epoch": 1.6990983434682323, + "grad_norm": 8.477230873653662, + "learning_rate": 8.344828969021563e-06, + "loss": 0.7552, + "step": 8103 + }, + { + "epoch": 1.6993080310337598, + "grad_norm": 6.366475115603839, + "learning_rate": 8.34259635280191e-06, + "loss": 0.7131, + "step": 8104 + }, + { + "epoch": 1.699517718599287, + "grad_norm": 6.732181579367404, + "learning_rate": 8.340363821527072e-06, + "loss": 0.7129, + "step": 8105 + }, + { + "epoch": 1.6997274061648144, + "grad_norm": 7.086881562592512, + "learning_rate": 8.338131375311468e-06, + "loss": 0.7122, + "step": 8106 + }, + { + "epoch": 1.6999370937303417, + "grad_norm": 8.21091843624606, + "learning_rate": 8.335899014269512e-06, + "loss": 0.7338, + "step": 8107 + }, + { + "epoch": 1.700146781295869, + "grad_norm": 6.414818131162577, + "learning_rate": 8.333666738515614e-06, + "loss": 0.5577, + "step": 8108 + }, + { + "epoch": 1.7003564688613966, + "grad_norm": 8.218206128148912, + "learning_rate": 8.331434548164193e-06, + "loss": 0.8518, + "step": 8109 + }, + { + "epoch": 1.7005661564269239, + "grad_norm": 7.091346840563901, + "learning_rate": 8.32920244332964e-06, + "loss": 0.7208, + "step": 8110 + }, + { + "epoch": 1.7007758439924512, + "grad_norm": 7.456917798361744, + "learning_rate": 8.32697042412636e-06, + "loss": 0.7951, + "step": 8111 + }, + { + "epoch": 1.7009855315579787, + "grad_norm": 7.007658690685822, + "learning_rate": 8.324738490668752e-06, + "loss": 0.9643, + "step": 8112 + }, + { + "epoch": 1.701195219123506, + "grad_norm": 8.779258400787134, + "learning_rate": 8.322506643071198e-06, + "loss": 0.9878, + "step": 8113 + }, + { + "epoch": 1.7014049066890333, + "grad_norm": 6.19855312425762, + "learning_rate": 8.320274881448087e-06, + "loss": 0.6184, + "step": 8114 + }, + { + "epoch": 1.7016145942545609, + "grad_norm": 8.37459822414262, + "learning_rate": 8.318043205913804e-06, + "loss": 0.933, + "step": 8115 + }, + { + "epoch": 1.701824281820088, + "grad_norm": 7.215723416654507, + "learning_rate": 8.315811616582724e-06, + "loss": 0.6966, + "step": 8116 + }, + { + "epoch": 1.7020339693856155, + "grad_norm": 8.40425324456381, + "learning_rate": 8.313580113569217e-06, + "loss": 1.0779, + "step": 8117 + }, + { + "epoch": 1.7022436569511428, + "grad_norm": 7.154046177409504, + "learning_rate": 8.311348696987658e-06, + "loss": 0.7226, + "step": 8118 + }, + { + "epoch": 1.70245334451667, + "grad_norm": 6.621859149944271, + "learning_rate": 8.309117366952402e-06, + "loss": 0.7242, + "step": 8119 + }, + { + "epoch": 1.7026630320821976, + "grad_norm": 6.535446153009052, + "learning_rate": 8.306886123577818e-06, + "loss": 0.8294, + "step": 8120 + }, + { + "epoch": 1.702872719647725, + "grad_norm": 7.997105941339061, + "learning_rate": 8.304654966978257e-06, + "loss": 0.8766, + "step": 8121 + }, + { + "epoch": 1.7030824072132522, + "grad_norm": 9.554788797922344, + "learning_rate": 8.302423897268067e-06, + "loss": 1.1145, + "step": 8122 + }, + { + "epoch": 1.7032920947787797, + "grad_norm": 6.7664162292072145, + "learning_rate": 8.300192914561596e-06, + "loss": 0.7203, + "step": 8123 + }, + { + "epoch": 1.7035017823443068, + "grad_norm": 6.436267864927759, + "learning_rate": 8.297962018973191e-06, + "loss": 0.7449, + "step": 8124 + }, + { + "epoch": 1.7037114699098344, + "grad_norm": 6.331257824372788, + "learning_rate": 8.295731210617179e-06, + "loss": 0.5993, + "step": 8125 + }, + { + "epoch": 1.7039211574753617, + "grad_norm": 5.517213105812385, + "learning_rate": 8.2935004896079e-06, + "loss": 0.6601, + "step": 8126 + }, + { + "epoch": 1.704130845040889, + "grad_norm": 7.7416230803384405, + "learning_rate": 8.291269856059682e-06, + "loss": 0.777, + "step": 8127 + }, + { + "epoch": 1.7043405326064165, + "grad_norm": 8.734991295828738, + "learning_rate": 8.289039310086849e-06, + "loss": 0.6993, + "step": 8128 + }, + { + "epoch": 1.7045502201719438, + "grad_norm": 6.777897873825169, + "learning_rate": 8.286808851803713e-06, + "loss": 0.653, + "step": 8129 + }, + { + "epoch": 1.704759907737471, + "grad_norm": 8.277010616846532, + "learning_rate": 8.284578481324602e-06, + "loss": 1.177, + "step": 8130 + }, + { + "epoch": 1.7049695953029986, + "grad_norm": 7.592793693392803, + "learning_rate": 8.282348198763813e-06, + "loss": 0.8078, + "step": 8131 + }, + { + "epoch": 1.705179282868526, + "grad_norm": 6.966209191661069, + "learning_rate": 8.280118004235661e-06, + "loss": 0.6955, + "step": 8132 + }, + { + "epoch": 1.7053889704340532, + "grad_norm": 7.5989210912196254, + "learning_rate": 8.277887897854445e-06, + "loss": 0.9461, + "step": 8133 + }, + { + "epoch": 1.7055986579995808, + "grad_norm": 7.206408451535142, + "learning_rate": 8.27565787973446e-06, + "loss": 0.624, + "step": 8134 + }, + { + "epoch": 1.7058083455651079, + "grad_norm": 7.383036104731705, + "learning_rate": 8.273427949989997e-06, + "loss": 0.7754, + "step": 8135 + }, + { + "epoch": 1.7060180331306354, + "grad_norm": 6.371152439583857, + "learning_rate": 8.27119810873535e-06, + "loss": 0.5685, + "step": 8136 + }, + { + "epoch": 1.7062277206961627, + "grad_norm": 6.657877386757704, + "learning_rate": 8.268968356084796e-06, + "loss": 0.6465, + "step": 8137 + }, + { + "epoch": 1.70643740826169, + "grad_norm": 8.939307188093435, + "learning_rate": 8.266738692152616e-06, + "loss": 0.9293, + "step": 8138 + }, + { + "epoch": 1.7066470958272175, + "grad_norm": 6.303317512184404, + "learning_rate": 8.264509117053087e-06, + "loss": 0.5717, + "step": 8139 + }, + { + "epoch": 1.7068567833927448, + "grad_norm": 7.035454710502382, + "learning_rate": 8.262279630900471e-06, + "loss": 0.9377, + "step": 8140 + }, + { + "epoch": 1.7070664709582721, + "grad_norm": 6.9056178319841, + "learning_rate": 8.260050233809041e-06, + "loss": 0.7148, + "step": 8141 + }, + { + "epoch": 1.7072761585237997, + "grad_norm": 6.948421428376395, + "learning_rate": 8.257820925893057e-06, + "loss": 0.8333, + "step": 8142 + }, + { + "epoch": 1.7074858460893267, + "grad_norm": 9.419293130135118, + "learning_rate": 8.255591707266768e-06, + "loss": 0.9382, + "step": 8143 + }, + { + "epoch": 1.7076955336548543, + "grad_norm": 7.973645708642495, + "learning_rate": 8.25336257804443e-06, + "loss": 0.8281, + "step": 8144 + }, + { + "epoch": 1.7079052212203818, + "grad_norm": 7.243946153051919, + "learning_rate": 8.251133538340292e-06, + "loss": 0.8125, + "step": 8145 + }, + { + "epoch": 1.7081149087859089, + "grad_norm": 7.524324331670648, + "learning_rate": 8.248904588268589e-06, + "loss": 1.016, + "step": 8146 + }, + { + "epoch": 1.7083245963514364, + "grad_norm": 7.6843625250358665, + "learning_rate": 8.246675727943566e-06, + "loss": 0.9363, + "step": 8147 + }, + { + "epoch": 1.7085342839169637, + "grad_norm": 7.174605121851173, + "learning_rate": 8.244446957479453e-06, + "loss": 0.6751, + "step": 8148 + }, + { + "epoch": 1.708743971482491, + "grad_norm": 5.679304368425895, + "learning_rate": 8.242218276990477e-06, + "loss": 0.5927, + "step": 8149 + }, + { + "epoch": 1.7089536590480185, + "grad_norm": 6.803803237682092, + "learning_rate": 8.23998968659086e-06, + "loss": 0.6211, + "step": 8150 + }, + { + "epoch": 1.7091633466135459, + "grad_norm": 5.493763154146191, + "learning_rate": 8.237761186394832e-06, + "loss": 0.6388, + "step": 8151 + }, + { + "epoch": 1.7093730341790732, + "grad_norm": 8.32410857997258, + "learning_rate": 8.235532776516594e-06, + "loss": 0.8094, + "step": 8152 + }, + { + "epoch": 1.7095827217446007, + "grad_norm": 5.994065547148515, + "learning_rate": 8.233304457070363e-06, + "loss": 0.7588, + "step": 8153 + }, + { + "epoch": 1.7097924093101278, + "grad_norm": 6.12817375365778, + "learning_rate": 8.231076228170343e-06, + "loss": 0.4896, + "step": 8154 + }, + { + "epoch": 1.7100020968756553, + "grad_norm": 6.115681697380893, + "learning_rate": 8.228848089930734e-06, + "loss": 0.527, + "step": 8155 + }, + { + "epoch": 1.7102117844411826, + "grad_norm": 7.440460060068651, + "learning_rate": 8.22662004246573e-06, + "loss": 0.7173, + "step": 8156 + }, + { + "epoch": 1.71042147200671, + "grad_norm": 6.8054841025024615, + "learning_rate": 8.22439208588953e-06, + "loss": 0.8208, + "step": 8157 + }, + { + "epoch": 1.7106311595722374, + "grad_norm": 7.56088015674912, + "learning_rate": 8.222164220316311e-06, + "loss": 0.7341, + "step": 8158 + }, + { + "epoch": 1.7108408471377647, + "grad_norm": 8.725436202836754, + "learning_rate": 8.21993644586026e-06, + "loss": 1.0063, + "step": 8159 + }, + { + "epoch": 1.711050534703292, + "grad_norm": 6.133455701981651, + "learning_rate": 8.217708762635556e-06, + "loss": 0.5859, + "step": 8160 + }, + { + "epoch": 1.7112602222688196, + "grad_norm": 7.3899053133477, + "learning_rate": 8.215481170756366e-06, + "loss": 0.896, + "step": 8161 + }, + { + "epoch": 1.7114699098343467, + "grad_norm": 6.244954221894262, + "learning_rate": 8.21325367033686e-06, + "loss": 0.6577, + "step": 8162 + }, + { + "epoch": 1.7116795973998742, + "grad_norm": 5.8680274880511245, + "learning_rate": 8.211026261491206e-06, + "loss": 0.7622, + "step": 8163 + }, + { + "epoch": 1.7118892849654017, + "grad_norm": 6.138147885366866, + "learning_rate": 8.208798944333556e-06, + "loss": 0.6731, + "step": 8164 + }, + { + "epoch": 1.7120989725309288, + "grad_norm": 6.348901416135137, + "learning_rate": 8.206571718978069e-06, + "loss": 0.6636, + "step": 8165 + }, + { + "epoch": 1.7123086600964563, + "grad_norm": 7.284893736187619, + "learning_rate": 8.204344585538888e-06, + "loss": 0.8092, + "step": 8166 + }, + { + "epoch": 1.7125183476619836, + "grad_norm": 6.449792419083231, + "learning_rate": 8.202117544130168e-06, + "loss": 0.6986, + "step": 8167 + }, + { + "epoch": 1.712728035227511, + "grad_norm": 6.627422944022606, + "learning_rate": 8.199890594866038e-06, + "loss": 0.7195, + "step": 8168 + }, + { + "epoch": 1.7129377227930385, + "grad_norm": 7.449214401703932, + "learning_rate": 8.197663737860638e-06, + "loss": 0.9217, + "step": 8169 + }, + { + "epoch": 1.7131474103585658, + "grad_norm": 7.430350347788243, + "learning_rate": 8.195436973228098e-06, + "loss": 0.8214, + "step": 8170 + }, + { + "epoch": 1.713357097924093, + "grad_norm": 6.325582353141731, + "learning_rate": 8.193210301082539e-06, + "loss": 0.635, + "step": 8171 + }, + { + "epoch": 1.7135667854896206, + "grad_norm": 8.111513429633575, + "learning_rate": 8.190983721538088e-06, + "loss": 0.7857, + "step": 8172 + }, + { + "epoch": 1.7137764730551477, + "grad_norm": 7.221307433607994, + "learning_rate": 8.188757234708863e-06, + "loss": 0.655, + "step": 8173 + }, + { + "epoch": 1.7139861606206752, + "grad_norm": 6.304837902808786, + "learning_rate": 8.186530840708968e-06, + "loss": 0.6685, + "step": 8174 + }, + { + "epoch": 1.7141958481862025, + "grad_norm": 6.126215867899674, + "learning_rate": 8.184304539652512e-06, + "loss": 0.6496, + "step": 8175 + }, + { + "epoch": 1.7144055357517298, + "grad_norm": 7.586864382644645, + "learning_rate": 8.1820783316536e-06, + "loss": 0.8729, + "step": 8176 + }, + { + "epoch": 1.7146152233172574, + "grad_norm": 5.807782788030051, + "learning_rate": 8.179852216826321e-06, + "loss": 0.6608, + "step": 8177 + }, + { + "epoch": 1.7148249108827847, + "grad_norm": 5.998664136034808, + "learning_rate": 8.177626195284776e-06, + "loss": 0.5823, + "step": 8178 + }, + { + "epoch": 1.715034598448312, + "grad_norm": 8.617672837822257, + "learning_rate": 8.175400267143053e-06, + "loss": 0.9128, + "step": 8179 + }, + { + "epoch": 1.7152442860138395, + "grad_norm": 6.285258396491024, + "learning_rate": 8.173174432515227e-06, + "loss": 0.6837, + "step": 8180 + }, + { + "epoch": 1.7154539735793668, + "grad_norm": 7.085117029459445, + "learning_rate": 8.170948691515379e-06, + "loss": 0.6301, + "step": 8181 + }, + { + "epoch": 1.715663661144894, + "grad_norm": 8.861707047417154, + "learning_rate": 8.168723044257587e-06, + "loss": 0.8713, + "step": 8182 + }, + { + "epoch": 1.7158733487104216, + "grad_norm": 10.404968662891305, + "learning_rate": 8.16649749085591e-06, + "loss": 1.0908, + "step": 8183 + }, + { + "epoch": 1.7160830362759487, + "grad_norm": 7.802812799100203, + "learning_rate": 8.164272031424418e-06, + "loss": 0.7555, + "step": 8184 + }, + { + "epoch": 1.7162927238414762, + "grad_norm": 5.549380862511597, + "learning_rate": 8.162046666077172e-06, + "loss": 0.6243, + "step": 8185 + }, + { + "epoch": 1.7165024114070035, + "grad_norm": 6.895903588544072, + "learning_rate": 8.15982139492822e-06, + "loss": 0.6109, + "step": 8186 + }, + { + "epoch": 1.7167120989725309, + "grad_norm": 6.02676419301265, + "learning_rate": 8.157596218091611e-06, + "loss": 0.6135, + "step": 8187 + }, + { + "epoch": 1.7169217865380584, + "grad_norm": 6.151604225126431, + "learning_rate": 8.155371135681394e-06, + "loss": 0.6399, + "step": 8188 + }, + { + "epoch": 1.7171314741035857, + "grad_norm": 8.367767517571766, + "learning_rate": 8.153146147811604e-06, + "loss": 0.8615, + "step": 8189 + }, + { + "epoch": 1.717341161669113, + "grad_norm": 7.191671002030148, + "learning_rate": 8.150921254596279e-06, + "loss": 0.9028, + "step": 8190 + }, + { + "epoch": 1.7175508492346405, + "grad_norm": 8.010597007260875, + "learning_rate": 8.148696456149446e-06, + "loss": 0.825, + "step": 8191 + }, + { + "epoch": 1.7177605368001676, + "grad_norm": 6.678832605098482, + "learning_rate": 8.14647175258513e-06, + "loss": 0.5557, + "step": 8192 + }, + { + "epoch": 1.7179702243656951, + "grad_norm": 8.134515799250186, + "learning_rate": 8.144247144017349e-06, + "loss": 0.7169, + "step": 8193 + }, + { + "epoch": 1.7181799119312224, + "grad_norm": 7.247201982357617, + "learning_rate": 8.142022630560125e-06, + "loss": 0.815, + "step": 8194 + }, + { + "epoch": 1.7183895994967497, + "grad_norm": 6.682812000129259, + "learning_rate": 8.13979821232746e-06, + "loss": 0.6736, + "step": 8195 + }, + { + "epoch": 1.7185992870622773, + "grad_norm": 6.322571647763374, + "learning_rate": 8.137573889433366e-06, + "loss": 0.6438, + "step": 8196 + }, + { + "epoch": 1.7188089746278046, + "grad_norm": 7.62503480365893, + "learning_rate": 8.135349661991841e-06, + "loss": 0.7344, + "step": 8197 + }, + { + "epoch": 1.7190186621933319, + "grad_norm": 5.639005474413709, + "learning_rate": 8.133125530116875e-06, + "loss": 0.6475, + "step": 8198 + }, + { + "epoch": 1.7192283497588594, + "grad_norm": 5.57458800454167, + "learning_rate": 8.130901493922467e-06, + "loss": 0.5328, + "step": 8199 + }, + { + "epoch": 1.7194380373243867, + "grad_norm": 7.614181660668505, + "learning_rate": 8.1286775535226e-06, + "loss": 0.7696, + "step": 8200 + }, + { + "epoch": 1.719647724889914, + "grad_norm": 8.524163589792032, + "learning_rate": 8.126453709031252e-06, + "loss": 0.8641, + "step": 8201 + }, + { + "epoch": 1.7198574124554415, + "grad_norm": 7.369601194228133, + "learning_rate": 8.1242299605624e-06, + "loss": 0.9053, + "step": 8202 + }, + { + "epoch": 1.7200671000209686, + "grad_norm": 6.542954888542638, + "learning_rate": 8.12200630823002e-06, + "loss": 0.6714, + "step": 8203 + }, + { + "epoch": 1.7202767875864962, + "grad_norm": 7.10957047605065, + "learning_rate": 8.11978275214807e-06, + "loss": 0.6898, + "step": 8204 + }, + { + "epoch": 1.7204864751520235, + "grad_norm": 5.81061884923644, + "learning_rate": 8.117559292430516e-06, + "loss": 0.5302, + "step": 8205 + }, + { + "epoch": 1.7206961627175508, + "grad_norm": 7.28781818100964, + "learning_rate": 8.115335929191315e-06, + "loss": 0.7449, + "step": 8206 + }, + { + "epoch": 1.7209058502830783, + "grad_norm": 5.916880409329518, + "learning_rate": 8.113112662544416e-06, + "loss": 0.5273, + "step": 8207 + }, + { + "epoch": 1.7211155378486056, + "grad_norm": 5.7786705332591035, + "learning_rate": 8.110889492603761e-06, + "loss": 0.6454, + "step": 8208 + }, + { + "epoch": 1.721325225414133, + "grad_norm": 5.786587315184699, + "learning_rate": 8.108666419483304e-06, + "loss": 0.6075, + "step": 8209 + }, + { + "epoch": 1.7215349129796604, + "grad_norm": 8.346565556440467, + "learning_rate": 8.106443443296968e-06, + "loss": 0.7603, + "step": 8210 + }, + { + "epoch": 1.7217446005451875, + "grad_norm": 9.860244573205568, + "learning_rate": 8.104220564158689e-06, + "loss": 0.832, + "step": 8211 + }, + { + "epoch": 1.721954288110715, + "grad_norm": 7.480555248009403, + "learning_rate": 8.101997782182399e-06, + "loss": 0.7123, + "step": 8212 + }, + { + "epoch": 1.7221639756762424, + "grad_norm": 7.180737630710306, + "learning_rate": 8.099775097482011e-06, + "loss": 0.8537, + "step": 8213 + }, + { + "epoch": 1.7223736632417697, + "grad_norm": 7.95985565892812, + "learning_rate": 8.097552510171443e-06, + "loss": 0.8298, + "step": 8214 + }, + { + "epoch": 1.7225833508072972, + "grad_norm": 7.272196253291843, + "learning_rate": 8.095330020364614e-06, + "loss": 0.7509, + "step": 8215 + }, + { + "epoch": 1.7227930383728245, + "grad_norm": 5.468905308425477, + "learning_rate": 8.09310762817542e-06, + "loss": 0.425, + "step": 8216 + }, + { + "epoch": 1.7230027259383518, + "grad_norm": 8.797484365560411, + "learning_rate": 8.090885333717769e-06, + "loss": 0.916, + "step": 8217 + }, + { + "epoch": 1.7232124135038793, + "grad_norm": 6.589387539124437, + "learning_rate": 8.088663137105559e-06, + "loss": 0.6779, + "step": 8218 + }, + { + "epoch": 1.7234221010694066, + "grad_norm": 9.934658094397228, + "learning_rate": 8.086441038452673e-06, + "loss": 1.254, + "step": 8219 + }, + { + "epoch": 1.723631788634934, + "grad_norm": 8.003808458455985, + "learning_rate": 8.084219037873005e-06, + "loss": 0.8663, + "step": 8220 + }, + { + "epoch": 1.7238414762004615, + "grad_norm": 6.405511111005282, + "learning_rate": 8.081997135480436e-06, + "loss": 0.7108, + "step": 8221 + }, + { + "epoch": 1.7240511637659885, + "grad_norm": 6.5854792333656995, + "learning_rate": 8.079775331388839e-06, + "loss": 0.7394, + "step": 8222 + }, + { + "epoch": 1.724260851331516, + "grad_norm": 6.735144110736056, + "learning_rate": 8.077553625712086e-06, + "loss": 0.7217, + "step": 8223 + }, + { + "epoch": 1.7244705388970434, + "grad_norm": 6.845757268998984, + "learning_rate": 8.075332018564046e-06, + "loss": 0.6625, + "step": 8224 + }, + { + "epoch": 1.7246802264625707, + "grad_norm": 6.672039650844115, + "learning_rate": 8.073110510058573e-06, + "loss": 0.7037, + "step": 8225 + }, + { + "epoch": 1.7248899140280982, + "grad_norm": 6.856333196103953, + "learning_rate": 8.070889100309532e-06, + "loss": 0.7383, + "step": 8226 + }, + { + "epoch": 1.7250996015936255, + "grad_norm": 9.615850524176137, + "learning_rate": 8.06866778943077e-06, + "loss": 1.1337, + "step": 8227 + }, + { + "epoch": 1.7253092891591528, + "grad_norm": 8.583647409102799, + "learning_rate": 8.066446577536137e-06, + "loss": 0.8598, + "step": 8228 + }, + { + "epoch": 1.7255189767246804, + "grad_norm": 4.539921241678876, + "learning_rate": 8.064225464739465e-06, + "loss": 0.504, + "step": 8229 + }, + { + "epoch": 1.7257286642902074, + "grad_norm": 7.949320314025649, + "learning_rate": 8.062004451154597e-06, + "loss": 0.871, + "step": 8230 + }, + { + "epoch": 1.725938351855735, + "grad_norm": 6.1514283000213466, + "learning_rate": 8.059783536895365e-06, + "loss": 0.6817, + "step": 8231 + }, + { + "epoch": 1.7261480394212623, + "grad_norm": 6.3662705175084415, + "learning_rate": 8.057562722075589e-06, + "loss": 0.7264, + "step": 8232 + }, + { + "epoch": 1.7263577269867896, + "grad_norm": 7.4341833140923095, + "learning_rate": 8.055342006809092e-06, + "loss": 0.6921, + "step": 8233 + }, + { + "epoch": 1.726567414552317, + "grad_norm": 7.01816945301242, + "learning_rate": 8.053121391209695e-06, + "loss": 0.8566, + "step": 8234 + }, + { + "epoch": 1.7267771021178444, + "grad_norm": 5.755900544814434, + "learning_rate": 8.050900875391196e-06, + "loss": 0.5714, + "step": 8235 + }, + { + "epoch": 1.7269867896833717, + "grad_norm": 7.229493471604383, + "learning_rate": 8.048680459467413e-06, + "loss": 0.8644, + "step": 8236 + }, + { + "epoch": 1.7271964772488992, + "grad_norm": 5.782282972841823, + "learning_rate": 8.046460143552141e-06, + "loss": 0.6117, + "step": 8237 + }, + { + "epoch": 1.7274061648144265, + "grad_norm": 6.481540888584115, + "learning_rate": 8.044239927759172e-06, + "loss": 0.8677, + "step": 8238 + }, + { + "epoch": 1.7276158523799539, + "grad_norm": 7.404074673348606, + "learning_rate": 8.042019812202297e-06, + "loss": 0.7362, + "step": 8239 + }, + { + "epoch": 1.7278255399454814, + "grad_norm": 7.867787277197765, + "learning_rate": 8.03979979699531e-06, + "loss": 0.7477, + "step": 8240 + }, + { + "epoch": 1.7280352275110085, + "grad_norm": 6.2950031051578845, + "learning_rate": 8.037579882251977e-06, + "loss": 0.6074, + "step": 8241 + }, + { + "epoch": 1.728244915076536, + "grad_norm": 5.675253449097033, + "learning_rate": 8.03536006808608e-06, + "loss": 0.4734, + "step": 8242 + }, + { + "epoch": 1.7284546026420633, + "grad_norm": 6.166922317284259, + "learning_rate": 8.033140354611389e-06, + "loss": 0.7166, + "step": 8243 + }, + { + "epoch": 1.7286642902075906, + "grad_norm": 7.680997787423135, + "learning_rate": 8.030920741941665e-06, + "loss": 0.8547, + "step": 8244 + }, + { + "epoch": 1.7288739777731181, + "grad_norm": 7.982062802826913, + "learning_rate": 8.028701230190664e-06, + "loss": 0.8488, + "step": 8245 + }, + { + "epoch": 1.7290836653386454, + "grad_norm": 5.451241781681055, + "learning_rate": 8.02648181947215e-06, + "loss": 0.5516, + "step": 8246 + }, + { + "epoch": 1.7292933529041727, + "grad_norm": 7.202149454924948, + "learning_rate": 8.024262509899863e-06, + "loss": 0.7099, + "step": 8247 + }, + { + "epoch": 1.7295030404697003, + "grad_norm": 6.2974388870488776, + "learning_rate": 8.02204330158755e-06, + "loss": 0.7021, + "step": 8248 + }, + { + "epoch": 1.7297127280352274, + "grad_norm": 7.5498901851334255, + "learning_rate": 8.019824194648949e-06, + "loss": 0.8218, + "step": 8249 + }, + { + "epoch": 1.7299224156007549, + "grad_norm": 6.581751741224917, + "learning_rate": 8.017605189197788e-06, + "loss": 0.7693, + "step": 8250 + }, + { + "epoch": 1.7301321031662822, + "grad_norm": 6.309746109822411, + "learning_rate": 8.015386285347803e-06, + "loss": 0.6963, + "step": 8251 + }, + { + "epoch": 1.7303417907318095, + "grad_norm": 6.303736070129266, + "learning_rate": 8.013167483212714e-06, + "loss": 0.8515, + "step": 8252 + }, + { + "epoch": 1.730551478297337, + "grad_norm": 5.350309776995385, + "learning_rate": 8.010948782906237e-06, + "loss": 0.4598, + "step": 8253 + }, + { + "epoch": 1.7307611658628643, + "grad_norm": 6.022426848843186, + "learning_rate": 8.008730184542084e-06, + "loss": 0.6121, + "step": 8254 + }, + { + "epoch": 1.7309708534283916, + "grad_norm": 7.191221966783318, + "learning_rate": 8.006511688233964e-06, + "loss": 0.5704, + "step": 8255 + }, + { + "epoch": 1.7311805409939192, + "grad_norm": 5.224210817696967, + "learning_rate": 8.004293294095573e-06, + "loss": 0.5177, + "step": 8256 + }, + { + "epoch": 1.7313902285594465, + "grad_norm": 7.402630925206769, + "learning_rate": 8.002075002240616e-06, + "loss": 0.7629, + "step": 8257 + }, + { + "epoch": 1.7315999161249738, + "grad_norm": 5.454537964849812, + "learning_rate": 7.99985681278278e-06, + "loss": 0.4808, + "step": 8258 + }, + { + "epoch": 1.7318096036905013, + "grad_norm": 7.15322158328722, + "learning_rate": 7.997638725835751e-06, + "loss": 0.8472, + "step": 8259 + }, + { + "epoch": 1.7320192912560284, + "grad_norm": 7.728920583018578, + "learning_rate": 7.995420741513207e-06, + "loss": 0.8911, + "step": 8260 + }, + { + "epoch": 1.732228978821556, + "grad_norm": 5.789871582707629, + "learning_rate": 7.993202859928835e-06, + "loss": 0.6883, + "step": 8261 + }, + { + "epoch": 1.7324386663870832, + "grad_norm": 7.34303947750113, + "learning_rate": 7.99098508119629e-06, + "loss": 0.6896, + "step": 8262 + }, + { + "epoch": 1.7326483539526105, + "grad_norm": 6.469918724805049, + "learning_rate": 7.988767405429246e-06, + "loss": 0.6805, + "step": 8263 + }, + { + "epoch": 1.732858041518138, + "grad_norm": 7.323054779387471, + "learning_rate": 7.986549832741361e-06, + "loss": 0.8674, + "step": 8264 + }, + { + "epoch": 1.7330677290836654, + "grad_norm": 6.990265006957256, + "learning_rate": 7.98433236324629e-06, + "loss": 0.8337, + "step": 8265 + }, + { + "epoch": 1.7332774166491927, + "grad_norm": 6.906869385986223, + "learning_rate": 7.982114997057677e-06, + "loss": 0.6747, + "step": 8266 + }, + { + "epoch": 1.7334871042147202, + "grad_norm": 6.795244900853352, + "learning_rate": 7.979897734289175e-06, + "loss": 0.6391, + "step": 8267 + }, + { + "epoch": 1.7336967917802473, + "grad_norm": 6.707399235163398, + "learning_rate": 7.977680575054417e-06, + "loss": 0.6898, + "step": 8268 + }, + { + "epoch": 1.7339064793457748, + "grad_norm": 6.471878022383615, + "learning_rate": 7.975463519467037e-06, + "loss": 0.7686, + "step": 8269 + }, + { + "epoch": 1.7341161669113023, + "grad_norm": 7.104654816408026, + "learning_rate": 7.973246567640664e-06, + "loss": 0.7643, + "step": 8270 + }, + { + "epoch": 1.7343258544768294, + "grad_norm": 6.38410107808737, + "learning_rate": 7.97102971968892e-06, + "loss": 0.7341, + "step": 8271 + }, + { + "epoch": 1.734535542042357, + "grad_norm": 7.444919822500842, + "learning_rate": 7.968812975725416e-06, + "loss": 0.7015, + "step": 8272 + }, + { + "epoch": 1.7347452296078842, + "grad_norm": 5.637188393391445, + "learning_rate": 7.966596335863776e-06, + "loss": 0.5601, + "step": 8273 + }, + { + "epoch": 1.7349549171734115, + "grad_norm": 7.110329623634496, + "learning_rate": 7.964379800217598e-06, + "loss": 0.6021, + "step": 8274 + }, + { + "epoch": 1.735164604738939, + "grad_norm": 6.581437988032651, + "learning_rate": 7.962163368900486e-06, + "loss": 0.7464, + "step": 8275 + }, + { + "epoch": 1.7353742923044664, + "grad_norm": 7.459196854556857, + "learning_rate": 7.959947042026038e-06, + "loss": 0.6217, + "step": 8276 + }, + { + "epoch": 1.7355839798699937, + "grad_norm": 9.355437499294768, + "learning_rate": 7.957730819707838e-06, + "loss": 0.9417, + "step": 8277 + }, + { + "epoch": 1.7357936674355212, + "grad_norm": 6.2292941407762035, + "learning_rate": 7.955514702059477e-06, + "loss": 0.5518, + "step": 8278 + }, + { + "epoch": 1.7360033550010483, + "grad_norm": 7.382120295776903, + "learning_rate": 7.953298689194538e-06, + "loss": 0.6794, + "step": 8279 + }, + { + "epoch": 1.7362130425665758, + "grad_norm": 6.56409708270257, + "learning_rate": 7.951082781226586e-06, + "loss": 0.5082, + "step": 8280 + }, + { + "epoch": 1.7364227301321031, + "grad_norm": 7.84310053467573, + "learning_rate": 7.948866978269194e-06, + "loss": 0.8546, + "step": 8281 + }, + { + "epoch": 1.7366324176976304, + "grad_norm": 6.534705237406541, + "learning_rate": 7.946651280435935e-06, + "loss": 0.7206, + "step": 8282 + }, + { + "epoch": 1.736842105263158, + "grad_norm": 8.990765198321185, + "learning_rate": 7.944435687840352e-06, + "loss": 0.7661, + "step": 8283 + }, + { + "epoch": 1.7370517928286853, + "grad_norm": 8.321480733497037, + "learning_rate": 7.942220200596007e-06, + "loss": 0.7348, + "step": 8284 + }, + { + "epoch": 1.7372614803942126, + "grad_norm": 7.884262947725674, + "learning_rate": 7.940004818816446e-06, + "loss": 0.8127, + "step": 8285 + }, + { + "epoch": 1.73747116795974, + "grad_norm": 6.328351740248241, + "learning_rate": 7.937789542615212e-06, + "loss": 0.5466, + "step": 8286 + }, + { + "epoch": 1.7376808555252672, + "grad_norm": 7.3128175928868275, + "learning_rate": 7.935574372105837e-06, + "loss": 0.7033, + "step": 8287 + }, + { + "epoch": 1.7378905430907947, + "grad_norm": 8.268222288544365, + "learning_rate": 7.933359307401857e-06, + "loss": 0.8022, + "step": 8288 + }, + { + "epoch": 1.7381002306563222, + "grad_norm": 7.459539807304485, + "learning_rate": 7.931144348616802e-06, + "loss": 0.7717, + "step": 8289 + }, + { + "epoch": 1.7383099182218493, + "grad_norm": 5.454302510356499, + "learning_rate": 7.928929495864182e-06, + "loss": 0.4903, + "step": 8290 + }, + { + "epoch": 1.7385196057873769, + "grad_norm": 6.635050422197754, + "learning_rate": 7.926714749257517e-06, + "loss": 0.7015, + "step": 8291 + }, + { + "epoch": 1.7387292933529042, + "grad_norm": 6.249368287288514, + "learning_rate": 7.924500108910324e-06, + "loss": 0.715, + "step": 8292 + }, + { + "epoch": 1.7389389809184315, + "grad_norm": 6.337758586607306, + "learning_rate": 7.922285574936092e-06, + "loss": 0.6696, + "step": 8293 + }, + { + "epoch": 1.739148668483959, + "grad_norm": 7.783572610343767, + "learning_rate": 7.920071147448332e-06, + "loss": 0.9077, + "step": 8294 + }, + { + "epoch": 1.7393583560494863, + "grad_norm": 7.261223859785903, + "learning_rate": 7.917856826560534e-06, + "loss": 0.8187, + "step": 8295 + }, + { + "epoch": 1.7395680436150136, + "grad_norm": 7.83582434330073, + "learning_rate": 7.915642612386184e-06, + "loss": 0.6169, + "step": 8296 + }, + { + "epoch": 1.7397777311805411, + "grad_norm": 6.18310439796562, + "learning_rate": 7.913428505038761e-06, + "loss": 0.5463, + "step": 8297 + }, + { + "epoch": 1.7399874187460682, + "grad_norm": 8.091328395805233, + "learning_rate": 7.911214504631752e-06, + "loss": 0.896, + "step": 8298 + }, + { + "epoch": 1.7401971063115957, + "grad_norm": 7.6694064551194066, + "learning_rate": 7.90900061127862e-06, + "loss": 0.7617, + "step": 8299 + }, + { + "epoch": 1.740406793877123, + "grad_norm": 7.627220139854269, + "learning_rate": 7.906786825092832e-06, + "loss": 0.7196, + "step": 8300 + }, + { + "epoch": 1.7406164814426504, + "grad_norm": 8.19599473420918, + "learning_rate": 7.904573146187854e-06, + "loss": 0.9337, + "step": 8301 + }, + { + "epoch": 1.7408261690081779, + "grad_norm": 7.594000218138649, + "learning_rate": 7.902359574677133e-06, + "loss": 0.6259, + "step": 8302 + }, + { + "epoch": 1.7410358565737052, + "grad_norm": 6.750478596740729, + "learning_rate": 7.900146110674121e-06, + "loss": 0.7981, + "step": 8303 + }, + { + "epoch": 1.7412455441392325, + "grad_norm": 5.659334536088653, + "learning_rate": 7.897932754292266e-06, + "loss": 0.5247, + "step": 8304 + }, + { + "epoch": 1.74145523170476, + "grad_norm": 6.9937158451341475, + "learning_rate": 7.895719505645002e-06, + "loss": 0.9145, + "step": 8305 + }, + { + "epoch": 1.7416649192702873, + "grad_norm": 7.731880666962982, + "learning_rate": 7.893506364845762e-06, + "loss": 0.755, + "step": 8306 + }, + { + "epoch": 1.7418746068358146, + "grad_norm": 7.491943245009911, + "learning_rate": 7.891293332007977e-06, + "loss": 0.7756, + "step": 8307 + }, + { + "epoch": 1.7420842944013422, + "grad_norm": 8.330455988579322, + "learning_rate": 7.889080407245063e-06, + "loss": 0.865, + "step": 8308 + }, + { + "epoch": 1.7422939819668692, + "grad_norm": 5.994755713686167, + "learning_rate": 7.88686759067044e-06, + "loss": 0.6359, + "step": 8309 + }, + { + "epoch": 1.7425036695323968, + "grad_norm": 6.9876751484953035, + "learning_rate": 7.88465488239752e-06, + "loss": 0.6817, + "step": 8310 + }, + { + "epoch": 1.742713357097924, + "grad_norm": 6.912174817659617, + "learning_rate": 7.882442282539704e-06, + "loss": 0.5292, + "step": 8311 + }, + { + "epoch": 1.7429230446634514, + "grad_norm": 6.139903703367255, + "learning_rate": 7.880229791210392e-06, + "loss": 0.7273, + "step": 8312 + }, + { + "epoch": 1.743132732228979, + "grad_norm": 7.244508436198552, + "learning_rate": 7.878017408522987e-06, + "loss": 0.6925, + "step": 8313 + }, + { + "epoch": 1.7433424197945062, + "grad_norm": 6.380479138070323, + "learning_rate": 7.875805134590862e-06, + "loss": 0.6611, + "step": 8314 + }, + { + "epoch": 1.7435521073600335, + "grad_norm": 6.18512297497648, + "learning_rate": 7.873592969527412e-06, + "loss": 0.5637, + "step": 8315 + }, + { + "epoch": 1.743761794925561, + "grad_norm": 9.061430051757958, + "learning_rate": 7.871380913446011e-06, + "loss": 0.8406, + "step": 8316 + }, + { + "epoch": 1.7439714824910881, + "grad_norm": 7.764680019996193, + "learning_rate": 7.869168966460029e-06, + "loss": 0.5375, + "step": 8317 + }, + { + "epoch": 1.7441811700566157, + "grad_norm": 6.566147811939327, + "learning_rate": 7.866957128682831e-06, + "loss": 0.6677, + "step": 8318 + }, + { + "epoch": 1.744390857622143, + "grad_norm": 5.717880569223168, + "learning_rate": 7.864745400227786e-06, + "loss": 0.5312, + "step": 8319 + }, + { + "epoch": 1.7446005451876703, + "grad_norm": 7.263312454491009, + "learning_rate": 7.862533781208236e-06, + "loss": 0.7405, + "step": 8320 + }, + { + "epoch": 1.7448102327531978, + "grad_norm": 8.001183117798696, + "learning_rate": 7.86032227173754e-06, + "loss": 0.7638, + "step": 8321 + }, + { + "epoch": 1.745019920318725, + "grad_norm": 6.025111626451926, + "learning_rate": 7.858110871929039e-06, + "loss": 0.5532, + "step": 8322 + }, + { + "epoch": 1.7452296078842524, + "grad_norm": 6.59798226897668, + "learning_rate": 7.855899581896069e-06, + "loss": 0.711, + "step": 8323 + }, + { + "epoch": 1.74543929544978, + "grad_norm": 6.733854705201149, + "learning_rate": 7.853688401751963e-06, + "loss": 0.6982, + "step": 8324 + }, + { + "epoch": 1.7456489830153072, + "grad_norm": 6.221975489853401, + "learning_rate": 7.851477331610053e-06, + "loss": 0.679, + "step": 8325 + }, + { + "epoch": 1.7458586705808345, + "grad_norm": 6.039367972478635, + "learning_rate": 7.849266371583655e-06, + "loss": 0.6311, + "step": 8326 + }, + { + "epoch": 1.746068358146362, + "grad_norm": 6.471004059676718, + "learning_rate": 7.847055521786085e-06, + "loss": 0.6347, + "step": 8327 + }, + { + "epoch": 1.7462780457118892, + "grad_norm": 7.30014313388475, + "learning_rate": 7.844844782330655e-06, + "loss": 0.7341, + "step": 8328 + }, + { + "epoch": 1.7464877332774167, + "grad_norm": 6.3888100699811226, + "learning_rate": 7.842634153330665e-06, + "loss": 0.8078, + "step": 8329 + }, + { + "epoch": 1.746697420842944, + "grad_norm": 6.62822689195477, + "learning_rate": 7.840423634899417e-06, + "loss": 0.6549, + "step": 8330 + }, + { + "epoch": 1.7469071084084713, + "grad_norm": 6.105384023106563, + "learning_rate": 7.838213227150206e-06, + "loss": 0.8198, + "step": 8331 + }, + { + "epoch": 1.7471167959739988, + "grad_norm": 6.462280147479494, + "learning_rate": 7.836002930196314e-06, + "loss": 0.6268, + "step": 8332 + }, + { + "epoch": 1.7473264835395261, + "grad_norm": 6.213619649004164, + "learning_rate": 7.833792744151025e-06, + "loss": 0.7088, + "step": 8333 + }, + { + "epoch": 1.7475361711050534, + "grad_norm": 7.678576966618167, + "learning_rate": 7.831582669127618e-06, + "loss": 0.7619, + "step": 8334 + }, + { + "epoch": 1.747745858670581, + "grad_norm": 6.367222886338224, + "learning_rate": 7.829372705239356e-06, + "loss": 0.7127, + "step": 8335 + }, + { + "epoch": 1.747955546236108, + "grad_norm": 6.479651165429848, + "learning_rate": 7.827162852599508e-06, + "loss": 0.5745, + "step": 8336 + }, + { + "epoch": 1.7481652338016356, + "grad_norm": 5.906467716301577, + "learning_rate": 7.824953111321336e-06, + "loss": 0.588, + "step": 8337 + }, + { + "epoch": 1.7483749213671629, + "grad_norm": 7.224072639068199, + "learning_rate": 7.822743481518086e-06, + "loss": 0.9651, + "step": 8338 + }, + { + "epoch": 1.7485846089326902, + "grad_norm": 9.966451982051009, + "learning_rate": 7.82053396330301e-06, + "loss": 0.8825, + "step": 8339 + }, + { + "epoch": 1.7487942964982177, + "grad_norm": 7.896737656740614, + "learning_rate": 7.818324556789352e-06, + "loss": 0.7555, + "step": 8340 + }, + { + "epoch": 1.749003984063745, + "grad_norm": 8.13569407162091, + "learning_rate": 7.816115262090338e-06, + "loss": 0.965, + "step": 8341 + }, + { + "epoch": 1.7492136716292723, + "grad_norm": 6.4256320791119235, + "learning_rate": 7.813906079319207e-06, + "loss": 0.5834, + "step": 8342 + }, + { + "epoch": 1.7494233591947999, + "grad_norm": 6.690790509656073, + "learning_rate": 7.811697008589184e-06, + "loss": 0.6692, + "step": 8343 + }, + { + "epoch": 1.7496330467603272, + "grad_norm": 5.985458221518401, + "learning_rate": 7.809488050013483e-06, + "loss": 0.5352, + "step": 8344 + }, + { + "epoch": 1.7498427343258545, + "grad_norm": 5.788203099395514, + "learning_rate": 7.807279203705315e-06, + "loss": 0.5902, + "step": 8345 + }, + { + "epoch": 1.750052421891382, + "grad_norm": 6.034995914847869, + "learning_rate": 7.805070469777895e-06, + "loss": 0.5335, + "step": 8346 + }, + { + "epoch": 1.750262109456909, + "grad_norm": 8.126886560964602, + "learning_rate": 7.80286184834442e-06, + "loss": 0.9237, + "step": 8347 + }, + { + "epoch": 1.7504717970224366, + "grad_norm": 6.999918407684177, + "learning_rate": 7.800653339518084e-06, + "loss": 0.7555, + "step": 8348 + }, + { + "epoch": 1.750681484587964, + "grad_norm": 6.107678680492182, + "learning_rate": 7.79844494341208e-06, + "loss": 0.543, + "step": 8349 + }, + { + "epoch": 1.7508911721534912, + "grad_norm": 6.979725930935626, + "learning_rate": 7.796236660139593e-06, + "loss": 0.7366, + "step": 8350 + }, + { + "epoch": 1.7511008597190187, + "grad_norm": 7.235925996085907, + "learning_rate": 7.794028489813797e-06, + "loss": 0.7713, + "step": 8351 + }, + { + "epoch": 1.751310547284546, + "grad_norm": 6.669378899441408, + "learning_rate": 7.791820432547866e-06, + "loss": 0.4604, + "step": 8352 + }, + { + "epoch": 1.7515202348500734, + "grad_norm": 7.337929885968389, + "learning_rate": 7.789612488454971e-06, + "loss": 0.7075, + "step": 8353 + }, + { + "epoch": 1.7517299224156009, + "grad_norm": 6.647050511430966, + "learning_rate": 7.78740465764827e-06, + "loss": 0.769, + "step": 8354 + }, + { + "epoch": 1.751939609981128, + "grad_norm": 6.602102977590646, + "learning_rate": 7.785196940240912e-06, + "loss": 0.7323, + "step": 8355 + }, + { + "epoch": 1.7521492975466555, + "grad_norm": 6.997494973515076, + "learning_rate": 7.782989336346058e-06, + "loss": 0.7627, + "step": 8356 + }, + { + "epoch": 1.7523589851121828, + "grad_norm": 7.03640027408144, + "learning_rate": 7.780781846076843e-06, + "loss": 0.5759, + "step": 8357 + }, + { + "epoch": 1.75256867267771, + "grad_norm": 5.8513813762683125, + "learning_rate": 7.778574469546408e-06, + "loss": 0.6003, + "step": 8358 + }, + { + "epoch": 1.7527783602432376, + "grad_norm": 8.212844802011938, + "learning_rate": 7.776367206867887e-06, + "loss": 0.734, + "step": 8359 + }, + { + "epoch": 1.752988047808765, + "grad_norm": 6.062959956204114, + "learning_rate": 7.774160058154399e-06, + "loss": 0.6097, + "step": 8360 + }, + { + "epoch": 1.7531977353742922, + "grad_norm": 6.775777736985307, + "learning_rate": 7.77195302351907e-06, + "loss": 0.6016, + "step": 8361 + }, + { + "epoch": 1.7534074229398198, + "grad_norm": 7.367775890017337, + "learning_rate": 7.769746103075016e-06, + "loss": 0.8402, + "step": 8362 + }, + { + "epoch": 1.753617110505347, + "grad_norm": 6.705186706317168, + "learning_rate": 7.767539296935338e-06, + "loss": 0.5781, + "step": 8363 + }, + { + "epoch": 1.7538267980708744, + "grad_norm": 6.315655877832037, + "learning_rate": 7.765332605213144e-06, + "loss": 0.6865, + "step": 8364 + }, + { + "epoch": 1.754036485636402, + "grad_norm": 8.795172077629546, + "learning_rate": 7.763126028021534e-06, + "loss": 0.7353, + "step": 8365 + }, + { + "epoch": 1.754246173201929, + "grad_norm": 6.139998738461982, + "learning_rate": 7.760919565473587e-06, + "loss": 0.5101, + "step": 8366 + }, + { + "epoch": 1.7544558607674565, + "grad_norm": 5.948526255729782, + "learning_rate": 7.7587132176824e-06, + "loss": 0.6087, + "step": 8367 + }, + { + "epoch": 1.7546655483329838, + "grad_norm": 10.357317149772832, + "learning_rate": 7.75650698476105e-06, + "loss": 0.964, + "step": 8368 + }, + { + "epoch": 1.7548752358985111, + "grad_norm": 9.807427891529505, + "learning_rate": 7.754300866822608e-06, + "loss": 0.8233, + "step": 8369 + }, + { + "epoch": 1.7550849234640387, + "grad_norm": 7.612976670472568, + "learning_rate": 7.752094863980137e-06, + "loss": 0.8632, + "step": 8370 + }, + { + "epoch": 1.755294611029566, + "grad_norm": 7.633476815554445, + "learning_rate": 7.749888976346708e-06, + "loss": 0.6741, + "step": 8371 + }, + { + "epoch": 1.7555042985950933, + "grad_norm": 7.804986759220151, + "learning_rate": 7.747683204035367e-06, + "loss": 0.7833, + "step": 8372 + }, + { + "epoch": 1.7557139861606208, + "grad_norm": 8.217312893624321, + "learning_rate": 7.74547754715917e-06, + "loss": 1.0026, + "step": 8373 + }, + { + "epoch": 1.7559236737261479, + "grad_norm": 6.21481258991733, + "learning_rate": 7.743272005831162e-06, + "loss": 0.5765, + "step": 8374 + }, + { + "epoch": 1.7561333612916754, + "grad_norm": 7.195771676684358, + "learning_rate": 7.741066580164377e-06, + "loss": 0.6951, + "step": 8375 + }, + { + "epoch": 1.7563430488572027, + "grad_norm": 9.018197081559544, + "learning_rate": 7.738861270271842e-06, + "loss": 1.2057, + "step": 8376 + }, + { + "epoch": 1.75655273642273, + "grad_norm": 6.906374772161643, + "learning_rate": 7.736656076266594e-06, + "loss": 0.6311, + "step": 8377 + }, + { + "epoch": 1.7567624239882575, + "grad_norm": 7.250030024493052, + "learning_rate": 7.734450998261647e-06, + "loss": 0.7661, + "step": 8378 + }, + { + "epoch": 1.7569721115537849, + "grad_norm": 5.089759870798284, + "learning_rate": 7.732246036370015e-06, + "loss": 0.5591, + "step": 8379 + }, + { + "epoch": 1.7571817991193122, + "grad_norm": 7.525142257922409, + "learning_rate": 7.730041190704709e-06, + "loss": 0.7682, + "step": 8380 + }, + { + "epoch": 1.7573914866848397, + "grad_norm": 8.417222828133237, + "learning_rate": 7.727836461378725e-06, + "loss": 1.142, + "step": 8381 + }, + { + "epoch": 1.757601174250367, + "grad_norm": 7.133122471626611, + "learning_rate": 7.725631848505063e-06, + "loss": 0.7348, + "step": 8382 + }, + { + "epoch": 1.7578108618158943, + "grad_norm": 7.138231293876203, + "learning_rate": 7.723427352196718e-06, + "loss": 0.7279, + "step": 8383 + }, + { + "epoch": 1.7580205493814218, + "grad_norm": 7.758087929314629, + "learning_rate": 7.721222972566667e-06, + "loss": 1.0003, + "step": 8384 + }, + { + "epoch": 1.758230236946949, + "grad_norm": 7.619428796410866, + "learning_rate": 7.71901870972789e-06, + "loss": 0.7514, + "step": 8385 + }, + { + "epoch": 1.7584399245124764, + "grad_norm": 8.465870547632365, + "learning_rate": 7.716814563793365e-06, + "loss": 0.9024, + "step": 8386 + }, + { + "epoch": 1.7586496120780037, + "grad_norm": 6.262868311063114, + "learning_rate": 7.714610534876047e-06, + "loss": 0.5254, + "step": 8387 + }, + { + "epoch": 1.758859299643531, + "grad_norm": 5.884486237055454, + "learning_rate": 7.712406623088905e-06, + "loss": 0.4959, + "step": 8388 + }, + { + "epoch": 1.7590689872090586, + "grad_norm": 6.544421002362297, + "learning_rate": 7.710202828544895e-06, + "loss": 0.7058, + "step": 8389 + }, + { + "epoch": 1.7592786747745859, + "grad_norm": 5.976999000740742, + "learning_rate": 7.707999151356958e-06, + "loss": 0.6256, + "step": 8390 + }, + { + "epoch": 1.7594883623401132, + "grad_norm": 6.5895410954666405, + "learning_rate": 7.70579559163804e-06, + "loss": 0.7145, + "step": 8391 + }, + { + "epoch": 1.7596980499056407, + "grad_norm": 6.7149770478249025, + "learning_rate": 7.70359214950108e-06, + "loss": 0.7279, + "step": 8392 + }, + { + "epoch": 1.7599077374711678, + "grad_norm": 8.015850038418254, + "learning_rate": 7.701388825059001e-06, + "loss": 0.8607, + "step": 8393 + }, + { + "epoch": 1.7601174250366953, + "grad_norm": 8.51327284603959, + "learning_rate": 7.699185618424733e-06, + "loss": 0.7749, + "step": 8394 + }, + { + "epoch": 1.7603271126022229, + "grad_norm": 5.754919899193569, + "learning_rate": 7.696982529711195e-06, + "loss": 0.7151, + "step": 8395 + }, + { + "epoch": 1.76053680016775, + "grad_norm": 8.94363215710054, + "learning_rate": 7.694779559031293e-06, + "loss": 1.0975, + "step": 8396 + }, + { + "epoch": 1.7607464877332775, + "grad_norm": 5.615187907482772, + "learning_rate": 7.692576706497936e-06, + "loss": 0.6542, + "step": 8397 + }, + { + "epoch": 1.7609561752988048, + "grad_norm": 6.06698202826116, + "learning_rate": 7.690373972224032e-06, + "loss": 0.629, + "step": 8398 + }, + { + "epoch": 1.761165862864332, + "grad_norm": 6.816281666341303, + "learning_rate": 7.68817135632246e-06, + "loss": 0.8312, + "step": 8399 + }, + { + "epoch": 1.7613755504298596, + "grad_norm": 6.529156810605829, + "learning_rate": 7.68596885890612e-06, + "loss": 0.7894, + "step": 8400 + }, + { + "epoch": 1.761585237995387, + "grad_norm": 6.5634570218672135, + "learning_rate": 7.68376648008789e-06, + "loss": 0.7903, + "step": 8401 + }, + { + "epoch": 1.7617949255609142, + "grad_norm": 4.6176445672439534, + "learning_rate": 7.681564219980642e-06, + "loss": 0.3915, + "step": 8402 + }, + { + "epoch": 1.7620046131264417, + "grad_norm": 6.586754072040553, + "learning_rate": 7.679362078697247e-06, + "loss": 0.695, + "step": 8403 + }, + { + "epoch": 1.7622143006919688, + "grad_norm": 7.4457966868721845, + "learning_rate": 7.677160056350577e-06, + "loss": 0.6014, + "step": 8404 + }, + { + "epoch": 1.7624239882574964, + "grad_norm": 7.122957288802633, + "learning_rate": 7.674958153053479e-06, + "loss": 0.758, + "step": 8405 + }, + { + "epoch": 1.7626336758230237, + "grad_norm": 7.0382553163305355, + "learning_rate": 7.672756368918809e-06, + "loss": 0.7179, + "step": 8406 + }, + { + "epoch": 1.762843363388551, + "grad_norm": 6.982633009966255, + "learning_rate": 7.670554704059412e-06, + "loss": 0.7201, + "step": 8407 + }, + { + "epoch": 1.7630530509540785, + "grad_norm": 7.463009917197567, + "learning_rate": 7.668353158588123e-06, + "loss": 0.7152, + "step": 8408 + }, + { + "epoch": 1.7632627385196058, + "grad_norm": 6.920167631942121, + "learning_rate": 7.666151732617782e-06, + "loss": 0.6441, + "step": 8409 + }, + { + "epoch": 1.763472426085133, + "grad_norm": 8.205821644911513, + "learning_rate": 7.66395042626121e-06, + "loss": 0.6917, + "step": 8410 + }, + { + "epoch": 1.7636821136506606, + "grad_norm": 8.140473842590335, + "learning_rate": 7.661749239631235e-06, + "loss": 0.6963, + "step": 8411 + }, + { + "epoch": 1.7638918012161877, + "grad_norm": 7.568298445425577, + "learning_rate": 7.659548172840664e-06, + "loss": 0.7371, + "step": 8412 + }, + { + "epoch": 1.7641014887817152, + "grad_norm": 9.164700518022064, + "learning_rate": 7.657347226002306e-06, + "loss": 0.9727, + "step": 8413 + }, + { + "epoch": 1.7643111763472428, + "grad_norm": 6.550522139752771, + "learning_rate": 7.655146399228971e-06, + "loss": 0.5577, + "step": 8414 + }, + { + "epoch": 1.7645208639127699, + "grad_norm": 7.4425018632890145, + "learning_rate": 7.652945692633446e-06, + "loss": 0.6697, + "step": 8415 + }, + { + "epoch": 1.7647305514782974, + "grad_norm": 7.050784670150012, + "learning_rate": 7.650745106328528e-06, + "loss": 0.6148, + "step": 8416 + }, + { + "epoch": 1.7649402390438247, + "grad_norm": 6.858320959137793, + "learning_rate": 7.648544640426998e-06, + "loss": 0.8089, + "step": 8417 + }, + { + "epoch": 1.765149926609352, + "grad_norm": 6.65943940491548, + "learning_rate": 7.646344295041629e-06, + "loss": 0.6146, + "step": 8418 + }, + { + "epoch": 1.7653596141748795, + "grad_norm": 7.544510056427483, + "learning_rate": 7.6441440702852e-06, + "loss": 1.0247, + "step": 8419 + }, + { + "epoch": 1.7655693017404068, + "grad_norm": 7.9584480877266355, + "learning_rate": 7.641943966270475e-06, + "loss": 0.6129, + "step": 8420 + }, + { + "epoch": 1.7657789893059341, + "grad_norm": 7.8336602517, + "learning_rate": 7.63974398311021e-06, + "loss": 0.8525, + "step": 8421 + }, + { + "epoch": 1.7659886768714617, + "grad_norm": 7.979725956016501, + "learning_rate": 7.637544120917158e-06, + "loss": 0.9821, + "step": 8422 + }, + { + "epoch": 1.7661983644369887, + "grad_norm": 8.379511918449996, + "learning_rate": 7.635344379804073e-06, + "loss": 0.8005, + "step": 8423 + }, + { + "epoch": 1.7664080520025163, + "grad_norm": 7.905665915337931, + "learning_rate": 7.633144759883682e-06, + "loss": 0.8621, + "step": 8424 + }, + { + "epoch": 1.7666177395680436, + "grad_norm": 5.8295354608155385, + "learning_rate": 7.63094526126873e-06, + "loss": 0.6837, + "step": 8425 + }, + { + "epoch": 1.7668274271335709, + "grad_norm": 6.966053558106857, + "learning_rate": 7.6287458840719465e-06, + "loss": 0.7568, + "step": 8426 + }, + { + "epoch": 1.7670371146990984, + "grad_norm": 5.533883951873656, + "learning_rate": 7.6265466284060455e-06, + "loss": 0.6848, + "step": 8427 + }, + { + "epoch": 1.7672468022646257, + "grad_norm": 6.438766029778968, + "learning_rate": 7.624347494383744e-06, + "loss": 0.7024, + "step": 8428 + }, + { + "epoch": 1.767456489830153, + "grad_norm": 7.177831894088845, + "learning_rate": 7.62214848211776e-06, + "loss": 0.7058, + "step": 8429 + }, + { + "epoch": 1.7676661773956805, + "grad_norm": 7.769997318571442, + "learning_rate": 7.619949591720784e-06, + "loss": 0.8546, + "step": 8430 + }, + { + "epoch": 1.7678758649612079, + "grad_norm": 6.663177887954277, + "learning_rate": 7.617750823305523e-06, + "loss": 0.5998, + "step": 8431 + }, + { + "epoch": 1.7680855525267352, + "grad_norm": 6.58023226204452, + "learning_rate": 7.615552176984665e-06, + "loss": 0.7618, + "step": 8432 + }, + { + "epoch": 1.7682952400922627, + "grad_norm": 7.277939778438598, + "learning_rate": 7.613353652870892e-06, + "loss": 0.8417, + "step": 8433 + }, + { + "epoch": 1.7685049276577898, + "grad_norm": 6.53106863751666, + "learning_rate": 7.611155251076881e-06, + "loss": 0.6965, + "step": 8434 + }, + { + "epoch": 1.7687146152233173, + "grad_norm": 7.065560428909784, + "learning_rate": 7.608956971715311e-06, + "loss": 0.8685, + "step": 8435 + }, + { + "epoch": 1.7689243027888446, + "grad_norm": 8.309969928855029, + "learning_rate": 7.606758814898841e-06, + "loss": 0.8374, + "step": 8436 + }, + { + "epoch": 1.769133990354372, + "grad_norm": 7.08047651998436, + "learning_rate": 7.604560780740134e-06, + "loss": 0.5909, + "step": 8437 + }, + { + "epoch": 1.7693436779198994, + "grad_norm": 5.723651254946721, + "learning_rate": 7.6023628693518415e-06, + "loss": 0.6721, + "step": 8438 + }, + { + "epoch": 1.7695533654854267, + "grad_norm": 6.973590855378079, + "learning_rate": 7.600165080846608e-06, + "loss": 0.6979, + "step": 8439 + }, + { + "epoch": 1.769763053050954, + "grad_norm": 6.368218673246091, + "learning_rate": 7.597967415337078e-06, + "loss": 0.5685, + "step": 8440 + }, + { + "epoch": 1.7699727406164816, + "grad_norm": 5.958216468128591, + "learning_rate": 7.595769872935885e-06, + "loss": 0.5501, + "step": 8441 + }, + { + "epoch": 1.7701824281820087, + "grad_norm": 6.087656295948984, + "learning_rate": 7.593572453755655e-06, + "loss": 0.6289, + "step": 8442 + }, + { + "epoch": 1.7703921157475362, + "grad_norm": 7.550056991387706, + "learning_rate": 7.5913751579090115e-06, + "loss": 0.8357, + "step": 8443 + }, + { + "epoch": 1.7706018033130635, + "grad_norm": 7.41018237105327, + "learning_rate": 7.589177985508569e-06, + "loss": 0.764, + "step": 8444 + }, + { + "epoch": 1.7708114908785908, + "grad_norm": 6.474432522528981, + "learning_rate": 7.586980936666932e-06, + "loss": 0.6047, + "step": 8445 + }, + { + "epoch": 1.7710211784441183, + "grad_norm": 6.0824007309177475, + "learning_rate": 7.584784011496711e-06, + "loss": 0.6894, + "step": 8446 + }, + { + "epoch": 1.7712308660096456, + "grad_norm": 7.246433112842389, + "learning_rate": 7.582587210110499e-06, + "loss": 0.6914, + "step": 8447 + }, + { + "epoch": 1.771440553575173, + "grad_norm": 7.42884640114442, + "learning_rate": 7.580390532620883e-06, + "loss": 0.6648, + "step": 8448 + }, + { + "epoch": 1.7716502411407005, + "grad_norm": 6.263779586785348, + "learning_rate": 7.5781939791404465e-06, + "loss": 0.5921, + "step": 8449 + }, + { + "epoch": 1.7718599287062278, + "grad_norm": 6.586723961996054, + "learning_rate": 7.575997549781777e-06, + "loss": 0.6502, + "step": 8450 + }, + { + "epoch": 1.772069616271755, + "grad_norm": 7.1385855725625404, + "learning_rate": 7.573801244657431e-06, + "loss": 0.6738, + "step": 8451 + }, + { + "epoch": 1.7722793038372826, + "grad_norm": 6.709043001698923, + "learning_rate": 7.571605063879979e-06, + "loss": 0.784, + "step": 8452 + }, + { + "epoch": 1.7724889914028097, + "grad_norm": 6.654812508432464, + "learning_rate": 7.569409007561983e-06, + "loss": 0.7903, + "step": 8453 + }, + { + "epoch": 1.7726986789683372, + "grad_norm": 7.08056384452814, + "learning_rate": 7.567213075815989e-06, + "loss": 0.6277, + "step": 8454 + }, + { + "epoch": 1.7729083665338645, + "grad_norm": 8.095824896675438, + "learning_rate": 7.565017268754542e-06, + "loss": 1.0345, + "step": 8455 + }, + { + "epoch": 1.7731180540993918, + "grad_norm": 6.6993076513191525, + "learning_rate": 7.5628215864901864e-06, + "loss": 0.4817, + "step": 8456 + }, + { + "epoch": 1.7733277416649194, + "grad_norm": 5.665278812728131, + "learning_rate": 7.56062602913545e-06, + "loss": 0.4828, + "step": 8457 + }, + { + "epoch": 1.7735374292304467, + "grad_norm": 6.361843127812987, + "learning_rate": 7.558430596802862e-06, + "loss": 0.6614, + "step": 8458 + }, + { + "epoch": 1.773747116795974, + "grad_norm": 8.029712324777721, + "learning_rate": 7.556235289604941e-06, + "loss": 0.7534, + "step": 8459 + }, + { + "epoch": 1.7739568043615015, + "grad_norm": 6.598077813373454, + "learning_rate": 7.554040107654199e-06, + "loss": 0.6324, + "step": 8460 + }, + { + "epoch": 1.7741664919270286, + "grad_norm": 7.326996972421166, + "learning_rate": 7.55184505106314e-06, + "loss": 0.6744, + "step": 8461 + }, + { + "epoch": 1.774376179492556, + "grad_norm": 7.442197958188835, + "learning_rate": 7.549650119944274e-06, + "loss": 0.7828, + "step": 8462 + }, + { + "epoch": 1.7745858670580834, + "grad_norm": 5.327338970528745, + "learning_rate": 7.547455314410087e-06, + "loss": 0.5452, + "step": 8463 + }, + { + "epoch": 1.7747955546236107, + "grad_norm": 8.010569178903205, + "learning_rate": 7.545260634573068e-06, + "loss": 0.9968, + "step": 8464 + }, + { + "epoch": 1.7750052421891382, + "grad_norm": 7.076586679055972, + "learning_rate": 7.543066080545702e-06, + "loss": 0.7472, + "step": 8465 + }, + { + "epoch": 1.7752149297546655, + "grad_norm": 7.388871805086191, + "learning_rate": 7.540871652440456e-06, + "loss": 0.9161, + "step": 8466 + }, + { + "epoch": 1.7754246173201929, + "grad_norm": 9.278049173740676, + "learning_rate": 7.538677350369805e-06, + "loss": 0.9707, + "step": 8467 + }, + { + "epoch": 1.7756343048857204, + "grad_norm": 5.224657037820614, + "learning_rate": 7.536483174446212e-06, + "loss": 0.5524, + "step": 8468 + }, + { + "epoch": 1.7758439924512477, + "grad_norm": 7.2835324080757635, + "learning_rate": 7.534289124782126e-06, + "loss": 0.6948, + "step": 8469 + }, + { + "epoch": 1.776053680016775, + "grad_norm": 9.122722851301939, + "learning_rate": 7.532095201489996e-06, + "loss": 0.838, + "step": 8470 + }, + { + "epoch": 1.7762633675823025, + "grad_norm": 7.281238872341526, + "learning_rate": 7.529901404682271e-06, + "loss": 0.8466, + "step": 8471 + }, + { + "epoch": 1.7764730551478296, + "grad_norm": 6.6302619988034355, + "learning_rate": 7.527707734471383e-06, + "loss": 0.6242, + "step": 8472 + }, + { + "epoch": 1.7766827427133571, + "grad_norm": 7.106371540684246, + "learning_rate": 7.5255141909697615e-06, + "loss": 0.6226, + "step": 8473 + }, + { + "epoch": 1.7768924302788844, + "grad_norm": 7.228130737592883, + "learning_rate": 7.52332077428983e-06, + "loss": 0.7849, + "step": 8474 + }, + { + "epoch": 1.7771021178444117, + "grad_norm": 7.109879141004516, + "learning_rate": 7.521127484544004e-06, + "loss": 0.7588, + "step": 8475 + }, + { + "epoch": 1.7773118054099393, + "grad_norm": 5.863221213162364, + "learning_rate": 7.518934321844692e-06, + "loss": 0.5756, + "step": 8476 + }, + { + "epoch": 1.7775214929754666, + "grad_norm": 7.288874584895362, + "learning_rate": 7.5167412863042996e-06, + "loss": 0.712, + "step": 8477 + }, + { + "epoch": 1.7777311805409939, + "grad_norm": 6.127213247144366, + "learning_rate": 7.514548378035227e-06, + "loss": 0.6488, + "step": 8478 + }, + { + "epoch": 1.7779408681065214, + "grad_norm": 7.693121770117709, + "learning_rate": 7.512355597149857e-06, + "loss": 0.7371, + "step": 8479 + }, + { + "epoch": 1.7781505556720485, + "grad_norm": 5.857298323106373, + "learning_rate": 7.510162943760576e-06, + "loss": 0.5392, + "step": 8480 + }, + { + "epoch": 1.778360243237576, + "grad_norm": 6.768684921200561, + "learning_rate": 7.507970417979769e-06, + "loss": 0.5919, + "step": 8481 + }, + { + "epoch": 1.7785699308031033, + "grad_norm": 5.632072122303574, + "learning_rate": 7.505778019919794e-06, + "loss": 0.5949, + "step": 8482 + }, + { + "epoch": 1.7787796183686306, + "grad_norm": 6.079162821020614, + "learning_rate": 7.503585749693023e-06, + "loss": 0.5917, + "step": 8483 + }, + { + "epoch": 1.7789893059341582, + "grad_norm": 8.336901285689663, + "learning_rate": 7.501393607411816e-06, + "loss": 0.848, + "step": 8484 + }, + { + "epoch": 1.7791989934996855, + "grad_norm": 6.9987330876139415, + "learning_rate": 7.499201593188517e-06, + "loss": 0.7346, + "step": 8485 + }, + { + "epoch": 1.7794086810652128, + "grad_norm": 6.937464921772398, + "learning_rate": 7.4970097071354695e-06, + "loss": 0.6266, + "step": 8486 + }, + { + "epoch": 1.7796183686307403, + "grad_norm": 7.906609192753642, + "learning_rate": 7.494817949365023e-06, + "loss": 0.9338, + "step": 8487 + }, + { + "epoch": 1.7798280561962676, + "grad_norm": 6.779320388485007, + "learning_rate": 7.4926263199894975e-06, + "loss": 0.5937, + "step": 8488 + }, + { + "epoch": 1.780037743761795, + "grad_norm": 7.1963417171913, + "learning_rate": 7.490434819121224e-06, + "loss": 0.6775, + "step": 8489 + }, + { + "epoch": 1.7802474313273224, + "grad_norm": 6.745052684731232, + "learning_rate": 7.4882434468725185e-06, + "loss": 0.6341, + "step": 8490 + }, + { + "epoch": 1.7804571188928495, + "grad_norm": 7.547785178381326, + "learning_rate": 7.486052203355692e-06, + "loss": 0.7576, + "step": 8491 + }, + { + "epoch": 1.780666806458377, + "grad_norm": 6.939044548480447, + "learning_rate": 7.483861088683047e-06, + "loss": 0.6648, + "step": 8492 + }, + { + "epoch": 1.7808764940239044, + "grad_norm": 7.130755297314795, + "learning_rate": 7.4816701029668915e-06, + "loss": 0.6636, + "step": 8493 + }, + { + "epoch": 1.7810861815894317, + "grad_norm": 7.173704550541052, + "learning_rate": 7.479479246319507e-06, + "loss": 0.7725, + "step": 8494 + }, + { + "epoch": 1.7812958691549592, + "grad_norm": 5.9263373906390635, + "learning_rate": 7.477288518853184e-06, + "loss": 0.6012, + "step": 8495 + }, + { + "epoch": 1.7815055567204865, + "grad_norm": 7.284666260970255, + "learning_rate": 7.475097920680201e-06, + "loss": 0.7472, + "step": 8496 + }, + { + "epoch": 1.7817152442860138, + "grad_norm": 6.493557186018949, + "learning_rate": 7.4729074519128255e-06, + "loss": 0.76, + "step": 8497 + }, + { + "epoch": 1.7819249318515413, + "grad_norm": 5.706050163951453, + "learning_rate": 7.470717112663328e-06, + "loss": 0.476, + "step": 8498 + }, + { + "epoch": 1.7821346194170684, + "grad_norm": 6.739242453364852, + "learning_rate": 7.4685269030439665e-06, + "loss": 0.7578, + "step": 8499 + }, + { + "epoch": 1.782344306982596, + "grad_norm": 7.925675552898601, + "learning_rate": 7.466336823166991e-06, + "loss": 0.6092, + "step": 8500 + }, + { + "epoch": 1.7825539945481232, + "grad_norm": 7.422968994624976, + "learning_rate": 7.464146873144644e-06, + "loss": 0.7752, + "step": 8501 + }, + { + "epoch": 1.7827636821136505, + "grad_norm": 7.101561734647468, + "learning_rate": 7.461957053089176e-06, + "loss": 0.6913, + "step": 8502 + }, + { + "epoch": 1.782973369679178, + "grad_norm": 6.913747249571515, + "learning_rate": 7.459767363112804e-06, + "loss": 0.7032, + "step": 8503 + }, + { + "epoch": 1.7831830572447054, + "grad_norm": 5.962879301941932, + "learning_rate": 7.457577803327762e-06, + "loss": 0.5643, + "step": 8504 + }, + { + "epoch": 1.7833927448102327, + "grad_norm": 6.20392450259821, + "learning_rate": 7.4553883738462705e-06, + "loss": 0.609, + "step": 8505 + }, + { + "epoch": 1.7836024323757602, + "grad_norm": 6.727553472947862, + "learning_rate": 7.453199074780535e-06, + "loss": 0.7763, + "step": 8506 + }, + { + "epoch": 1.7838121199412875, + "grad_norm": 8.43277264679798, + "learning_rate": 7.4510099062427635e-06, + "loss": 0.6382, + "step": 8507 + }, + { + "epoch": 1.7840218075068148, + "grad_norm": 7.862389121160747, + "learning_rate": 7.4488208683451605e-06, + "loss": 0.896, + "step": 8508 + }, + { + "epoch": 1.7842314950723424, + "grad_norm": 8.268207906810831, + "learning_rate": 7.446631961199906e-06, + "loss": 0.8226, + "step": 8509 + }, + { + "epoch": 1.7844411826378694, + "grad_norm": 5.973492408226178, + "learning_rate": 7.444443184919194e-06, + "loss": 0.6757, + "step": 8510 + }, + { + "epoch": 1.784650870203397, + "grad_norm": 7.019481034117828, + "learning_rate": 7.4422545396152045e-06, + "loss": 0.6544, + "step": 8511 + }, + { + "epoch": 1.7848605577689243, + "grad_norm": 6.89496597724163, + "learning_rate": 7.440066025400104e-06, + "loss": 0.6717, + "step": 8512 + }, + { + "epoch": 1.7850702453344516, + "grad_norm": 6.399418888646156, + "learning_rate": 7.437877642386057e-06, + "loss": 0.5971, + "step": 8513 + }, + { + "epoch": 1.785279932899979, + "grad_norm": 4.718987828578632, + "learning_rate": 7.435689390685229e-06, + "loss": 0.4631, + "step": 8514 + }, + { + "epoch": 1.7854896204655064, + "grad_norm": 7.042004414252741, + "learning_rate": 7.433501270409764e-06, + "loss": 0.8621, + "step": 8515 + }, + { + "epoch": 1.7856993080310337, + "grad_norm": 6.3338115064525695, + "learning_rate": 7.431313281671813e-06, + "loss": 0.7602, + "step": 8516 + }, + { + "epoch": 1.7859089955965612, + "grad_norm": 7.500686755005799, + "learning_rate": 7.429125424583511e-06, + "loss": 0.7085, + "step": 8517 + }, + { + "epoch": 1.7861186831620883, + "grad_norm": 7.239593402366131, + "learning_rate": 7.426937699256986e-06, + "loss": 0.666, + "step": 8518 + }, + { + "epoch": 1.7863283707276159, + "grad_norm": 5.803147653713699, + "learning_rate": 7.424750105804371e-06, + "loss": 0.6201, + "step": 8519 + }, + { + "epoch": 1.7865380582931434, + "grad_norm": 8.358827457366093, + "learning_rate": 7.4225626443377784e-06, + "loss": 0.8998, + "step": 8520 + }, + { + "epoch": 1.7867477458586705, + "grad_norm": 8.465956629514688, + "learning_rate": 7.420375314969322e-06, + "loss": 0.7821, + "step": 8521 + }, + { + "epoch": 1.786957433424198, + "grad_norm": 7.5536516282182715, + "learning_rate": 7.418188117811102e-06, + "loss": 0.9381, + "step": 8522 + }, + { + "epoch": 1.7871671209897253, + "grad_norm": 5.523300575558262, + "learning_rate": 7.416001052975221e-06, + "loss": 0.4752, + "step": 8523 + }, + { + "epoch": 1.7873768085552526, + "grad_norm": 7.004006369662888, + "learning_rate": 7.413814120573765e-06, + "loss": 0.7069, + "step": 8524 + }, + { + "epoch": 1.7875864961207801, + "grad_norm": 7.270733864256641, + "learning_rate": 7.411627320718824e-06, + "loss": 0.726, + "step": 8525 + }, + { + "epoch": 1.7877961836863074, + "grad_norm": 7.92000082896209, + "learning_rate": 7.4094406535224725e-06, + "loss": 0.7617, + "step": 8526 + }, + { + "epoch": 1.7880058712518347, + "grad_norm": 7.920088044009678, + "learning_rate": 7.407254119096778e-06, + "loss": 0.8987, + "step": 8527 + }, + { + "epoch": 1.7882155588173623, + "grad_norm": 7.152442108771772, + "learning_rate": 7.4050677175538055e-06, + "loss": 0.7, + "step": 8528 + }, + { + "epoch": 1.7884252463828894, + "grad_norm": 6.0669959777087294, + "learning_rate": 7.40288144900562e-06, + "loss": 0.5454, + "step": 8529 + }, + { + "epoch": 1.7886349339484169, + "grad_norm": 7.482284654172083, + "learning_rate": 7.4006953135642565e-06, + "loss": 0.708, + "step": 8530 + }, + { + "epoch": 1.7888446215139442, + "grad_norm": 8.556294684349144, + "learning_rate": 7.39850931134177e-06, + "loss": 0.9896, + "step": 8531 + }, + { + "epoch": 1.7890543090794715, + "grad_norm": 5.806225283537285, + "learning_rate": 7.396323442450192e-06, + "loss": 0.6357, + "step": 8532 + }, + { + "epoch": 1.789263996644999, + "grad_norm": 6.031312566176968, + "learning_rate": 7.394137707001558e-06, + "loss": 0.7069, + "step": 8533 + }, + { + "epoch": 1.7894736842105263, + "grad_norm": 6.9112563648421395, + "learning_rate": 7.3919521051078814e-06, + "loss": 0.7064, + "step": 8534 + }, + { + "epoch": 1.7896833717760536, + "grad_norm": 6.784390587780009, + "learning_rate": 7.389766636881184e-06, + "loss": 0.7397, + "step": 8535 + }, + { + "epoch": 1.7898930593415812, + "grad_norm": 6.9086898918562705, + "learning_rate": 7.387581302433476e-06, + "loss": 0.8434, + "step": 8536 + }, + { + "epoch": 1.7901027469071082, + "grad_norm": 7.160303378245445, + "learning_rate": 7.385396101876756e-06, + "loss": 0.8045, + "step": 8537 + }, + { + "epoch": 1.7903124344726358, + "grad_norm": 7.389959177808936, + "learning_rate": 7.383211035323017e-06, + "loss": 0.7965, + "step": 8538 + }, + { + "epoch": 1.7905221220381633, + "grad_norm": 6.4625515188848945, + "learning_rate": 7.3810261028842585e-06, + "loss": 0.5236, + "step": 8539 + }, + { + "epoch": 1.7907318096036904, + "grad_norm": 7.619907526106467, + "learning_rate": 7.378841304672448e-06, + "loss": 0.7905, + "step": 8540 + }, + { + "epoch": 1.790941497169218, + "grad_norm": 5.609252433248018, + "learning_rate": 7.376656640799569e-06, + "loss": 0.375, + "step": 8541 + }, + { + "epoch": 1.7911511847347452, + "grad_norm": 6.638066150542521, + "learning_rate": 7.374472111377588e-06, + "loss": 0.7529, + "step": 8542 + }, + { + "epoch": 1.7913608723002725, + "grad_norm": 8.690569909822019, + "learning_rate": 7.372287716518465e-06, + "loss": 1.0888, + "step": 8543 + }, + { + "epoch": 1.7915705598658, + "grad_norm": 7.669644012449896, + "learning_rate": 7.37010345633415e-06, + "loss": 0.5809, + "step": 8544 + }, + { + "epoch": 1.7917802474313274, + "grad_norm": 5.680955272776126, + "learning_rate": 7.367919330936599e-06, + "loss": 0.5131, + "step": 8545 + }, + { + "epoch": 1.7919899349968547, + "grad_norm": 7.934008735150557, + "learning_rate": 7.365735340437745e-06, + "loss": 0.7781, + "step": 8546 + }, + { + "epoch": 1.7921996225623822, + "grad_norm": 7.231129444804259, + "learning_rate": 7.363551484949525e-06, + "loss": 0.8326, + "step": 8547 + }, + { + "epoch": 1.7924093101279093, + "grad_norm": 5.933860113921983, + "learning_rate": 7.361367764583866e-06, + "loss": 0.6042, + "step": 8548 + }, + { + "epoch": 1.7926189976934368, + "grad_norm": 7.787922775329241, + "learning_rate": 7.359184179452681e-06, + "loss": 0.7345, + "step": 8549 + }, + { + "epoch": 1.792828685258964, + "grad_norm": 6.746325515798119, + "learning_rate": 7.357000729667888e-06, + "loss": 0.6081, + "step": 8550 + }, + { + "epoch": 1.7930383728244914, + "grad_norm": 6.325859698100907, + "learning_rate": 7.354817415341395e-06, + "loss": 0.7407, + "step": 8551 + }, + { + "epoch": 1.793248060390019, + "grad_norm": 7.8733365791769865, + "learning_rate": 7.352634236585095e-06, + "loss": 0.7032, + "step": 8552 + }, + { + "epoch": 1.7934577479555462, + "grad_norm": 7.993327796391577, + "learning_rate": 7.350451193510882e-06, + "loss": 0.8547, + "step": 8553 + }, + { + "epoch": 1.7936674355210735, + "grad_norm": 4.796742340359671, + "learning_rate": 7.348268286230642e-06, + "loss": 0.4277, + "step": 8554 + }, + { + "epoch": 1.793877123086601, + "grad_norm": 6.6736381070831285, + "learning_rate": 7.346085514856248e-06, + "loss": 0.9108, + "step": 8555 + }, + { + "epoch": 1.7940868106521284, + "grad_norm": 5.696958726823761, + "learning_rate": 7.343902879499576e-06, + "loss": 0.5378, + "step": 8556 + }, + { + "epoch": 1.7942964982176557, + "grad_norm": 7.895400209230349, + "learning_rate": 7.3417203802724915e-06, + "loss": 0.8163, + "step": 8557 + }, + { + "epoch": 1.7945061857831832, + "grad_norm": 5.722944233519205, + "learning_rate": 7.339538017286844e-06, + "loss": 0.4031, + "step": 8558 + }, + { + "epoch": 1.7947158733487103, + "grad_norm": 6.369467306371739, + "learning_rate": 7.337355790654488e-06, + "loss": 0.6541, + "step": 8559 + }, + { + "epoch": 1.7949255609142378, + "grad_norm": 7.154725611267172, + "learning_rate": 7.3351737004872705e-06, + "loss": 0.7313, + "step": 8560 + }, + { + "epoch": 1.7951352484797651, + "grad_norm": 5.957504054520665, + "learning_rate": 7.332991746897016e-06, + "loss": 0.4537, + "step": 8561 + }, + { + "epoch": 1.7953449360452924, + "grad_norm": 8.967212137890364, + "learning_rate": 7.330809929995564e-06, + "loss": 1.1113, + "step": 8562 + }, + { + "epoch": 1.79555462361082, + "grad_norm": 6.882344923905063, + "learning_rate": 7.3286282498947335e-06, + "loss": 0.8033, + "step": 8563 + }, + { + "epoch": 1.7957643111763473, + "grad_norm": 8.147665361420723, + "learning_rate": 7.326446706706336e-06, + "loss": 0.6587, + "step": 8564 + }, + { + "epoch": 1.7959739987418746, + "grad_norm": 6.515093185369705, + "learning_rate": 7.32426530054218e-06, + "loss": 0.6989, + "step": 8565 + }, + { + "epoch": 1.796183686307402, + "grad_norm": 6.400558489849805, + "learning_rate": 7.322084031514073e-06, + "loss": 0.6788, + "step": 8566 + }, + { + "epoch": 1.7963933738729292, + "grad_norm": 7.266156517860445, + "learning_rate": 7.319902899733802e-06, + "loss": 0.715, + "step": 8567 + }, + { + "epoch": 1.7966030614384567, + "grad_norm": 7.804231603393305, + "learning_rate": 7.317721905313155e-06, + "loss": 0.8506, + "step": 8568 + }, + { + "epoch": 1.796812749003984, + "grad_norm": 7.07112405608429, + "learning_rate": 7.3155410483639145e-06, + "loss": 0.6164, + "step": 8569 + }, + { + "epoch": 1.7970224365695113, + "grad_norm": 7.749679480450499, + "learning_rate": 7.313360328997849e-06, + "loss": 0.7337, + "step": 8570 + }, + { + "epoch": 1.7972321241350389, + "grad_norm": 7.732527425478706, + "learning_rate": 7.311179747326725e-06, + "loss": 0.6781, + "step": 8571 + }, + { + "epoch": 1.7974418117005662, + "grad_norm": 9.026050454145706, + "learning_rate": 7.3089993034623076e-06, + "loss": 1.0107, + "step": 8572 + }, + { + "epoch": 1.7976514992660935, + "grad_norm": 7.459617566269735, + "learning_rate": 7.306818997516338e-06, + "loss": 0.8722, + "step": 8573 + }, + { + "epoch": 1.797861186831621, + "grad_norm": 7.744838682897569, + "learning_rate": 7.304638829600569e-06, + "loss": 0.774, + "step": 8574 + }, + { + "epoch": 1.7980708743971483, + "grad_norm": 10.090243215329245, + "learning_rate": 7.302458799826735e-06, + "loss": 0.9973, + "step": 8575 + }, + { + "epoch": 1.7982805619626756, + "grad_norm": 6.0432982918975675, + "learning_rate": 7.300278908306564e-06, + "loss": 0.6413, + "step": 8576 + }, + { + "epoch": 1.7984902495282031, + "grad_norm": 7.019722322045745, + "learning_rate": 7.298099155151781e-06, + "loss": 0.7167, + "step": 8577 + }, + { + "epoch": 1.7986999370937302, + "grad_norm": 5.509042752792497, + "learning_rate": 7.2959195404741056e-06, + "loss": 0.5491, + "step": 8578 + }, + { + "epoch": 1.7989096246592577, + "grad_norm": 5.629109441172555, + "learning_rate": 7.293740064385242e-06, + "loss": 0.6278, + "step": 8579 + }, + { + "epoch": 1.799119312224785, + "grad_norm": 7.553742932804851, + "learning_rate": 7.291560726996892e-06, + "loss": 0.8958, + "step": 8580 + }, + { + "epoch": 1.7993289997903124, + "grad_norm": 5.150973779562384, + "learning_rate": 7.2893815284207585e-06, + "loss": 0.5672, + "step": 8581 + }, + { + "epoch": 1.7995386873558399, + "grad_norm": 7.259753547167847, + "learning_rate": 7.287202468768516e-06, + "loss": 0.6715, + "step": 8582 + }, + { + "epoch": 1.7997483749213672, + "grad_norm": 8.96453074114518, + "learning_rate": 7.285023548151855e-06, + "loss": 0.9256, + "step": 8583 + }, + { + "epoch": 1.7999580624868945, + "grad_norm": 6.542165717479757, + "learning_rate": 7.282844766682448e-06, + "loss": 0.8384, + "step": 8584 + }, + { + "epoch": 1.800167750052422, + "grad_norm": 6.186439156804989, + "learning_rate": 7.2806661244719565e-06, + "loss": 0.6021, + "step": 8585 + }, + { + "epoch": 1.800377437617949, + "grad_norm": 6.095291406134879, + "learning_rate": 7.278487621632041e-06, + "loss": 0.837, + "step": 8586 + }, + { + "epoch": 1.8005871251834766, + "grad_norm": 6.684685476075956, + "learning_rate": 7.276309258274361e-06, + "loss": 0.6958, + "step": 8587 + }, + { + "epoch": 1.800796812749004, + "grad_norm": 8.153883895466917, + "learning_rate": 7.27413103451055e-06, + "loss": 0.6804, + "step": 8588 + }, + { + "epoch": 1.8010065003145312, + "grad_norm": 5.842255463706033, + "learning_rate": 7.271952950452253e-06, + "loss": 0.714, + "step": 8589 + }, + { + "epoch": 1.8012161878800588, + "grad_norm": 6.158669257345333, + "learning_rate": 7.269775006211101e-06, + "loss": 0.6882, + "step": 8590 + }, + { + "epoch": 1.801425875445586, + "grad_norm": 5.356304311678659, + "learning_rate": 7.267597201898713e-06, + "loss": 0.5209, + "step": 8591 + }, + { + "epoch": 1.8016355630111134, + "grad_norm": 6.198955110233281, + "learning_rate": 7.265419537626705e-06, + "loss": 0.561, + "step": 8592 + }, + { + "epoch": 1.801845250576641, + "grad_norm": 6.3374765446001255, + "learning_rate": 7.263242013506692e-06, + "loss": 0.6499, + "step": 8593 + }, + { + "epoch": 1.8020549381421682, + "grad_norm": 7.687225966880493, + "learning_rate": 7.261064629650274e-06, + "loss": 0.8481, + "step": 8594 + }, + { + "epoch": 1.8022646257076955, + "grad_norm": 7.761273975616362, + "learning_rate": 7.258887386169041e-06, + "loss": 0.7833, + "step": 8595 + }, + { + "epoch": 1.802474313273223, + "grad_norm": 8.124199805722295, + "learning_rate": 7.2567102831745824e-06, + "loss": 0.83, + "step": 8596 + }, + { + "epoch": 1.8026840008387501, + "grad_norm": 6.028072353623083, + "learning_rate": 7.2545333207784854e-06, + "loss": 0.6831, + "step": 8597 + }, + { + "epoch": 1.8028936884042777, + "grad_norm": 7.268137658346249, + "learning_rate": 7.252356499092312e-06, + "loss": 0.5859, + "step": 8598 + }, + { + "epoch": 1.803103375969805, + "grad_norm": 8.369194385574005, + "learning_rate": 7.250179818227636e-06, + "loss": 0.8506, + "step": 8599 + }, + { + "epoch": 1.8033130635353323, + "grad_norm": 6.2456208268317415, + "learning_rate": 7.248003278296016e-06, + "loss": 0.7053, + "step": 8600 + }, + { + "epoch": 1.8035227511008598, + "grad_norm": 5.850378126886383, + "learning_rate": 7.245826879408998e-06, + "loss": 0.5834, + "step": 8601 + }, + { + "epoch": 1.803732438666387, + "grad_norm": 5.33436037814003, + "learning_rate": 7.243650621678128e-06, + "loss": 0.5472, + "step": 8602 + }, + { + "epoch": 1.8039421262319144, + "grad_norm": 7.252917637319995, + "learning_rate": 7.2414745052149475e-06, + "loss": 0.6878, + "step": 8603 + }, + { + "epoch": 1.804151813797442, + "grad_norm": 6.694583813107809, + "learning_rate": 7.239298530130982e-06, + "loss": 0.7088, + "step": 8604 + }, + { + "epoch": 1.804361501362969, + "grad_norm": 6.728995746208721, + "learning_rate": 7.237122696537754e-06, + "loss": 0.5775, + "step": 8605 + }, + { + "epoch": 1.8045711889284966, + "grad_norm": 5.606599004951529, + "learning_rate": 7.234947004546782e-06, + "loss": 0.526, + "step": 8606 + }, + { + "epoch": 1.8047808764940239, + "grad_norm": 6.641404477821367, + "learning_rate": 7.232771454269567e-06, + "loss": 0.7342, + "step": 8607 + }, + { + "epoch": 1.8049905640595512, + "grad_norm": 6.442118483887747, + "learning_rate": 7.230596045817619e-06, + "loss": 0.5325, + "step": 8608 + }, + { + "epoch": 1.8052002516250787, + "grad_norm": 7.141080158515472, + "learning_rate": 7.228420779302427e-06, + "loss": 0.7336, + "step": 8609 + }, + { + "epoch": 1.805409939190606, + "grad_norm": 7.729646547936568, + "learning_rate": 7.226245654835475e-06, + "loss": 0.7979, + "step": 8610 + }, + { + "epoch": 1.8056196267561333, + "grad_norm": 9.018149517894344, + "learning_rate": 7.224070672528243e-06, + "loss": 1.1419, + "step": 8611 + }, + { + "epoch": 1.8058293143216608, + "grad_norm": 6.684690663582744, + "learning_rate": 7.2218958324922095e-06, + "loss": 0.7942, + "step": 8612 + }, + { + "epoch": 1.8060390018871881, + "grad_norm": 6.892969088223079, + "learning_rate": 7.219721134838826e-06, + "loss": 0.7447, + "step": 8613 + }, + { + "epoch": 1.8062486894527154, + "grad_norm": 9.89638321122598, + "learning_rate": 7.21754657967956e-06, + "loss": 0.884, + "step": 8614 + }, + { + "epoch": 1.806458377018243, + "grad_norm": 7.211392984439015, + "learning_rate": 7.215372167125858e-06, + "loss": 0.7147, + "step": 8615 + }, + { + "epoch": 1.80666806458377, + "grad_norm": 5.659392056170881, + "learning_rate": 7.213197897289161e-06, + "loss": 0.5751, + "step": 8616 + }, + { + "epoch": 1.8068777521492976, + "grad_norm": 6.36876674444938, + "learning_rate": 7.211023770280903e-06, + "loss": 0.6965, + "step": 8617 + }, + { + "epoch": 1.8070874397148249, + "grad_norm": 8.624524251542926, + "learning_rate": 7.208849786212519e-06, + "loss": 0.9057, + "step": 8618 + }, + { + "epoch": 1.8072971272803522, + "grad_norm": 6.646663165931561, + "learning_rate": 7.20667594519542e-06, + "loss": 0.6309, + "step": 8619 + }, + { + "epoch": 1.8075068148458797, + "grad_norm": 6.0118187343916185, + "learning_rate": 7.204502247341025e-06, + "loss": 0.5764, + "step": 8620 + }, + { + "epoch": 1.807716502411407, + "grad_norm": 5.407080188304323, + "learning_rate": 7.202328692760741e-06, + "loss": 0.47, + "step": 8621 + }, + { + "epoch": 1.8079261899769343, + "grad_norm": 6.98178398633119, + "learning_rate": 7.200155281565961e-06, + "loss": 0.8821, + "step": 8622 + }, + { + "epoch": 1.8081358775424619, + "grad_norm": 8.601075894487982, + "learning_rate": 7.1979820138680786e-06, + "loss": 1.015, + "step": 8623 + }, + { + "epoch": 1.808345565107989, + "grad_norm": 7.087302266044096, + "learning_rate": 7.195808889778479e-06, + "loss": 0.7439, + "step": 8624 + }, + { + "epoch": 1.8085552526735165, + "grad_norm": 6.338547635960465, + "learning_rate": 7.193635909408538e-06, + "loss": 0.6436, + "step": 8625 + }, + { + "epoch": 1.8087649402390438, + "grad_norm": 5.580416864208404, + "learning_rate": 7.191463072869624e-06, + "loss": 0.5888, + "step": 8626 + }, + { + "epoch": 1.808974627804571, + "grad_norm": 7.260647629594801, + "learning_rate": 7.1892903802731e-06, + "loss": 0.7591, + "step": 8627 + }, + { + "epoch": 1.8091843153700986, + "grad_norm": 7.203299896522841, + "learning_rate": 7.187117831730314e-06, + "loss": 0.8843, + "step": 8628 + }, + { + "epoch": 1.809394002935626, + "grad_norm": 9.300361403530278, + "learning_rate": 7.18494542735262e-06, + "loss": 0.9758, + "step": 8629 + }, + { + "epoch": 1.8096036905011532, + "grad_norm": 5.66501252921601, + "learning_rate": 7.182773167251358e-06, + "loss": 0.6254, + "step": 8630 + }, + { + "epoch": 1.8098133780666807, + "grad_norm": 6.593871328516726, + "learning_rate": 7.180601051537854e-06, + "loss": 0.5968, + "step": 8631 + }, + { + "epoch": 1.810023065632208, + "grad_norm": 6.824785349043687, + "learning_rate": 7.178429080323438e-06, + "loss": 0.5883, + "step": 8632 + }, + { + "epoch": 1.8102327531977354, + "grad_norm": 5.438131640321373, + "learning_rate": 7.1762572537194275e-06, + "loss": 0.5901, + "step": 8633 + }, + { + "epoch": 1.8104424407632629, + "grad_norm": 7.6265716733312665, + "learning_rate": 7.1740855718371234e-06, + "loss": 0.8291, + "step": 8634 + }, + { + "epoch": 1.81065212832879, + "grad_norm": 4.538021245207589, + "learning_rate": 7.17191403478784e-06, + "loss": 0.31, + "step": 8635 + }, + { + "epoch": 1.8108618158943175, + "grad_norm": 7.231341794372251, + "learning_rate": 7.169742642682868e-06, + "loss": 0.7058, + "step": 8636 + }, + { + "epoch": 1.8110715034598448, + "grad_norm": 7.303722398924656, + "learning_rate": 7.167571395633494e-06, + "loss": 0.5058, + "step": 8637 + }, + { + "epoch": 1.811281191025372, + "grad_norm": 5.750226412531893, + "learning_rate": 7.165400293750995e-06, + "loss": 0.6363, + "step": 8638 + }, + { + "epoch": 1.8114908785908996, + "grad_norm": 9.850912588086373, + "learning_rate": 7.163229337146653e-06, + "loss": 1.0093, + "step": 8639 + }, + { + "epoch": 1.811700566156427, + "grad_norm": 6.690438615366756, + "learning_rate": 7.161058525931724e-06, + "loss": 0.6675, + "step": 8640 + }, + { + "epoch": 1.8119102537219542, + "grad_norm": 6.394408418976392, + "learning_rate": 7.1588878602174696e-06, + "loss": 0.7519, + "step": 8641 + }, + { + "epoch": 1.8121199412874818, + "grad_norm": 5.242323509816903, + "learning_rate": 7.1567173401151425e-06, + "loss": 0.4907, + "step": 8642 + }, + { + "epoch": 1.8123296288530089, + "grad_norm": 8.752254718827217, + "learning_rate": 7.15454696573598e-06, + "loss": 0.7236, + "step": 8643 + }, + { + "epoch": 1.8125393164185364, + "grad_norm": 8.620133056919888, + "learning_rate": 7.15237673719122e-06, + "loss": 1.0653, + "step": 8644 + }, + { + "epoch": 1.812749003984064, + "grad_norm": 8.172353012991103, + "learning_rate": 7.150206654592094e-06, + "loss": 0.9265, + "step": 8645 + }, + { + "epoch": 1.812958691549591, + "grad_norm": 6.468407315325875, + "learning_rate": 7.148036718049819e-06, + "loss": 0.6829, + "step": 8646 + }, + { + "epoch": 1.8131683791151185, + "grad_norm": 5.919430984380698, + "learning_rate": 7.1458669276756075e-06, + "loss": 0.745, + "step": 8647 + }, + { + "epoch": 1.8133780666806458, + "grad_norm": 5.721880332891165, + "learning_rate": 7.1436972835806685e-06, + "loss": 0.5213, + "step": 8648 + }, + { + "epoch": 1.8135877542461731, + "grad_norm": 8.489978423340053, + "learning_rate": 7.141527785876197e-06, + "loss": 0.8967, + "step": 8649 + }, + { + "epoch": 1.8137974418117007, + "grad_norm": 6.625590841525813, + "learning_rate": 7.139358434673381e-06, + "loss": 0.7818, + "step": 8650 + }, + { + "epoch": 1.814007129377228, + "grad_norm": 6.432927693583469, + "learning_rate": 7.13718923008341e-06, + "loss": 0.7661, + "step": 8651 + }, + { + "epoch": 1.8142168169427553, + "grad_norm": 9.05009793219841, + "learning_rate": 7.1350201722174564e-06, + "loss": 1.0373, + "step": 8652 + }, + { + "epoch": 1.8144265045082828, + "grad_norm": 6.20000010135193, + "learning_rate": 7.132851261186686e-06, + "loss": 0.4625, + "step": 8653 + }, + { + "epoch": 1.8146361920738099, + "grad_norm": 9.989354991151128, + "learning_rate": 7.13068249710226e-06, + "loss": 0.9588, + "step": 8654 + }, + { + "epoch": 1.8148458796393374, + "grad_norm": 6.881295863547874, + "learning_rate": 7.128513880075338e-06, + "loss": 0.7785, + "step": 8655 + }, + { + "epoch": 1.8150555672048647, + "grad_norm": 7.630944376028559, + "learning_rate": 7.126345410217056e-06, + "loss": 0.7834, + "step": 8656 + }, + { + "epoch": 1.815265254770392, + "grad_norm": 6.386635629013877, + "learning_rate": 7.124177087638556e-06, + "loss": 0.7218, + "step": 8657 + }, + { + "epoch": 1.8154749423359196, + "grad_norm": 7.718779493624984, + "learning_rate": 7.12200891245097e-06, + "loss": 0.7711, + "step": 8658 + }, + { + "epoch": 1.8156846299014469, + "grad_norm": 8.103924425147584, + "learning_rate": 7.119840884765417e-06, + "loss": 0.8802, + "step": 8659 + }, + { + "epoch": 1.8158943174669742, + "grad_norm": 7.0877105041431925, + "learning_rate": 7.117673004693014e-06, + "loss": 0.5965, + "step": 8660 + }, + { + "epoch": 1.8161040050325017, + "grad_norm": 9.442081468503812, + "learning_rate": 7.1155052723448715e-06, + "loss": 0.9724, + "step": 8661 + }, + { + "epoch": 1.8163136925980288, + "grad_norm": 6.303786693800357, + "learning_rate": 7.113337687832085e-06, + "loss": 0.5789, + "step": 8662 + }, + { + "epoch": 1.8165233801635563, + "grad_norm": 5.963784002220839, + "learning_rate": 7.111170251265748e-06, + "loss": 0.5568, + "step": 8663 + }, + { + "epoch": 1.8167330677290838, + "grad_norm": 6.82532624503764, + "learning_rate": 7.109002962756948e-06, + "loss": 0.7707, + "step": 8664 + }, + { + "epoch": 1.816942755294611, + "grad_norm": 7.4172184917790105, + "learning_rate": 7.1068358224167555e-06, + "loss": 0.8225, + "step": 8665 + }, + { + "epoch": 1.8171524428601384, + "grad_norm": 6.176463429518696, + "learning_rate": 7.104668830356249e-06, + "loss": 0.5975, + "step": 8666 + }, + { + "epoch": 1.8173621304256657, + "grad_norm": 6.385399875021124, + "learning_rate": 7.1025019866864876e-06, + "loss": 0.724, + "step": 8667 + }, + { + "epoch": 1.817571817991193, + "grad_norm": 6.635831513129939, + "learning_rate": 7.100335291518523e-06, + "loss": 0.7083, + "step": 8668 + }, + { + "epoch": 1.8177815055567206, + "grad_norm": 7.51733592573119, + "learning_rate": 7.098168744963401e-06, + "loss": 0.7331, + "step": 8669 + }, + { + "epoch": 1.8179911931222479, + "grad_norm": 7.3677078143731825, + "learning_rate": 7.0960023471321715e-06, + "loss": 0.8664, + "step": 8670 + }, + { + "epoch": 1.8182008806877752, + "grad_norm": 7.287418692482427, + "learning_rate": 7.093836098135851e-06, + "loss": 0.725, + "step": 8671 + }, + { + "epoch": 1.8184105682533027, + "grad_norm": 6.158429707255441, + "learning_rate": 7.091669998085473e-06, + "loss": 0.6579, + "step": 8672 + }, + { + "epoch": 1.8186202558188298, + "grad_norm": 7.119260868056512, + "learning_rate": 7.089504047092055e-06, + "loss": 0.7561, + "step": 8673 + }, + { + "epoch": 1.8188299433843573, + "grad_norm": 8.33814240104753, + "learning_rate": 7.0873382452666e-06, + "loss": 0.8211, + "step": 8674 + }, + { + "epoch": 1.8190396309498846, + "grad_norm": 7.602171652949819, + "learning_rate": 7.0851725927201085e-06, + "loss": 0.5861, + "step": 8675 + }, + { + "epoch": 1.819249318515412, + "grad_norm": 6.837141994899123, + "learning_rate": 7.083007089563583e-06, + "loss": 0.6277, + "step": 8676 + }, + { + "epoch": 1.8194590060809395, + "grad_norm": 6.6498267403844, + "learning_rate": 7.080841735908002e-06, + "loss": 0.6819, + "step": 8677 + }, + { + "epoch": 1.8196686936464668, + "grad_norm": 7.066677932566973, + "learning_rate": 7.078676531864344e-06, + "loss": 0.7498, + "step": 8678 + }, + { + "epoch": 1.819878381211994, + "grad_norm": 6.335117026203028, + "learning_rate": 7.076511477543582e-06, + "loss": 0.6251, + "step": 8679 + }, + { + "epoch": 1.8200880687775216, + "grad_norm": 6.909444284393642, + "learning_rate": 7.074346573056674e-06, + "loss": 0.6151, + "step": 8680 + }, + { + "epoch": 1.820297756343049, + "grad_norm": 7.378694225817318, + "learning_rate": 7.072181818514578e-06, + "loss": 0.8965, + "step": 8681 + }, + { + "epoch": 1.8205074439085762, + "grad_norm": 7.753541982295481, + "learning_rate": 7.070017214028246e-06, + "loss": 0.7945, + "step": 8682 + }, + { + "epoch": 1.8207171314741037, + "grad_norm": 7.027446948571806, + "learning_rate": 7.067852759708611e-06, + "loss": 0.5932, + "step": 8683 + }, + { + "epoch": 1.8209268190396308, + "grad_norm": 9.056426725274095, + "learning_rate": 7.065688455666608e-06, + "loss": 0.9956, + "step": 8684 + }, + { + "epoch": 1.8211365066051584, + "grad_norm": 6.506502036862909, + "learning_rate": 7.063524302013164e-06, + "loss": 0.7481, + "step": 8685 + }, + { + "epoch": 1.8213461941706857, + "grad_norm": 7.794916094646638, + "learning_rate": 7.0613602988591875e-06, + "loss": 0.7796, + "step": 8686 + }, + { + "epoch": 1.821555881736213, + "grad_norm": 7.433348913014491, + "learning_rate": 7.0591964463155944e-06, + "loss": 0.7806, + "step": 8687 + }, + { + "epoch": 1.8217655693017405, + "grad_norm": 6.566015970731753, + "learning_rate": 7.057032744493286e-06, + "loss": 0.5266, + "step": 8688 + }, + { + "epoch": 1.8219752568672678, + "grad_norm": 6.985525387980135, + "learning_rate": 7.054869193503152e-06, + "loss": 0.8061, + "step": 8689 + }, + { + "epoch": 1.822184944432795, + "grad_norm": 6.51650743246444, + "learning_rate": 7.0527057934560785e-06, + "loss": 0.6479, + "step": 8690 + }, + { + "epoch": 1.8223946319983226, + "grad_norm": 6.6337575682891075, + "learning_rate": 7.050542544462951e-06, + "loss": 0.7169, + "step": 8691 + }, + { + "epoch": 1.8226043195638497, + "grad_norm": 8.21266523067796, + "learning_rate": 7.048379446634625e-06, + "loss": 0.9336, + "step": 8692 + }, + { + "epoch": 1.8228140071293772, + "grad_norm": 7.012486241821403, + "learning_rate": 7.046216500081976e-06, + "loss": 0.7059, + "step": 8693 + }, + { + "epoch": 1.8230236946949046, + "grad_norm": 7.23715081602742, + "learning_rate": 7.044053704915857e-06, + "loss": 0.6544, + "step": 8694 + }, + { + "epoch": 1.8232333822604319, + "grad_norm": 6.672387298864582, + "learning_rate": 7.041891061247108e-06, + "loss": 0.7263, + "step": 8695 + }, + { + "epoch": 1.8234430698259594, + "grad_norm": 8.220262527893999, + "learning_rate": 7.039728569186572e-06, + "loss": 1.0388, + "step": 8696 + }, + { + "epoch": 1.8236527573914867, + "grad_norm": 6.360117539432224, + "learning_rate": 7.037566228845087e-06, + "loss": 0.7424, + "step": 8697 + }, + { + "epoch": 1.823862444957014, + "grad_norm": 7.8922182913005665, + "learning_rate": 7.0354040403334666e-06, + "loss": 0.7306, + "step": 8698 + }, + { + "epoch": 1.8240721325225415, + "grad_norm": 7.8680786219552, + "learning_rate": 7.03324200376253e-06, + "loss": 0.9052, + "step": 8699 + }, + { + "epoch": 1.8242818200880688, + "grad_norm": 5.514099947851406, + "learning_rate": 7.031080119243091e-06, + "loss": 0.6161, + "step": 8700 + }, + { + "epoch": 1.8244915076535961, + "grad_norm": 6.2165817198942595, + "learning_rate": 7.028918386885942e-06, + "loss": 0.7789, + "step": 8701 + }, + { + "epoch": 1.8247011952191237, + "grad_norm": 6.87572875143172, + "learning_rate": 7.026756806801876e-06, + "loss": 0.5704, + "step": 8702 + }, + { + "epoch": 1.8249108827846507, + "grad_norm": 7.222352085187118, + "learning_rate": 7.024595379101686e-06, + "loss": 0.7873, + "step": 8703 + }, + { + "epoch": 1.8251205703501783, + "grad_norm": 7.076588332161067, + "learning_rate": 7.022434103896141e-06, + "loss": 0.7913, + "step": 8704 + }, + { + "epoch": 1.8253302579157056, + "grad_norm": 5.741026646510769, + "learning_rate": 7.020272981296012e-06, + "loss": 0.6904, + "step": 8705 + }, + { + "epoch": 1.8255399454812329, + "grad_norm": 6.513183262584279, + "learning_rate": 7.018112011412064e-06, + "loss": 0.6948, + "step": 8706 + }, + { + "epoch": 1.8257496330467604, + "grad_norm": 8.135417303975887, + "learning_rate": 7.0159511943550415e-06, + "loss": 0.7501, + "step": 8707 + }, + { + "epoch": 1.8259593206122877, + "grad_norm": 6.2538296083070355, + "learning_rate": 7.013790530235698e-06, + "loss": 0.5833, + "step": 8708 + }, + { + "epoch": 1.826169008177815, + "grad_norm": 5.773500697999223, + "learning_rate": 7.011630019164772e-06, + "loss": 0.5896, + "step": 8709 + }, + { + "epoch": 1.8263786957433426, + "grad_norm": 6.9028336742922916, + "learning_rate": 7.009469661252988e-06, + "loss": 0.6876, + "step": 8710 + }, + { + "epoch": 1.8265883833088696, + "grad_norm": 6.001094809190476, + "learning_rate": 7.007309456611071e-06, + "loss": 0.6823, + "step": 8711 + }, + { + "epoch": 1.8267980708743972, + "grad_norm": 5.912935784258116, + "learning_rate": 7.005149405349737e-06, + "loss": 0.627, + "step": 8712 + }, + { + "epoch": 1.8270077584399245, + "grad_norm": 6.797852956637171, + "learning_rate": 7.002989507579686e-06, + "loss": 0.9309, + "step": 8713 + }, + { + "epoch": 1.8272174460054518, + "grad_norm": 8.351978425462695, + "learning_rate": 7.000829763411623e-06, + "loss": 1.0725, + "step": 8714 + }, + { + "epoch": 1.8274271335709793, + "grad_norm": 7.437460364809131, + "learning_rate": 6.998670172956237e-06, + "loss": 0.9832, + "step": 8715 + }, + { + "epoch": 1.8276368211365066, + "grad_norm": 7.246507728580772, + "learning_rate": 6.996510736324211e-06, + "loss": 0.7723, + "step": 8716 + }, + { + "epoch": 1.827846508702034, + "grad_norm": 5.553491283028539, + "learning_rate": 6.994351453626216e-06, + "loss": 0.5879, + "step": 8717 + }, + { + "epoch": 1.8280561962675614, + "grad_norm": 7.428693580281207, + "learning_rate": 6.992192324972923e-06, + "loss": 0.6408, + "step": 8718 + }, + { + "epoch": 1.8282658838330887, + "grad_norm": 7.578631914773462, + "learning_rate": 6.990033350474993e-06, + "loss": 0.7565, + "step": 8719 + }, + { + "epoch": 1.828475571398616, + "grad_norm": 7.618102656433465, + "learning_rate": 6.987874530243073e-06, + "loss": 0.6942, + "step": 8720 + }, + { + "epoch": 1.8286852589641436, + "grad_norm": 6.297917652123478, + "learning_rate": 6.985715864387803e-06, + "loss": 0.7384, + "step": 8721 + }, + { + "epoch": 1.8288949465296707, + "grad_norm": 7.140646263436267, + "learning_rate": 6.983557353019832e-06, + "loss": 0.7352, + "step": 8722 + }, + { + "epoch": 1.8291046340951982, + "grad_norm": 6.511628378719846, + "learning_rate": 6.9813989962497705e-06, + "loss": 0.8202, + "step": 8723 + }, + { + "epoch": 1.8293143216607255, + "grad_norm": 7.300741156278667, + "learning_rate": 6.979240794188249e-06, + "loss": 0.7353, + "step": 8724 + }, + { + "epoch": 1.8295240092262528, + "grad_norm": 5.860088581820466, + "learning_rate": 6.977082746945877e-06, + "loss": 0.5208, + "step": 8725 + }, + { + "epoch": 1.8297336967917803, + "grad_norm": 6.587206167243888, + "learning_rate": 6.974924854633256e-06, + "loss": 0.668, + "step": 8726 + }, + { + "epoch": 1.8299433843573076, + "grad_norm": 7.5204339568045055, + "learning_rate": 6.9727671173609805e-06, + "loss": 0.7291, + "step": 8727 + }, + { + "epoch": 1.830153071922835, + "grad_norm": 8.918988742637003, + "learning_rate": 6.970609535239646e-06, + "loss": 1.0647, + "step": 8728 + }, + { + "epoch": 1.8303627594883625, + "grad_norm": 5.766941862861837, + "learning_rate": 6.968452108379822e-06, + "loss": 0.6599, + "step": 8729 + }, + { + "epoch": 1.8305724470538896, + "grad_norm": 6.717247972623266, + "learning_rate": 6.966294836892087e-06, + "loss": 0.6757, + "step": 8730 + }, + { + "epoch": 1.830782134619417, + "grad_norm": 8.134540076637986, + "learning_rate": 6.964137720887006e-06, + "loss": 0.8276, + "step": 8731 + }, + { + "epoch": 1.8309918221849444, + "grad_norm": 5.714469471093912, + "learning_rate": 6.96198076047513e-06, + "loss": 0.547, + "step": 8732 + }, + { + "epoch": 1.8312015097504717, + "grad_norm": 8.071601498851622, + "learning_rate": 6.9598239557670066e-06, + "loss": 0.7075, + "step": 8733 + }, + { + "epoch": 1.8314111973159992, + "grad_norm": 8.804782226132483, + "learning_rate": 6.9576673068731835e-06, + "loss": 0.996, + "step": 8734 + }, + { + "epoch": 1.8316208848815265, + "grad_norm": 7.167396000735304, + "learning_rate": 6.955510813904185e-06, + "loss": 0.6136, + "step": 8735 + }, + { + "epoch": 1.8318305724470538, + "grad_norm": 7.678504133295939, + "learning_rate": 6.953354476970538e-06, + "loss": 0.8345, + "step": 8736 + }, + { + "epoch": 1.8320402600125814, + "grad_norm": 6.467309340291007, + "learning_rate": 6.9511982961827594e-06, + "loss": 0.6597, + "step": 8737 + }, + { + "epoch": 1.8322499475781087, + "grad_norm": 6.311879074958667, + "learning_rate": 6.949042271651352e-06, + "loss": 0.5755, + "step": 8738 + }, + { + "epoch": 1.832459635143636, + "grad_norm": 5.4778145734243475, + "learning_rate": 6.9468864034868214e-06, + "loss": 0.6009, + "step": 8739 + }, + { + "epoch": 1.8326693227091635, + "grad_norm": 8.820167555095335, + "learning_rate": 6.944730691799661e-06, + "loss": 0.7453, + "step": 8740 + }, + { + "epoch": 1.8328790102746906, + "grad_norm": 6.646315838742104, + "learning_rate": 6.942575136700349e-06, + "loss": 0.7186, + "step": 8741 + }, + { + "epoch": 1.833088697840218, + "grad_norm": 6.379388715476397, + "learning_rate": 6.940419738299365e-06, + "loss": 0.6422, + "step": 8742 + }, + { + "epoch": 1.8332983854057454, + "grad_norm": 6.994016120924202, + "learning_rate": 6.938264496707178e-06, + "loss": 0.6554, + "step": 8743 + }, + { + "epoch": 1.8335080729712727, + "grad_norm": 6.4505305781359485, + "learning_rate": 6.936109412034241e-06, + "loss": 0.6176, + "step": 8744 + }, + { + "epoch": 1.8337177605368002, + "grad_norm": 6.384236307978722, + "learning_rate": 6.933954484391014e-06, + "loss": 0.5764, + "step": 8745 + }, + { + "epoch": 1.8339274481023276, + "grad_norm": 7.161755463750637, + "learning_rate": 6.931799713887937e-06, + "loss": 0.5691, + "step": 8746 + }, + { + "epoch": 1.8341371356678549, + "grad_norm": 7.594024279432976, + "learning_rate": 6.929645100635447e-06, + "loss": 0.8112, + "step": 8747 + }, + { + "epoch": 1.8343468232333824, + "grad_norm": 7.3238262559629055, + "learning_rate": 6.927490644743967e-06, + "loss": 0.7137, + "step": 8748 + }, + { + "epoch": 1.8345565107989095, + "grad_norm": 6.442704289435585, + "learning_rate": 6.925336346323927e-06, + "loss": 0.545, + "step": 8749 + }, + { + "epoch": 1.834766198364437, + "grad_norm": 6.673162729139319, + "learning_rate": 6.923182205485726e-06, + "loss": 0.526, + "step": 8750 + }, + { + "epoch": 1.8349758859299643, + "grad_norm": 8.109778447028734, + "learning_rate": 6.921028222339776e-06, + "loss": 0.7791, + "step": 8751 + }, + { + "epoch": 1.8351855734954916, + "grad_norm": 7.371887924357651, + "learning_rate": 6.918874396996471e-06, + "loss": 0.7395, + "step": 8752 + }, + { + "epoch": 1.8353952610610191, + "grad_norm": 7.4048190975448955, + "learning_rate": 6.916720729566195e-06, + "loss": 0.7401, + "step": 8753 + }, + { + "epoch": 1.8356049486265464, + "grad_norm": 6.224903839293317, + "learning_rate": 6.914567220159327e-06, + "loss": 0.5301, + "step": 8754 + }, + { + "epoch": 1.8358146361920737, + "grad_norm": 6.720279593371625, + "learning_rate": 6.9124138688862455e-06, + "loss": 0.6154, + "step": 8755 + }, + { + "epoch": 1.8360243237576013, + "grad_norm": 5.906003996573714, + "learning_rate": 6.910260675857305e-06, + "loss": 0.4938, + "step": 8756 + }, + { + "epoch": 1.8362340113231286, + "grad_norm": 7.424182895286606, + "learning_rate": 6.908107641182866e-06, + "loss": 0.8045, + "step": 8757 + }, + { + "epoch": 1.8364436988886559, + "grad_norm": 6.305857360326501, + "learning_rate": 6.9059547649732725e-06, + "loss": 0.8079, + "step": 8758 + }, + { + "epoch": 1.8366533864541834, + "grad_norm": 7.76574959309599, + "learning_rate": 6.903802047338863e-06, + "loss": 0.7858, + "step": 8759 + }, + { + "epoch": 1.8368630740197105, + "grad_norm": 5.981770952663717, + "learning_rate": 6.901649488389967e-06, + "loss": 0.7034, + "step": 8760 + }, + { + "epoch": 1.837072761585238, + "grad_norm": 7.8155487370364165, + "learning_rate": 6.899497088236913e-06, + "loss": 0.9212, + "step": 8761 + }, + { + "epoch": 1.8372824491507653, + "grad_norm": 6.767690774054371, + "learning_rate": 6.8973448469900065e-06, + "loss": 0.7651, + "step": 8762 + }, + { + "epoch": 1.8374921367162926, + "grad_norm": 7.49527758659679, + "learning_rate": 6.89519276475956e-06, + "loss": 0.7666, + "step": 8763 + }, + { + "epoch": 1.8377018242818202, + "grad_norm": 7.307603947944684, + "learning_rate": 6.8930408416558705e-06, + "loss": 0.7377, + "step": 8764 + }, + { + "epoch": 1.8379115118473475, + "grad_norm": 6.804909240447105, + "learning_rate": 6.8908890777892226e-06, + "loss": 0.7796, + "step": 8765 + }, + { + "epoch": 1.8381211994128748, + "grad_norm": 9.027313663196484, + "learning_rate": 6.888737473269904e-06, + "loss": 0.9549, + "step": 8766 + }, + { + "epoch": 1.8383308869784023, + "grad_norm": 5.179351990680766, + "learning_rate": 6.886586028208188e-06, + "loss": 0.4716, + "step": 8767 + }, + { + "epoch": 1.8385405745439294, + "grad_norm": 7.662277965893355, + "learning_rate": 6.884434742714336e-06, + "loss": 0.8249, + "step": 8768 + }, + { + "epoch": 1.838750262109457, + "grad_norm": 5.643284308001007, + "learning_rate": 6.882283616898604e-06, + "loss": 0.4252, + "step": 8769 + }, + { + "epoch": 1.8389599496749844, + "grad_norm": 8.367960931523038, + "learning_rate": 6.880132650871252e-06, + "loss": 0.8399, + "step": 8770 + }, + { + "epoch": 1.8391696372405115, + "grad_norm": 7.8850147648764874, + "learning_rate": 6.877981844742505e-06, + "loss": 0.7551, + "step": 8771 + }, + { + "epoch": 1.839379324806039, + "grad_norm": 6.774967576821188, + "learning_rate": 6.875831198622607e-06, + "loss": 0.737, + "step": 8772 + }, + { + "epoch": 1.8395890123715664, + "grad_norm": 6.511907095093191, + "learning_rate": 6.87368071262178e-06, + "loss": 0.6485, + "step": 8773 + }, + { + "epoch": 1.8397986999370937, + "grad_norm": 6.8029007747581485, + "learning_rate": 6.871530386850236e-06, + "loss": 0.8299, + "step": 8774 + }, + { + "epoch": 1.8400083875026212, + "grad_norm": 6.47913204045999, + "learning_rate": 6.869380221418185e-06, + "loss": 0.68, + "step": 8775 + }, + { + "epoch": 1.8402180750681485, + "grad_norm": 6.56204407618195, + "learning_rate": 6.867230216435829e-06, + "loss": 0.7734, + "step": 8776 + }, + { + "epoch": 1.8404277626336758, + "grad_norm": 7.944365517174927, + "learning_rate": 6.8650803720133595e-06, + "loss": 0.9319, + "step": 8777 + }, + { + "epoch": 1.8406374501992033, + "grad_norm": 7.68424845709979, + "learning_rate": 6.862930688260956e-06, + "loss": 0.9491, + "step": 8778 + }, + { + "epoch": 1.8408471377647304, + "grad_norm": 8.251331752428593, + "learning_rate": 6.860781165288795e-06, + "loss": 0.7613, + "step": 8779 + }, + { + "epoch": 1.841056825330258, + "grad_norm": 8.783563510218759, + "learning_rate": 6.85863180320705e-06, + "loss": 0.9162, + "step": 8780 + }, + { + "epoch": 1.8412665128957852, + "grad_norm": 6.006705760242939, + "learning_rate": 6.856482602125865e-06, + "loss": 0.5131, + "step": 8781 + }, + { + "epoch": 1.8414762004613126, + "grad_norm": 6.4566421755038155, + "learning_rate": 6.854333562155403e-06, + "loss": 0.5553, + "step": 8782 + }, + { + "epoch": 1.84168588802684, + "grad_norm": 7.623377075686887, + "learning_rate": 6.852184683405804e-06, + "loss": 0.739, + "step": 8783 + }, + { + "epoch": 1.8418955755923674, + "grad_norm": 6.777429353134894, + "learning_rate": 6.850035965987196e-06, + "loss": 0.833, + "step": 8784 + }, + { + "epoch": 1.8421052631578947, + "grad_norm": 7.218277731403332, + "learning_rate": 6.847887410009707e-06, + "loss": 0.6661, + "step": 8785 + }, + { + "epoch": 1.8423149507234222, + "grad_norm": 8.087248565514475, + "learning_rate": 6.845739015583459e-06, + "loss": 0.7049, + "step": 8786 + }, + { + "epoch": 1.8425246382889493, + "grad_norm": 8.209382688074394, + "learning_rate": 6.843590782818556e-06, + "loss": 0.9034, + "step": 8787 + }, + { + "epoch": 1.8427343258544768, + "grad_norm": 7.292294602632365, + "learning_rate": 6.841442711825099e-06, + "loss": 0.8407, + "step": 8788 + }, + { + "epoch": 1.8429440134200044, + "grad_norm": 7.587416640934165, + "learning_rate": 6.839294802713183e-06, + "loss": 0.9503, + "step": 8789 + }, + { + "epoch": 1.8431537009855314, + "grad_norm": 6.88882586775115, + "learning_rate": 6.8371470555928895e-06, + "loss": 0.6873, + "step": 8790 + }, + { + "epoch": 1.843363388551059, + "grad_norm": 6.504081474555663, + "learning_rate": 6.834999470574293e-06, + "loss": 0.7416, + "step": 8791 + }, + { + "epoch": 1.8435730761165863, + "grad_norm": 6.339350714406659, + "learning_rate": 6.8328520477674665e-06, + "loss": 0.6479, + "step": 8792 + }, + { + "epoch": 1.8437827636821136, + "grad_norm": 6.153805017716203, + "learning_rate": 6.830704787282464e-06, + "loss": 0.761, + "step": 8793 + }, + { + "epoch": 1.843992451247641, + "grad_norm": 7.497326873830622, + "learning_rate": 6.828557689229338e-06, + "loss": 0.8069, + "step": 8794 + }, + { + "epoch": 1.8442021388131684, + "grad_norm": 8.551723296542427, + "learning_rate": 6.826410753718133e-06, + "loss": 0.7128, + "step": 8795 + }, + { + "epoch": 1.8444118263786957, + "grad_norm": 5.815277958899624, + "learning_rate": 6.824263980858876e-06, + "loss": 0.6506, + "step": 8796 + }, + { + "epoch": 1.8446215139442232, + "grad_norm": 6.736973141979157, + "learning_rate": 6.822117370761602e-06, + "loss": 0.5724, + "step": 8797 + }, + { + "epoch": 1.8448312015097503, + "grad_norm": 6.703066376560904, + "learning_rate": 6.819970923536326e-06, + "loss": 0.6001, + "step": 8798 + }, + { + "epoch": 1.8450408890752779, + "grad_norm": 6.358485872619281, + "learning_rate": 6.8178246392930525e-06, + "loss": 0.5928, + "step": 8799 + }, + { + "epoch": 1.8452505766408052, + "grad_norm": 7.17451323497089, + "learning_rate": 6.815678518141784e-06, + "loss": 0.7811, + "step": 8800 + }, + { + "epoch": 1.8454602642063325, + "grad_norm": 7.184417984195259, + "learning_rate": 6.813532560192521e-06, + "loss": 0.7148, + "step": 8801 + }, + { + "epoch": 1.84566995177186, + "grad_norm": 6.622713259594123, + "learning_rate": 6.811386765555234e-06, + "loss": 0.6251, + "step": 8802 + }, + { + "epoch": 1.8458796393373873, + "grad_norm": 6.371339231769335, + "learning_rate": 6.809241134339908e-06, + "loss": 0.6485, + "step": 8803 + }, + { + "epoch": 1.8460893269029146, + "grad_norm": 6.1606406876775095, + "learning_rate": 6.8070956666565104e-06, + "loss": 0.8153, + "step": 8804 + }, + { + "epoch": 1.8462990144684421, + "grad_norm": 6.58912325404316, + "learning_rate": 6.804950362614995e-06, + "loss": 0.6781, + "step": 8805 + }, + { + "epoch": 1.8465087020339694, + "grad_norm": 7.765294050690382, + "learning_rate": 6.802805222325312e-06, + "loss": 0.8856, + "step": 8806 + }, + { + "epoch": 1.8467183895994967, + "grad_norm": 6.370699939040486, + "learning_rate": 6.800660245897416e-06, + "loss": 0.664, + "step": 8807 + }, + { + "epoch": 1.8469280771650243, + "grad_norm": 7.865284213620316, + "learning_rate": 6.798515433441223e-06, + "loss": 0.7884, + "step": 8808 + }, + { + "epoch": 1.8471377647305514, + "grad_norm": 6.894833373347385, + "learning_rate": 6.796370785066668e-06, + "loss": 0.8296, + "step": 8809 + }, + { + "epoch": 1.8473474522960789, + "grad_norm": 7.823675164610121, + "learning_rate": 6.794226300883671e-06, + "loss": 0.9087, + "step": 8810 + }, + { + "epoch": 1.8475571398616062, + "grad_norm": 6.600198464859498, + "learning_rate": 6.792081981002132e-06, + "loss": 0.6011, + "step": 8811 + }, + { + "epoch": 1.8477668274271335, + "grad_norm": 7.041773279406263, + "learning_rate": 6.789937825531955e-06, + "loss": 0.7141, + "step": 8812 + }, + { + "epoch": 1.847976514992661, + "grad_norm": 7.289098017637719, + "learning_rate": 6.7877938345830356e-06, + "loss": 0.5964, + "step": 8813 + }, + { + "epoch": 1.8481862025581883, + "grad_norm": 6.708332441824258, + "learning_rate": 6.785650008265251e-06, + "loss": 0.5387, + "step": 8814 + }, + { + "epoch": 1.8483958901237156, + "grad_norm": 5.513481877202209, + "learning_rate": 6.783506346688478e-06, + "loss": 0.5649, + "step": 8815 + }, + { + "epoch": 1.8486055776892432, + "grad_norm": 6.764574080232439, + "learning_rate": 6.781362849962587e-06, + "loss": 0.6651, + "step": 8816 + }, + { + "epoch": 1.8488152652547702, + "grad_norm": 6.468305105710468, + "learning_rate": 6.7792195181974265e-06, + "loss": 0.7264, + "step": 8817 + }, + { + "epoch": 1.8490249528202978, + "grad_norm": 5.934825396928831, + "learning_rate": 6.777076351502856e-06, + "loss": 0.6363, + "step": 8818 + }, + { + "epoch": 1.849234640385825, + "grad_norm": 6.802112973861438, + "learning_rate": 6.774933349988713e-06, + "loss": 0.6218, + "step": 8819 + }, + { + "epoch": 1.8494443279513524, + "grad_norm": 9.508540778568484, + "learning_rate": 6.772790513764828e-06, + "loss": 1.0334, + "step": 8820 + }, + { + "epoch": 1.84965401551688, + "grad_norm": 7.876813362824488, + "learning_rate": 6.770647842941025e-06, + "loss": 0.7504, + "step": 8821 + }, + { + "epoch": 1.8498637030824072, + "grad_norm": 6.410012445806904, + "learning_rate": 6.768505337627124e-06, + "loss": 0.5726, + "step": 8822 + }, + { + "epoch": 1.8500733906479345, + "grad_norm": 6.381425316930798, + "learning_rate": 6.766362997932926e-06, + "loss": 0.7145, + "step": 8823 + }, + { + "epoch": 1.850283078213462, + "grad_norm": 6.525651035404871, + "learning_rate": 6.764220823968234e-06, + "loss": 0.6312, + "step": 8824 + }, + { + "epoch": 1.8504927657789894, + "grad_norm": 8.822498025423728, + "learning_rate": 6.762078815842838e-06, + "loss": 0.7329, + "step": 8825 + }, + { + "epoch": 1.8507024533445167, + "grad_norm": 7.032901590605517, + "learning_rate": 6.759936973666519e-06, + "loss": 0.8169, + "step": 8826 + }, + { + "epoch": 1.8509121409100442, + "grad_norm": 7.553829844647925, + "learning_rate": 6.757795297549046e-06, + "loss": 0.7136, + "step": 8827 + }, + { + "epoch": 1.8511218284755713, + "grad_norm": 5.721436796682656, + "learning_rate": 6.755653787600194e-06, + "loss": 0.5972, + "step": 8828 + }, + { + "epoch": 1.8513315160410988, + "grad_norm": 7.3646695472519355, + "learning_rate": 6.753512443929706e-06, + "loss": 0.7422, + "step": 8829 + }, + { + "epoch": 1.851541203606626, + "grad_norm": 6.442225788840229, + "learning_rate": 6.751371266647338e-06, + "loss": 0.591, + "step": 8830 + }, + { + "epoch": 1.8517508911721534, + "grad_norm": 6.935939436228216, + "learning_rate": 6.74923025586283e-06, + "loss": 0.7439, + "step": 8831 + }, + { + "epoch": 1.851960578737681, + "grad_norm": 6.913249673586981, + "learning_rate": 6.747089411685906e-06, + "loss": 0.6602, + "step": 8832 + }, + { + "epoch": 1.8521702663032082, + "grad_norm": 6.198042614676255, + "learning_rate": 6.744948734226291e-06, + "loss": 0.4786, + "step": 8833 + }, + { + "epoch": 1.8523799538687356, + "grad_norm": 6.838416173057786, + "learning_rate": 6.742808223593702e-06, + "loss": 0.7624, + "step": 8834 + }, + { + "epoch": 1.852589641434263, + "grad_norm": 8.222283824496552, + "learning_rate": 6.740667879897839e-06, + "loss": 0.8665, + "step": 8835 + }, + { + "epoch": 1.8527993289997902, + "grad_norm": 8.083539683029457, + "learning_rate": 6.738527703248401e-06, + "loss": 0.8253, + "step": 8836 + }, + { + "epoch": 1.8530090165653177, + "grad_norm": 7.2470422542960815, + "learning_rate": 6.736387693755071e-06, + "loss": 0.8012, + "step": 8837 + }, + { + "epoch": 1.853218704130845, + "grad_norm": 7.345132051571992, + "learning_rate": 6.73424785152754e-06, + "loss": 0.8605, + "step": 8838 + }, + { + "epoch": 1.8534283916963723, + "grad_norm": 7.557256944245459, + "learning_rate": 6.732108176675465e-06, + "loss": 0.8796, + "step": 8839 + }, + { + "epoch": 1.8536380792618998, + "grad_norm": 7.2110434487702335, + "learning_rate": 6.7299686693085145e-06, + "loss": 0.7316, + "step": 8840 + }, + { + "epoch": 1.8538477668274271, + "grad_norm": 7.696283194998683, + "learning_rate": 6.727829329536344e-06, + "loss": 0.7346, + "step": 8841 + }, + { + "epoch": 1.8540574543929544, + "grad_norm": 7.563602074849365, + "learning_rate": 6.725690157468593e-06, + "loss": 0.8762, + "step": 8842 + }, + { + "epoch": 1.854267141958482, + "grad_norm": 7.536290936394836, + "learning_rate": 6.723551153214898e-06, + "loss": 0.7293, + "step": 8843 + }, + { + "epoch": 1.8544768295240093, + "grad_norm": 6.332324054485579, + "learning_rate": 6.721412316884894e-06, + "loss": 0.6841, + "step": 8844 + }, + { + "epoch": 1.8546865170895366, + "grad_norm": 5.7310436681325525, + "learning_rate": 6.7192736485881935e-06, + "loss": 0.5599, + "step": 8845 + }, + { + "epoch": 1.854896204655064, + "grad_norm": 5.592282695165193, + "learning_rate": 6.717135148434408e-06, + "loss": 0.4635, + "step": 8846 + }, + { + "epoch": 1.8551058922205912, + "grad_norm": 6.046011829560615, + "learning_rate": 6.714996816533142e-06, + "loss": 0.675, + "step": 8847 + }, + { + "epoch": 1.8553155797861187, + "grad_norm": 7.67778962726067, + "learning_rate": 6.712858652993983e-06, + "loss": 0.7734, + "step": 8848 + }, + { + "epoch": 1.855525267351646, + "grad_norm": 7.731736382347402, + "learning_rate": 6.710720657926522e-06, + "loss": 0.7663, + "step": 8849 + }, + { + "epoch": 1.8557349549171733, + "grad_norm": 6.985793787001254, + "learning_rate": 6.708582831440333e-06, + "loss": 0.73, + "step": 8850 + }, + { + "epoch": 1.8559446424827009, + "grad_norm": 7.283955032171695, + "learning_rate": 6.706445173644982e-06, + "loss": 0.8422, + "step": 8851 + }, + { + "epoch": 1.8561543300482282, + "grad_norm": 7.225940031616147, + "learning_rate": 6.7043076846500265e-06, + "loss": 0.6935, + "step": 8852 + }, + { + "epoch": 1.8563640176137555, + "grad_norm": 8.839704772895123, + "learning_rate": 6.702170364565021e-06, + "loss": 1.027, + "step": 8853 + }, + { + "epoch": 1.856573705179283, + "grad_norm": 7.613208424947482, + "learning_rate": 6.700033213499501e-06, + "loss": 0.7765, + "step": 8854 + }, + { + "epoch": 1.85678339274481, + "grad_norm": 5.391501685739063, + "learning_rate": 6.697896231563004e-06, + "loss": 0.6015, + "step": 8855 + }, + { + "epoch": 1.8569930803103376, + "grad_norm": 6.934836968867341, + "learning_rate": 6.695759418865053e-06, + "loss": 0.673, + "step": 8856 + }, + { + "epoch": 1.857202767875865, + "grad_norm": 6.391277890209812, + "learning_rate": 6.69362277551516e-06, + "loss": 0.7344, + "step": 8857 + }, + { + "epoch": 1.8574124554413922, + "grad_norm": 5.942802575330476, + "learning_rate": 6.691486301622835e-06, + "loss": 0.5278, + "step": 8858 + }, + { + "epoch": 1.8576221430069197, + "grad_norm": 6.460421602427836, + "learning_rate": 6.689349997297581e-06, + "loss": 0.6349, + "step": 8859 + }, + { + "epoch": 1.857831830572447, + "grad_norm": 6.506632623211253, + "learning_rate": 6.687213862648874e-06, + "loss": 0.7203, + "step": 8860 + }, + { + "epoch": 1.8580415181379744, + "grad_norm": 6.799073737315293, + "learning_rate": 6.685077897786205e-06, + "loss": 0.9028, + "step": 8861 + }, + { + "epoch": 1.8582512057035019, + "grad_norm": 6.9433676570730265, + "learning_rate": 6.682942102819046e-06, + "loss": 0.5904, + "step": 8862 + }, + { + "epoch": 1.8584608932690292, + "grad_norm": 6.019819731424689, + "learning_rate": 6.680806477856854e-06, + "loss": 0.6146, + "step": 8863 + }, + { + "epoch": 1.8586705808345565, + "grad_norm": 7.724605108421488, + "learning_rate": 6.678671023009087e-06, + "loss": 0.7486, + "step": 8864 + }, + { + "epoch": 1.858880268400084, + "grad_norm": 7.088456477643206, + "learning_rate": 6.676535738385193e-06, + "loss": 0.6422, + "step": 8865 + }, + { + "epoch": 1.859089955965611, + "grad_norm": 6.901438102881417, + "learning_rate": 6.674400624094603e-06, + "loss": 0.6493, + "step": 8866 + }, + { + "epoch": 1.8592996435311386, + "grad_norm": 6.917424890996764, + "learning_rate": 6.672265680246752e-06, + "loss": 0.6539, + "step": 8867 + }, + { + "epoch": 1.859509331096666, + "grad_norm": 6.470614612018865, + "learning_rate": 6.670130906951058e-06, + "loss": 0.7055, + "step": 8868 + }, + { + "epoch": 1.8597190186621932, + "grad_norm": 5.4230921355498385, + "learning_rate": 6.667996304316927e-06, + "loss": 0.4805, + "step": 8869 + }, + { + "epoch": 1.8599287062277208, + "grad_norm": 7.365996521134372, + "learning_rate": 6.665861872453763e-06, + "loss": 0.8637, + "step": 8870 + }, + { + "epoch": 1.860138393793248, + "grad_norm": 6.301453200704071, + "learning_rate": 6.663727611470966e-06, + "loss": 0.7539, + "step": 8871 + }, + { + "epoch": 1.8603480813587754, + "grad_norm": 6.032662558705327, + "learning_rate": 6.661593521477912e-06, + "loss": 0.5953, + "step": 8872 + }, + { + "epoch": 1.860557768924303, + "grad_norm": 4.725083525844369, + "learning_rate": 6.659459602583981e-06, + "loss": 0.4363, + "step": 8873 + }, + { + "epoch": 1.86076745648983, + "grad_norm": 7.5574122382743525, + "learning_rate": 6.657325854898541e-06, + "loss": 0.8328, + "step": 8874 + }, + { + "epoch": 1.8609771440553575, + "grad_norm": 6.092646076242667, + "learning_rate": 6.655192278530944e-06, + "loss": 0.6048, + "step": 8875 + }, + { + "epoch": 1.8611868316208848, + "grad_norm": 9.732242866407068, + "learning_rate": 6.653058873590546e-06, + "loss": 1.0371, + "step": 8876 + }, + { + "epoch": 1.8613965191864121, + "grad_norm": 8.130190455484051, + "learning_rate": 6.6509256401866886e-06, + "loss": 0.913, + "step": 8877 + }, + { + "epoch": 1.8616062067519397, + "grad_norm": 7.719126143441198, + "learning_rate": 6.6487925784286975e-06, + "loss": 0.8452, + "step": 8878 + }, + { + "epoch": 1.861815894317467, + "grad_norm": 7.64850545581207, + "learning_rate": 6.646659688425898e-06, + "loss": 0.8829, + "step": 8879 + }, + { + "epoch": 1.8620255818829943, + "grad_norm": 6.577181418264686, + "learning_rate": 6.6445269702876124e-06, + "loss": 0.7318, + "step": 8880 + }, + { + "epoch": 1.8622352694485218, + "grad_norm": 5.890659130279615, + "learning_rate": 6.642394424123133e-06, + "loss": 0.6811, + "step": 8881 + }, + { + "epoch": 1.862444957014049, + "grad_norm": 7.503770995820039, + "learning_rate": 6.640262050041764e-06, + "loss": 0.8002, + "step": 8882 + }, + { + "epoch": 1.8626546445795764, + "grad_norm": 5.20649126303187, + "learning_rate": 6.6381298481527945e-06, + "loss": 0.5716, + "step": 8883 + }, + { + "epoch": 1.862864332145104, + "grad_norm": 7.544733486050402, + "learning_rate": 6.635997818565499e-06, + "loss": 0.7659, + "step": 8884 + }, + { + "epoch": 1.863074019710631, + "grad_norm": 7.205893693741829, + "learning_rate": 6.633865961389148e-06, + "loss": 0.7655, + "step": 8885 + }, + { + "epoch": 1.8632837072761586, + "grad_norm": 5.343533088892523, + "learning_rate": 6.63173427673301e-06, + "loss": 0.6497, + "step": 8886 + }, + { + "epoch": 1.8634933948416859, + "grad_norm": 8.980788269182918, + "learning_rate": 6.629602764706327e-06, + "loss": 0.9527, + "step": 8887 + }, + { + "epoch": 1.8637030824072132, + "grad_norm": 6.421780517531628, + "learning_rate": 6.62747142541835e-06, + "loss": 0.6544, + "step": 8888 + }, + { + "epoch": 1.8639127699727407, + "grad_norm": 5.150681066153522, + "learning_rate": 6.625340258978313e-06, + "loss": 0.4267, + "step": 8889 + }, + { + "epoch": 1.864122457538268, + "grad_norm": 5.260612581794034, + "learning_rate": 6.62320926549544e-06, + "loss": 0.4719, + "step": 8890 + }, + { + "epoch": 1.8643321451037953, + "grad_norm": 6.612865620264495, + "learning_rate": 6.621078445078946e-06, + "loss": 0.7476, + "step": 8891 + }, + { + "epoch": 1.8645418326693228, + "grad_norm": 5.7134223514473526, + "learning_rate": 6.618947797838047e-06, + "loss": 0.4969, + "step": 8892 + }, + { + "epoch": 1.86475152023485, + "grad_norm": 5.860707185936737, + "learning_rate": 6.616817323881934e-06, + "loss": 0.5803, + "step": 8893 + }, + { + "epoch": 1.8649612078003774, + "grad_norm": 5.806977242869526, + "learning_rate": 6.614687023319804e-06, + "loss": 0.42, + "step": 8894 + }, + { + "epoch": 1.865170895365905, + "grad_norm": 7.599407840184925, + "learning_rate": 6.612556896260835e-06, + "loss": 0.9014, + "step": 8895 + }, + { + "epoch": 1.865380582931432, + "grad_norm": 4.525484553125013, + "learning_rate": 6.610426942814197e-06, + "loss": 0.379, + "step": 8896 + }, + { + "epoch": 1.8655902704969596, + "grad_norm": 9.280598504186766, + "learning_rate": 6.608297163089058e-06, + "loss": 0.9187, + "step": 8897 + }, + { + "epoch": 1.8657999580624869, + "grad_norm": 6.573195469618976, + "learning_rate": 6.606167557194574e-06, + "loss": 0.6383, + "step": 8898 + }, + { + "epoch": 1.8660096456280142, + "grad_norm": 7.854213203816996, + "learning_rate": 6.604038125239889e-06, + "loss": 0.8125, + "step": 8899 + }, + { + "epoch": 1.8662193331935417, + "grad_norm": 5.8858506164386375, + "learning_rate": 6.601908867334139e-06, + "loss": 0.6233, + "step": 8900 + }, + { + "epoch": 1.866429020759069, + "grad_norm": 7.595317667476934, + "learning_rate": 6.599779783586451e-06, + "loss": 0.6713, + "step": 8901 + }, + { + "epoch": 1.8666387083245963, + "grad_norm": 6.614381674667473, + "learning_rate": 6.597650874105951e-06, + "loss": 0.5873, + "step": 8902 + }, + { + "epoch": 1.8668483958901239, + "grad_norm": 7.864690117658409, + "learning_rate": 6.595522139001742e-06, + "loss": 0.7024, + "step": 8903 + }, + { + "epoch": 1.867058083455651, + "grad_norm": 6.684074803413309, + "learning_rate": 6.59339357838293e-06, + "loss": 0.6293, + "step": 8904 + }, + { + "epoch": 1.8672677710211785, + "grad_norm": 7.279803440112739, + "learning_rate": 6.591265192358605e-06, + "loss": 0.7423, + "step": 8905 + }, + { + "epoch": 1.8674774585867058, + "grad_norm": 7.306437518354194, + "learning_rate": 6.589136981037849e-06, + "loss": 0.7373, + "step": 8906 + }, + { + "epoch": 1.867687146152233, + "grad_norm": 7.122241310280439, + "learning_rate": 6.5870089445297395e-06, + "loss": 0.5429, + "step": 8907 + }, + { + "epoch": 1.8678968337177606, + "grad_norm": 6.248476416581585, + "learning_rate": 6.584881082943344e-06, + "loss": 0.5769, + "step": 8908 + }, + { + "epoch": 1.868106521283288, + "grad_norm": 6.976369911997928, + "learning_rate": 6.582753396387713e-06, + "loss": 0.7296, + "step": 8909 + }, + { + "epoch": 1.8683162088488152, + "grad_norm": 7.477179307019196, + "learning_rate": 6.580625884971896e-06, + "loss": 0.9902, + "step": 8910 + }, + { + "epoch": 1.8685258964143427, + "grad_norm": 8.760351687062585, + "learning_rate": 6.578498548804939e-06, + "loss": 1.032, + "step": 8911 + }, + { + "epoch": 1.8687355839798698, + "grad_norm": 6.207798507838127, + "learning_rate": 6.576371387995859e-06, + "loss": 0.7004, + "step": 8912 + }, + { + "epoch": 1.8689452715453974, + "grad_norm": 5.4493282698491905, + "learning_rate": 6.574244402653687e-06, + "loss": 0.5798, + "step": 8913 + }, + { + "epoch": 1.8691549591109249, + "grad_norm": 6.83261381738648, + "learning_rate": 6.572117592887433e-06, + "loss": 0.7441, + "step": 8914 + }, + { + "epoch": 1.869364646676452, + "grad_norm": 8.4741185941393, + "learning_rate": 6.5699909588060935e-06, + "loss": 0.7635, + "step": 8915 + }, + { + "epoch": 1.8695743342419795, + "grad_norm": 6.144122827067461, + "learning_rate": 6.567864500518666e-06, + "loss": 0.687, + "step": 8916 + }, + { + "epoch": 1.8697840218075068, + "grad_norm": 7.075808395191733, + "learning_rate": 6.565738218134141e-06, + "loss": 0.631, + "step": 8917 + }, + { + "epoch": 1.869993709373034, + "grad_norm": 8.969218875262396, + "learning_rate": 6.563612111761484e-06, + "loss": 1.0271, + "step": 8918 + }, + { + "epoch": 1.8702033969385616, + "grad_norm": 6.363895689850094, + "learning_rate": 6.561486181509666e-06, + "loss": 0.6306, + "step": 8919 + }, + { + "epoch": 1.870413084504089, + "grad_norm": 6.578826797911811, + "learning_rate": 6.559360427487649e-06, + "loss": 0.6168, + "step": 8920 + }, + { + "epoch": 1.8706227720696162, + "grad_norm": 7.048244666740947, + "learning_rate": 6.557234849804375e-06, + "loss": 0.7544, + "step": 8921 + }, + { + "epoch": 1.8708324596351438, + "grad_norm": 6.909507666577538, + "learning_rate": 6.555109448568784e-06, + "loss": 0.7355, + "step": 8922 + }, + { + "epoch": 1.8710421472006709, + "grad_norm": 6.043428641168836, + "learning_rate": 6.552984223889813e-06, + "loss": 0.6072, + "step": 8923 + }, + { + "epoch": 1.8712518347661984, + "grad_norm": 6.927432719928769, + "learning_rate": 6.550859175876376e-06, + "loss": 0.7859, + "step": 8924 + }, + { + "epoch": 1.8714615223317257, + "grad_norm": 5.750444733062134, + "learning_rate": 6.548734304637389e-06, + "loss": 0.5718, + "step": 8925 + }, + { + "epoch": 1.871671209897253, + "grad_norm": 6.7915582670454775, + "learning_rate": 6.546609610281758e-06, + "loss": 0.8597, + "step": 8926 + }, + { + "epoch": 1.8718808974627805, + "grad_norm": 6.122602169339049, + "learning_rate": 6.544485092918369e-06, + "loss": 0.6887, + "step": 8927 + }, + { + "epoch": 1.8720905850283078, + "grad_norm": 6.425600311955218, + "learning_rate": 6.542360752656114e-06, + "loss": 0.6948, + "step": 8928 + }, + { + "epoch": 1.8723002725938351, + "grad_norm": 8.614202927995255, + "learning_rate": 6.54023658960387e-06, + "loss": 0.9709, + "step": 8929 + }, + { + "epoch": 1.8725099601593627, + "grad_norm": 8.476560605176372, + "learning_rate": 6.538112603870499e-06, + "loss": 0.9771, + "step": 8930 + }, + { + "epoch": 1.87271964772489, + "grad_norm": 6.954862038486184, + "learning_rate": 6.5359887955648624e-06, + "loss": 0.6574, + "step": 8931 + }, + { + "epoch": 1.8729293352904173, + "grad_norm": 6.222568665802242, + "learning_rate": 6.53386516479581e-06, + "loss": 0.5438, + "step": 8932 + }, + { + "epoch": 1.8731390228559448, + "grad_norm": 4.959865571404766, + "learning_rate": 6.531741711672175e-06, + "loss": 0.4911, + "step": 8933 + }, + { + "epoch": 1.8733487104214719, + "grad_norm": 5.805135323788341, + "learning_rate": 6.529618436302794e-06, + "loss": 0.5746, + "step": 8934 + }, + { + "epoch": 1.8735583979869994, + "grad_norm": 7.684195461237127, + "learning_rate": 6.52749533879649e-06, + "loss": 0.7282, + "step": 8935 + }, + { + "epoch": 1.8737680855525267, + "grad_norm": 6.288204704747225, + "learning_rate": 6.5253724192620705e-06, + "loss": 0.7443, + "step": 8936 + }, + { + "epoch": 1.873977773118054, + "grad_norm": 5.583061637872511, + "learning_rate": 6.523249677808339e-06, + "loss": 0.6514, + "step": 8937 + }, + { + "epoch": 1.8741874606835816, + "grad_norm": 7.115169158875418, + "learning_rate": 6.521127114544097e-06, + "loss": 0.6332, + "step": 8938 + }, + { + "epoch": 1.8743971482491089, + "grad_norm": 6.803543682295695, + "learning_rate": 6.519004729578119e-06, + "loss": 0.7904, + "step": 8939 + }, + { + "epoch": 1.8746068358146362, + "grad_norm": 6.8270308356618585, + "learning_rate": 6.516882523019188e-06, + "loss": 0.6484, + "step": 8940 + }, + { + "epoch": 1.8748165233801637, + "grad_norm": 7.705124357154914, + "learning_rate": 6.51476049497607e-06, + "loss": 0.7667, + "step": 8941 + }, + { + "epoch": 1.8750262109456908, + "grad_norm": 6.5294676255434005, + "learning_rate": 6.512638645557521e-06, + "loss": 0.6332, + "step": 8942 + }, + { + "epoch": 1.8752358985112183, + "grad_norm": 5.0371993155952905, + "learning_rate": 6.510516974872286e-06, + "loss": 0.5078, + "step": 8943 + }, + { + "epoch": 1.8754455860767456, + "grad_norm": 6.004225104743536, + "learning_rate": 6.5083954830291126e-06, + "loss": 0.6194, + "step": 8944 + }, + { + "epoch": 1.875655273642273, + "grad_norm": 6.083101894981867, + "learning_rate": 6.506274170136726e-06, + "loss": 0.6356, + "step": 8945 + }, + { + "epoch": 1.8758649612078004, + "grad_norm": 7.193023671217823, + "learning_rate": 6.5041530363038475e-06, + "loss": 0.8813, + "step": 8946 + }, + { + "epoch": 1.8760746487733277, + "grad_norm": 6.702513618808895, + "learning_rate": 6.50203208163919e-06, + "loss": 0.7228, + "step": 8947 + }, + { + "epoch": 1.876284336338855, + "grad_norm": 7.0157992711968555, + "learning_rate": 6.499911306251454e-06, + "loss": 0.7464, + "step": 8948 + }, + { + "epoch": 1.8764940239043826, + "grad_norm": 7.387384453011469, + "learning_rate": 6.497790710249332e-06, + "loss": 0.7829, + "step": 8949 + }, + { + "epoch": 1.8767037114699099, + "grad_norm": 6.294736181396978, + "learning_rate": 6.495670293741514e-06, + "loss": 0.7263, + "step": 8950 + }, + { + "epoch": 1.8769133990354372, + "grad_norm": 6.967283317206903, + "learning_rate": 6.49355005683667e-06, + "loss": 0.603, + "step": 8951 + }, + { + "epoch": 1.8771230866009647, + "grad_norm": 6.950041571853424, + "learning_rate": 6.491429999643468e-06, + "loss": 0.4412, + "step": 8952 + }, + { + "epoch": 1.8773327741664918, + "grad_norm": 8.902568674112553, + "learning_rate": 6.489310122270564e-06, + "loss": 0.8974, + "step": 8953 + }, + { + "epoch": 1.8775424617320193, + "grad_norm": 7.679624562088054, + "learning_rate": 6.487190424826601e-06, + "loss": 0.6639, + "step": 8954 + }, + { + "epoch": 1.8777521492975466, + "grad_norm": 7.473587370459821, + "learning_rate": 6.485070907420225e-06, + "loss": 0.6834, + "step": 8955 + }, + { + "epoch": 1.877961836863074, + "grad_norm": 6.036893945044784, + "learning_rate": 6.482951570160063e-06, + "loss": 0.7007, + "step": 8956 + }, + { + "epoch": 1.8781715244286015, + "grad_norm": 7.252405544719615, + "learning_rate": 6.480832413154732e-06, + "loss": 0.7026, + "step": 8957 + }, + { + "epoch": 1.8783812119941288, + "grad_norm": 8.155894741500735, + "learning_rate": 6.47871343651284e-06, + "loss": 0.9691, + "step": 8958 + }, + { + "epoch": 1.878590899559656, + "grad_norm": 6.890982968764576, + "learning_rate": 6.476594640342993e-06, + "loss": 0.6606, + "step": 8959 + }, + { + "epoch": 1.8788005871251836, + "grad_norm": 8.726550712893387, + "learning_rate": 6.4744760247537855e-06, + "loss": 0.9447, + "step": 8960 + }, + { + "epoch": 1.8790102746907107, + "grad_norm": 7.240029471173216, + "learning_rate": 6.472357589853793e-06, + "loss": 0.6056, + "step": 8961 + }, + { + "epoch": 1.8792199622562382, + "grad_norm": 7.882731099837918, + "learning_rate": 6.470239335751594e-06, + "loss": 0.9855, + "step": 8962 + }, + { + "epoch": 1.8794296498217655, + "grad_norm": 7.565005743444662, + "learning_rate": 6.468121262555752e-06, + "loss": 0.8214, + "step": 8963 + }, + { + "epoch": 1.8796393373872928, + "grad_norm": 6.6689163902282305, + "learning_rate": 6.466003370374817e-06, + "loss": 0.8282, + "step": 8964 + }, + { + "epoch": 1.8798490249528204, + "grad_norm": 6.993420022084548, + "learning_rate": 6.4638856593173416e-06, + "loss": 0.6934, + "step": 8965 + }, + { + "epoch": 1.8800587125183477, + "grad_norm": 5.133658428228081, + "learning_rate": 6.46176812949186e-06, + "loss": 0.4979, + "step": 8966 + }, + { + "epoch": 1.880268400083875, + "grad_norm": 7.017300011416189, + "learning_rate": 6.459650781006897e-06, + "loss": 0.7037, + "step": 8967 + }, + { + "epoch": 1.8804780876494025, + "grad_norm": 6.638147808313806, + "learning_rate": 6.45753361397097e-06, + "loss": 0.6873, + "step": 8968 + }, + { + "epoch": 1.8806877752149298, + "grad_norm": 8.36345720195135, + "learning_rate": 6.455416628492596e-06, + "loss": 1.0301, + "step": 8969 + }, + { + "epoch": 1.880897462780457, + "grad_norm": 5.87060066920337, + "learning_rate": 6.4532998246802605e-06, + "loss": 0.576, + "step": 8970 + }, + { + "epoch": 1.8811071503459846, + "grad_norm": 6.642853366081228, + "learning_rate": 6.451183202642463e-06, + "loss": 0.6483, + "step": 8971 + }, + { + "epoch": 1.8813168379115117, + "grad_norm": 6.3328560021998355, + "learning_rate": 6.449066762487684e-06, + "loss": 0.6491, + "step": 8972 + }, + { + "epoch": 1.8815265254770392, + "grad_norm": 4.989911108874801, + "learning_rate": 6.4469505043243885e-06, + "loss": 0.4519, + "step": 8973 + }, + { + "epoch": 1.8817362130425666, + "grad_norm": 6.32969834054392, + "learning_rate": 6.444834428261042e-06, + "loss": 0.6474, + "step": 8974 + }, + { + "epoch": 1.8819459006080939, + "grad_norm": 7.57485872892723, + "learning_rate": 6.4427185344061e-06, + "loss": 0.6824, + "step": 8975 + }, + { + "epoch": 1.8821555881736214, + "grad_norm": 6.728962601042951, + "learning_rate": 6.4406028228680005e-06, + "loss": 0.6553, + "step": 8976 + }, + { + "epoch": 1.8823652757391487, + "grad_norm": 6.691419482983556, + "learning_rate": 6.4384872937551805e-06, + "loss": 0.5701, + "step": 8977 + }, + { + "epoch": 1.882574963304676, + "grad_norm": 7.614783066258959, + "learning_rate": 6.436371947176066e-06, + "loss": 0.9495, + "step": 8978 + }, + { + "epoch": 1.8827846508702035, + "grad_norm": 7.511072995463747, + "learning_rate": 6.434256783239068e-06, + "loss": 0.7676, + "step": 8979 + }, + { + "epoch": 1.8829943384357306, + "grad_norm": 6.835069916438593, + "learning_rate": 6.432141802052592e-06, + "loss": 0.7322, + "step": 8980 + }, + { + "epoch": 1.8832040260012581, + "grad_norm": 7.529890368778924, + "learning_rate": 6.430027003725041e-06, + "loss": 0.9392, + "step": 8981 + }, + { + "epoch": 1.8834137135667854, + "grad_norm": 5.697224226491581, + "learning_rate": 6.427912388364796e-06, + "loss": 0.5872, + "step": 8982 + }, + { + "epoch": 1.8836234011323127, + "grad_norm": 6.838932593203239, + "learning_rate": 6.425797956080236e-06, + "loss": 0.8455, + "step": 8983 + }, + { + "epoch": 1.8838330886978403, + "grad_norm": 5.921217453019628, + "learning_rate": 6.423683706979733e-06, + "loss": 0.5809, + "step": 8984 + }, + { + "epoch": 1.8840427762633676, + "grad_norm": 5.499811032758023, + "learning_rate": 6.421569641171638e-06, + "loss": 0.6489, + "step": 8985 + }, + { + "epoch": 1.8842524638288949, + "grad_norm": 5.324100269166418, + "learning_rate": 6.4194557587643066e-06, + "loss": 0.5409, + "step": 8986 + }, + { + "epoch": 1.8844621513944224, + "grad_norm": 5.601625625036533, + "learning_rate": 6.417342059866081e-06, + "loss": 0.6347, + "step": 8987 + }, + { + "epoch": 1.8846718389599497, + "grad_norm": 6.789689469661332, + "learning_rate": 6.415228544585286e-06, + "loss": 0.7457, + "step": 8988 + }, + { + "epoch": 1.884881526525477, + "grad_norm": 7.68154280789634, + "learning_rate": 6.413115213030243e-06, + "loss": 0.7901, + "step": 8989 + }, + { + "epoch": 1.8850912140910046, + "grad_norm": 6.549744468811412, + "learning_rate": 6.411002065309274e-06, + "loss": 0.611, + "step": 8990 + }, + { + "epoch": 1.8853009016565316, + "grad_norm": 7.204557346747703, + "learning_rate": 6.408889101530666e-06, + "loss": 0.8142, + "step": 8991 + }, + { + "epoch": 1.8855105892220592, + "grad_norm": 6.9965841679486935, + "learning_rate": 6.406776321802725e-06, + "loss": 0.737, + "step": 8992 + }, + { + "epoch": 1.8857202767875865, + "grad_norm": 8.716060398973092, + "learning_rate": 6.404663726233728e-06, + "loss": 0.9585, + "step": 8993 + }, + { + "epoch": 1.8859299643531138, + "grad_norm": 5.231424842197686, + "learning_rate": 6.402551314931952e-06, + "loss": 0.5039, + "step": 8994 + }, + { + "epoch": 1.8861396519186413, + "grad_norm": 6.459207999825532, + "learning_rate": 6.4004390880056566e-06, + "loss": 0.6143, + "step": 8995 + }, + { + "epoch": 1.8863493394841686, + "grad_norm": 7.482139476773287, + "learning_rate": 6.398327045563108e-06, + "loss": 0.8055, + "step": 8996 + }, + { + "epoch": 1.886559027049696, + "grad_norm": 6.874487298845809, + "learning_rate": 6.3962151877125385e-06, + "loss": 0.727, + "step": 8997 + }, + { + "epoch": 1.8867687146152234, + "grad_norm": 6.786037002893921, + "learning_rate": 6.394103514562193e-06, + "loss": 0.5984, + "step": 8998 + }, + { + "epoch": 1.8869784021807505, + "grad_norm": 8.457215285996815, + "learning_rate": 6.391992026220299e-06, + "loss": 0.8675, + "step": 8999 + }, + { + "epoch": 1.887188089746278, + "grad_norm": 7.716848580581355, + "learning_rate": 6.38988072279507e-06, + "loss": 0.7905, + "step": 9000 + }, + { + "epoch": 1.8873977773118054, + "grad_norm": 5.32908579320883, + "learning_rate": 6.387769604394713e-06, + "loss": 0.5501, + "step": 9001 + }, + { + "epoch": 1.8876074648773327, + "grad_norm": 6.497847530791774, + "learning_rate": 6.385658671127433e-06, + "loss": 0.6941, + "step": 9002 + }, + { + "epoch": 1.8878171524428602, + "grad_norm": 6.1307103975041, + "learning_rate": 6.383547923101412e-06, + "loss": 0.5342, + "step": 9003 + }, + { + "epoch": 1.8880268400083875, + "grad_norm": 7.765491888854994, + "learning_rate": 6.3814373604248315e-06, + "loss": 0.7923, + "step": 9004 + }, + { + "epoch": 1.8882365275739148, + "grad_norm": 6.143227148109602, + "learning_rate": 6.3793269832058665e-06, + "loss": 0.5646, + "step": 9005 + }, + { + "epoch": 1.8884462151394423, + "grad_norm": 8.097047907607376, + "learning_rate": 6.377216791552668e-06, + "loss": 0.8071, + "step": 9006 + }, + { + "epoch": 1.8886559027049696, + "grad_norm": 6.084921774434358, + "learning_rate": 6.375106785573395e-06, + "loss": 0.5932, + "step": 9007 + }, + { + "epoch": 1.888865590270497, + "grad_norm": 6.538946075069554, + "learning_rate": 6.372996965376188e-06, + "loss": 0.5712, + "step": 9008 + }, + { + "epoch": 1.8890752778360245, + "grad_norm": 7.200732521790574, + "learning_rate": 6.370887331069176e-06, + "loss": 0.8045, + "step": 9009 + }, + { + "epoch": 1.8892849654015516, + "grad_norm": 8.601416704395712, + "learning_rate": 6.368777882760482e-06, + "loss": 0.7995, + "step": 9010 + }, + { + "epoch": 1.889494652967079, + "grad_norm": 7.653629587078774, + "learning_rate": 6.366668620558221e-06, + "loss": 0.8363, + "step": 9011 + }, + { + "epoch": 1.8897043405326064, + "grad_norm": 5.238812819565244, + "learning_rate": 6.364559544570491e-06, + "loss": 0.4146, + "step": 9012 + }, + { + "epoch": 1.8899140280981337, + "grad_norm": 6.398572992222365, + "learning_rate": 6.362450654905392e-06, + "loss": 0.6192, + "step": 9013 + }, + { + "epoch": 1.8901237156636612, + "grad_norm": 7.108089892596612, + "learning_rate": 6.360341951671009e-06, + "loss": 0.7132, + "step": 9014 + }, + { + "epoch": 1.8903334032291885, + "grad_norm": 7.204043090405134, + "learning_rate": 6.35823343497541e-06, + "loss": 0.8402, + "step": 9015 + }, + { + "epoch": 1.8905430907947158, + "grad_norm": 6.306860209044986, + "learning_rate": 6.356125104926663e-06, + "loss": 0.5375, + "step": 9016 + }, + { + "epoch": 1.8907527783602434, + "grad_norm": 6.794073080627332, + "learning_rate": 6.35401696163283e-06, + "loss": 0.77, + "step": 9017 + }, + { + "epoch": 1.8909624659257704, + "grad_norm": 7.1303014602296475, + "learning_rate": 6.351909005201944e-06, + "loss": 0.7385, + "step": 9018 + }, + { + "epoch": 1.891172153491298, + "grad_norm": 6.500577542633357, + "learning_rate": 6.349801235742053e-06, + "loss": 0.6575, + "step": 9019 + }, + { + "epoch": 1.8913818410568253, + "grad_norm": 5.9824188618803396, + "learning_rate": 6.347693653361176e-06, + "loss": 0.593, + "step": 9020 + }, + { + "epoch": 1.8915915286223526, + "grad_norm": 7.816785221501974, + "learning_rate": 6.34558625816734e-06, + "loss": 0.7875, + "step": 9021 + }, + { + "epoch": 1.89180121618788, + "grad_norm": 7.005855691641357, + "learning_rate": 6.343479050268542e-06, + "loss": 0.6356, + "step": 9022 + }, + { + "epoch": 1.8920109037534074, + "grad_norm": 7.156502321273942, + "learning_rate": 6.3413720297727855e-06, + "loss": 0.6618, + "step": 9023 + }, + { + "epoch": 1.8922205913189347, + "grad_norm": 6.894294714449833, + "learning_rate": 6.33926519678806e-06, + "loss": 0.802, + "step": 9024 + }, + { + "epoch": 1.8924302788844622, + "grad_norm": 7.455603504332197, + "learning_rate": 6.33715855142234e-06, + "loss": 0.7907, + "step": 9025 + }, + { + "epoch": 1.8926399664499896, + "grad_norm": 6.271804481529485, + "learning_rate": 6.3350520937835956e-06, + "loss": 0.6858, + "step": 9026 + }, + { + "epoch": 1.8928496540155169, + "grad_norm": 5.484111224763741, + "learning_rate": 6.332945823979795e-06, + "loss": 0.62, + "step": 9027 + }, + { + "epoch": 1.8930593415810444, + "grad_norm": 6.78551177892579, + "learning_rate": 6.330839742118873e-06, + "loss": 0.7759, + "step": 9028 + }, + { + "epoch": 1.8932690291465715, + "grad_norm": 6.993557637116799, + "learning_rate": 6.328733848308781e-06, + "loss": 0.6785, + "step": 9029 + }, + { + "epoch": 1.893478716712099, + "grad_norm": 5.126076341952603, + "learning_rate": 6.3266281426574495e-06, + "loss": 0.478, + "step": 9030 + }, + { + "epoch": 1.8936884042776263, + "grad_norm": 7.624656798502881, + "learning_rate": 6.324522625272793e-06, + "loss": 0.793, + "step": 9031 + }, + { + "epoch": 1.8938980918431536, + "grad_norm": 6.193206999668794, + "learning_rate": 6.322417296262726e-06, + "loss": 0.7363, + "step": 9032 + }, + { + "epoch": 1.8941077794086811, + "grad_norm": 6.639816099996433, + "learning_rate": 6.320312155735156e-06, + "loss": 0.6514, + "step": 9033 + }, + { + "epoch": 1.8943174669742084, + "grad_norm": 5.321986323911192, + "learning_rate": 6.318207203797967e-06, + "loss": 0.541, + "step": 9034 + }, + { + "epoch": 1.8945271545397357, + "grad_norm": 7.759774728467715, + "learning_rate": 6.316102440559044e-06, + "loss": 0.7734, + "step": 9035 + }, + { + "epoch": 1.8947368421052633, + "grad_norm": 7.11020066043924, + "learning_rate": 6.3139978661262635e-06, + "loss": 0.5538, + "step": 9036 + }, + { + "epoch": 1.8949465296707904, + "grad_norm": 7.362698812494282, + "learning_rate": 6.311893480607481e-06, + "loss": 0.7707, + "step": 9037 + }, + { + "epoch": 1.8951562172363179, + "grad_norm": 5.28953207342459, + "learning_rate": 6.309789284110555e-06, + "loss": 0.5376, + "step": 9038 + }, + { + "epoch": 1.8953659048018454, + "grad_norm": 6.50100437371527, + "learning_rate": 6.3076852767433325e-06, + "loss": 0.6038, + "step": 9039 + }, + { + "epoch": 1.8955755923673725, + "grad_norm": 6.21203372925816, + "learning_rate": 6.305581458613641e-06, + "loss": 0.5111, + "step": 9040 + }, + { + "epoch": 1.8957852799329, + "grad_norm": 10.037091138073055, + "learning_rate": 6.303477829829305e-06, + "loss": 1.0874, + "step": 9041 + }, + { + "epoch": 1.8959949674984273, + "grad_norm": 6.714104149538097, + "learning_rate": 6.301374390498147e-06, + "loss": 0.8119, + "step": 9042 + }, + { + "epoch": 1.8962046550639546, + "grad_norm": 6.127757367450675, + "learning_rate": 6.299271140727962e-06, + "loss": 0.5786, + "step": 9043 + }, + { + "epoch": 1.8964143426294822, + "grad_norm": 7.763536760845901, + "learning_rate": 6.29716808062655e-06, + "loss": 0.7182, + "step": 9044 + }, + { + "epoch": 1.8966240301950095, + "grad_norm": 5.608766144361198, + "learning_rate": 6.295065210301699e-06, + "loss": 0.5903, + "step": 9045 + }, + { + "epoch": 1.8968337177605368, + "grad_norm": 8.022382414718525, + "learning_rate": 6.29296252986118e-06, + "loss": 0.8858, + "step": 9046 + }, + { + "epoch": 1.8970434053260643, + "grad_norm": 6.3978998577802235, + "learning_rate": 6.290860039412758e-06, + "loss": 0.6672, + "step": 9047 + }, + { + "epoch": 1.8972530928915914, + "grad_norm": 6.595551143512827, + "learning_rate": 6.2887577390642e-06, + "loss": 0.7484, + "step": 9048 + }, + { + "epoch": 1.897462780457119, + "grad_norm": 6.693288000848857, + "learning_rate": 6.286655628923237e-06, + "loss": 0.7635, + "step": 9049 + }, + { + "epoch": 1.8976724680226462, + "grad_norm": 6.703232260131536, + "learning_rate": 6.284553709097616e-06, + "loss": 0.6711, + "step": 9050 + }, + { + "epoch": 1.8978821555881735, + "grad_norm": 7.136691085561368, + "learning_rate": 6.282451979695063e-06, + "loss": 0.6758, + "step": 9051 + }, + { + "epoch": 1.898091843153701, + "grad_norm": 6.155548977241294, + "learning_rate": 6.280350440823292e-06, + "loss": 0.6813, + "step": 9052 + }, + { + "epoch": 1.8983015307192284, + "grad_norm": 7.471725647580179, + "learning_rate": 6.278249092590009e-06, + "loss": 0.7158, + "step": 9053 + }, + { + "epoch": 1.8985112182847557, + "grad_norm": 6.973128331656198, + "learning_rate": 6.276147935102922e-06, + "loss": 0.6108, + "step": 9054 + }, + { + "epoch": 1.8987209058502832, + "grad_norm": 5.605320781489347, + "learning_rate": 6.274046968469709e-06, + "loss": 0.5382, + "step": 9055 + }, + { + "epoch": 1.8989305934158105, + "grad_norm": 6.120858741148775, + "learning_rate": 6.271946192798051e-06, + "loss": 0.522, + "step": 9056 + }, + { + "epoch": 1.8991402809813378, + "grad_norm": 8.199832552751719, + "learning_rate": 6.269845608195617e-06, + "loss": 0.7982, + "step": 9057 + }, + { + "epoch": 1.8993499685468653, + "grad_norm": 8.509136923994387, + "learning_rate": 6.267745214770065e-06, + "loss": 0.9397, + "step": 9058 + }, + { + "epoch": 1.8995596561123924, + "grad_norm": 6.5282795742878035, + "learning_rate": 6.265645012629042e-06, + "loss": 0.7041, + "step": 9059 + }, + { + "epoch": 1.89976934367792, + "grad_norm": 6.399206347320185, + "learning_rate": 6.263545001880192e-06, + "loss": 0.6985, + "step": 9060 + }, + { + "epoch": 1.8999790312434472, + "grad_norm": 7.4663703412715945, + "learning_rate": 6.261445182631141e-06, + "loss": 0.7214, + "step": 9061 + }, + { + "epoch": 1.9001887188089746, + "grad_norm": 8.085544298811916, + "learning_rate": 6.259345554989507e-06, + "loss": 0.8153, + "step": 9062 + }, + { + "epoch": 1.900398406374502, + "grad_norm": 5.400880713411227, + "learning_rate": 6.257246119062903e-06, + "loss": 0.5791, + "step": 9063 + }, + { + "epoch": 1.9006080939400294, + "grad_norm": 5.738390552435382, + "learning_rate": 6.2551468749589235e-06, + "loss": 0.6487, + "step": 9064 + }, + { + "epoch": 1.9008177815055567, + "grad_norm": 7.129336315403556, + "learning_rate": 6.253047822785163e-06, + "loss": 0.8305, + "step": 9065 + }, + { + "epoch": 1.9010274690710842, + "grad_norm": 6.626744924057898, + "learning_rate": 6.250948962649203e-06, + "loss": 0.6631, + "step": 9066 + }, + { + "epoch": 1.9012371566366113, + "grad_norm": 6.854769677739551, + "learning_rate": 6.24885029465861e-06, + "loss": 0.7244, + "step": 9067 + }, + { + "epoch": 1.9014468442021388, + "grad_norm": 9.352684845972671, + "learning_rate": 6.246751818920942e-06, + "loss": 0.7809, + "step": 9068 + }, + { + "epoch": 1.9016565317676661, + "grad_norm": 6.71503066118461, + "learning_rate": 6.244653535543761e-06, + "loss": 0.7525, + "step": 9069 + }, + { + "epoch": 1.9018662193331934, + "grad_norm": 7.055347972244156, + "learning_rate": 6.242555444634592e-06, + "loss": 0.664, + "step": 9070 + }, + { + "epoch": 1.902075906898721, + "grad_norm": 7.641179472018461, + "learning_rate": 6.240457546300975e-06, + "loss": 0.7026, + "step": 9071 + }, + { + "epoch": 1.9022855944642483, + "grad_norm": 6.216153024907154, + "learning_rate": 6.238359840650432e-06, + "loss": 0.6951, + "step": 9072 + }, + { + "epoch": 1.9024952820297756, + "grad_norm": 5.107328948895102, + "learning_rate": 6.23626232779047e-06, + "loss": 0.4799, + "step": 9073 + }, + { + "epoch": 1.902704969595303, + "grad_norm": 7.679785441145659, + "learning_rate": 6.234165007828587e-06, + "loss": 0.9029, + "step": 9074 + }, + { + "epoch": 1.9029146571608304, + "grad_norm": 8.234386008190128, + "learning_rate": 6.232067880872288e-06, + "loss": 0.6897, + "step": 9075 + }, + { + "epoch": 1.9031243447263577, + "grad_norm": 5.983965735223116, + "learning_rate": 6.229970947029038e-06, + "loss": 0.583, + "step": 9076 + }, + { + "epoch": 1.9033340322918852, + "grad_norm": 6.5115682304780185, + "learning_rate": 6.2278742064063166e-06, + "loss": 0.5923, + "step": 9077 + }, + { + "epoch": 1.9035437198574123, + "grad_norm": 8.604485664282944, + "learning_rate": 6.225777659111589e-06, + "loss": 0.8574, + "step": 9078 + }, + { + "epoch": 1.9037534074229399, + "grad_norm": 4.553535797560089, + "learning_rate": 6.2236813052522986e-06, + "loss": 0.4361, + "step": 9079 + }, + { + "epoch": 1.9039630949884672, + "grad_norm": 7.746790973335353, + "learning_rate": 6.221585144935889e-06, + "loss": 0.6398, + "step": 9080 + }, + { + "epoch": 1.9041727825539945, + "grad_norm": 6.303396338421388, + "learning_rate": 6.219489178269798e-06, + "loss": 0.7479, + "step": 9081 + }, + { + "epoch": 1.904382470119522, + "grad_norm": 6.791186542845121, + "learning_rate": 6.217393405361443e-06, + "loss": 0.7714, + "step": 9082 + }, + { + "epoch": 1.9045921576850493, + "grad_norm": 6.06417220320398, + "learning_rate": 6.215297826318236e-06, + "loss": 0.5004, + "step": 9083 + }, + { + "epoch": 1.9048018452505766, + "grad_norm": 8.237523514351546, + "learning_rate": 6.213202441247577e-06, + "loss": 0.7543, + "step": 9084 + }, + { + "epoch": 1.9050115328161041, + "grad_norm": 7.2654754123363094, + "learning_rate": 6.211107250256865e-06, + "loss": 0.7512, + "step": 9085 + }, + { + "epoch": 1.9052212203816312, + "grad_norm": 8.224161068590927, + "learning_rate": 6.2090122534534765e-06, + "loss": 0.8016, + "step": 9086 + }, + { + "epoch": 1.9054309079471587, + "grad_norm": 7.429953410677938, + "learning_rate": 6.206917450944785e-06, + "loss": 0.924, + "step": 9087 + }, + { + "epoch": 1.905640595512686, + "grad_norm": 6.927919649637011, + "learning_rate": 6.204822842838154e-06, + "loss": 0.7153, + "step": 9088 + }, + { + "epoch": 1.9058502830782134, + "grad_norm": 8.425167478977086, + "learning_rate": 6.202728429240931e-06, + "loss": 0.8736, + "step": 9089 + }, + { + "epoch": 1.9060599706437409, + "grad_norm": 7.377057161625221, + "learning_rate": 6.200634210260464e-06, + "loss": 0.8848, + "step": 9090 + }, + { + "epoch": 1.9062696582092682, + "grad_norm": 6.766126372115526, + "learning_rate": 6.198540186004086e-06, + "loss": 0.6415, + "step": 9091 + }, + { + "epoch": 1.9064793457747955, + "grad_norm": 8.728347245976163, + "learning_rate": 6.196446356579115e-06, + "loss": 1.0288, + "step": 9092 + }, + { + "epoch": 1.906689033340323, + "grad_norm": 6.619654272517235, + "learning_rate": 6.194352722092864e-06, + "loss": 0.7075, + "step": 9093 + }, + { + "epoch": 1.9068987209058503, + "grad_norm": 6.564929765525703, + "learning_rate": 6.192259282652639e-06, + "loss": 0.7521, + "step": 9094 + }, + { + "epoch": 1.9071084084713776, + "grad_norm": 6.853229650318318, + "learning_rate": 6.190166038365725e-06, + "loss": 0.6715, + "step": 9095 + }, + { + "epoch": 1.9073180960369052, + "grad_norm": 6.357485968036707, + "learning_rate": 6.188072989339414e-06, + "loss": 0.6732, + "step": 9096 + }, + { + "epoch": 1.9075277836024322, + "grad_norm": 7.76001776586375, + "learning_rate": 6.185980135680975e-06, + "loss": 0.9084, + "step": 9097 + }, + { + "epoch": 1.9077374711679598, + "grad_norm": 6.895661462428787, + "learning_rate": 6.183887477497665e-06, + "loss": 0.6244, + "step": 9098 + }, + { + "epoch": 1.907947158733487, + "grad_norm": 6.450003192705835, + "learning_rate": 6.181795014896741e-06, + "loss": 0.5019, + "step": 9099 + }, + { + "epoch": 1.9081568462990144, + "grad_norm": 5.571363813014145, + "learning_rate": 6.17970274798545e-06, + "loss": 0.4562, + "step": 9100 + }, + { + "epoch": 1.908366533864542, + "grad_norm": 7.382282277145855, + "learning_rate": 6.177610676871012e-06, + "loss": 0.823, + "step": 9101 + }, + { + "epoch": 1.9085762214300692, + "grad_norm": 6.927303336290081, + "learning_rate": 6.175518801660659e-06, + "loss": 0.6553, + "step": 9102 + }, + { + "epoch": 1.9087859089955965, + "grad_norm": 6.114409144164231, + "learning_rate": 6.173427122461604e-06, + "loss": 0.6102, + "step": 9103 + }, + { + "epoch": 1.908995596561124, + "grad_norm": 6.386235760732453, + "learning_rate": 6.171335639381042e-06, + "loss": 0.6282, + "step": 9104 + }, + { + "epoch": 1.9092052841266511, + "grad_norm": 7.410881420993415, + "learning_rate": 6.169244352526167e-06, + "loss": 0.7806, + "step": 9105 + }, + { + "epoch": 1.9094149716921787, + "grad_norm": 7.0515819374956, + "learning_rate": 6.167153262004171e-06, + "loss": 0.7283, + "step": 9106 + }, + { + "epoch": 1.909624659257706, + "grad_norm": 7.802412724221096, + "learning_rate": 6.16506236792221e-06, + "loss": 0.683, + "step": 9107 + }, + { + "epoch": 1.9098343468232333, + "grad_norm": 6.207539766096078, + "learning_rate": 6.162971670387457e-06, + "loss": 0.5298, + "step": 9108 + }, + { + "epoch": 1.9100440343887608, + "grad_norm": 5.405935969148804, + "learning_rate": 6.160881169507062e-06, + "loss": 0.5624, + "step": 9109 + }, + { + "epoch": 1.910253721954288, + "grad_norm": 6.652201333484288, + "learning_rate": 6.158790865388164e-06, + "loss": 0.6425, + "step": 9110 + }, + { + "epoch": 1.9104634095198154, + "grad_norm": 6.6782549269991875, + "learning_rate": 6.156700758137895e-06, + "loss": 0.6581, + "step": 9111 + }, + { + "epoch": 1.910673097085343, + "grad_norm": 7.728669079965136, + "learning_rate": 6.154610847863383e-06, + "loss": 0.7246, + "step": 9112 + }, + { + "epoch": 1.9108827846508702, + "grad_norm": 6.87586869730982, + "learning_rate": 6.15252113467173e-06, + "loss": 0.6548, + "step": 9113 + }, + { + "epoch": 1.9110924722163976, + "grad_norm": 6.60880623956218, + "learning_rate": 6.150431618670043e-06, + "loss": 0.7322, + "step": 9114 + }, + { + "epoch": 1.911302159781925, + "grad_norm": 6.818183484154013, + "learning_rate": 6.148342299965415e-06, + "loss": 0.9061, + "step": 9115 + }, + { + "epoch": 1.9115118473474522, + "grad_norm": 7.52681811926945, + "learning_rate": 6.14625317866492e-06, + "loss": 0.7357, + "step": 9116 + }, + { + "epoch": 1.9117215349129797, + "grad_norm": 6.348985101548476, + "learning_rate": 6.144164254875635e-06, + "loss": 0.7264, + "step": 9117 + }, + { + "epoch": 1.911931222478507, + "grad_norm": 6.527068126025374, + "learning_rate": 6.142075528704623e-06, + "loss": 0.6643, + "step": 9118 + }, + { + "epoch": 1.9121409100440343, + "grad_norm": 8.975990707763494, + "learning_rate": 6.139987000258928e-06, + "loss": 0.8361, + "step": 9119 + }, + { + "epoch": 1.9123505976095618, + "grad_norm": 10.790468247628679, + "learning_rate": 6.137898669645592e-06, + "loss": 1.1709, + "step": 9120 + }, + { + "epoch": 1.9125602851750891, + "grad_norm": 6.688955881057216, + "learning_rate": 6.135810536971655e-06, + "loss": 0.7099, + "step": 9121 + }, + { + "epoch": 1.9127699727406164, + "grad_norm": 7.025398559290096, + "learning_rate": 6.133722602344123e-06, + "loss": 0.8218, + "step": 9122 + }, + { + "epoch": 1.912979660306144, + "grad_norm": 6.929604564781155, + "learning_rate": 6.131634865870015e-06, + "loss": 0.7349, + "step": 9123 + }, + { + "epoch": 1.913189347871671, + "grad_norm": 6.135038794538746, + "learning_rate": 6.129547327656333e-06, + "loss": 0.5375, + "step": 9124 + }, + { + "epoch": 1.9133990354371986, + "grad_norm": 6.986841285048638, + "learning_rate": 6.127459987810061e-06, + "loss": 0.8427, + "step": 9125 + }, + { + "epoch": 1.9136087230027259, + "grad_norm": 9.019622473092278, + "learning_rate": 6.1253728464381776e-06, + "loss": 0.7606, + "step": 9126 + }, + { + "epoch": 1.9138184105682532, + "grad_norm": 5.529049538362865, + "learning_rate": 6.123285903647665e-06, + "loss": 0.4932, + "step": 9127 + }, + { + "epoch": 1.9140280981337807, + "grad_norm": 5.395914430097848, + "learning_rate": 6.121199159545465e-06, + "loss": 0.4906, + "step": 9128 + }, + { + "epoch": 1.914237785699308, + "grad_norm": 5.139612158245559, + "learning_rate": 6.119112614238538e-06, + "loss": 0.6098, + "step": 9129 + }, + { + "epoch": 1.9144474732648353, + "grad_norm": 6.256769229062346, + "learning_rate": 6.117026267833823e-06, + "loss": 0.6793, + "step": 9130 + }, + { + "epoch": 1.9146571608303629, + "grad_norm": 6.025461317973854, + "learning_rate": 6.114940120438244e-06, + "loss": 0.735, + "step": 9131 + }, + { + "epoch": 1.9148668483958902, + "grad_norm": 8.668052171174216, + "learning_rate": 6.112854172158721e-06, + "loss": 0.7462, + "step": 9132 + }, + { + "epoch": 1.9150765359614175, + "grad_norm": 6.422776824565058, + "learning_rate": 6.110768423102167e-06, + "loss": 0.668, + "step": 9133 + }, + { + "epoch": 1.915286223526945, + "grad_norm": 4.580057136796635, + "learning_rate": 6.108682873375476e-06, + "loss": 0.4537, + "step": 9134 + }, + { + "epoch": 1.915495911092472, + "grad_norm": 6.231359392442732, + "learning_rate": 6.106597523085535e-06, + "loss": 0.6518, + "step": 9135 + }, + { + "epoch": 1.9157055986579996, + "grad_norm": 7.056009146623417, + "learning_rate": 6.104512372339227e-06, + "loss": 0.6164, + "step": 9136 + }, + { + "epoch": 1.915915286223527, + "grad_norm": 7.517923545632127, + "learning_rate": 6.102427421243411e-06, + "loss": 0.8008, + "step": 9137 + }, + { + "epoch": 1.9161249737890542, + "grad_norm": 6.809358798940432, + "learning_rate": 6.100342669904953e-06, + "loss": 0.6515, + "step": 9138 + }, + { + "epoch": 1.9163346613545817, + "grad_norm": 6.937425089981554, + "learning_rate": 6.098258118430698e-06, + "loss": 0.7662, + "step": 9139 + }, + { + "epoch": 1.916544348920109, + "grad_norm": 6.123212770255885, + "learning_rate": 6.09617376692748e-06, + "loss": 0.6688, + "step": 9140 + }, + { + "epoch": 1.9167540364856364, + "grad_norm": 5.599001937833183, + "learning_rate": 6.094089615502127e-06, + "loss": 0.6122, + "step": 9141 + }, + { + "epoch": 1.9169637240511639, + "grad_norm": 6.732902342326816, + "learning_rate": 6.092005664261453e-06, + "loss": 0.6461, + "step": 9142 + }, + { + "epoch": 1.917173411616691, + "grad_norm": 7.902813312695726, + "learning_rate": 6.089921913312272e-06, + "loss": 0.8317, + "step": 9143 + }, + { + "epoch": 1.9173830991822185, + "grad_norm": 6.641439167348805, + "learning_rate": 6.087838362761372e-06, + "loss": 0.6143, + "step": 9144 + }, + { + "epoch": 1.9175927867477458, + "grad_norm": 6.354099916116643, + "learning_rate": 6.085755012715542e-06, + "loss": 0.7456, + "step": 9145 + }, + { + "epoch": 1.917802474313273, + "grad_norm": 7.355514843776557, + "learning_rate": 6.083671863281557e-06, + "loss": 0.6493, + "step": 9146 + }, + { + "epoch": 1.9180121618788006, + "grad_norm": 7.357720894809341, + "learning_rate": 6.081588914566179e-06, + "loss": 0.7073, + "step": 9147 + }, + { + "epoch": 1.918221849444328, + "grad_norm": 6.268968389622958, + "learning_rate": 6.079506166676168e-06, + "loss": 0.7148, + "step": 9148 + }, + { + "epoch": 1.9184315370098552, + "grad_norm": 8.406200567218905, + "learning_rate": 6.077423619718266e-06, + "loss": 0.8671, + "step": 9149 + }, + { + "epoch": 1.9186412245753828, + "grad_norm": 6.133066460945635, + "learning_rate": 6.075341273799205e-06, + "loss": 0.5847, + "step": 9150 + }, + { + "epoch": 1.91885091214091, + "grad_norm": 5.655222529902011, + "learning_rate": 6.073259129025709e-06, + "loss": 0.5088, + "step": 9151 + }, + { + "epoch": 1.9190605997064374, + "grad_norm": 7.623272778112802, + "learning_rate": 6.0711771855045e-06, + "loss": 0.5804, + "step": 9152 + }, + { + "epoch": 1.919270287271965, + "grad_norm": 6.80615048826173, + "learning_rate": 6.069095443342268e-06, + "loss": 0.7016, + "step": 9153 + }, + { + "epoch": 1.919479974837492, + "grad_norm": 6.442726991262663, + "learning_rate": 6.067013902645713e-06, + "loss": 0.6521, + "step": 9154 + }, + { + "epoch": 1.9196896624030195, + "grad_norm": 8.394638438048586, + "learning_rate": 6.06493256352152e-06, + "loss": 0.7845, + "step": 9155 + }, + { + "epoch": 1.9198993499685468, + "grad_norm": 6.531901041349821, + "learning_rate": 6.062851426076356e-06, + "loss": 0.7459, + "step": 9156 + }, + { + "epoch": 1.9201090375340741, + "grad_norm": 7.287788582350326, + "learning_rate": 6.060770490416883e-06, + "loss": 0.7302, + "step": 9157 + }, + { + "epoch": 1.9203187250996017, + "grad_norm": 6.288854307673709, + "learning_rate": 6.0586897566497605e-06, + "loss": 0.568, + "step": 9158 + }, + { + "epoch": 1.920528412665129, + "grad_norm": 7.136806794945663, + "learning_rate": 6.056609224881619e-06, + "loss": 0.7276, + "step": 9159 + }, + { + "epoch": 1.9207381002306563, + "grad_norm": 6.2292382259840915, + "learning_rate": 6.054528895219095e-06, + "loss": 0.6155, + "step": 9160 + }, + { + "epoch": 1.9209477877961838, + "grad_norm": 7.176081663970161, + "learning_rate": 6.0524487677688106e-06, + "loss": 0.7813, + "step": 9161 + }, + { + "epoch": 1.9211574753617109, + "grad_norm": 6.875262491677104, + "learning_rate": 6.050368842637371e-06, + "loss": 0.7492, + "step": 9162 + }, + { + "epoch": 1.9213671629272384, + "grad_norm": 6.00566162621194, + "learning_rate": 6.0482891199313765e-06, + "loss": 0.5621, + "step": 9163 + }, + { + "epoch": 1.921576850492766, + "grad_norm": 6.856406708752517, + "learning_rate": 6.046209599757423e-06, + "loss": 0.5873, + "step": 9164 + }, + { + "epoch": 1.921786538058293, + "grad_norm": 7.977876291141394, + "learning_rate": 6.0441302822220825e-06, + "loss": 0.8558, + "step": 9165 + }, + { + "epoch": 1.9219962256238206, + "grad_norm": 6.498539129163264, + "learning_rate": 6.042051167431928e-06, + "loss": 0.6293, + "step": 9166 + }, + { + "epoch": 1.9222059131893479, + "grad_norm": 9.38095639710982, + "learning_rate": 6.039972255493516e-06, + "loss": 0.8867, + "step": 9167 + }, + { + "epoch": 1.9224156007548752, + "grad_norm": 8.256527493500366, + "learning_rate": 6.03789354651339e-06, + "loss": 0.8981, + "step": 9168 + }, + { + "epoch": 1.9226252883204027, + "grad_norm": 6.636979131940089, + "learning_rate": 6.0358150405980944e-06, + "loss": 0.7353, + "step": 9169 + }, + { + "epoch": 1.92283497588593, + "grad_norm": 8.909903698117736, + "learning_rate": 6.033736737854157e-06, + "loss": 0.8786, + "step": 9170 + }, + { + "epoch": 1.9230446634514573, + "grad_norm": 8.225122566219394, + "learning_rate": 6.031658638388087e-06, + "loss": 0.7735, + "step": 9171 + }, + { + "epoch": 1.9232543510169848, + "grad_norm": 6.640383690488738, + "learning_rate": 6.0295807423063935e-06, + "loss": 0.6074, + "step": 9172 + }, + { + "epoch": 1.923464038582512, + "grad_norm": 7.302975957444926, + "learning_rate": 6.0275030497155775e-06, + "loss": 0.6412, + "step": 9173 + }, + { + "epoch": 1.9236737261480394, + "grad_norm": 5.311622440560782, + "learning_rate": 6.025425560722115e-06, + "loss": 0.4873, + "step": 9174 + }, + { + "epoch": 1.9238834137135667, + "grad_norm": 8.505035951352397, + "learning_rate": 6.023348275432488e-06, + "loss": 0.7454, + "step": 9175 + }, + { + "epoch": 1.924093101279094, + "grad_norm": 7.182864645319939, + "learning_rate": 6.021271193953162e-06, + "loss": 0.5788, + "step": 9176 + }, + { + "epoch": 1.9243027888446216, + "grad_norm": 7.365906933789078, + "learning_rate": 6.019194316390586e-06, + "loss": 0.7115, + "step": 9177 + }, + { + "epoch": 1.9245124764101489, + "grad_norm": 6.874371495020125, + "learning_rate": 6.017117642851203e-06, + "loss": 0.6019, + "step": 9178 + }, + { + "epoch": 1.9247221639756762, + "grad_norm": 5.652872581992205, + "learning_rate": 6.015041173441455e-06, + "loss": 0.5893, + "step": 9179 + }, + { + "epoch": 1.9249318515412037, + "grad_norm": 8.205954617252523, + "learning_rate": 6.012964908267752e-06, + "loss": 0.8125, + "step": 9180 + }, + { + "epoch": 1.925141539106731, + "grad_norm": 6.051682343565646, + "learning_rate": 6.010888847436516e-06, + "loss": 0.5219, + "step": 9181 + }, + { + "epoch": 1.9253512266722583, + "grad_norm": 6.202107529977549, + "learning_rate": 6.008812991054145e-06, + "loss": 0.7788, + "step": 9182 + }, + { + "epoch": 1.9255609142377859, + "grad_norm": 8.115459329767129, + "learning_rate": 6.00673733922703e-06, + "loss": 0.8739, + "step": 9183 + }, + { + "epoch": 1.925770601803313, + "grad_norm": 5.62990360487765, + "learning_rate": 6.004661892061549e-06, + "loss": 0.5218, + "step": 9184 + }, + { + "epoch": 1.9259802893688405, + "grad_norm": 6.9834884602544625, + "learning_rate": 6.002586649664084e-06, + "loss": 0.7865, + "step": 9185 + }, + { + "epoch": 1.9261899769343678, + "grad_norm": 7.1938287846980495, + "learning_rate": 6.000511612140979e-06, + "loss": 0.6954, + "step": 9186 + }, + { + "epoch": 1.926399664499895, + "grad_norm": 6.332940468764646, + "learning_rate": 5.998436779598593e-06, + "loss": 0.6354, + "step": 9187 + }, + { + "epoch": 1.9266093520654226, + "grad_norm": 7.798446494274803, + "learning_rate": 5.996362152143264e-06, + "loss": 0.6095, + "step": 9188 + }, + { + "epoch": 1.92681903963095, + "grad_norm": 9.065076775109011, + "learning_rate": 5.994287729881316e-06, + "loss": 1.0213, + "step": 9189 + }, + { + "epoch": 1.9270287271964772, + "grad_norm": 9.302297919705035, + "learning_rate": 5.99221351291907e-06, + "loss": 0.8142, + "step": 9190 + }, + { + "epoch": 1.9272384147620047, + "grad_norm": 6.5626044903679, + "learning_rate": 5.990139501362834e-06, + "loss": 0.5153, + "step": 9191 + }, + { + "epoch": 1.9274481023275318, + "grad_norm": 6.240116425117373, + "learning_rate": 5.988065695318902e-06, + "loss": 0.5713, + "step": 9192 + }, + { + "epoch": 1.9276577898930594, + "grad_norm": 6.130376204925679, + "learning_rate": 5.9859920948935624e-06, + "loss": 0.5022, + "step": 9193 + }, + { + "epoch": 1.9278674774585867, + "grad_norm": 7.957710147729744, + "learning_rate": 5.98391870019309e-06, + "loss": 0.5867, + "step": 9194 + }, + { + "epoch": 1.928077165024114, + "grad_norm": 7.014091963790038, + "learning_rate": 5.9818455113237475e-06, + "loss": 0.5082, + "step": 9195 + }, + { + "epoch": 1.9282868525896415, + "grad_norm": 7.8688242984343955, + "learning_rate": 5.979772528391793e-06, + "loss": 0.909, + "step": 9196 + }, + { + "epoch": 1.9284965401551688, + "grad_norm": 9.609443192317569, + "learning_rate": 5.9776997515034705e-06, + "loss": 0.8842, + "step": 9197 + }, + { + "epoch": 1.928706227720696, + "grad_norm": 4.919427128259844, + "learning_rate": 5.975627180765011e-06, + "loss": 0.4792, + "step": 9198 + }, + { + "epoch": 1.9289159152862236, + "grad_norm": 6.855620883275216, + "learning_rate": 5.9735548162826365e-06, + "loss": 0.5523, + "step": 9199 + }, + { + "epoch": 1.929125602851751, + "grad_norm": 8.82945972792503, + "learning_rate": 5.971482658162566e-06, + "loss": 0.8178, + "step": 9200 + }, + { + "epoch": 1.9293352904172782, + "grad_norm": 6.8894152690381425, + "learning_rate": 5.969410706510992e-06, + "loss": 0.6986, + "step": 9201 + }, + { + "epoch": 1.9295449779828058, + "grad_norm": 6.184587592156968, + "learning_rate": 5.96733896143411e-06, + "loss": 0.5876, + "step": 9202 + }, + { + "epoch": 1.9297546655483329, + "grad_norm": 5.809723403344811, + "learning_rate": 5.965267423038103e-06, + "loss": 0.5792, + "step": 9203 + }, + { + "epoch": 1.9299643531138604, + "grad_norm": 6.877603480359358, + "learning_rate": 5.96319609142914e-06, + "loss": 0.6381, + "step": 9204 + }, + { + "epoch": 1.9301740406793877, + "grad_norm": 9.635570910052625, + "learning_rate": 5.961124966713373e-06, + "loss": 0.8593, + "step": 9205 + }, + { + "epoch": 1.930383728244915, + "grad_norm": 8.527611782223218, + "learning_rate": 5.95905404899696e-06, + "loss": 0.7332, + "step": 9206 + }, + { + "epoch": 1.9305934158104425, + "grad_norm": 7.99365304815592, + "learning_rate": 5.956983338386036e-06, + "loss": 0.8465, + "step": 9207 + }, + { + "epoch": 1.9308031033759698, + "grad_norm": 5.610793084051084, + "learning_rate": 5.954912834986728e-06, + "loss": 0.5189, + "step": 9208 + }, + { + "epoch": 1.9310127909414971, + "grad_norm": 5.853094618494927, + "learning_rate": 5.95284253890515e-06, + "loss": 0.4658, + "step": 9209 + }, + { + "epoch": 1.9312224785070247, + "grad_norm": 6.546092581967956, + "learning_rate": 5.950772450247417e-06, + "loss": 0.694, + "step": 9210 + }, + { + "epoch": 1.9314321660725517, + "grad_norm": 7.602615479787789, + "learning_rate": 5.948702569119615e-06, + "loss": 0.7681, + "step": 9211 + }, + { + "epoch": 1.9316418536380793, + "grad_norm": 7.713090827354776, + "learning_rate": 5.946632895627833e-06, + "loss": 0.7491, + "step": 9212 + }, + { + "epoch": 1.9318515412036066, + "grad_norm": 6.00339158010869, + "learning_rate": 5.944563429878147e-06, + "loss": 0.5661, + "step": 9213 + }, + { + "epoch": 1.9320612287691339, + "grad_norm": 8.213204873385854, + "learning_rate": 5.942494171976619e-06, + "loss": 0.8473, + "step": 9214 + }, + { + "epoch": 1.9322709163346614, + "grad_norm": 8.983965592800304, + "learning_rate": 5.940425122029298e-06, + "loss": 0.9672, + "step": 9215 + }, + { + "epoch": 1.9324806039001887, + "grad_norm": 6.871460281070904, + "learning_rate": 5.938356280142237e-06, + "loss": 0.7107, + "step": 9216 + }, + { + "epoch": 1.932690291465716, + "grad_norm": 6.421179836039351, + "learning_rate": 5.936287646421457e-06, + "loss": 0.7952, + "step": 9217 + }, + { + "epoch": 1.9328999790312436, + "grad_norm": 6.965772093420556, + "learning_rate": 5.934219220972984e-06, + "loss": 0.664, + "step": 9218 + }, + { + "epoch": 1.9331096665967709, + "grad_norm": 7.618222413794964, + "learning_rate": 5.9321510039028295e-06, + "loss": 0.8827, + "step": 9219 + }, + { + "epoch": 1.9333193541622982, + "grad_norm": 6.847746469688917, + "learning_rate": 5.930082995316991e-06, + "loss": 0.7411, + "step": 9220 + }, + { + "epoch": 1.9335290417278257, + "grad_norm": 7.2907641957869345, + "learning_rate": 5.928015195321452e-06, + "loss": 0.699, + "step": 9221 + }, + { + "epoch": 1.9337387292933528, + "grad_norm": 5.128945889977109, + "learning_rate": 5.925947604022204e-06, + "loss": 0.4736, + "step": 9222 + }, + { + "epoch": 1.9339484168588803, + "grad_norm": 6.745419119802339, + "learning_rate": 5.923880221525205e-06, + "loss": 0.7443, + "step": 9223 + }, + { + "epoch": 1.9341581044244076, + "grad_norm": 6.944667290314115, + "learning_rate": 5.921813047936413e-06, + "loss": 0.6644, + "step": 9224 + }, + { + "epoch": 1.934367791989935, + "grad_norm": 7.084972351611858, + "learning_rate": 5.9197460833617794e-06, + "loss": 0.7261, + "step": 9225 + }, + { + "epoch": 1.9345774795554624, + "grad_norm": 7.819354598503194, + "learning_rate": 5.917679327907231e-06, + "loss": 0.6477, + "step": 9226 + }, + { + "epoch": 1.9347871671209897, + "grad_norm": 7.209069175810552, + "learning_rate": 5.915612781678699e-06, + "loss": 0.6908, + "step": 9227 + }, + { + "epoch": 1.934996854686517, + "grad_norm": 7.757144290188985, + "learning_rate": 5.913546444782099e-06, + "loss": 1.0458, + "step": 9228 + }, + { + "epoch": 1.9352065422520446, + "grad_norm": 7.267495411671543, + "learning_rate": 5.911480317323328e-06, + "loss": 0.6362, + "step": 9229 + }, + { + "epoch": 1.9354162298175717, + "grad_norm": 6.449304460548617, + "learning_rate": 5.90941439940828e-06, + "loss": 0.7882, + "step": 9230 + }, + { + "epoch": 1.9356259173830992, + "grad_norm": 5.854659580799673, + "learning_rate": 5.907348691142846e-06, + "loss": 0.4732, + "step": 9231 + }, + { + "epoch": 1.9358356049486265, + "grad_norm": 7.108478836022061, + "learning_rate": 5.905283192632883e-06, + "loss": 0.8041, + "step": 9232 + }, + { + "epoch": 1.9360452925141538, + "grad_norm": 6.318351076259204, + "learning_rate": 5.90321790398426e-06, + "loss": 0.5548, + "step": 9233 + }, + { + "epoch": 1.9362549800796813, + "grad_norm": 5.040833899471434, + "learning_rate": 5.901152825302827e-06, + "loss": 0.6743, + "step": 9234 + }, + { + "epoch": 1.9364646676452086, + "grad_norm": 7.828821237424376, + "learning_rate": 5.89908795669442e-06, + "loss": 0.8771, + "step": 9235 + }, + { + "epoch": 1.936674355210736, + "grad_norm": 6.578201393173153, + "learning_rate": 5.8970232982648656e-06, + "loss": 0.702, + "step": 9236 + }, + { + "epoch": 1.9368840427762635, + "grad_norm": 5.794155517280722, + "learning_rate": 5.894958850119988e-06, + "loss": 0.6955, + "step": 9237 + }, + { + "epoch": 1.9370937303417908, + "grad_norm": 5.915526856954934, + "learning_rate": 5.892894612365585e-06, + "loss": 0.7092, + "step": 9238 + }, + { + "epoch": 1.937303417907318, + "grad_norm": 6.153325669480489, + "learning_rate": 5.890830585107456e-06, + "loss": 0.5228, + "step": 9239 + }, + { + "epoch": 1.9375131054728456, + "grad_norm": 5.805033618586204, + "learning_rate": 5.888766768451391e-06, + "loss": 0.5226, + "step": 9240 + }, + { + "epoch": 1.9377227930383727, + "grad_norm": 8.544604226883976, + "learning_rate": 5.886703162503158e-06, + "loss": 1.0777, + "step": 9241 + }, + { + "epoch": 1.9379324806039002, + "grad_norm": 7.108878802967099, + "learning_rate": 5.884639767368518e-06, + "loss": 0.7181, + "step": 9242 + }, + { + "epoch": 1.9381421681694275, + "grad_norm": 8.500185130010607, + "learning_rate": 5.882576583153232e-06, + "loss": 0.9749, + "step": 9243 + }, + { + "epoch": 1.9383518557349548, + "grad_norm": 5.7689104669556075, + "learning_rate": 5.880513609963037e-06, + "loss": 0.5329, + "step": 9244 + }, + { + "epoch": 1.9385615433004824, + "grad_norm": 5.938595326934034, + "learning_rate": 5.878450847903663e-06, + "loss": 0.6059, + "step": 9245 + }, + { + "epoch": 1.9387712308660097, + "grad_norm": 7.310343659542488, + "learning_rate": 5.876388297080834e-06, + "loss": 0.7572, + "step": 9246 + }, + { + "epoch": 1.938980918431537, + "grad_norm": 6.459458859136461, + "learning_rate": 5.874325957600253e-06, + "loss": 0.4941, + "step": 9247 + }, + { + "epoch": 1.9391906059970645, + "grad_norm": 6.409904485497111, + "learning_rate": 5.8722638295676235e-06, + "loss": 0.4399, + "step": 9248 + }, + { + "epoch": 1.9394002935625916, + "grad_norm": 7.121771038556411, + "learning_rate": 5.870201913088634e-06, + "loss": 0.812, + "step": 9249 + }, + { + "epoch": 1.939609981128119, + "grad_norm": 7.164893122717032, + "learning_rate": 5.8681402082689575e-06, + "loss": 0.6704, + "step": 9250 + }, + { + "epoch": 1.9398196686936464, + "grad_norm": 7.410330048570561, + "learning_rate": 5.8660787152142625e-06, + "loss": 0.5786, + "step": 9251 + }, + { + "epoch": 1.9400293562591737, + "grad_norm": 7.087249885624948, + "learning_rate": 5.864017434030205e-06, + "loss": 0.7156, + "step": 9252 + }, + { + "epoch": 1.9402390438247012, + "grad_norm": 8.665001107189395, + "learning_rate": 5.861956364822422e-06, + "loss": 0.921, + "step": 9253 + }, + { + "epoch": 1.9404487313902286, + "grad_norm": 5.720479568659965, + "learning_rate": 5.859895507696556e-06, + "loss": 0.4832, + "step": 9254 + }, + { + "epoch": 1.9406584189557559, + "grad_norm": 7.731598221467886, + "learning_rate": 5.857834862758227e-06, + "loss": 0.7504, + "step": 9255 + }, + { + "epoch": 1.9408681065212834, + "grad_norm": 5.8858587203793, + "learning_rate": 5.855774430113044e-06, + "loss": 0.4916, + "step": 9256 + }, + { + "epoch": 1.9410777940868107, + "grad_norm": 7.611795451936627, + "learning_rate": 5.853714209866608e-06, + "loss": 0.6805, + "step": 9257 + }, + { + "epoch": 1.941287481652338, + "grad_norm": 6.872140220252903, + "learning_rate": 5.851654202124515e-06, + "loss": 0.7804, + "step": 9258 + }, + { + "epoch": 1.9414971692178655, + "grad_norm": 6.3589864748054925, + "learning_rate": 5.849594406992331e-06, + "loss": 0.784, + "step": 9259 + }, + { + "epoch": 1.9417068567833926, + "grad_norm": 7.527512992625092, + "learning_rate": 5.847534824575638e-06, + "loss": 0.7265, + "step": 9260 + }, + { + "epoch": 1.9419165443489201, + "grad_norm": 6.626663692999666, + "learning_rate": 5.845475454979988e-06, + "loss": 0.6825, + "step": 9261 + }, + { + "epoch": 1.9421262319144474, + "grad_norm": 6.604748541123709, + "learning_rate": 5.843416298310924e-06, + "loss": 0.6192, + "step": 9262 + }, + { + "epoch": 1.9423359194799747, + "grad_norm": 7.740332536414257, + "learning_rate": 5.841357354673982e-06, + "loss": 0.6147, + "step": 9263 + }, + { + "epoch": 1.9425456070455023, + "grad_norm": 6.724110318904424, + "learning_rate": 5.839298624174691e-06, + "loss": 0.6427, + "step": 9264 + }, + { + "epoch": 1.9427552946110296, + "grad_norm": 7.046734516927857, + "learning_rate": 5.837240106918563e-06, + "loss": 0.6331, + "step": 9265 + }, + { + "epoch": 1.9429649821765569, + "grad_norm": 6.946113395963471, + "learning_rate": 5.835181803011096e-06, + "loss": 0.6876, + "step": 9266 + }, + { + "epoch": 1.9431746697420844, + "grad_norm": 7.581627960819508, + "learning_rate": 5.833123712557789e-06, + "loss": 0.6814, + "step": 9267 + }, + { + "epoch": 1.9433843573076115, + "grad_norm": 6.150674179847868, + "learning_rate": 5.831065835664119e-06, + "loss": 0.6461, + "step": 9268 + }, + { + "epoch": 1.943594044873139, + "grad_norm": 9.355387978414148, + "learning_rate": 5.829008172435552e-06, + "loss": 0.7282, + "step": 9269 + }, + { + "epoch": 1.9438037324386663, + "grad_norm": 8.010429471087958, + "learning_rate": 5.826950722977549e-06, + "loss": 0.5761, + "step": 9270 + }, + { + "epoch": 1.9440134200041936, + "grad_norm": 5.61438119145265, + "learning_rate": 5.824893487395564e-06, + "loss": 0.648, + "step": 9271 + }, + { + "epoch": 1.9442231075697212, + "grad_norm": 7.487553321826831, + "learning_rate": 5.82283646579503e-06, + "loss": 0.6758, + "step": 9272 + }, + { + "epoch": 1.9444327951352485, + "grad_norm": 7.0716788083506135, + "learning_rate": 5.820779658281368e-06, + "loss": 0.7614, + "step": 9273 + }, + { + "epoch": 1.9446424827007758, + "grad_norm": 7.927492651645763, + "learning_rate": 5.818723064959999e-06, + "loss": 0.8617, + "step": 9274 + }, + { + "epoch": 1.9448521702663033, + "grad_norm": 6.242842619586923, + "learning_rate": 5.816666685936327e-06, + "loss": 0.5955, + "step": 9275 + }, + { + "epoch": 1.9450618578318306, + "grad_norm": 6.785903902923679, + "learning_rate": 5.814610521315738e-06, + "loss": 0.6331, + "step": 9276 + }, + { + "epoch": 1.945271545397358, + "grad_norm": 6.107795187285666, + "learning_rate": 5.812554571203624e-06, + "loss": 0.6826, + "step": 9277 + }, + { + "epoch": 1.9454812329628854, + "grad_norm": 6.040836956269254, + "learning_rate": 5.810498835705345e-06, + "loss": 0.6376, + "step": 9278 + }, + { + "epoch": 1.9456909205284125, + "grad_norm": 5.780480782298735, + "learning_rate": 5.808443314926273e-06, + "loss": 0.5165, + "step": 9279 + }, + { + "epoch": 1.94590060809394, + "grad_norm": 6.8258817826654985, + "learning_rate": 5.806388008971751e-06, + "loss": 0.6663, + "step": 9280 + }, + { + "epoch": 1.9461102956594674, + "grad_norm": 6.690171162849062, + "learning_rate": 5.804332917947111e-06, + "loss": 0.696, + "step": 9281 + }, + { + "epoch": 1.9463199832249947, + "grad_norm": 6.12625218474809, + "learning_rate": 5.802278041957687e-06, + "loss": 0.6677, + "step": 9282 + }, + { + "epoch": 1.9465296707905222, + "grad_norm": 8.743377118394523, + "learning_rate": 5.800223381108798e-06, + "loss": 0.9713, + "step": 9283 + }, + { + "epoch": 1.9467393583560495, + "grad_norm": 7.910693649827986, + "learning_rate": 5.798168935505746e-06, + "loss": 0.8692, + "step": 9284 + }, + { + "epoch": 1.9469490459215768, + "grad_norm": 6.457286587355363, + "learning_rate": 5.796114705253818e-06, + "loss": 0.545, + "step": 9285 + }, + { + "epoch": 1.9471587334871043, + "grad_norm": 8.304300458453238, + "learning_rate": 5.794060690458307e-06, + "loss": 0.7866, + "step": 9286 + }, + { + "epoch": 1.9473684210526314, + "grad_norm": 6.86688387061758, + "learning_rate": 5.792006891224478e-06, + "loss": 0.7037, + "step": 9287 + }, + { + "epoch": 1.947578108618159, + "grad_norm": 5.957818819577157, + "learning_rate": 5.789953307657596e-06, + "loss": 0.5539, + "step": 9288 + }, + { + "epoch": 1.9477877961836865, + "grad_norm": 6.766338458887155, + "learning_rate": 5.78789993986291e-06, + "loss": 0.7346, + "step": 9289 + }, + { + "epoch": 1.9479974837492136, + "grad_norm": 8.245215517592019, + "learning_rate": 5.785846787945654e-06, + "loss": 0.8964, + "step": 9290 + }, + { + "epoch": 1.948207171314741, + "grad_norm": 7.5080481369699275, + "learning_rate": 5.783793852011058e-06, + "loss": 0.7699, + "step": 9291 + }, + { + "epoch": 1.9484168588802684, + "grad_norm": 6.883068163996783, + "learning_rate": 5.781741132164349e-06, + "loss": 0.6528, + "step": 9292 + }, + { + "epoch": 1.9486265464457957, + "grad_norm": 5.283084538234154, + "learning_rate": 5.779688628510713e-06, + "loss": 0.344, + "step": 9293 + }, + { + "epoch": 1.9488362340113232, + "grad_norm": 7.53107655912143, + "learning_rate": 5.777636341155357e-06, + "loss": 0.8135, + "step": 9294 + }, + { + "epoch": 1.9490459215768505, + "grad_norm": 6.193966469472814, + "learning_rate": 5.775584270203465e-06, + "loss": 0.4428, + "step": 9295 + }, + { + "epoch": 1.9492556091423778, + "grad_norm": 5.562705661522564, + "learning_rate": 5.7735324157602055e-06, + "loss": 0.6155, + "step": 9296 + }, + { + "epoch": 1.9494652967079054, + "grad_norm": 6.83694466746269, + "learning_rate": 5.771480777930738e-06, + "loss": 0.5822, + "step": 9297 + }, + { + "epoch": 1.9496749842734324, + "grad_norm": 6.3695744723782735, + "learning_rate": 5.769429356820218e-06, + "loss": 0.5949, + "step": 9298 + }, + { + "epoch": 1.94988467183896, + "grad_norm": 6.262326135822616, + "learning_rate": 5.767378152533776e-06, + "loss": 0.5755, + "step": 9299 + }, + { + "epoch": 1.9500943594044873, + "grad_norm": 5.654409756464303, + "learning_rate": 5.765327165176549e-06, + "loss": 0.6237, + "step": 9300 + }, + { + "epoch": 1.9503040469700146, + "grad_norm": 7.2591537590281225, + "learning_rate": 5.763276394853648e-06, + "loss": 0.7722, + "step": 9301 + }, + { + "epoch": 1.950513734535542, + "grad_norm": 6.127519413257496, + "learning_rate": 5.7612258416701784e-06, + "loss": 0.6137, + "step": 9302 + }, + { + "epoch": 1.9507234221010694, + "grad_norm": 6.328713334956077, + "learning_rate": 5.759175505731236e-06, + "loss": 0.4834, + "step": 9303 + }, + { + "epoch": 1.9509331096665967, + "grad_norm": 6.165626741839007, + "learning_rate": 5.757125387141906e-06, + "loss": 0.6624, + "step": 9304 + }, + { + "epoch": 1.9511427972321242, + "grad_norm": 6.215042525426812, + "learning_rate": 5.755075486007261e-06, + "loss": 0.7724, + "step": 9305 + }, + { + "epoch": 1.9513524847976516, + "grad_norm": 7.124243081454236, + "learning_rate": 5.753025802432355e-06, + "loss": 0.7672, + "step": 9306 + }, + { + "epoch": 1.9515621723631789, + "grad_norm": 7.188388350388214, + "learning_rate": 5.750976336522247e-06, + "loss": 0.7827, + "step": 9307 + }, + { + "epoch": 1.9517718599287064, + "grad_norm": 9.647800390544647, + "learning_rate": 5.748927088381967e-06, + "loss": 0.8735, + "step": 9308 + }, + { + "epoch": 1.9519815474942335, + "grad_norm": 7.625115787006105, + "learning_rate": 5.7468780581165515e-06, + "loss": 0.763, + "step": 9309 + }, + { + "epoch": 1.952191235059761, + "grad_norm": 5.929156517893515, + "learning_rate": 5.744829245831012e-06, + "loss": 0.6244, + "step": 9310 + }, + { + "epoch": 1.9524009226252883, + "grad_norm": 6.857416179978395, + "learning_rate": 5.742780651630349e-06, + "loss": 0.6931, + "step": 9311 + }, + { + "epoch": 1.9526106101908156, + "grad_norm": 9.552718969179411, + "learning_rate": 5.740732275619562e-06, + "loss": 0.8407, + "step": 9312 + }, + { + "epoch": 1.9528202977563431, + "grad_norm": 7.119021256418084, + "learning_rate": 5.7386841179036415e-06, + "loss": 0.6902, + "step": 9313 + }, + { + "epoch": 1.9530299853218704, + "grad_norm": 5.993071399822086, + "learning_rate": 5.736636178587543e-06, + "loss": 0.5794, + "step": 9314 + }, + { + "epoch": 1.9532396728873977, + "grad_norm": 5.755348810075273, + "learning_rate": 5.734588457776234e-06, + "loss": 0.4426, + "step": 9315 + }, + { + "epoch": 1.9534493604529253, + "grad_norm": 5.516629429167927, + "learning_rate": 5.732540955574668e-06, + "loss": 0.589, + "step": 9316 + }, + { + "epoch": 1.9536590480184524, + "grad_norm": 5.857164654262589, + "learning_rate": 5.73049367208778e-06, + "loss": 0.5204, + "step": 9317 + }, + { + "epoch": 1.9538687355839799, + "grad_norm": 6.070086157996333, + "learning_rate": 5.728446607420493e-06, + "loss": 0.6715, + "step": 9318 + }, + { + "epoch": 1.9540784231495072, + "grad_norm": 6.321055237505776, + "learning_rate": 5.72639976167773e-06, + "loss": 0.5754, + "step": 9319 + }, + { + "epoch": 1.9542881107150345, + "grad_norm": 6.832625902490554, + "learning_rate": 5.724353134964386e-06, + "loss": 0.8166, + "step": 9320 + }, + { + "epoch": 1.954497798280562, + "grad_norm": 6.9326533414556515, + "learning_rate": 5.722306727385365e-06, + "loss": 0.5492, + "step": 9321 + }, + { + "epoch": 1.9547074858460893, + "grad_norm": 5.343943579308795, + "learning_rate": 5.7202605390455435e-06, + "loss": 0.5598, + "step": 9322 + }, + { + "epoch": 1.9549171734116166, + "grad_norm": 6.967273725248564, + "learning_rate": 5.718214570049787e-06, + "loss": 0.6867, + "step": 9323 + }, + { + "epoch": 1.9551268609771442, + "grad_norm": 6.960116021922867, + "learning_rate": 5.716168820502961e-06, + "loss": 0.6135, + "step": 9324 + }, + { + "epoch": 1.9553365485426715, + "grad_norm": 6.7628145982709, + "learning_rate": 5.714123290509918e-06, + "loss": 0.5242, + "step": 9325 + }, + { + "epoch": 1.9555462361081988, + "grad_norm": 7.541825819569992, + "learning_rate": 5.7120779801754875e-06, + "loss": 0.63, + "step": 9326 + }, + { + "epoch": 1.9557559236737263, + "grad_norm": 5.936818211712879, + "learning_rate": 5.710032889604496e-06, + "loss": 0.5356, + "step": 9327 + }, + { + "epoch": 1.9559656112392534, + "grad_norm": 7.847554958832894, + "learning_rate": 5.707988018901758e-06, + "loss": 0.7911, + "step": 9328 + }, + { + "epoch": 1.956175298804781, + "grad_norm": 6.765685472702689, + "learning_rate": 5.705943368172082e-06, + "loss": 0.6274, + "step": 9329 + }, + { + "epoch": 1.9563849863703082, + "grad_norm": 7.040261921243772, + "learning_rate": 5.703898937520256e-06, + "loss": 0.6799, + "step": 9330 + }, + { + "epoch": 1.9565946739358355, + "grad_norm": 6.47968987046206, + "learning_rate": 5.701854727051057e-06, + "loss": 0.5799, + "step": 9331 + }, + { + "epoch": 1.956804361501363, + "grad_norm": 6.363966215617241, + "learning_rate": 5.6998107368692625e-06, + "loss": 0.6411, + "step": 9332 + }, + { + "epoch": 1.9570140490668904, + "grad_norm": 6.877135785062415, + "learning_rate": 5.697766967079621e-06, + "loss": 0.6187, + "step": 9333 + }, + { + "epoch": 1.9572237366324177, + "grad_norm": 7.13610607904063, + "learning_rate": 5.6957234177868895e-06, + "loss": 0.9326, + "step": 9334 + }, + { + "epoch": 1.9574334241979452, + "grad_norm": 7.6776117481462265, + "learning_rate": 5.693680089095798e-06, + "loss": 0.8441, + "step": 9335 + }, + { + "epoch": 1.9576431117634723, + "grad_norm": 5.710194240747184, + "learning_rate": 5.691636981111066e-06, + "loss": 0.4383, + "step": 9336 + }, + { + "epoch": 1.9578527993289998, + "grad_norm": 7.237677067987588, + "learning_rate": 5.689594093937416e-06, + "loss": 0.6966, + "step": 9337 + }, + { + "epoch": 1.958062486894527, + "grad_norm": 5.363887846678818, + "learning_rate": 5.687551427679545e-06, + "loss": 0.4774, + "step": 9338 + }, + { + "epoch": 1.9582721744600544, + "grad_norm": 7.902001910469446, + "learning_rate": 5.685508982442138e-06, + "loss": 0.7774, + "step": 9339 + }, + { + "epoch": 1.958481862025582, + "grad_norm": 7.4645462613137195, + "learning_rate": 5.6834667583298785e-06, + "loss": 0.6452, + "step": 9340 + }, + { + "epoch": 1.9586915495911092, + "grad_norm": 6.735411050386404, + "learning_rate": 5.681424755447439e-06, + "loss": 0.5758, + "step": 9341 + }, + { + "epoch": 1.9589012371566366, + "grad_norm": 7.766642635260357, + "learning_rate": 5.679382973899471e-06, + "loss": 0.8949, + "step": 9342 + }, + { + "epoch": 1.959110924722164, + "grad_norm": 6.037821367055008, + "learning_rate": 5.6773414137906155e-06, + "loss": 0.6325, + "step": 9343 + }, + { + "epoch": 1.9593206122876914, + "grad_norm": 7.673878220044886, + "learning_rate": 5.675300075225514e-06, + "loss": 0.8598, + "step": 9344 + }, + { + "epoch": 1.9595302998532187, + "grad_norm": 7.122512581610538, + "learning_rate": 5.673258958308778e-06, + "loss": 0.7313, + "step": 9345 + }, + { + "epoch": 1.9597399874187462, + "grad_norm": 6.773324047280245, + "learning_rate": 5.671218063145032e-06, + "loss": 0.6981, + "step": 9346 + }, + { + "epoch": 1.9599496749842733, + "grad_norm": 6.289635908748517, + "learning_rate": 5.669177389838867e-06, + "loss": 0.6058, + "step": 9347 + }, + { + "epoch": 1.9601593625498008, + "grad_norm": 7.868882676605081, + "learning_rate": 5.6671369384948674e-06, + "loss": 0.7898, + "step": 9348 + }, + { + "epoch": 1.9603690501153281, + "grad_norm": 7.604711655325037, + "learning_rate": 5.665096709217615e-06, + "loss": 0.8801, + "step": 9349 + }, + { + "epoch": 1.9605787376808554, + "grad_norm": 6.917368239753009, + "learning_rate": 5.66305670211168e-06, + "loss": 0.4461, + "step": 9350 + }, + { + "epoch": 1.960788425246383, + "grad_norm": 5.825341802707215, + "learning_rate": 5.66101691728161e-06, + "loss": 0.523, + "step": 9351 + }, + { + "epoch": 1.9609981128119103, + "grad_norm": 5.8033050447809496, + "learning_rate": 5.658977354831944e-06, + "loss": 0.4775, + "step": 9352 + }, + { + "epoch": 1.9612078003774376, + "grad_norm": 6.794858279510377, + "learning_rate": 5.656938014867223e-06, + "loss": 0.6737, + "step": 9353 + }, + { + "epoch": 1.961417487942965, + "grad_norm": 5.606029073277059, + "learning_rate": 5.6548988974919605e-06, + "loss": 0.6143, + "step": 9354 + }, + { + "epoch": 1.9616271755084922, + "grad_norm": 7.377239987409064, + "learning_rate": 5.6528600028106614e-06, + "loss": 0.7319, + "step": 9355 + }, + { + "epoch": 1.9618368630740197, + "grad_norm": 6.598389338215506, + "learning_rate": 5.650821330927832e-06, + "loss": 0.718, + "step": 9356 + }, + { + "epoch": 1.962046550639547, + "grad_norm": 7.09458677360249, + "learning_rate": 5.648782881947948e-06, + "loss": 0.4531, + "step": 9357 + }, + { + "epoch": 1.9622562382050743, + "grad_norm": 6.265265433061276, + "learning_rate": 5.646744655975491e-06, + "loss": 0.5947, + "step": 9358 + }, + { + "epoch": 1.9624659257706019, + "grad_norm": 6.906000673046209, + "learning_rate": 5.644706653114923e-06, + "loss": 0.7007, + "step": 9359 + }, + { + "epoch": 1.9626756133361292, + "grad_norm": 6.440475060835994, + "learning_rate": 5.642668873470687e-06, + "loss": 0.6435, + "step": 9360 + }, + { + "epoch": 1.9628853009016565, + "grad_norm": 6.82922668184217, + "learning_rate": 5.640631317147229e-06, + "loss": 0.5335, + "step": 9361 + }, + { + "epoch": 1.963094988467184, + "grad_norm": 7.171982925019574, + "learning_rate": 5.638593984248983e-06, + "loss": 0.7807, + "step": 9362 + }, + { + "epoch": 1.9633046760327113, + "grad_norm": 8.62092906352122, + "learning_rate": 5.636556874880358e-06, + "loss": 0.7998, + "step": 9363 + }, + { + "epoch": 1.9635143635982386, + "grad_norm": 7.999146266688588, + "learning_rate": 5.634519989145757e-06, + "loss": 0.8201, + "step": 9364 + }, + { + "epoch": 1.9637240511637661, + "grad_norm": 6.076879299069951, + "learning_rate": 5.632483327149583e-06, + "loss": 0.625, + "step": 9365 + }, + { + "epoch": 1.9639337387292932, + "grad_norm": 6.251812390313838, + "learning_rate": 5.630446888996208e-06, + "loss": 0.5772, + "step": 9366 + }, + { + "epoch": 1.9641434262948207, + "grad_norm": 6.138863755946371, + "learning_rate": 5.628410674790015e-06, + "loss": 0.6098, + "step": 9367 + }, + { + "epoch": 1.964353113860348, + "grad_norm": 6.603710533719285, + "learning_rate": 5.626374684635354e-06, + "loss": 0.7069, + "step": 9368 + }, + { + "epoch": 1.9645628014258754, + "grad_norm": 6.7745432168439725, + "learning_rate": 5.624338918636574e-06, + "loss": 0.6011, + "step": 9369 + }, + { + "epoch": 1.9647724889914029, + "grad_norm": 5.227262983509193, + "learning_rate": 5.622303376898013e-06, + "loss": 0.4877, + "step": 9370 + }, + { + "epoch": 1.9649821765569302, + "grad_norm": 7.412895579602011, + "learning_rate": 5.6202680595240056e-06, + "loss": 0.578, + "step": 9371 + }, + { + "epoch": 1.9651918641224575, + "grad_norm": 8.143893219406824, + "learning_rate": 5.618232966618848e-06, + "loss": 0.9202, + "step": 9372 + }, + { + "epoch": 1.965401551687985, + "grad_norm": 6.731643576160104, + "learning_rate": 5.6161980982868505e-06, + "loss": 0.7872, + "step": 9373 + }, + { + "epoch": 1.965611239253512, + "grad_norm": 6.837902938038457, + "learning_rate": 5.614163454632308e-06, + "loss": 0.6965, + "step": 9374 + }, + { + "epoch": 1.9658209268190396, + "grad_norm": 6.374086728022143, + "learning_rate": 5.612129035759495e-06, + "loss": 0.6395, + "step": 9375 + }, + { + "epoch": 1.966030614384567, + "grad_norm": 6.633101033280499, + "learning_rate": 5.610094841772675e-06, + "loss": 0.702, + "step": 9376 + }, + { + "epoch": 1.9662403019500942, + "grad_norm": 5.402698443485467, + "learning_rate": 5.608060872776112e-06, + "loss": 0.4333, + "step": 9377 + }, + { + "epoch": 1.9664499895156218, + "grad_norm": 7.339949664691134, + "learning_rate": 5.6060271288740435e-06, + "loss": 0.8198, + "step": 9378 + }, + { + "epoch": 1.966659677081149, + "grad_norm": 7.953023414300289, + "learning_rate": 5.60399361017071e-06, + "loss": 0.8272, + "step": 9379 + }, + { + "epoch": 1.9668693646466764, + "grad_norm": 9.21246066271867, + "learning_rate": 5.601960316770328e-06, + "loss": 1.0815, + "step": 9380 + }, + { + "epoch": 1.967079052212204, + "grad_norm": 6.2376399289821105, + "learning_rate": 5.599927248777103e-06, + "loss": 0.7165, + "step": 9381 + }, + { + "epoch": 1.9672887397777312, + "grad_norm": 8.07526508600555, + "learning_rate": 5.597894406295239e-06, + "loss": 0.9383, + "step": 9382 + }, + { + "epoch": 1.9674984273432585, + "grad_norm": 5.783708449822075, + "learning_rate": 5.595861789428925e-06, + "loss": 0.4718, + "step": 9383 + }, + { + "epoch": 1.967708114908786, + "grad_norm": 7.007941281927883, + "learning_rate": 5.593829398282334e-06, + "loss": 0.6255, + "step": 9384 + }, + { + "epoch": 1.9679178024743131, + "grad_norm": 6.962596699983334, + "learning_rate": 5.591797232959623e-06, + "loss": 0.7037, + "step": 9385 + }, + { + "epoch": 1.9681274900398407, + "grad_norm": 7.8042862770405375, + "learning_rate": 5.58976529356495e-06, + "loss": 0.7684, + "step": 9386 + }, + { + "epoch": 1.968337177605368, + "grad_norm": 7.786136202570437, + "learning_rate": 5.587733580202459e-06, + "loss": 0.6926, + "step": 9387 + }, + { + "epoch": 1.9685468651708953, + "grad_norm": 9.177226742386363, + "learning_rate": 5.585702092976275e-06, + "loss": 0.9397, + "step": 9388 + }, + { + "epoch": 1.9687565527364228, + "grad_norm": 6.969590138568324, + "learning_rate": 5.583670831990511e-06, + "loss": 0.6873, + "step": 9389 + }, + { + "epoch": 1.96896624030195, + "grad_norm": 9.650787107579669, + "learning_rate": 5.5816397973492794e-06, + "loss": 1.1807, + "step": 9390 + }, + { + "epoch": 1.9691759278674774, + "grad_norm": 6.902674016731259, + "learning_rate": 5.579608989156668e-06, + "loss": 0.6546, + "step": 9391 + }, + { + "epoch": 1.969385615433005, + "grad_norm": 6.2798509356303915, + "learning_rate": 5.577578407516768e-06, + "loss": 0.5278, + "step": 9392 + }, + { + "epoch": 1.969595302998532, + "grad_norm": 8.447894334273606, + "learning_rate": 5.575548052533644e-06, + "loss": 0.9003, + "step": 9393 + }, + { + "epoch": 1.9698049905640596, + "grad_norm": 6.726723638569409, + "learning_rate": 5.573517924311351e-06, + "loss": 0.8201, + "step": 9394 + }, + { + "epoch": 1.9700146781295869, + "grad_norm": 8.783838241709548, + "learning_rate": 5.571488022953944e-06, + "loss": 0.8921, + "step": 9395 + }, + { + "epoch": 1.9702243656951142, + "grad_norm": 6.45283198817106, + "learning_rate": 5.569458348565463e-06, + "loss": 0.6829, + "step": 9396 + }, + { + "epoch": 1.9704340532606417, + "grad_norm": 6.0174351594428455, + "learning_rate": 5.567428901249919e-06, + "loss": 0.5207, + "step": 9397 + }, + { + "epoch": 1.970643740826169, + "grad_norm": 7.699177638108156, + "learning_rate": 5.565399681111331e-06, + "loss": 0.6518, + "step": 9398 + }, + { + "epoch": 1.9708534283916963, + "grad_norm": 6.335525619839974, + "learning_rate": 5.5633706882537045e-06, + "loss": 0.4691, + "step": 9399 + }, + { + "epoch": 1.9710631159572238, + "grad_norm": 5.174179431174651, + "learning_rate": 5.561341922781026e-06, + "loss": 0.5739, + "step": 9400 + }, + { + "epoch": 1.9712728035227511, + "grad_norm": 6.324741584455501, + "learning_rate": 5.559313384797268e-06, + "loss": 0.5169, + "step": 9401 + }, + { + "epoch": 1.9714824910882784, + "grad_norm": 6.788546184522243, + "learning_rate": 5.557285074406405e-06, + "loss": 0.5962, + "step": 9402 + }, + { + "epoch": 1.971692178653806, + "grad_norm": 7.4878767110235716, + "learning_rate": 5.555256991712383e-06, + "loss": 0.7694, + "step": 9403 + }, + { + "epoch": 1.971901866219333, + "grad_norm": 6.381684228076002, + "learning_rate": 5.553229136819154e-06, + "loss": 0.5776, + "step": 9404 + }, + { + "epoch": 1.9721115537848606, + "grad_norm": 7.285354227541252, + "learning_rate": 5.5512015098306435e-06, + "loss": 0.7065, + "step": 9405 + }, + { + "epoch": 1.9723212413503879, + "grad_norm": 7.472143171805601, + "learning_rate": 5.549174110850769e-06, + "loss": 1.0457, + "step": 9406 + }, + { + "epoch": 1.9725309289159152, + "grad_norm": 9.481669256434031, + "learning_rate": 5.54714693998344e-06, + "loss": 1.0122, + "step": 9407 + }, + { + "epoch": 1.9727406164814427, + "grad_norm": 5.4602505292051875, + "learning_rate": 5.545119997332558e-06, + "loss": 0.4985, + "step": 9408 + }, + { + "epoch": 1.97295030404697, + "grad_norm": 8.207387419883343, + "learning_rate": 5.543093283002001e-06, + "loss": 0.9234, + "step": 9409 + }, + { + "epoch": 1.9731599916124973, + "grad_norm": 6.2840237412150675, + "learning_rate": 5.541066797095641e-06, + "loss": 0.7049, + "step": 9410 + }, + { + "epoch": 1.9733696791780249, + "grad_norm": 8.583895324791946, + "learning_rate": 5.539040539717345e-06, + "loss": 0.7651, + "step": 9411 + }, + { + "epoch": 1.973579366743552, + "grad_norm": 4.752872794226008, + "learning_rate": 5.537014510970955e-06, + "loss": 0.4306, + "step": 9412 + }, + { + "epoch": 1.9737890543090795, + "grad_norm": 6.442977046781971, + "learning_rate": 5.534988710960314e-06, + "loss": 0.7468, + "step": 9413 + }, + { + "epoch": 1.973998741874607, + "grad_norm": 7.563168597369974, + "learning_rate": 5.532963139789247e-06, + "loss": 0.671, + "step": 9414 + }, + { + "epoch": 1.974208429440134, + "grad_norm": 6.237503327962937, + "learning_rate": 5.530937797561564e-06, + "loss": 0.6813, + "step": 9415 + }, + { + "epoch": 1.9744181170056616, + "grad_norm": 4.912833147401991, + "learning_rate": 5.528912684381071e-06, + "loss": 0.3851, + "step": 9416 + }, + { + "epoch": 1.974627804571189, + "grad_norm": 7.374596284417961, + "learning_rate": 5.526887800351561e-06, + "loss": 0.7471, + "step": 9417 + }, + { + "epoch": 1.9748374921367162, + "grad_norm": 5.144064355994955, + "learning_rate": 5.524863145576804e-06, + "loss": 0.4548, + "step": 9418 + }, + { + "epoch": 1.9750471797022437, + "grad_norm": 7.442101373770801, + "learning_rate": 5.522838720160573e-06, + "loss": 0.7283, + "step": 9419 + }, + { + "epoch": 1.975256867267771, + "grad_norm": 6.016025671682098, + "learning_rate": 5.5208145242066255e-06, + "loss": 0.6271, + "step": 9420 + }, + { + "epoch": 1.9754665548332984, + "grad_norm": 7.407011519084812, + "learning_rate": 5.518790557818703e-06, + "loss": 0.7151, + "step": 9421 + }, + { + "epoch": 1.9756762423988259, + "grad_norm": 7.230074007706148, + "learning_rate": 5.516766821100532e-06, + "loss": 0.7484, + "step": 9422 + }, + { + "epoch": 1.975885929964353, + "grad_norm": 5.956696107692165, + "learning_rate": 5.514743314155842e-06, + "loss": 0.7203, + "step": 9423 + }, + { + "epoch": 1.9760956175298805, + "grad_norm": 5.081451357091475, + "learning_rate": 5.512720037088332e-06, + "loss": 0.4171, + "step": 9424 + }, + { + "epoch": 1.9763053050954078, + "grad_norm": 6.999221019875437, + "learning_rate": 5.510696990001709e-06, + "loss": 0.6734, + "step": 9425 + }, + { + "epoch": 1.976514992660935, + "grad_norm": 9.391370219701086, + "learning_rate": 5.50867417299965e-06, + "loss": 0.9458, + "step": 9426 + }, + { + "epoch": 1.9767246802264626, + "grad_norm": 8.891903065739188, + "learning_rate": 5.506651586185825e-06, + "loss": 1.0335, + "step": 9427 + }, + { + "epoch": 1.97693436779199, + "grad_norm": 7.184276652326688, + "learning_rate": 5.5046292296639e-06, + "loss": 0.7992, + "step": 9428 + }, + { + "epoch": 1.9771440553575172, + "grad_norm": 6.483110092348531, + "learning_rate": 5.502607103537528e-06, + "loss": 0.6756, + "step": 9429 + }, + { + "epoch": 1.9773537429230448, + "grad_norm": 6.608294088329845, + "learning_rate": 5.500585207910342e-06, + "loss": 0.6605, + "step": 9430 + }, + { + "epoch": 1.977563430488572, + "grad_norm": 7.008185365043187, + "learning_rate": 5.498563542885964e-06, + "loss": 0.7652, + "step": 9431 + }, + { + "epoch": 1.9777731180540994, + "grad_norm": 7.706006103979323, + "learning_rate": 5.496542108568015e-06, + "loss": 0.759, + "step": 9432 + }, + { + "epoch": 1.977982805619627, + "grad_norm": 6.63354643857184, + "learning_rate": 5.494520905060095e-06, + "loss": 0.6688, + "step": 9433 + }, + { + "epoch": 1.978192493185154, + "grad_norm": 7.450388954247213, + "learning_rate": 5.492499932465788e-06, + "loss": 0.7653, + "step": 9434 + }, + { + "epoch": 1.9784021807506815, + "grad_norm": 7.324945295029092, + "learning_rate": 5.490479190888683e-06, + "loss": 0.6122, + "step": 9435 + }, + { + "epoch": 1.9786118683162088, + "grad_norm": 6.963614042749927, + "learning_rate": 5.488458680432333e-06, + "loss": 0.7877, + "step": 9436 + }, + { + "epoch": 1.9788215558817361, + "grad_norm": 7.272216273213749, + "learning_rate": 5.486438401200309e-06, + "loss": 0.6951, + "step": 9437 + }, + { + "epoch": 1.9790312434472637, + "grad_norm": 6.290803817800356, + "learning_rate": 5.4844183532961426e-06, + "loss": 0.6761, + "step": 9438 + }, + { + "epoch": 1.979240931012791, + "grad_norm": 7.270464472644718, + "learning_rate": 5.482398536823364e-06, + "loss": 0.7697, + "step": 9439 + }, + { + "epoch": 1.9794506185783183, + "grad_norm": 6.65241455448122, + "learning_rate": 5.480378951885496e-06, + "loss": 0.6053, + "step": 9440 + }, + { + "epoch": 1.9796603061438458, + "grad_norm": 6.47877162068758, + "learning_rate": 5.4783595985860495e-06, + "loss": 0.7568, + "step": 9441 + }, + { + "epoch": 1.9798699937093729, + "grad_norm": 5.33570126000777, + "learning_rate": 5.476340477028517e-06, + "loss": 0.5061, + "step": 9442 + }, + { + "epoch": 1.9800796812749004, + "grad_norm": 8.035842186557062, + "learning_rate": 5.474321587316376e-06, + "loss": 0.8157, + "step": 9443 + }, + { + "epoch": 1.9802893688404277, + "grad_norm": 8.60835021249437, + "learning_rate": 5.4723029295531075e-06, + "loss": 0.7627, + "step": 9444 + }, + { + "epoch": 1.980499056405955, + "grad_norm": 6.672032054994563, + "learning_rate": 5.470284503842163e-06, + "loss": 0.6409, + "step": 9445 + }, + { + "epoch": 1.9807087439714826, + "grad_norm": 8.440472147140861, + "learning_rate": 5.468266310286996e-06, + "loss": 0.7713, + "step": 9446 + }, + { + "epoch": 1.9809184315370099, + "grad_norm": 7.330758440570189, + "learning_rate": 5.466248348991039e-06, + "loss": 0.8678, + "step": 9447 + }, + { + "epoch": 1.9811281191025372, + "grad_norm": 7.548305564203889, + "learning_rate": 5.46423062005772e-06, + "loss": 0.8558, + "step": 9448 + }, + { + "epoch": 1.9813378066680647, + "grad_norm": 6.738726490624901, + "learning_rate": 5.462213123590445e-06, + "loss": 0.7448, + "step": 9449 + }, + { + "epoch": 1.981547494233592, + "grad_norm": 5.880309786655555, + "learning_rate": 5.460195859692621e-06, + "loss": 0.6144, + "step": 9450 + }, + { + "epoch": 1.9817571817991193, + "grad_norm": 7.170694123341599, + "learning_rate": 5.458178828467635e-06, + "loss": 0.7332, + "step": 9451 + }, + { + "epoch": 1.9819668693646468, + "grad_norm": 6.788659596038343, + "learning_rate": 5.456162030018856e-06, + "loss": 0.7349, + "step": 9452 + }, + { + "epoch": 1.982176556930174, + "grad_norm": 7.652891320562545, + "learning_rate": 5.454145464449654e-06, + "loss": 0.8881, + "step": 9453 + }, + { + "epoch": 1.9823862444957014, + "grad_norm": 8.598729244950007, + "learning_rate": 5.45212913186339e-06, + "loss": 0.8452, + "step": 9454 + }, + { + "epoch": 1.9825959320612287, + "grad_norm": 5.928767362202632, + "learning_rate": 5.450113032363387e-06, + "loss": 0.5574, + "step": 9455 + }, + { + "epoch": 1.982805619626756, + "grad_norm": 6.9519064631136365, + "learning_rate": 5.448097166052983e-06, + "loss": 0.7053, + "step": 9456 + }, + { + "epoch": 1.9830153071922836, + "grad_norm": 8.443731131257264, + "learning_rate": 5.4460815330354975e-06, + "loss": 0.8789, + "step": 9457 + }, + { + "epoch": 1.9832249947578109, + "grad_norm": 6.1762552647965325, + "learning_rate": 5.444066133414232e-06, + "loss": 0.4977, + "step": 9458 + }, + { + "epoch": 1.9834346823233382, + "grad_norm": 8.120962267109922, + "learning_rate": 5.442050967292476e-06, + "loss": 0.8862, + "step": 9459 + }, + { + "epoch": 1.9836443698888657, + "grad_norm": 7.02294902222002, + "learning_rate": 5.440036034773517e-06, + "loss": 0.7794, + "step": 9460 + }, + { + "epoch": 1.9838540574543928, + "grad_norm": 6.166623535071192, + "learning_rate": 5.438021335960615e-06, + "loss": 0.6707, + "step": 9461 + }, + { + "epoch": 1.9840637450199203, + "grad_norm": 6.545968616484437, + "learning_rate": 5.436006870957037e-06, + "loss": 0.5821, + "step": 9462 + }, + { + "epoch": 1.9842734325854476, + "grad_norm": 8.752112825788625, + "learning_rate": 5.433992639866022e-06, + "loss": 0.914, + "step": 9463 + }, + { + "epoch": 1.984483120150975, + "grad_norm": 5.213425043088967, + "learning_rate": 5.4319786427908e-06, + "loss": 0.4919, + "step": 9464 + }, + { + "epoch": 1.9846928077165025, + "grad_norm": 6.504175622476662, + "learning_rate": 5.429964879834595e-06, + "loss": 0.6517, + "step": 9465 + }, + { + "epoch": 1.9849024952820298, + "grad_norm": 6.6495877872249505, + "learning_rate": 5.42795135110062e-06, + "loss": 0.7559, + "step": 9466 + }, + { + "epoch": 1.985112182847557, + "grad_norm": 6.598786953938955, + "learning_rate": 5.425938056692068e-06, + "loss": 0.6777, + "step": 9467 + }, + { + "epoch": 1.9853218704130846, + "grad_norm": 7.346209749862573, + "learning_rate": 5.423924996712122e-06, + "loss": 0.6072, + "step": 9468 + }, + { + "epoch": 1.985531557978612, + "grad_norm": 6.339433350768146, + "learning_rate": 5.421912171263959e-06, + "loss": 0.6662, + "step": 9469 + }, + { + "epoch": 1.9857412455441392, + "grad_norm": 8.24782553851633, + "learning_rate": 5.4198995804507335e-06, + "loss": 0.8687, + "step": 9470 + }, + { + "epoch": 1.9859509331096667, + "grad_norm": 6.104598991681448, + "learning_rate": 5.417887224375602e-06, + "loss": 0.5329, + "step": 9471 + }, + { + "epoch": 1.9861606206751938, + "grad_norm": 7.071163539617937, + "learning_rate": 5.415875103141698e-06, + "loss": 0.7198, + "step": 9472 + }, + { + "epoch": 1.9863703082407214, + "grad_norm": 6.686773708184286, + "learning_rate": 5.413863216852142e-06, + "loss": 0.4949, + "step": 9473 + }, + { + "epoch": 1.9865799958062487, + "grad_norm": 6.42064073661928, + "learning_rate": 5.41185156561005e-06, + "loss": 0.548, + "step": 9474 + }, + { + "epoch": 1.986789683371776, + "grad_norm": 6.0919823130629425, + "learning_rate": 5.409840149518531e-06, + "loss": 0.72, + "step": 9475 + }, + { + "epoch": 1.9869993709373035, + "grad_norm": 6.618788605416218, + "learning_rate": 5.407828968680657e-06, + "loss": 0.5999, + "step": 9476 + }, + { + "epoch": 1.9872090585028308, + "grad_norm": 7.0881444991206575, + "learning_rate": 5.405818023199514e-06, + "loss": 0.773, + "step": 9477 + }, + { + "epoch": 1.987418746068358, + "grad_norm": 6.443849337988713, + "learning_rate": 5.403807313178169e-06, + "loss": 0.6718, + "step": 9478 + }, + { + "epoch": 1.9876284336338856, + "grad_norm": 6.907495589688916, + "learning_rate": 5.401796838719672e-06, + "loss": 0.7668, + "step": 9479 + }, + { + "epoch": 1.9878381211994127, + "grad_norm": 8.319902880082024, + "learning_rate": 5.399786599927057e-06, + "loss": 0.748, + "step": 9480 + }, + { + "epoch": 1.9880478087649402, + "grad_norm": 7.085893351681256, + "learning_rate": 5.397776596903361e-06, + "loss": 0.7148, + "step": 9481 + }, + { + "epoch": 1.9882574963304676, + "grad_norm": 6.175848088575357, + "learning_rate": 5.395766829751594e-06, + "loss": 0.5696, + "step": 9482 + }, + { + "epoch": 1.9884671838959949, + "grad_norm": 7.25650782017251, + "learning_rate": 5.393757298574767e-06, + "loss": 0.7646, + "step": 9483 + }, + { + "epoch": 1.9886768714615224, + "grad_norm": 7.794097950516992, + "learning_rate": 5.391748003475865e-06, + "loss": 0.8639, + "step": 9484 + }, + { + "epoch": 1.9888865590270497, + "grad_norm": 6.552283306202356, + "learning_rate": 5.389738944557867e-06, + "loss": 0.6833, + "step": 9485 + }, + { + "epoch": 1.989096246592577, + "grad_norm": 6.8115769814195, + "learning_rate": 5.387730121923747e-06, + "loss": 0.6566, + "step": 9486 + }, + { + "epoch": 1.9893059341581045, + "grad_norm": 7.080468123382386, + "learning_rate": 5.385721535676459e-06, + "loss": 0.8411, + "step": 9487 + }, + { + "epoch": 1.9895156217236318, + "grad_norm": 5.651009838868676, + "learning_rate": 5.3837131859189466e-06, + "loss": 0.4571, + "step": 9488 + }, + { + "epoch": 1.9897253092891591, + "grad_norm": 7.8272547683109, + "learning_rate": 5.381705072754136e-06, + "loss": 0.5843, + "step": 9489 + }, + { + "epoch": 1.9899349968546867, + "grad_norm": 6.49177208006313, + "learning_rate": 5.3796971962849545e-06, + "loss": 0.6053, + "step": 9490 + }, + { + "epoch": 1.9901446844202137, + "grad_norm": 6.859427276740151, + "learning_rate": 5.3776895566143005e-06, + "loss": 0.7397, + "step": 9491 + }, + { + "epoch": 1.9903543719857413, + "grad_norm": 8.95844756938858, + "learning_rate": 5.375682153845078e-06, + "loss": 0.9381, + "step": 9492 + }, + { + "epoch": 1.9905640595512686, + "grad_norm": 5.995766473747759, + "learning_rate": 5.373674988080164e-06, + "loss": 0.5716, + "step": 9493 + }, + { + "epoch": 1.9907737471167959, + "grad_norm": 6.592947171193219, + "learning_rate": 5.371668059422429e-06, + "loss": 0.658, + "step": 9494 + }, + { + "epoch": 1.9909834346823234, + "grad_norm": 7.130471565893441, + "learning_rate": 5.3696613679747365e-06, + "loss": 0.7266, + "step": 9495 + }, + { + "epoch": 1.9911931222478507, + "grad_norm": 9.731677914761875, + "learning_rate": 5.36765491383993e-06, + "loss": 1.1766, + "step": 9496 + }, + { + "epoch": 1.991402809813378, + "grad_norm": 5.580330408713471, + "learning_rate": 5.36564869712084e-06, + "loss": 0.522, + "step": 9497 + }, + { + "epoch": 1.9916124973789056, + "grad_norm": 6.558839137341501, + "learning_rate": 5.3636427179202924e-06, + "loss": 0.6995, + "step": 9498 + }, + { + "epoch": 1.9918221849444326, + "grad_norm": 7.914830619142934, + "learning_rate": 5.3616369763411e-06, + "loss": 0.657, + "step": 9499 + }, + { + "epoch": 1.9920318725099602, + "grad_norm": 5.802051472066316, + "learning_rate": 5.359631472486057e-06, + "loss": 0.529, + "step": 9500 + }, + { + "epoch": 1.9922415600754875, + "grad_norm": 6.144464228550104, + "learning_rate": 5.357626206457946e-06, + "loss": 0.6986, + "step": 9501 + }, + { + "epoch": 1.9924512476410148, + "grad_norm": 6.582266032364133, + "learning_rate": 5.355621178359549e-06, + "loss": 0.6467, + "step": 9502 + }, + { + "epoch": 1.9926609352065423, + "grad_norm": 6.687776499674018, + "learning_rate": 5.353616388293617e-06, + "loss": 0.485, + "step": 9503 + }, + { + "epoch": 1.9928706227720696, + "grad_norm": 6.362933304209302, + "learning_rate": 5.351611836362907e-06, + "loss": 0.6516, + "step": 9504 + }, + { + "epoch": 1.993080310337597, + "grad_norm": 7.344398220094289, + "learning_rate": 5.3496075226701525e-06, + "loss": 0.6535, + "step": 9505 + }, + { + "epoch": 1.9932899979031244, + "grad_norm": 5.859689075476522, + "learning_rate": 5.347603447318075e-06, + "loss": 0.5416, + "step": 9506 + }, + { + "epoch": 1.9934996854686517, + "grad_norm": 5.985380369338966, + "learning_rate": 5.345599610409389e-06, + "loss": 0.6754, + "step": 9507 + }, + { + "epoch": 1.993709373034179, + "grad_norm": 7.494279636468731, + "learning_rate": 5.3435960120468014e-06, + "loss": 0.827, + "step": 9508 + }, + { + "epoch": 1.9939190605997066, + "grad_norm": 7.187364253286043, + "learning_rate": 5.341592652332992e-06, + "loss": 0.5401, + "step": 9509 + }, + { + "epoch": 1.9941287481652337, + "grad_norm": 7.121691402763271, + "learning_rate": 5.339589531370637e-06, + "loss": 0.9629, + "step": 9510 + }, + { + "epoch": 1.9943384357307612, + "grad_norm": 7.3734510550967505, + "learning_rate": 5.3375866492624e-06, + "loss": 0.6937, + "step": 9511 + }, + { + "epoch": 1.9945481232962885, + "grad_norm": 8.1002367787133, + "learning_rate": 5.3355840061109435e-06, + "loss": 0.7852, + "step": 9512 + }, + { + "epoch": 1.9947578108618158, + "grad_norm": 8.492345264078343, + "learning_rate": 5.333581602018887e-06, + "loss": 0.7809, + "step": 9513 + }, + { + "epoch": 1.9949674984273433, + "grad_norm": 6.378506011343575, + "learning_rate": 5.331579437088867e-06, + "loss": 0.5481, + "step": 9514 + }, + { + "epoch": 1.9951771859928706, + "grad_norm": 8.355202408906475, + "learning_rate": 5.329577511423503e-06, + "loss": 0.9749, + "step": 9515 + }, + { + "epoch": 1.995386873558398, + "grad_norm": 6.531685533871646, + "learning_rate": 5.327575825125391e-06, + "loss": 0.5367, + "step": 9516 + }, + { + "epoch": 1.9955965611239255, + "grad_norm": 6.219300373022308, + "learning_rate": 5.325574378297118e-06, + "loss": 0.5536, + "step": 9517 + }, + { + "epoch": 1.9958062486894526, + "grad_norm": 7.491805274937188, + "learning_rate": 5.3235731710412695e-06, + "loss": 0.795, + "step": 9518 + }, + { + "epoch": 1.99601593625498, + "grad_norm": 5.109495819920621, + "learning_rate": 5.321572203460403e-06, + "loss": 0.4679, + "step": 9519 + }, + { + "epoch": 1.9962256238205074, + "grad_norm": 8.200383703092333, + "learning_rate": 5.31957147565708e-06, + "loss": 0.6264, + "step": 9520 + }, + { + "epoch": 1.9964353113860347, + "grad_norm": 6.5908987510039125, + "learning_rate": 5.317570987733836e-06, + "loss": 0.6694, + "step": 9521 + }, + { + "epoch": 1.9966449989515622, + "grad_norm": 8.0699981547914, + "learning_rate": 5.315570739793194e-06, + "loss": 0.7175, + "step": 9522 + }, + { + "epoch": 1.9968546865170895, + "grad_norm": 6.843659335835738, + "learning_rate": 5.313570731937679e-06, + "loss": 0.5744, + "step": 9523 + }, + { + "epoch": 1.9970643740826168, + "grad_norm": 6.540647539926049, + "learning_rate": 5.311570964269796e-06, + "loss": 0.7388, + "step": 9524 + }, + { + "epoch": 1.9972740616481444, + "grad_norm": 7.399494533194387, + "learning_rate": 5.3095714368920325e-06, + "loss": 0.7449, + "step": 9525 + }, + { + "epoch": 1.9974837492136717, + "grad_norm": 7.474271122216735, + "learning_rate": 5.3075721499068634e-06, + "loss": 0.679, + "step": 9526 + }, + { + "epoch": 1.997693436779199, + "grad_norm": 7.012617566297008, + "learning_rate": 5.305573103416766e-06, + "loss": 0.8424, + "step": 9527 + }, + { + "epoch": 1.9979031243447265, + "grad_norm": 6.1609432156821935, + "learning_rate": 5.3035742975241836e-06, + "loss": 0.4282, + "step": 9528 + }, + { + "epoch": 1.9981128119102536, + "grad_norm": 5.870847821874881, + "learning_rate": 5.3015757323315675e-06, + "loss": 0.4788, + "step": 9529 + }, + { + "epoch": 1.9983224994757811, + "grad_norm": 7.395338042292884, + "learning_rate": 5.299577407941345e-06, + "loss": 0.6664, + "step": 9530 + }, + { + "epoch": 1.9985321870413084, + "grad_norm": 9.68710053816261, + "learning_rate": 5.297579324455928e-06, + "loss": 1.1013, + "step": 9531 + }, + { + "epoch": 1.9987418746068357, + "grad_norm": 8.797282530913737, + "learning_rate": 5.295581481977727e-06, + "loss": 0.6876, + "step": 9532 + }, + { + "epoch": 1.9989515621723632, + "grad_norm": 8.530586612719365, + "learning_rate": 5.293583880609143e-06, + "loss": 0.7476, + "step": 9533 + }, + { + "epoch": 1.9991612497378906, + "grad_norm": 6.792998468077389, + "learning_rate": 5.291586520452541e-06, + "loss": 0.5898, + "step": 9534 + }, + { + "epoch": 1.9993709373034179, + "grad_norm": 7.300385259550495, + "learning_rate": 5.289589401610293e-06, + "loss": 0.6956, + "step": 9535 + }, + { + "epoch": 1.9995806248689454, + "grad_norm": 7.400259138093475, + "learning_rate": 5.2875925241847645e-06, + "loss": 0.6737, + "step": 9536 + }, + { + "epoch": 1.9997903124344725, + "grad_norm": 6.629400462192665, + "learning_rate": 5.285595888278292e-06, + "loss": 0.6111, + "step": 9537 + }, + { + "epoch": 2.0, + "grad_norm": 5.767978948009052, + "learning_rate": 5.283599493993202e-06, + "loss": 0.5602, + "step": 9538 + }, + { + "epoch": 2.0002096875655275, + "grad_norm": 3.598477005270664, + "learning_rate": 5.281603341431822e-06, + "loss": 0.2145, + "step": 9539 + }, + { + "epoch": 2.0004193751310546, + "grad_norm": 3.533707912011714, + "learning_rate": 5.279607430696452e-06, + "loss": 0.2322, + "step": 9540 + }, + { + "epoch": 2.000629062696582, + "grad_norm": 3.9390980060928014, + "learning_rate": 5.277611761889392e-06, + "loss": 0.1959, + "step": 9541 + }, + { + "epoch": 2.0008387502621097, + "grad_norm": 3.4720062287390703, + "learning_rate": 5.275616335112918e-06, + "loss": 0.1642, + "step": 9542 + }, + { + "epoch": 2.0010484378276368, + "grad_norm": 3.2003169542283163, + "learning_rate": 5.273621150469299e-06, + "loss": 0.1834, + "step": 9543 + }, + { + "epoch": 2.0012581253931643, + "grad_norm": 4.379293570544043, + "learning_rate": 5.271626208060791e-06, + "loss": 0.2045, + "step": 9544 + }, + { + "epoch": 2.0014678129586914, + "grad_norm": 3.9151749958400313, + "learning_rate": 5.269631507989647e-06, + "loss": 0.1967, + "step": 9545 + }, + { + "epoch": 2.001677500524219, + "grad_norm": 3.7454938122245536, + "learning_rate": 5.267637050358092e-06, + "loss": 0.2195, + "step": 9546 + }, + { + "epoch": 2.0018871880897464, + "grad_norm": 4.391417624413541, + "learning_rate": 5.265642835268343e-06, + "loss": 0.254, + "step": 9547 + }, + { + "epoch": 2.0020968756552735, + "grad_norm": 3.7194747169797924, + "learning_rate": 5.2636488628226125e-06, + "loss": 0.1743, + "step": 9548 + }, + { + "epoch": 2.002306563220801, + "grad_norm": 3.2097374997450947, + "learning_rate": 5.261655133123089e-06, + "loss": 0.1752, + "step": 9549 + }, + { + "epoch": 2.0025162507863286, + "grad_norm": 2.915400488410725, + "learning_rate": 5.259661646271962e-06, + "loss": 0.152, + "step": 9550 + }, + { + "epoch": 2.0027259383518556, + "grad_norm": 3.4782649504915537, + "learning_rate": 5.257668402371398e-06, + "loss": 0.2002, + "step": 9551 + }, + { + "epoch": 2.002935625917383, + "grad_norm": 3.226478019123892, + "learning_rate": 5.255675401523549e-06, + "loss": 0.1956, + "step": 9552 + }, + { + "epoch": 2.0031453134829103, + "grad_norm": 2.49423828490175, + "learning_rate": 5.253682643830564e-06, + "loss": 0.1435, + "step": 9553 + }, + { + "epoch": 2.0033550010484378, + "grad_norm": 3.1133456986312553, + "learning_rate": 5.2516901293945845e-06, + "loss": 0.1664, + "step": 9554 + }, + { + "epoch": 2.0035646886139653, + "grad_norm": 2.8309237238346334, + "learning_rate": 5.249697858317713e-06, + "loss": 0.1373, + "step": 9555 + }, + { + "epoch": 2.0037743761794924, + "grad_norm": 3.3596507953111403, + "learning_rate": 5.247705830702065e-06, + "loss": 0.1886, + "step": 9556 + }, + { + "epoch": 2.00398406374502, + "grad_norm": 2.9423103088570683, + "learning_rate": 5.2457140466497394e-06, + "loss": 0.159, + "step": 9557 + }, + { + "epoch": 2.0041937513105474, + "grad_norm": 3.5576378153921007, + "learning_rate": 5.243722506262815e-06, + "loss": 0.1744, + "step": 9558 + }, + { + "epoch": 2.0044034388760745, + "grad_norm": 3.5724011372558224, + "learning_rate": 5.241731209643357e-06, + "loss": 0.1909, + "step": 9559 + }, + { + "epoch": 2.004613126441602, + "grad_norm": 2.922771159607419, + "learning_rate": 5.239740156893432e-06, + "loss": 0.1276, + "step": 9560 + }, + { + "epoch": 2.0048228140071296, + "grad_norm": 4.05732032089272, + "learning_rate": 5.237749348115077e-06, + "loss": 0.2023, + "step": 9561 + }, + { + "epoch": 2.0050325015726567, + "grad_norm": 4.255949172512091, + "learning_rate": 5.23575878341033e-06, + "loss": 0.2401, + "step": 9562 + }, + { + "epoch": 2.005242189138184, + "grad_norm": 4.236095703928406, + "learning_rate": 5.233768462881209e-06, + "loss": 0.1943, + "step": 9563 + }, + { + "epoch": 2.0054518767037113, + "grad_norm": 3.7567133654819527, + "learning_rate": 5.231778386629718e-06, + "loss": 0.1779, + "step": 9564 + }, + { + "epoch": 2.005661564269239, + "grad_norm": 4.030923733611952, + "learning_rate": 5.229788554757853e-06, + "loss": 0.1835, + "step": 9565 + }, + { + "epoch": 2.0058712518347663, + "grad_norm": 4.397573243341452, + "learning_rate": 5.227798967367603e-06, + "loss": 0.2215, + "step": 9566 + }, + { + "epoch": 2.0060809394002934, + "grad_norm": 3.5191453731177518, + "learning_rate": 5.225809624560933e-06, + "loss": 0.1435, + "step": 9567 + }, + { + "epoch": 2.006290626965821, + "grad_norm": 3.239360549296363, + "learning_rate": 5.223820526439796e-06, + "loss": 0.1657, + "step": 9568 + }, + { + "epoch": 2.0065003145313485, + "grad_norm": 3.9533902462593975, + "learning_rate": 5.221831673106146e-06, + "loss": 0.1784, + "step": 9569 + }, + { + "epoch": 2.0067100020968756, + "grad_norm": 4.150861028379902, + "learning_rate": 5.2198430646619045e-06, + "loss": 0.2011, + "step": 9570 + }, + { + "epoch": 2.006919689662403, + "grad_norm": 2.589652690827136, + "learning_rate": 5.217854701209001e-06, + "loss": 0.1212, + "step": 9571 + }, + { + "epoch": 2.00712937722793, + "grad_norm": 2.9636329584249, + "learning_rate": 5.215866582849334e-06, + "loss": 0.1359, + "step": 9572 + }, + { + "epoch": 2.0073390647934577, + "grad_norm": 3.119259996002108, + "learning_rate": 5.2138787096848066e-06, + "loss": 0.1286, + "step": 9573 + }, + { + "epoch": 2.0075487523589852, + "grad_norm": 3.1734878688501724, + "learning_rate": 5.211891081817297e-06, + "loss": 0.1376, + "step": 9574 + }, + { + "epoch": 2.0077584399245123, + "grad_norm": 4.126856300646213, + "learning_rate": 5.209903699348668e-06, + "loss": 0.1677, + "step": 9575 + }, + { + "epoch": 2.00796812749004, + "grad_norm": 3.3645318805528084, + "learning_rate": 5.207916562380789e-06, + "loss": 0.1249, + "step": 9576 + }, + { + "epoch": 2.0081778150555674, + "grad_norm": 3.515844320237842, + "learning_rate": 5.205929671015492e-06, + "loss": 0.174, + "step": 9577 + }, + { + "epoch": 2.0083875026210944, + "grad_norm": 3.297452062501551, + "learning_rate": 5.20394302535462e-06, + "loss": 0.1168, + "step": 9578 + }, + { + "epoch": 2.008597190186622, + "grad_norm": 3.5917045677727297, + "learning_rate": 5.201956625499984e-06, + "loss": 0.1512, + "step": 9579 + }, + { + "epoch": 2.0088068777521495, + "grad_norm": 4.16183215891395, + "learning_rate": 5.199970471553391e-06, + "loss": 0.1607, + "step": 9580 + }, + { + "epoch": 2.0090165653176766, + "grad_norm": 3.5389679135331074, + "learning_rate": 5.197984563616635e-06, + "loss": 0.1506, + "step": 9581 + }, + { + "epoch": 2.009226252883204, + "grad_norm": 3.638484210935406, + "learning_rate": 5.195998901791503e-06, + "loss": 0.1664, + "step": 9582 + }, + { + "epoch": 2.009435940448731, + "grad_norm": 3.903128448172513, + "learning_rate": 5.19401348617976e-06, + "loss": 0.1511, + "step": 9583 + }, + { + "epoch": 2.0096456280142587, + "grad_norm": 4.753730943679167, + "learning_rate": 5.1920283168831574e-06, + "loss": 0.1572, + "step": 9584 + }, + { + "epoch": 2.0098553155797863, + "grad_norm": 6.516551821184562, + "learning_rate": 5.190043394003446e-06, + "loss": 0.2198, + "step": 9585 + }, + { + "epoch": 2.0100650031453133, + "grad_norm": 3.79207372958416, + "learning_rate": 5.188058717642348e-06, + "loss": 0.1632, + "step": 9586 + }, + { + "epoch": 2.010274690710841, + "grad_norm": 4.551769207676404, + "learning_rate": 5.186074287901593e-06, + "loss": 0.1667, + "step": 9587 + }, + { + "epoch": 2.0104843782763684, + "grad_norm": 3.657065323431733, + "learning_rate": 5.184090104882878e-06, + "loss": 0.1747, + "step": 9588 + }, + { + "epoch": 2.0106940658418955, + "grad_norm": 3.4201333464222854, + "learning_rate": 5.182106168687895e-06, + "loss": 0.1429, + "step": 9589 + }, + { + "epoch": 2.010903753407423, + "grad_norm": 4.005225802675114, + "learning_rate": 5.180122479418328e-06, + "loss": 0.1998, + "step": 9590 + }, + { + "epoch": 2.01111344097295, + "grad_norm": 3.2020629341771842, + "learning_rate": 5.17813903717585e-06, + "loss": 0.1196, + "step": 9591 + }, + { + "epoch": 2.0113231285384776, + "grad_norm": 4.431591184129306, + "learning_rate": 5.176155842062101e-06, + "loss": 0.1832, + "step": 9592 + }, + { + "epoch": 2.011532816104005, + "grad_norm": 3.2884369822231387, + "learning_rate": 5.174172894178733e-06, + "loss": 0.1653, + "step": 9593 + }, + { + "epoch": 2.0117425036695322, + "grad_norm": 4.546683063257566, + "learning_rate": 5.1721901936273775e-06, + "loss": 0.1459, + "step": 9594 + }, + { + "epoch": 2.0119521912350598, + "grad_norm": 3.9704349508650614, + "learning_rate": 5.1702077405096485e-06, + "loss": 0.1487, + "step": 9595 + }, + { + "epoch": 2.0121618788005873, + "grad_norm": 3.705160865880354, + "learning_rate": 5.168225534927145e-06, + "loss": 0.1258, + "step": 9596 + }, + { + "epoch": 2.0123715663661144, + "grad_norm": 4.843432841377177, + "learning_rate": 5.1662435769814675e-06, + "loss": 0.1806, + "step": 9597 + }, + { + "epoch": 2.012581253931642, + "grad_norm": 4.586769806280941, + "learning_rate": 5.1642618667741875e-06, + "loss": 0.1645, + "step": 9598 + }, + { + "epoch": 2.0127909414971694, + "grad_norm": 3.0805542999739552, + "learning_rate": 5.162280404406877e-06, + "loss": 0.1414, + "step": 9599 + }, + { + "epoch": 2.0130006290626965, + "grad_norm": 3.656694495761225, + "learning_rate": 5.160299189981086e-06, + "loss": 0.135, + "step": 9600 + }, + { + "epoch": 2.013210316628224, + "grad_norm": 3.827620723898737, + "learning_rate": 5.158318223598351e-06, + "loss": 0.1267, + "step": 9601 + }, + { + "epoch": 2.013420004193751, + "grad_norm": 4.397724670627419, + "learning_rate": 5.156337505360205e-06, + "loss": 0.1492, + "step": 9602 + }, + { + "epoch": 2.0136296917592786, + "grad_norm": 5.218183819209082, + "learning_rate": 5.154357035368167e-06, + "loss": 0.193, + "step": 9603 + }, + { + "epoch": 2.013839379324806, + "grad_norm": 4.0734098527295295, + "learning_rate": 5.152376813723736e-06, + "loss": 0.1749, + "step": 9604 + }, + { + "epoch": 2.0140490668903333, + "grad_norm": 3.7934669970046246, + "learning_rate": 5.150396840528396e-06, + "loss": 0.1183, + "step": 9605 + }, + { + "epoch": 2.014258754455861, + "grad_norm": 4.311910532620614, + "learning_rate": 5.148417115883634e-06, + "loss": 0.1438, + "step": 9606 + }, + { + "epoch": 2.0144684420213883, + "grad_norm": 3.6661508302291677, + "learning_rate": 5.146437639890904e-06, + "loss": 0.1415, + "step": 9607 + }, + { + "epoch": 2.0146781295869154, + "grad_norm": 4.233076823204496, + "learning_rate": 5.14445841265167e-06, + "loss": 0.1784, + "step": 9608 + }, + { + "epoch": 2.014887817152443, + "grad_norm": 4.024010762906287, + "learning_rate": 5.14247943426736e-06, + "loss": 0.1484, + "step": 9609 + }, + { + "epoch": 2.0150975047179704, + "grad_norm": 3.3112016937504576, + "learning_rate": 5.140500704839401e-06, + "loss": 0.1086, + "step": 9610 + }, + { + "epoch": 2.0153071922834975, + "grad_norm": 7.712004509953738, + "learning_rate": 5.1385222244692115e-06, + "loss": 0.1419, + "step": 9611 + }, + { + "epoch": 2.015516879849025, + "grad_norm": 4.648598272183988, + "learning_rate": 5.1365439932581965e-06, + "loss": 0.1805, + "step": 9612 + }, + { + "epoch": 2.015726567414552, + "grad_norm": 2.771395711399631, + "learning_rate": 5.134566011307729e-06, + "loss": 0.1147, + "step": 9613 + }, + { + "epoch": 2.0159362549800797, + "grad_norm": 3.6040552159791424, + "learning_rate": 5.132588278719193e-06, + "loss": 0.0988, + "step": 9614 + }, + { + "epoch": 2.016145942545607, + "grad_norm": 4.810532742168623, + "learning_rate": 5.130610795593952e-06, + "loss": 0.1422, + "step": 9615 + }, + { + "epoch": 2.0163556301111343, + "grad_norm": 4.10575577449865, + "learning_rate": 5.1286335620333535e-06, + "loss": 0.1617, + "step": 9616 + }, + { + "epoch": 2.016565317676662, + "grad_norm": 4.153823863048835, + "learning_rate": 5.1266565781387315e-06, + "loss": 0.152, + "step": 9617 + }, + { + "epoch": 2.0167750052421893, + "grad_norm": 2.791959776811991, + "learning_rate": 5.124679844011415e-06, + "loss": 0.0966, + "step": 9618 + }, + { + "epoch": 2.0169846928077164, + "grad_norm": 3.4449761606579012, + "learning_rate": 5.1227033597527075e-06, + "loss": 0.1271, + "step": 9619 + }, + { + "epoch": 2.017194380373244, + "grad_norm": 5.433131439449977, + "learning_rate": 5.120727125463918e-06, + "loss": 0.1284, + "step": 9620 + }, + { + "epoch": 2.017404067938771, + "grad_norm": 4.168742365151846, + "learning_rate": 5.118751141246325e-06, + "loss": 0.1673, + "step": 9621 + }, + { + "epoch": 2.0176137555042986, + "grad_norm": 3.901538213960578, + "learning_rate": 5.1167754072011976e-06, + "loss": 0.1569, + "step": 9622 + }, + { + "epoch": 2.017823443069826, + "grad_norm": 3.69980330390423, + "learning_rate": 5.114799923429802e-06, + "loss": 0.1466, + "step": 9623 + }, + { + "epoch": 2.018033130635353, + "grad_norm": 3.67441934422368, + "learning_rate": 5.112824690033386e-06, + "loss": 0.1707, + "step": 9624 + }, + { + "epoch": 2.0182428182008807, + "grad_norm": 4.336406741655112, + "learning_rate": 5.110849707113182e-06, + "loss": 0.167, + "step": 9625 + }, + { + "epoch": 2.0184525057664082, + "grad_norm": 4.513238140846731, + "learning_rate": 5.108874974770404e-06, + "loss": 0.1787, + "step": 9626 + }, + { + "epoch": 2.0186621933319353, + "grad_norm": 3.571342308267117, + "learning_rate": 5.106900493106272e-06, + "loss": 0.1363, + "step": 9627 + }, + { + "epoch": 2.018871880897463, + "grad_norm": 3.086493993330371, + "learning_rate": 5.1049262622219716e-06, + "loss": 0.1429, + "step": 9628 + }, + { + "epoch": 2.0190815684629904, + "grad_norm": 4.210058631020082, + "learning_rate": 5.102952282218696e-06, + "loss": 0.2052, + "step": 9629 + }, + { + "epoch": 2.0192912560285174, + "grad_norm": 3.602835000121995, + "learning_rate": 5.100978553197609e-06, + "loss": 0.133, + "step": 9630 + }, + { + "epoch": 2.019500943594045, + "grad_norm": 4.354146716021864, + "learning_rate": 5.099005075259864e-06, + "loss": 0.1449, + "step": 9631 + }, + { + "epoch": 2.019710631159572, + "grad_norm": 3.876350521511767, + "learning_rate": 5.097031848506609e-06, + "loss": 0.15, + "step": 9632 + }, + { + "epoch": 2.0199203187250996, + "grad_norm": 3.8037142031292004, + "learning_rate": 5.095058873038979e-06, + "loss": 0.1284, + "step": 9633 + }, + { + "epoch": 2.020130006290627, + "grad_norm": 3.859768656536526, + "learning_rate": 5.093086148958089e-06, + "loss": 0.1606, + "step": 9634 + }, + { + "epoch": 2.020339693856154, + "grad_norm": 5.196297389355179, + "learning_rate": 5.091113676365042e-06, + "loss": 0.2104, + "step": 9635 + }, + { + "epoch": 2.0205493814216817, + "grad_norm": 3.8644041731888263, + "learning_rate": 5.089141455360935e-06, + "loss": 0.1751, + "step": 9636 + }, + { + "epoch": 2.0207590689872093, + "grad_norm": 3.2820119277574427, + "learning_rate": 5.087169486046847e-06, + "loss": 0.152, + "step": 9637 + }, + { + "epoch": 2.0209687565527363, + "grad_norm": 4.014960921051778, + "learning_rate": 5.0851977685238395e-06, + "loss": 0.1781, + "step": 9638 + }, + { + "epoch": 2.021178444118264, + "grad_norm": 3.9988857477306294, + "learning_rate": 5.0832263028929716e-06, + "loss": 0.1352, + "step": 9639 + }, + { + "epoch": 2.021388131683791, + "grad_norm": 4.310035989869975, + "learning_rate": 5.0812550892552856e-06, + "loss": 0.1798, + "step": 9640 + }, + { + "epoch": 2.0215978192493185, + "grad_norm": 4.541028360738635, + "learning_rate": 5.079284127711809e-06, + "loss": 0.168, + "step": 9641 + }, + { + "epoch": 2.021807506814846, + "grad_norm": 3.925784069031799, + "learning_rate": 5.07731341836355e-06, + "loss": 0.1562, + "step": 9642 + }, + { + "epoch": 2.022017194380373, + "grad_norm": 3.8410072306637577, + "learning_rate": 5.075342961311522e-06, + "loss": 0.1566, + "step": 9643 + }, + { + "epoch": 2.0222268819459006, + "grad_norm": 3.7500665263891064, + "learning_rate": 5.073372756656702e-06, + "loss": 0.1776, + "step": 9644 + }, + { + "epoch": 2.022436569511428, + "grad_norm": 3.8304556143148574, + "learning_rate": 5.07140280450008e-06, + "loss": 0.1292, + "step": 9645 + }, + { + "epoch": 2.0226462570769552, + "grad_norm": 4.252951200630093, + "learning_rate": 5.069433104942611e-06, + "loss": 0.1411, + "step": 9646 + }, + { + "epoch": 2.0228559446424828, + "grad_norm": 5.062434457922006, + "learning_rate": 5.067463658085243e-06, + "loss": 0.1811, + "step": 9647 + }, + { + "epoch": 2.0230656322080103, + "grad_norm": 3.8351394665434957, + "learning_rate": 5.065494464028917e-06, + "loss": 0.135, + "step": 9648 + }, + { + "epoch": 2.0232753197735374, + "grad_norm": 3.965814489117527, + "learning_rate": 5.063525522874564e-06, + "loss": 0.1653, + "step": 9649 + }, + { + "epoch": 2.023485007339065, + "grad_norm": 3.928235644928943, + "learning_rate": 5.061556834723089e-06, + "loss": 0.175, + "step": 9650 + }, + { + "epoch": 2.023694694904592, + "grad_norm": 3.749121003522189, + "learning_rate": 5.0595883996753866e-06, + "loss": 0.148, + "step": 9651 + }, + { + "epoch": 2.0239043824701195, + "grad_norm": 3.441236476290837, + "learning_rate": 5.057620217832353e-06, + "loss": 0.1677, + "step": 9652 + }, + { + "epoch": 2.024114070035647, + "grad_norm": 3.6638049897352682, + "learning_rate": 5.0556522892948545e-06, + "loss": 0.1304, + "step": 9653 + }, + { + "epoch": 2.024323757601174, + "grad_norm": 3.124557263408565, + "learning_rate": 5.0536846141637474e-06, + "loss": 0.1193, + "step": 9654 + }, + { + "epoch": 2.0245334451667016, + "grad_norm": 3.654230789002932, + "learning_rate": 5.051717192539887e-06, + "loss": 0.1876, + "step": 9655 + }, + { + "epoch": 2.024743132732229, + "grad_norm": 4.848428850887715, + "learning_rate": 5.0497500245240986e-06, + "loss": 0.2077, + "step": 9656 + }, + { + "epoch": 2.0249528202977563, + "grad_norm": 3.0322540494652768, + "learning_rate": 5.0477831102172116e-06, + "loss": 0.1141, + "step": 9657 + }, + { + "epoch": 2.025162507863284, + "grad_norm": 3.7332693679355295, + "learning_rate": 5.045816449720029e-06, + "loss": 0.1523, + "step": 9658 + }, + { + "epoch": 2.025372195428811, + "grad_norm": 4.0242393627446065, + "learning_rate": 5.043850043133343e-06, + "loss": 0.1354, + "step": 9659 + }, + { + "epoch": 2.0255818829943384, + "grad_norm": 4.281777340609352, + "learning_rate": 5.041883890557937e-06, + "loss": 0.1314, + "step": 9660 + }, + { + "epoch": 2.025791570559866, + "grad_norm": 4.322734741048664, + "learning_rate": 5.0399179920945855e-06, + "loss": 0.1852, + "step": 9661 + }, + { + "epoch": 2.026001258125393, + "grad_norm": 4.100995785980219, + "learning_rate": 5.0379523478440394e-06, + "loss": 0.1615, + "step": 9662 + }, + { + "epoch": 2.0262109456909205, + "grad_norm": 3.5853989549038685, + "learning_rate": 5.035986957907039e-06, + "loss": 0.154, + "step": 9663 + }, + { + "epoch": 2.026420633256448, + "grad_norm": 4.473961295130404, + "learning_rate": 5.0340218223843185e-06, + "loss": 0.1587, + "step": 9664 + }, + { + "epoch": 2.026630320821975, + "grad_norm": 3.417181978681296, + "learning_rate": 5.032056941376591e-06, + "loss": 0.1554, + "step": 9665 + }, + { + "epoch": 2.0268400083875027, + "grad_norm": 3.616791410469849, + "learning_rate": 5.030092314984565e-06, + "loss": 0.1261, + "step": 9666 + }, + { + "epoch": 2.02704969595303, + "grad_norm": 3.65784884265805, + "learning_rate": 5.028127943308927e-06, + "loss": 0.1548, + "step": 9667 + }, + { + "epoch": 2.0272593835185573, + "grad_norm": 4.784307433963424, + "learning_rate": 5.026163826450351e-06, + "loss": 0.1979, + "step": 9668 + }, + { + "epoch": 2.027469071084085, + "grad_norm": 4.591937761993029, + "learning_rate": 5.024199964509506e-06, + "loss": 0.1376, + "step": 9669 + }, + { + "epoch": 2.027678758649612, + "grad_norm": 3.882193314038145, + "learning_rate": 5.022236357587051e-06, + "loss": 0.14, + "step": 9670 + }, + { + "epoch": 2.0278884462151394, + "grad_norm": 4.840160751948868, + "learning_rate": 5.020273005783609e-06, + "loss": 0.1118, + "step": 9671 + }, + { + "epoch": 2.028098133780667, + "grad_norm": 4.450026740831961, + "learning_rate": 5.018309909199811e-06, + "loss": 0.1506, + "step": 9672 + }, + { + "epoch": 2.028307821346194, + "grad_norm": 3.6572217134003013, + "learning_rate": 5.016347067936275e-06, + "loss": 0.1395, + "step": 9673 + }, + { + "epoch": 2.0285175089117216, + "grad_norm": 3.752521424318331, + "learning_rate": 5.014384482093595e-06, + "loss": 0.1166, + "step": 9674 + }, + { + "epoch": 2.028727196477249, + "grad_norm": 5.949694644663068, + "learning_rate": 5.0124221517723535e-06, + "loss": 0.2236, + "step": 9675 + }, + { + "epoch": 2.028936884042776, + "grad_norm": 4.317470239454295, + "learning_rate": 5.010460077073132e-06, + "loss": 0.1875, + "step": 9676 + }, + { + "epoch": 2.0291465716083037, + "grad_norm": 3.51429979514136, + "learning_rate": 5.008498258096481e-06, + "loss": 0.1248, + "step": 9677 + }, + { + "epoch": 2.029356259173831, + "grad_norm": 4.995472775571437, + "learning_rate": 5.006536694942956e-06, + "loss": 0.1787, + "step": 9678 + }, + { + "epoch": 2.0295659467393583, + "grad_norm": 3.3343069323315184, + "learning_rate": 5.004575387713087e-06, + "loss": 0.1603, + "step": 9679 + }, + { + "epoch": 2.029775634304886, + "grad_norm": 4.934297369703559, + "learning_rate": 5.002614336507388e-06, + "loss": 0.1622, + "step": 9680 + }, + { + "epoch": 2.029985321870413, + "grad_norm": 4.564986126409634, + "learning_rate": 5.000653541426371e-06, + "loss": 0.1701, + "step": 9681 + }, + { + "epoch": 2.0301950094359404, + "grad_norm": 4.4370152060040695, + "learning_rate": 4.998693002570536e-06, + "loss": 0.118, + "step": 9682 + }, + { + "epoch": 2.030404697001468, + "grad_norm": 3.8685425209905264, + "learning_rate": 4.9967327200403595e-06, + "loss": 0.1618, + "step": 9683 + }, + { + "epoch": 2.030614384566995, + "grad_norm": 5.61159471794053, + "learning_rate": 4.994772693936304e-06, + "loss": 0.1696, + "step": 9684 + }, + { + "epoch": 2.0308240721325226, + "grad_norm": 3.67717618411063, + "learning_rate": 4.992812924358833e-06, + "loss": 0.1199, + "step": 9685 + }, + { + "epoch": 2.03103375969805, + "grad_norm": 3.057613163850684, + "learning_rate": 4.9908534114083804e-06, + "loss": 0.1133, + "step": 9686 + }, + { + "epoch": 2.031243447263577, + "grad_norm": 5.112729928585523, + "learning_rate": 4.988894155185383e-06, + "loss": 0.1617, + "step": 9687 + }, + { + "epoch": 2.0314531348291047, + "grad_norm": 4.454841502439337, + "learning_rate": 4.98693515579025e-06, + "loss": 0.1599, + "step": 9688 + }, + { + "epoch": 2.031662822394632, + "grad_norm": 4.702200159440042, + "learning_rate": 4.984976413323382e-06, + "loss": 0.1726, + "step": 9689 + }, + { + "epoch": 2.0318725099601593, + "grad_norm": 4.117285605362018, + "learning_rate": 4.983017927885172e-06, + "loss": 0.1379, + "step": 9690 + }, + { + "epoch": 2.032082197525687, + "grad_norm": 4.460421659721673, + "learning_rate": 4.981059699576001e-06, + "loss": 0.1443, + "step": 9691 + }, + { + "epoch": 2.032291885091214, + "grad_norm": 3.7528911494793444, + "learning_rate": 4.979101728496217e-06, + "loss": 0.1447, + "step": 9692 + }, + { + "epoch": 2.0325015726567415, + "grad_norm": 4.655677877870464, + "learning_rate": 4.977144014746179e-06, + "loss": 0.1544, + "step": 9693 + }, + { + "epoch": 2.032711260222269, + "grad_norm": 4.931239573723924, + "learning_rate": 4.975186558426222e-06, + "loss": 0.1482, + "step": 9694 + }, + { + "epoch": 2.032920947787796, + "grad_norm": 4.157377861062008, + "learning_rate": 4.973229359636677e-06, + "loss": 0.1268, + "step": 9695 + }, + { + "epoch": 2.0331306353533236, + "grad_norm": 4.061055616565977, + "learning_rate": 4.971272418477837e-06, + "loss": 0.1322, + "step": 9696 + }, + { + "epoch": 2.0333403229188507, + "grad_norm": 4.7175996389496255, + "learning_rate": 4.969315735050008e-06, + "loss": 0.2018, + "step": 9697 + }, + { + "epoch": 2.0335500104843782, + "grad_norm": 4.254970893928762, + "learning_rate": 4.967359309453477e-06, + "loss": 0.1315, + "step": 9698 + }, + { + "epoch": 2.0337596980499058, + "grad_norm": 3.999630750226817, + "learning_rate": 4.965403141788508e-06, + "loss": 0.1164, + "step": 9699 + }, + { + "epoch": 2.033969385615433, + "grad_norm": 6.287055805916647, + "learning_rate": 4.9634472321553586e-06, + "loss": 0.2057, + "step": 9700 + }, + { + "epoch": 2.0341790731809604, + "grad_norm": 4.379526313468788, + "learning_rate": 4.961491580654277e-06, + "loss": 0.1865, + "step": 9701 + }, + { + "epoch": 2.034388760746488, + "grad_norm": 4.804854996308676, + "learning_rate": 4.959536187385488e-06, + "loss": 0.1581, + "step": 9702 + }, + { + "epoch": 2.034598448312015, + "grad_norm": 3.7674044495623313, + "learning_rate": 4.957581052449213e-06, + "loss": 0.1376, + "step": 9703 + }, + { + "epoch": 2.0348081358775425, + "grad_norm": 4.544457370068342, + "learning_rate": 4.955626175945657e-06, + "loss": 0.134, + "step": 9704 + }, + { + "epoch": 2.03501782344307, + "grad_norm": 3.871084734591505, + "learning_rate": 4.953671557975004e-06, + "loss": 0.1259, + "step": 9705 + }, + { + "epoch": 2.035227511008597, + "grad_norm": 3.6286614216617328, + "learning_rate": 4.951717198637436e-06, + "loss": 0.1278, + "step": 9706 + }, + { + "epoch": 2.0354371985741246, + "grad_norm": 3.7150033181097735, + "learning_rate": 4.94976309803312e-06, + "loss": 0.1277, + "step": 9707 + }, + { + "epoch": 2.0356468861396517, + "grad_norm": 5.25235936780869, + "learning_rate": 4.947809256262206e-06, + "loss": 0.1781, + "step": 9708 + }, + { + "epoch": 2.0358565737051793, + "grad_norm": 3.8211949856235448, + "learning_rate": 4.945855673424825e-06, + "loss": 0.1342, + "step": 9709 + }, + { + "epoch": 2.036066261270707, + "grad_norm": 6.102145700753543, + "learning_rate": 4.9439023496211125e-06, + "loss": 0.195, + "step": 9710 + }, + { + "epoch": 2.036275948836234, + "grad_norm": 4.857021782272797, + "learning_rate": 4.9419492849511686e-06, + "loss": 0.2308, + "step": 9711 + }, + { + "epoch": 2.0364856364017614, + "grad_norm": 4.775266027482804, + "learning_rate": 4.939996479515101e-06, + "loss": 0.1639, + "step": 9712 + }, + { + "epoch": 2.036695323967289, + "grad_norm": 4.080888562450899, + "learning_rate": 4.9380439334129896e-06, + "loss": 0.1645, + "step": 9713 + }, + { + "epoch": 2.036905011532816, + "grad_norm": 4.343808832816453, + "learning_rate": 4.936091646744902e-06, + "loss": 0.1431, + "step": 9714 + }, + { + "epoch": 2.0371146990983435, + "grad_norm": 3.536102113061114, + "learning_rate": 4.9341396196109045e-06, + "loss": 0.1301, + "step": 9715 + }, + { + "epoch": 2.0373243866638706, + "grad_norm": 5.193094418937569, + "learning_rate": 4.932187852111039e-06, + "loss": 0.1531, + "step": 9716 + }, + { + "epoch": 2.037534074229398, + "grad_norm": 4.895702018534499, + "learning_rate": 4.930236344345331e-06, + "loss": 0.1369, + "step": 9717 + }, + { + "epoch": 2.0377437617949257, + "grad_norm": 3.423019109997735, + "learning_rate": 4.928285096413803e-06, + "loss": 0.1192, + "step": 9718 + }, + { + "epoch": 2.0379534493604528, + "grad_norm": 4.230739263443211, + "learning_rate": 4.926334108416464e-06, + "loss": 0.1392, + "step": 9719 + }, + { + "epoch": 2.0381631369259803, + "grad_norm": 4.696825001644658, + "learning_rate": 4.9243833804533025e-06, + "loss": 0.1739, + "step": 9720 + }, + { + "epoch": 2.038372824491508, + "grad_norm": 4.588169001557765, + "learning_rate": 4.922432912624292e-06, + "loss": 0.1538, + "step": 9721 + }, + { + "epoch": 2.038582512057035, + "grad_norm": 3.8433952645173965, + "learning_rate": 4.920482705029405e-06, + "loss": 0.1445, + "step": 9722 + }, + { + "epoch": 2.0387921996225624, + "grad_norm": 5.986940461029213, + "learning_rate": 4.9185327577685856e-06, + "loss": 0.1726, + "step": 9723 + }, + { + "epoch": 2.03900188718809, + "grad_norm": 4.715264503870053, + "learning_rate": 4.91658307094178e-06, + "loss": 0.1465, + "step": 9724 + }, + { + "epoch": 2.039211574753617, + "grad_norm": 4.957169834451293, + "learning_rate": 4.914633644648908e-06, + "loss": 0.1826, + "step": 9725 + }, + { + "epoch": 2.0394212623191446, + "grad_norm": 3.5697092278553204, + "learning_rate": 4.912684478989878e-06, + "loss": 0.1258, + "step": 9726 + }, + { + "epoch": 2.0396309498846716, + "grad_norm": 4.2516744340078185, + "learning_rate": 4.910735574064592e-06, + "loss": 0.1435, + "step": 9727 + }, + { + "epoch": 2.039840637450199, + "grad_norm": 3.5410851949368687, + "learning_rate": 4.908786929972938e-06, + "loss": 0.1322, + "step": 9728 + }, + { + "epoch": 2.0400503250157267, + "grad_norm": 4.552925139619608, + "learning_rate": 4.906838546814785e-06, + "loss": 0.2012, + "step": 9729 + }, + { + "epoch": 2.040260012581254, + "grad_norm": 4.005678051532541, + "learning_rate": 4.9048904246899865e-06, + "loss": 0.1157, + "step": 9730 + }, + { + "epoch": 2.0404697001467813, + "grad_norm": 4.651973172814157, + "learning_rate": 4.902942563698394e-06, + "loss": 0.1711, + "step": 9731 + }, + { + "epoch": 2.040679387712309, + "grad_norm": 4.7361127375757315, + "learning_rate": 4.900994963939835e-06, + "loss": 0.1604, + "step": 9732 + }, + { + "epoch": 2.040889075277836, + "grad_norm": 4.101375486560881, + "learning_rate": 4.899047625514125e-06, + "loss": 0.1507, + "step": 9733 + }, + { + "epoch": 2.0410987628433634, + "grad_norm": 3.4864299214351524, + "learning_rate": 4.897100548521075e-06, + "loss": 0.1285, + "step": 9734 + }, + { + "epoch": 2.0413084504088905, + "grad_norm": 4.828958347326046, + "learning_rate": 4.89515373306047e-06, + "loss": 0.2053, + "step": 9735 + }, + { + "epoch": 2.041518137974418, + "grad_norm": 5.321312523641549, + "learning_rate": 4.893207179232091e-06, + "loss": 0.1776, + "step": 9736 + }, + { + "epoch": 2.0417278255399456, + "grad_norm": 5.064349778548283, + "learning_rate": 4.8912608871357025e-06, + "loss": 0.198, + "step": 9737 + }, + { + "epoch": 2.0419375131054727, + "grad_norm": 3.549737635953889, + "learning_rate": 4.889314856871051e-06, + "loss": 0.1566, + "step": 9738 + }, + { + "epoch": 2.042147200671, + "grad_norm": 3.391221718259378, + "learning_rate": 4.887369088537877e-06, + "loss": 0.1202, + "step": 9739 + }, + { + "epoch": 2.0423568882365277, + "grad_norm": 4.126576377111557, + "learning_rate": 4.885423582235909e-06, + "loss": 0.1524, + "step": 9740 + }, + { + "epoch": 2.042566575802055, + "grad_norm": 3.7244952351384835, + "learning_rate": 4.883478338064851e-06, + "loss": 0.1657, + "step": 9741 + }, + { + "epoch": 2.0427762633675823, + "grad_norm": 3.8013591666363133, + "learning_rate": 4.8815333561244e-06, + "loss": 0.114, + "step": 9742 + }, + { + "epoch": 2.04298595093311, + "grad_norm": 4.469455178598115, + "learning_rate": 4.879588636514246e-06, + "loss": 0.1448, + "step": 9743 + }, + { + "epoch": 2.043195638498637, + "grad_norm": 4.279595719914409, + "learning_rate": 4.877644179334051e-06, + "loss": 0.1781, + "step": 9744 + }, + { + "epoch": 2.0434053260641645, + "grad_norm": 4.290524092660325, + "learning_rate": 4.875699984683481e-06, + "loss": 0.1569, + "step": 9745 + }, + { + "epoch": 2.0436150136296916, + "grad_norm": 4.721326246909336, + "learning_rate": 4.873756052662173e-06, + "loss": 0.154, + "step": 9746 + }, + { + "epoch": 2.043824701195219, + "grad_norm": 4.140175251669526, + "learning_rate": 4.871812383369755e-06, + "loss": 0.1556, + "step": 9747 + }, + { + "epoch": 2.0440343887607466, + "grad_norm": 5.5721019646631005, + "learning_rate": 4.869868976905847e-06, + "loss": 0.187, + "step": 9748 + }, + { + "epoch": 2.0442440763262737, + "grad_norm": 3.110542674309111, + "learning_rate": 4.867925833370058e-06, + "loss": 0.1296, + "step": 9749 + }, + { + "epoch": 2.0444537638918012, + "grad_norm": 3.6780221504132427, + "learning_rate": 4.865982952861965e-06, + "loss": 0.1284, + "step": 9750 + }, + { + "epoch": 2.0446634514573288, + "grad_norm": 3.241275803154536, + "learning_rate": 4.86404033548115e-06, + "loss": 0.118, + "step": 9751 + }, + { + "epoch": 2.044873139022856, + "grad_norm": 3.5889762481026013, + "learning_rate": 4.86209798132718e-06, + "loss": 0.1745, + "step": 9752 + }, + { + "epoch": 2.0450828265883834, + "grad_norm": 5.162952073405409, + "learning_rate": 4.860155890499598e-06, + "loss": 0.148, + "step": 9753 + }, + { + "epoch": 2.045292514153911, + "grad_norm": 5.159580770033513, + "learning_rate": 4.858214063097939e-06, + "loss": 0.1767, + "step": 9754 + }, + { + "epoch": 2.045502201719438, + "grad_norm": 5.295814069975944, + "learning_rate": 4.856272499221727e-06, + "loss": 0.2044, + "step": 9755 + }, + { + "epoch": 2.0457118892849655, + "grad_norm": 3.7763549369453497, + "learning_rate": 4.854331198970474e-06, + "loss": 0.1236, + "step": 9756 + }, + { + "epoch": 2.0459215768504926, + "grad_norm": 4.764330352553426, + "learning_rate": 4.852390162443672e-06, + "loss": 0.1438, + "step": 9757 + }, + { + "epoch": 2.04613126441602, + "grad_norm": 4.303369723345257, + "learning_rate": 4.850449389740798e-06, + "loss": 0.1714, + "step": 9758 + }, + { + "epoch": 2.0463409519815476, + "grad_norm": 3.6617425814999245, + "learning_rate": 4.848508880961329e-06, + "loss": 0.1238, + "step": 9759 + }, + { + "epoch": 2.0465506395470747, + "grad_norm": 3.7887024712096538, + "learning_rate": 4.8465686362047105e-06, + "loss": 0.1401, + "step": 9760 + }, + { + "epoch": 2.0467603271126023, + "grad_norm": 5.122948427632381, + "learning_rate": 4.84462865557039e-06, + "loss": 0.1731, + "step": 9761 + }, + { + "epoch": 2.04697001467813, + "grad_norm": 5.1304353174633786, + "learning_rate": 4.842688939157795e-06, + "loss": 0.1949, + "step": 9762 + }, + { + "epoch": 2.047179702243657, + "grad_norm": 4.8860890133668775, + "learning_rate": 4.840749487066331e-06, + "loss": 0.156, + "step": 9763 + }, + { + "epoch": 2.0473893898091844, + "grad_norm": 3.5049700590622486, + "learning_rate": 4.838810299395405e-06, + "loss": 0.1161, + "step": 9764 + }, + { + "epoch": 2.0475990773747115, + "grad_norm": 5.596071627268399, + "learning_rate": 4.836871376244405e-06, + "loss": 0.2077, + "step": 9765 + }, + { + "epoch": 2.047808764940239, + "grad_norm": 4.456633883171991, + "learning_rate": 4.834932717712704e-06, + "loss": 0.1598, + "step": 9766 + }, + { + "epoch": 2.0480184525057665, + "grad_norm": 4.957496731684181, + "learning_rate": 4.832994323899655e-06, + "loss": 0.1734, + "step": 9767 + }, + { + "epoch": 2.0482281400712936, + "grad_norm": 4.111405757886916, + "learning_rate": 4.831056194904612e-06, + "loss": 0.1697, + "step": 9768 + }, + { + "epoch": 2.048437827636821, + "grad_norm": 4.466436643882388, + "learning_rate": 4.829118330826901e-06, + "loss": 0.123, + "step": 9769 + }, + { + "epoch": 2.0486475152023487, + "grad_norm": 3.5683220093800783, + "learning_rate": 4.827180731765847e-06, + "loss": 0.1398, + "step": 9770 + }, + { + "epoch": 2.0488572027678758, + "grad_norm": 4.294467718239291, + "learning_rate": 4.8252433978207535e-06, + "loss": 0.1477, + "step": 9771 + }, + { + "epoch": 2.0490668903334033, + "grad_norm": 4.339898188495926, + "learning_rate": 4.823306329090906e-06, + "loss": 0.1257, + "step": 9772 + }, + { + "epoch": 2.049276577898931, + "grad_norm": 4.910078758039469, + "learning_rate": 4.821369525675588e-06, + "loss": 0.1577, + "step": 9773 + }, + { + "epoch": 2.049486265464458, + "grad_norm": 4.550252070991739, + "learning_rate": 4.819432987674072e-06, + "loss": 0.165, + "step": 9774 + }, + { + "epoch": 2.0496959530299854, + "grad_norm": 4.217995165035445, + "learning_rate": 4.817496715185593e-06, + "loss": 0.1548, + "step": 9775 + }, + { + "epoch": 2.0499056405955125, + "grad_norm": 3.617185027114443, + "learning_rate": 4.815560708309394e-06, + "loss": 0.1528, + "step": 9776 + }, + { + "epoch": 2.05011532816104, + "grad_norm": 4.196934124657236, + "learning_rate": 4.813624967144707e-06, + "loss": 0.1628, + "step": 9777 + }, + { + "epoch": 2.0503250157265676, + "grad_norm": 3.9782240726376563, + "learning_rate": 4.811689491790734e-06, + "loss": 0.1388, + "step": 9778 + }, + { + "epoch": 2.0505347032920946, + "grad_norm": 3.9864284535697143, + "learning_rate": 4.8097542823466695e-06, + "loss": 0.1419, + "step": 9779 + }, + { + "epoch": 2.050744390857622, + "grad_norm": 3.90096987191959, + "learning_rate": 4.807819338911704e-06, + "loss": 0.1499, + "step": 9780 + }, + { + "epoch": 2.0509540784231497, + "grad_norm": 4.889395632797826, + "learning_rate": 4.805884661585e-06, + "loss": 0.1528, + "step": 9781 + }, + { + "epoch": 2.051163765988677, + "grad_norm": 3.903589573525906, + "learning_rate": 4.803950250465719e-06, + "loss": 0.1374, + "step": 9782 + }, + { + "epoch": 2.0513734535542043, + "grad_norm": 5.266024356669548, + "learning_rate": 4.802016105652999e-06, + "loss": 0.1686, + "step": 9783 + }, + { + "epoch": 2.0515831411197314, + "grad_norm": 4.2575823995114925, + "learning_rate": 4.8000822272459665e-06, + "loss": 0.1347, + "step": 9784 + }, + { + "epoch": 2.051792828685259, + "grad_norm": 4.789495569809839, + "learning_rate": 4.7981486153437375e-06, + "loss": 0.1468, + "step": 9785 + }, + { + "epoch": 2.0520025162507864, + "grad_norm": 3.881873722552294, + "learning_rate": 4.796215270045419e-06, + "loss": 0.1699, + "step": 9786 + }, + { + "epoch": 2.0522122038163135, + "grad_norm": 3.4816791161654623, + "learning_rate": 4.7942821914500924e-06, + "loss": 0.122, + "step": 9787 + }, + { + "epoch": 2.052421891381841, + "grad_norm": 5.230403549857729, + "learning_rate": 4.7923493796568295e-06, + "loss": 0.1586, + "step": 9788 + }, + { + "epoch": 2.0526315789473686, + "grad_norm": 3.509921936939052, + "learning_rate": 4.790416834764696e-06, + "loss": 0.125, + "step": 9789 + }, + { + "epoch": 2.0528412665128957, + "grad_norm": 4.613713760192309, + "learning_rate": 4.788484556872732e-06, + "loss": 0.1282, + "step": 9790 + }, + { + "epoch": 2.053050954078423, + "grad_norm": 4.474865594477442, + "learning_rate": 4.786552546079978e-06, + "loss": 0.1604, + "step": 9791 + }, + { + "epoch": 2.0532606416439507, + "grad_norm": 4.297074097442312, + "learning_rate": 4.784620802485447e-06, + "loss": 0.1671, + "step": 9792 + }, + { + "epoch": 2.053470329209478, + "grad_norm": 3.974078148528853, + "learning_rate": 4.7826893261881425e-06, + "loss": 0.1696, + "step": 9793 + }, + { + "epoch": 2.0536800167750053, + "grad_norm": 4.259664670862086, + "learning_rate": 4.7807581172870625e-06, + "loss": 0.1467, + "step": 9794 + }, + { + "epoch": 2.0538897043405324, + "grad_norm": 4.8450665149936825, + "learning_rate": 4.77882717588118e-06, + "loss": 0.1589, + "step": 9795 + }, + { + "epoch": 2.05409939190606, + "grad_norm": 4.729706117036031, + "learning_rate": 4.776896502069458e-06, + "loss": 0.1701, + "step": 9796 + }, + { + "epoch": 2.0543090794715875, + "grad_norm": 4.402136589844036, + "learning_rate": 4.774966095950849e-06, + "loss": 0.1163, + "step": 9797 + }, + { + "epoch": 2.0545187670371146, + "grad_norm": 4.311885643127705, + "learning_rate": 4.773035957624294e-06, + "loss": 0.1338, + "step": 9798 + }, + { + "epoch": 2.054728454602642, + "grad_norm": 4.553215427546349, + "learning_rate": 4.771106087188712e-06, + "loss": 0.1582, + "step": 9799 + }, + { + "epoch": 2.0549381421681696, + "grad_norm": 3.9721270021601525, + "learning_rate": 4.769176484743008e-06, + "loss": 0.1405, + "step": 9800 + }, + { + "epoch": 2.0551478297336967, + "grad_norm": 3.4334814984468074, + "learning_rate": 4.767247150386085e-06, + "loss": 0.1412, + "step": 9801 + }, + { + "epoch": 2.0553575172992242, + "grad_norm": 5.467359886772754, + "learning_rate": 4.7653180842168174e-06, + "loss": 0.1861, + "step": 9802 + }, + { + "epoch": 2.0555672048647513, + "grad_norm": 3.3055167609799705, + "learning_rate": 4.763389286334081e-06, + "loss": 0.1199, + "step": 9803 + }, + { + "epoch": 2.055776892430279, + "grad_norm": 5.645059321460999, + "learning_rate": 4.7614607568367275e-06, + "loss": 0.1718, + "step": 9804 + }, + { + "epoch": 2.0559865799958064, + "grad_norm": 3.8038382911663797, + "learning_rate": 4.759532495823591e-06, + "loss": 0.1377, + "step": 9805 + }, + { + "epoch": 2.0561962675613334, + "grad_norm": 3.794227323765912, + "learning_rate": 4.7576045033935035e-06, + "loss": 0.1236, + "step": 9806 + }, + { + "epoch": 2.056405955126861, + "grad_norm": 4.274609286417137, + "learning_rate": 4.7556767796452805e-06, + "loss": 0.1293, + "step": 9807 + }, + { + "epoch": 2.0566156426923885, + "grad_norm": 3.891406885181315, + "learning_rate": 4.75374932467772e-06, + "loss": 0.1597, + "step": 9808 + }, + { + "epoch": 2.0568253302579156, + "grad_norm": 4.425901164713951, + "learning_rate": 4.7518221385896014e-06, + "loss": 0.1799, + "step": 9809 + }, + { + "epoch": 2.057035017823443, + "grad_norm": 4.3101988235130415, + "learning_rate": 4.749895221479703e-06, + "loss": 0.1557, + "step": 9810 + }, + { + "epoch": 2.0572447053889706, + "grad_norm": 4.509880468806988, + "learning_rate": 4.747968573446783e-06, + "loss": 0.116, + "step": 9811 + }, + { + "epoch": 2.0574543929544977, + "grad_norm": 3.862237420804464, + "learning_rate": 4.746042194589578e-06, + "loss": 0.16, + "step": 9812 + }, + { + "epoch": 2.0576640805200253, + "grad_norm": 3.3057082533171704, + "learning_rate": 4.744116085006826e-06, + "loss": 0.1321, + "step": 9813 + }, + { + "epoch": 2.0578737680855523, + "grad_norm": 4.291659391518171, + "learning_rate": 4.742190244797237e-06, + "loss": 0.1892, + "step": 9814 + }, + { + "epoch": 2.05808345565108, + "grad_norm": 4.935304095135732, + "learning_rate": 4.740264674059521e-06, + "loss": 0.1756, + "step": 9815 + }, + { + "epoch": 2.0582931432166074, + "grad_norm": 3.7112552541916908, + "learning_rate": 4.738339372892359e-06, + "loss": 0.1678, + "step": 9816 + }, + { + "epoch": 2.0585028307821345, + "grad_norm": 3.398814387291517, + "learning_rate": 4.736414341394434e-06, + "loss": 0.1435, + "step": 9817 + }, + { + "epoch": 2.058712518347662, + "grad_norm": 5.072759137732459, + "learning_rate": 4.7344895796644e-06, + "loss": 0.1513, + "step": 9818 + }, + { + "epoch": 2.0589222059131895, + "grad_norm": 3.9355020514602628, + "learning_rate": 4.7325650878009115e-06, + "loss": 0.1688, + "step": 9819 + }, + { + "epoch": 2.0591318934787166, + "grad_norm": 4.388975638193464, + "learning_rate": 4.730640865902597e-06, + "loss": 0.1661, + "step": 9820 + }, + { + "epoch": 2.059341581044244, + "grad_norm": 5.006149333268862, + "learning_rate": 4.728716914068075e-06, + "loss": 0.1719, + "step": 9821 + }, + { + "epoch": 2.0595512686097712, + "grad_norm": 5.162142712719607, + "learning_rate": 4.726793232395952e-06, + "loss": 0.1624, + "step": 9822 + }, + { + "epoch": 2.0597609561752988, + "grad_norm": 4.5253183458291275, + "learning_rate": 4.724869820984826e-06, + "loss": 0.1908, + "step": 9823 + }, + { + "epoch": 2.0599706437408263, + "grad_norm": 3.5961990092210754, + "learning_rate": 4.7229466799332725e-06, + "loss": 0.1499, + "step": 9824 + }, + { + "epoch": 2.0601803313063534, + "grad_norm": 3.161093635157218, + "learning_rate": 4.721023809339849e-06, + "loss": 0.1113, + "step": 9825 + }, + { + "epoch": 2.060390018871881, + "grad_norm": 4.33355494499541, + "learning_rate": 4.719101209303115e-06, + "loss": 0.1466, + "step": 9826 + }, + { + "epoch": 2.0605997064374084, + "grad_norm": 4.535528047199118, + "learning_rate": 4.7171788799216e-06, + "loss": 0.1553, + "step": 9827 + }, + { + "epoch": 2.0608093940029355, + "grad_norm": 4.317305053367913, + "learning_rate": 4.7152568212938336e-06, + "loss": 0.1803, + "step": 9828 + }, + { + "epoch": 2.061019081568463, + "grad_norm": 3.5967097815860796, + "learning_rate": 4.713335033518321e-06, + "loss": 0.1493, + "step": 9829 + }, + { + "epoch": 2.0612287691339906, + "grad_norm": 5.5806160733606776, + "learning_rate": 4.711413516693554e-06, + "loss": 0.2129, + "step": 9830 + }, + { + "epoch": 2.0614384566995176, + "grad_norm": 3.1217119338915467, + "learning_rate": 4.709492270918016e-06, + "loss": 0.12, + "step": 9831 + }, + { + "epoch": 2.061648144265045, + "grad_norm": 4.184075015744736, + "learning_rate": 4.707571296290182e-06, + "loss": 0.1701, + "step": 9832 + }, + { + "epoch": 2.0618578318305723, + "grad_norm": 4.655865407348202, + "learning_rate": 4.705650592908492e-06, + "loss": 0.1619, + "step": 9833 + }, + { + "epoch": 2.0620675193961, + "grad_norm": 3.4016707751298623, + "learning_rate": 4.703730160871392e-06, + "loss": 0.1484, + "step": 9834 + }, + { + "epoch": 2.0622772069616273, + "grad_norm": 4.1713595286173835, + "learning_rate": 4.70181000027731e-06, + "loss": 0.1523, + "step": 9835 + }, + { + "epoch": 2.0624868945271544, + "grad_norm": 4.062228782362276, + "learning_rate": 4.699890111224655e-06, + "loss": 0.1446, + "step": 9836 + }, + { + "epoch": 2.062696582092682, + "grad_norm": 4.173052503513639, + "learning_rate": 4.697970493811821e-06, + "loss": 0.1736, + "step": 9837 + }, + { + "epoch": 2.0629062696582094, + "grad_norm": 4.410452885530461, + "learning_rate": 4.696051148137198e-06, + "loss": 0.1321, + "step": 9838 + }, + { + "epoch": 2.0631159572237365, + "grad_norm": 6.680127587795301, + "learning_rate": 4.694132074299149e-06, + "loss": 0.1961, + "step": 9839 + }, + { + "epoch": 2.063325644789264, + "grad_norm": 3.6277563751302626, + "learning_rate": 4.692213272396039e-06, + "loss": 0.1517, + "step": 9840 + }, + { + "epoch": 2.0635353323547916, + "grad_norm": 4.977892836811028, + "learning_rate": 4.6902947425262035e-06, + "loss": 0.1358, + "step": 9841 + }, + { + "epoch": 2.0637450199203187, + "grad_norm": 3.4455307834054785, + "learning_rate": 4.688376484787969e-06, + "loss": 0.1167, + "step": 9842 + }, + { + "epoch": 2.063954707485846, + "grad_norm": 5.285245420263358, + "learning_rate": 4.68645849927965e-06, + "loss": 0.1755, + "step": 9843 + }, + { + "epoch": 2.0641643950513733, + "grad_norm": 3.9498095787337597, + "learning_rate": 4.684540786099555e-06, + "loss": 0.169, + "step": 9844 + }, + { + "epoch": 2.064374082616901, + "grad_norm": 4.306700814187674, + "learning_rate": 4.682623345345962e-06, + "loss": 0.1309, + "step": 9845 + }, + { + "epoch": 2.0645837701824283, + "grad_norm": 4.2602422546839485, + "learning_rate": 4.680706177117143e-06, + "loss": 0.1299, + "step": 9846 + }, + { + "epoch": 2.0647934577479554, + "grad_norm": 3.85664829718295, + "learning_rate": 4.678789281511361e-06, + "loss": 0.1347, + "step": 9847 + }, + { + "epoch": 2.065003145313483, + "grad_norm": 4.302621917990932, + "learning_rate": 4.676872658626854e-06, + "loss": 0.1583, + "step": 9848 + }, + { + "epoch": 2.0652128328790105, + "grad_norm": 4.7393011265663265, + "learning_rate": 4.674956308561859e-06, + "loss": 0.1927, + "step": 9849 + }, + { + "epoch": 2.0654225204445376, + "grad_norm": 4.7959998082606035, + "learning_rate": 4.6730402314145895e-06, + "loss": 0.1449, + "step": 9850 + }, + { + "epoch": 2.065632208010065, + "grad_norm": 4.5838107467083855, + "learning_rate": 4.671124427283243e-06, + "loss": 0.137, + "step": 9851 + }, + { + "epoch": 2.065841895575592, + "grad_norm": 4.402381285534479, + "learning_rate": 4.6692088962660134e-06, + "loss": 0.1683, + "step": 9852 + }, + { + "epoch": 2.0660515831411197, + "grad_norm": 4.600697316242818, + "learning_rate": 4.66729363846108e-06, + "loss": 0.1595, + "step": 9853 + }, + { + "epoch": 2.0662612707066472, + "grad_norm": 3.5785812220866093, + "learning_rate": 4.665378653966591e-06, + "loss": 0.1223, + "step": 9854 + }, + { + "epoch": 2.0664709582721743, + "grad_norm": 4.913849900889719, + "learning_rate": 4.663463942880696e-06, + "loss": 0.2068, + "step": 9855 + }, + { + "epoch": 2.066680645837702, + "grad_norm": 3.676429624335855, + "learning_rate": 4.661549505301533e-06, + "loss": 0.1387, + "step": 9856 + }, + { + "epoch": 2.0668903334032294, + "grad_norm": 4.841741721034167, + "learning_rate": 4.659635341327219e-06, + "loss": 0.1829, + "step": 9857 + }, + { + "epoch": 2.0671000209687564, + "grad_norm": 5.388231917468368, + "learning_rate": 4.657721451055852e-06, + "loss": 0.1335, + "step": 9858 + }, + { + "epoch": 2.067309708534284, + "grad_norm": 4.6134229582954855, + "learning_rate": 4.65580783458553e-06, + "loss": 0.1716, + "step": 9859 + }, + { + "epoch": 2.0675193960998115, + "grad_norm": 3.838022936901379, + "learning_rate": 4.653894492014321e-06, + "loss": 0.1722, + "step": 9860 + }, + { + "epoch": 2.0677290836653386, + "grad_norm": 4.25482259127271, + "learning_rate": 4.651981423440296e-06, + "loss": 0.1707, + "step": 9861 + }, + { + "epoch": 2.067938771230866, + "grad_norm": 5.0728955024655376, + "learning_rate": 4.650068628961499e-06, + "loss": 0.2098, + "step": 9862 + }, + { + "epoch": 2.068148458796393, + "grad_norm": 4.511049454244188, + "learning_rate": 4.64815610867596e-06, + "loss": 0.1499, + "step": 9863 + }, + { + "epoch": 2.0683581463619207, + "grad_norm": 3.762999021380038, + "learning_rate": 4.646243862681702e-06, + "loss": 0.1231, + "step": 9864 + }, + { + "epoch": 2.0685678339274483, + "grad_norm": 4.892388772060436, + "learning_rate": 4.644331891076736e-06, + "loss": 0.169, + "step": 9865 + }, + { + "epoch": 2.0687775214929753, + "grad_norm": 4.394890335882914, + "learning_rate": 4.642420193959049e-06, + "loss": 0.1457, + "step": 9866 + }, + { + "epoch": 2.068987209058503, + "grad_norm": 3.1566105636787505, + "learning_rate": 4.640508771426616e-06, + "loss": 0.1316, + "step": 9867 + }, + { + "epoch": 2.0691968966240304, + "grad_norm": 4.13807875272563, + "learning_rate": 4.6385976235774074e-06, + "loss": 0.1587, + "step": 9868 + }, + { + "epoch": 2.0694065841895575, + "grad_norm": 5.010358970112285, + "learning_rate": 4.636686750509366e-06, + "loss": 0.1607, + "step": 9869 + }, + { + "epoch": 2.069616271755085, + "grad_norm": 3.6861103172486995, + "learning_rate": 4.634776152320435e-06, + "loss": 0.1493, + "step": 9870 + }, + { + "epoch": 2.069825959320612, + "grad_norm": 3.3430396165443814, + "learning_rate": 4.632865829108532e-06, + "loss": 0.1546, + "step": 9871 + }, + { + "epoch": 2.0700356468861396, + "grad_norm": 3.4497702429447634, + "learning_rate": 4.630955780971559e-06, + "loss": 0.1405, + "step": 9872 + }, + { + "epoch": 2.070245334451667, + "grad_norm": 3.979309856941839, + "learning_rate": 4.629046008007418e-06, + "loss": 0.1586, + "step": 9873 + }, + { + "epoch": 2.0704550220171942, + "grad_norm": 4.326740114553557, + "learning_rate": 4.627136510313986e-06, + "loss": 0.1258, + "step": 9874 + }, + { + "epoch": 2.0706647095827218, + "grad_norm": 3.4459829598510385, + "learning_rate": 4.6252272879891225e-06, + "loss": 0.1297, + "step": 9875 + }, + { + "epoch": 2.0708743971482493, + "grad_norm": 3.717413220616672, + "learning_rate": 4.623318341130682e-06, + "loss": 0.142, + "step": 9876 + }, + { + "epoch": 2.0710840847137764, + "grad_norm": 4.312745311703431, + "learning_rate": 4.621409669836505e-06, + "loss": 0.1741, + "step": 9877 + }, + { + "epoch": 2.071293772279304, + "grad_norm": 4.855753641294632, + "learning_rate": 4.619501274204412e-06, + "loss": 0.1343, + "step": 9878 + }, + { + "epoch": 2.0715034598448314, + "grad_norm": 5.193885223637972, + "learning_rate": 4.617593154332207e-06, + "loss": 0.1436, + "step": 9879 + }, + { + "epoch": 2.0717131474103585, + "grad_norm": 4.0394066322979585, + "learning_rate": 4.615685310317688e-06, + "loss": 0.1435, + "step": 9880 + }, + { + "epoch": 2.071922834975886, + "grad_norm": 3.8073050398408, + "learning_rate": 4.6137777422586395e-06, + "loss": 0.1384, + "step": 9881 + }, + { + "epoch": 2.072132522541413, + "grad_norm": 5.276837357466768, + "learning_rate": 4.611870450252823e-06, + "loss": 0.1864, + "step": 9882 + }, + { + "epoch": 2.0723422101069406, + "grad_norm": 4.653807942958038, + "learning_rate": 4.609963434397989e-06, + "loss": 0.1766, + "step": 9883 + }, + { + "epoch": 2.072551897672468, + "grad_norm": 4.718646870600865, + "learning_rate": 4.608056694791879e-06, + "loss": 0.181, + "step": 9884 + }, + { + "epoch": 2.0727615852379953, + "grad_norm": 4.627845992212738, + "learning_rate": 4.6061502315322124e-06, + "loss": 0.1795, + "step": 9885 + }, + { + "epoch": 2.072971272803523, + "grad_norm": 4.516202143251554, + "learning_rate": 4.604244044716706e-06, + "loss": 0.1688, + "step": 9886 + }, + { + "epoch": 2.0731809603690503, + "grad_norm": 3.2774125610284326, + "learning_rate": 4.60233813444305e-06, + "loss": 0.1009, + "step": 9887 + }, + { + "epoch": 2.0733906479345774, + "grad_norm": 3.228192073765901, + "learning_rate": 4.600432500808924e-06, + "loss": 0.1448, + "step": 9888 + }, + { + "epoch": 2.073600335500105, + "grad_norm": 3.2864884333600557, + "learning_rate": 4.598527143911995e-06, + "loss": 0.1285, + "step": 9889 + }, + { + "epoch": 2.073810023065632, + "grad_norm": 4.850350734629302, + "learning_rate": 4.596622063849927e-06, + "loss": 0.1386, + "step": 9890 + }, + { + "epoch": 2.0740197106311595, + "grad_norm": 6.47227884411318, + "learning_rate": 4.594717260720343e-06, + "loss": 0.2033, + "step": 9891 + }, + { + "epoch": 2.074229398196687, + "grad_norm": 4.810446587143289, + "learning_rate": 4.592812734620874e-06, + "loss": 0.1771, + "step": 9892 + }, + { + "epoch": 2.074439085762214, + "grad_norm": 3.0483492659857463, + "learning_rate": 4.5909084856491336e-06, + "loss": 0.1004, + "step": 9893 + }, + { + "epoch": 2.0746487733277417, + "grad_norm": 4.033444521840543, + "learning_rate": 4.589004513902716e-06, + "loss": 0.1394, + "step": 9894 + }, + { + "epoch": 2.074858460893269, + "grad_norm": 4.196624074670315, + "learning_rate": 4.587100819479198e-06, + "loss": 0.1446, + "step": 9895 + }, + { + "epoch": 2.0750681484587963, + "grad_norm": 3.7960259992944523, + "learning_rate": 4.585197402476155e-06, + "loss": 0.1301, + "step": 9896 + }, + { + "epoch": 2.075277836024324, + "grad_norm": 3.243663392528349, + "learning_rate": 4.583294262991132e-06, + "loss": 0.1278, + "step": 9897 + }, + { + "epoch": 2.0754875235898513, + "grad_norm": 4.534585056756628, + "learning_rate": 4.5813914011216775e-06, + "loss": 0.1475, + "step": 9898 + }, + { + "epoch": 2.0756972111553784, + "grad_norm": 4.629777471828104, + "learning_rate": 4.579488816965312e-06, + "loss": 0.1521, + "step": 9899 + }, + { + "epoch": 2.075906898720906, + "grad_norm": 4.478844623276444, + "learning_rate": 4.577586510619541e-06, + "loss": 0.1685, + "step": 9900 + }, + { + "epoch": 2.076116586286433, + "grad_norm": 5.330111625981263, + "learning_rate": 4.575684482181867e-06, + "loss": 0.1854, + "step": 9901 + }, + { + "epoch": 2.0763262738519606, + "grad_norm": 3.4182647480861466, + "learning_rate": 4.573782731749774e-06, + "loss": 0.1285, + "step": 9902 + }, + { + "epoch": 2.076535961417488, + "grad_norm": 6.0336691125956605, + "learning_rate": 4.571881259420728e-06, + "loss": 0.2079, + "step": 9903 + }, + { + "epoch": 2.076745648983015, + "grad_norm": 4.139161254488462, + "learning_rate": 4.569980065292179e-06, + "loss": 0.1285, + "step": 9904 + }, + { + "epoch": 2.0769553365485427, + "grad_norm": 4.946687594002389, + "learning_rate": 4.568079149461573e-06, + "loss": 0.155, + "step": 9905 + }, + { + "epoch": 2.0771650241140702, + "grad_norm": 3.3972881555687384, + "learning_rate": 4.566178512026328e-06, + "loss": 0.1448, + "step": 9906 + }, + { + "epoch": 2.0773747116795973, + "grad_norm": 4.418225157946723, + "learning_rate": 4.564278153083862e-06, + "loss": 0.163, + "step": 9907 + }, + { + "epoch": 2.077584399245125, + "grad_norm": 3.39876899006876, + "learning_rate": 4.562378072731568e-06, + "loss": 0.1162, + "step": 9908 + }, + { + "epoch": 2.077794086810652, + "grad_norm": 4.768248416768718, + "learning_rate": 4.560478271066826e-06, + "loss": 0.141, + "step": 9909 + }, + { + "epoch": 2.0780037743761794, + "grad_norm": 4.672334172216667, + "learning_rate": 4.558578748187007e-06, + "loss": 0.1995, + "step": 9910 + }, + { + "epoch": 2.078213461941707, + "grad_norm": 4.788753924688709, + "learning_rate": 4.556679504189472e-06, + "loss": 0.1901, + "step": 9911 + }, + { + "epoch": 2.078423149507234, + "grad_norm": 3.891542290172759, + "learning_rate": 4.554780539171546e-06, + "loss": 0.1998, + "step": 9912 + }, + { + "epoch": 2.0786328370727616, + "grad_norm": 4.623149770055773, + "learning_rate": 4.552881853230562e-06, + "loss": 0.1946, + "step": 9913 + }, + { + "epoch": 2.078842524638289, + "grad_norm": 3.870663598053487, + "learning_rate": 4.550983446463834e-06, + "loss": 0.1294, + "step": 9914 + }, + { + "epoch": 2.079052212203816, + "grad_norm": 3.6186756055560627, + "learning_rate": 4.549085318968655e-06, + "loss": 0.154, + "step": 9915 + }, + { + "epoch": 2.0792618997693437, + "grad_norm": 3.135469715721393, + "learning_rate": 4.547187470842304e-06, + "loss": 0.1257, + "step": 9916 + }, + { + "epoch": 2.0794715873348713, + "grad_norm": 2.862892687302075, + "learning_rate": 4.545289902182055e-06, + "loss": 0.0979, + "step": 9917 + }, + { + "epoch": 2.0796812749003983, + "grad_norm": 4.638551248688577, + "learning_rate": 4.543392613085157e-06, + "loss": 0.1669, + "step": 9918 + }, + { + "epoch": 2.079890962465926, + "grad_norm": 2.9819532555855925, + "learning_rate": 4.541495603648854e-06, + "loss": 0.112, + "step": 9919 + }, + { + "epoch": 2.080100650031453, + "grad_norm": 4.422434782975081, + "learning_rate": 4.5395988739703704e-06, + "loss": 0.1794, + "step": 9920 + }, + { + "epoch": 2.0803103375969805, + "grad_norm": 4.232637175638705, + "learning_rate": 4.53770242414691e-06, + "loss": 0.1386, + "step": 9921 + }, + { + "epoch": 2.080520025162508, + "grad_norm": 5.084814883120491, + "learning_rate": 4.535806254275675e-06, + "loss": 0.2158, + "step": 9922 + }, + { + "epoch": 2.080729712728035, + "grad_norm": 3.775504506821414, + "learning_rate": 4.533910364453851e-06, + "loss": 0.1392, + "step": 9923 + }, + { + "epoch": 2.0809394002935626, + "grad_norm": 4.084945256902815, + "learning_rate": 4.532014754778599e-06, + "loss": 0.1307, + "step": 9924 + }, + { + "epoch": 2.08114908785909, + "grad_norm": 4.605619107162844, + "learning_rate": 4.530119425347072e-06, + "loss": 0.1627, + "step": 9925 + }, + { + "epoch": 2.0813587754246172, + "grad_norm": 3.751223623359653, + "learning_rate": 4.528224376256416e-06, + "loss": 0.1507, + "step": 9926 + }, + { + "epoch": 2.0815684629901448, + "grad_norm": 4.258587929674082, + "learning_rate": 4.526329607603748e-06, + "loss": 0.1673, + "step": 9927 + }, + { + "epoch": 2.081778150555672, + "grad_norm": 3.6932002262239427, + "learning_rate": 4.524435119486183e-06, + "loss": 0.1369, + "step": 9928 + }, + { + "epoch": 2.0819878381211994, + "grad_norm": 3.218559333127295, + "learning_rate": 4.522540912000815e-06, + "loss": 0.1166, + "step": 9929 + }, + { + "epoch": 2.082197525686727, + "grad_norm": 3.838510271577625, + "learning_rate": 4.520646985244723e-06, + "loss": 0.1485, + "step": 9930 + }, + { + "epoch": 2.082407213252254, + "grad_norm": 4.3088347799024165, + "learning_rate": 4.518753339314976e-06, + "loss": 0.1358, + "step": 9931 + }, + { + "epoch": 2.0826169008177815, + "grad_norm": 4.658867148489126, + "learning_rate": 4.516859974308634e-06, + "loss": 0.1494, + "step": 9932 + }, + { + "epoch": 2.082826588383309, + "grad_norm": 4.68076465028237, + "learning_rate": 4.514966890322722e-06, + "loss": 0.1411, + "step": 9933 + }, + { + "epoch": 2.083036275948836, + "grad_norm": 4.824966080112759, + "learning_rate": 4.513074087454269e-06, + "loss": 0.1891, + "step": 9934 + }, + { + "epoch": 2.0832459635143636, + "grad_norm": 4.055925530419778, + "learning_rate": 4.51118156580029e-06, + "loss": 0.1674, + "step": 9935 + }, + { + "epoch": 2.083455651079891, + "grad_norm": 5.672455172940869, + "learning_rate": 4.509289325457775e-06, + "loss": 0.169, + "step": 9936 + }, + { + "epoch": 2.0836653386454183, + "grad_norm": 4.639172933621322, + "learning_rate": 4.507397366523703e-06, + "loss": 0.163, + "step": 9937 + }, + { + "epoch": 2.083875026210946, + "grad_norm": 3.776853790233513, + "learning_rate": 4.5055056890950414e-06, + "loss": 0.1124, + "step": 9938 + }, + { + "epoch": 2.084084713776473, + "grad_norm": 4.771092267580428, + "learning_rate": 4.503614293268745e-06, + "loss": 0.1483, + "step": 9939 + }, + { + "epoch": 2.0842944013420004, + "grad_norm": 3.218889847542544, + "learning_rate": 4.501723179141752e-06, + "loss": 0.1249, + "step": 9940 + }, + { + "epoch": 2.084504088907528, + "grad_norm": 5.0095150094804035, + "learning_rate": 4.499832346810976e-06, + "loss": 0.2037, + "step": 9941 + }, + { + "epoch": 2.084713776473055, + "grad_norm": 4.525541148791993, + "learning_rate": 4.4979417963733376e-06, + "loss": 0.1759, + "step": 9942 + }, + { + "epoch": 2.0849234640385825, + "grad_norm": 4.718125371059388, + "learning_rate": 4.49605152792572e-06, + "loss": 0.1529, + "step": 9943 + }, + { + "epoch": 2.08513315160411, + "grad_norm": 4.975112082928179, + "learning_rate": 4.494161541565012e-06, + "loss": 0.1735, + "step": 9944 + }, + { + "epoch": 2.085342839169637, + "grad_norm": 3.6215337723336014, + "learning_rate": 4.4922718373880745e-06, + "loss": 0.1261, + "step": 9945 + }, + { + "epoch": 2.0855525267351647, + "grad_norm": 3.7784406929232937, + "learning_rate": 4.490382415491754e-06, + "loss": 0.1398, + "step": 9946 + }, + { + "epoch": 2.0857622143006918, + "grad_norm": 3.2708266560007107, + "learning_rate": 4.488493275972891e-06, + "loss": 0.1237, + "step": 9947 + }, + { + "epoch": 2.0859719018662193, + "grad_norm": 3.6794083731370413, + "learning_rate": 4.48660441892831e-06, + "loss": 0.1611, + "step": 9948 + }, + { + "epoch": 2.086181589431747, + "grad_norm": 4.29806996072104, + "learning_rate": 4.484715844454816e-06, + "loss": 0.1537, + "step": 9949 + }, + { + "epoch": 2.086391276997274, + "grad_norm": 3.7580847087667655, + "learning_rate": 4.482827552649196e-06, + "loss": 0.1493, + "step": 9950 + }, + { + "epoch": 2.0866009645628014, + "grad_norm": 4.61325294169398, + "learning_rate": 4.480939543608237e-06, + "loss": 0.1512, + "step": 9951 + }, + { + "epoch": 2.086810652128329, + "grad_norm": 3.4669147318968245, + "learning_rate": 4.479051817428698e-06, + "loss": 0.1152, + "step": 9952 + }, + { + "epoch": 2.087020339693856, + "grad_norm": 3.747501903564905, + "learning_rate": 4.477164374207326e-06, + "loss": 0.1498, + "step": 9953 + }, + { + "epoch": 2.0872300272593836, + "grad_norm": 3.8311283290221527, + "learning_rate": 4.475277214040863e-06, + "loss": 0.1421, + "step": 9954 + }, + { + "epoch": 2.087439714824911, + "grad_norm": 4.038068032023206, + "learning_rate": 4.47339033702602e-06, + "loss": 0.1276, + "step": 9955 + }, + { + "epoch": 2.087649402390438, + "grad_norm": 4.719244594934842, + "learning_rate": 4.471503743259512e-06, + "loss": 0.1604, + "step": 9956 + }, + { + "epoch": 2.0878590899559657, + "grad_norm": 4.67082924505463, + "learning_rate": 4.469617432838026e-06, + "loss": 0.1336, + "step": 9957 + }, + { + "epoch": 2.088068777521493, + "grad_norm": 3.834657396475897, + "learning_rate": 4.467731405858234e-06, + "loss": 0.1451, + "step": 9958 + }, + { + "epoch": 2.0882784650870203, + "grad_norm": 3.870934619890367, + "learning_rate": 4.465845662416803e-06, + "loss": 0.1136, + "step": 9959 + }, + { + "epoch": 2.088488152652548, + "grad_norm": 4.569532634046341, + "learning_rate": 4.463960202610386e-06, + "loss": 0.1494, + "step": 9960 + }, + { + "epoch": 2.088697840218075, + "grad_norm": 4.156412597170068, + "learning_rate": 4.4620750265356084e-06, + "loss": 0.1248, + "step": 9961 + }, + { + "epoch": 2.0889075277836024, + "grad_norm": 4.2684462532874905, + "learning_rate": 4.4601901342890865e-06, + "loss": 0.1333, + "step": 9962 + }, + { + "epoch": 2.08911721534913, + "grad_norm": 4.085800677013555, + "learning_rate": 4.458305525967433e-06, + "loss": 0.138, + "step": 9963 + }, + { + "epoch": 2.089326902914657, + "grad_norm": 3.805517237321966, + "learning_rate": 4.456421201667229e-06, + "loss": 0.1663, + "step": 9964 + }, + { + "epoch": 2.0895365904801846, + "grad_norm": 4.455267385299155, + "learning_rate": 4.4545371614850575e-06, + "loss": 0.14, + "step": 9965 + }, + { + "epoch": 2.0897462780457117, + "grad_norm": 3.6559890372218855, + "learning_rate": 4.4526534055174745e-06, + "loss": 0.1236, + "step": 9966 + }, + { + "epoch": 2.089955965611239, + "grad_norm": 4.467697483521923, + "learning_rate": 4.4507699338610224e-06, + "loss": 0.1625, + "step": 9967 + }, + { + "epoch": 2.0901656531767667, + "grad_norm": 5.261975311164107, + "learning_rate": 4.448886746612235e-06, + "loss": 0.1985, + "step": 9968 + }, + { + "epoch": 2.090375340742294, + "grad_norm": 3.928433702441015, + "learning_rate": 4.4470038438676375e-06, + "loss": 0.1284, + "step": 9969 + }, + { + "epoch": 2.0905850283078213, + "grad_norm": 3.6287231985567, + "learning_rate": 4.445121225723716e-06, + "loss": 0.1163, + "step": 9970 + }, + { + "epoch": 2.090794715873349, + "grad_norm": 3.6665430717139262, + "learning_rate": 4.443238892276967e-06, + "loss": 0.1323, + "step": 9971 + }, + { + "epoch": 2.091004403438876, + "grad_norm": 5.564435999506864, + "learning_rate": 4.4413568436238675e-06, + "loss": 0.184, + "step": 9972 + }, + { + "epoch": 2.0912140910044035, + "grad_norm": 5.66839727961225, + "learning_rate": 4.439475079860869e-06, + "loss": 0.2088, + "step": 9973 + }, + { + "epoch": 2.091423778569931, + "grad_norm": 3.4699683617138626, + "learning_rate": 4.437593601084413e-06, + "loss": 0.1254, + "step": 9974 + }, + { + "epoch": 2.091633466135458, + "grad_norm": 4.247013788116846, + "learning_rate": 4.435712407390937e-06, + "loss": 0.1628, + "step": 9975 + }, + { + "epoch": 2.0918431537009856, + "grad_norm": 4.587708598293912, + "learning_rate": 4.433831498876847e-06, + "loss": 0.1708, + "step": 9976 + }, + { + "epoch": 2.0920528412665127, + "grad_norm": 3.8954986759594097, + "learning_rate": 4.431950875638551e-06, + "loss": 0.1274, + "step": 9977 + }, + { + "epoch": 2.0922625288320402, + "grad_norm": 5.066123601761499, + "learning_rate": 4.430070537772428e-06, + "loss": 0.1762, + "step": 9978 + }, + { + "epoch": 2.0924722163975678, + "grad_norm": 4.534376848446051, + "learning_rate": 4.428190485374849e-06, + "loss": 0.1489, + "step": 9979 + }, + { + "epoch": 2.092681903963095, + "grad_norm": 3.806205809818997, + "learning_rate": 4.42631071854217e-06, + "loss": 0.1564, + "step": 9980 + }, + { + "epoch": 2.0928915915286224, + "grad_norm": 4.3962552628625975, + "learning_rate": 4.424431237370738e-06, + "loss": 0.1671, + "step": 9981 + }, + { + "epoch": 2.09310127909415, + "grad_norm": 4.254370730699198, + "learning_rate": 4.422552041956876e-06, + "loss": 0.1473, + "step": 9982 + }, + { + "epoch": 2.093310966659677, + "grad_norm": 4.013037289042238, + "learning_rate": 4.420673132396891e-06, + "loss": 0.1097, + "step": 9983 + }, + { + "epoch": 2.0935206542252045, + "grad_norm": 4.72473926245404, + "learning_rate": 4.418794508787088e-06, + "loss": 0.1456, + "step": 9984 + }, + { + "epoch": 2.0937303417907316, + "grad_norm": 5.274460152966914, + "learning_rate": 4.416916171223744e-06, + "loss": 0.1625, + "step": 9985 + }, + { + "epoch": 2.093940029356259, + "grad_norm": 6.006779094458719, + "learning_rate": 4.415038119803132e-06, + "loss": 0.1896, + "step": 9986 + }, + { + "epoch": 2.0941497169217866, + "grad_norm": 4.742135297336328, + "learning_rate": 4.413160354621504e-06, + "loss": 0.1732, + "step": 9987 + }, + { + "epoch": 2.0943594044873137, + "grad_norm": 3.8943255812833786, + "learning_rate": 4.411282875775094e-06, + "loss": 0.1594, + "step": 9988 + }, + { + "epoch": 2.0945690920528413, + "grad_norm": 3.1545151698454053, + "learning_rate": 4.40940568336013e-06, + "loss": 0.1121, + "step": 9989 + }, + { + "epoch": 2.094778779618369, + "grad_norm": 4.1747396683788605, + "learning_rate": 4.407528777472827e-06, + "loss": 0.1495, + "step": 9990 + }, + { + "epoch": 2.094988467183896, + "grad_norm": 3.3640843012403865, + "learning_rate": 4.405652158209368e-06, + "loss": 0.1366, + "step": 9991 + }, + { + "epoch": 2.0951981547494234, + "grad_norm": 3.381026101229787, + "learning_rate": 4.403775825665938e-06, + "loss": 0.1262, + "step": 9992 + }, + { + "epoch": 2.095407842314951, + "grad_norm": 3.9147047962059527, + "learning_rate": 4.401899779938707e-06, + "loss": 0.1644, + "step": 9993 + }, + { + "epoch": 2.095617529880478, + "grad_norm": 3.3383443449668775, + "learning_rate": 4.400024021123822e-06, + "loss": 0.133, + "step": 9994 + }, + { + "epoch": 2.0958272174460055, + "grad_norm": 5.169996080921641, + "learning_rate": 4.398148549317414e-06, + "loss": 0.184, + "step": 9995 + }, + { + "epoch": 2.0960369050115326, + "grad_norm": 5.169805831312984, + "learning_rate": 4.396273364615613e-06, + "loss": 0.1314, + "step": 9996 + }, + { + "epoch": 2.09624659257706, + "grad_norm": 3.8376495245557405, + "learning_rate": 4.394398467114518e-06, + "loss": 0.1656, + "step": 9997 + }, + { + "epoch": 2.0964562801425877, + "grad_norm": 4.1428463030487634, + "learning_rate": 4.392523856910227e-06, + "loss": 0.173, + "step": 9998 + }, + { + "epoch": 2.0966659677081148, + "grad_norm": 4.86425368801479, + "learning_rate": 4.390649534098811e-06, + "loss": 0.1633, + "step": 9999 + }, + { + "epoch": 2.0968756552736423, + "grad_norm": 4.2475605699460255, + "learning_rate": 4.388775498776338e-06, + "loss": 0.1181, + "step": 10000 + }, + { + "epoch": 2.09708534283917, + "grad_norm": 4.19297273951789, + "learning_rate": 4.38690175103885e-06, + "loss": 0.1253, + "step": 10001 + }, + { + "epoch": 2.097295030404697, + "grad_norm": 4.134736487489113, + "learning_rate": 4.385028290982386e-06, + "loss": 0.1401, + "step": 10002 + }, + { + "epoch": 2.0975047179702244, + "grad_norm": 4.949889406507961, + "learning_rate": 4.383155118702961e-06, + "loss": 0.1513, + "step": 10003 + }, + { + "epoch": 2.0977144055357515, + "grad_norm": 4.730273154717187, + "learning_rate": 4.381282234296575e-06, + "loss": 0.1527, + "step": 10004 + }, + { + "epoch": 2.097924093101279, + "grad_norm": 4.900188113642948, + "learning_rate": 4.379409637859218e-06, + "loss": 0.1763, + "step": 10005 + }, + { + "epoch": 2.0981337806668066, + "grad_norm": 5.012756724253606, + "learning_rate": 4.3775373294868705e-06, + "loss": 0.1767, + "step": 10006 + }, + { + "epoch": 2.0983434682323336, + "grad_norm": 4.596162706064852, + "learning_rate": 4.375665309275485e-06, + "loss": 0.1955, + "step": 10007 + }, + { + "epoch": 2.098553155797861, + "grad_norm": 3.98645206954477, + "learning_rate": 4.3737935773210046e-06, + "loss": 0.1545, + "step": 10008 + }, + { + "epoch": 2.0987628433633887, + "grad_norm": 3.130811775476905, + "learning_rate": 4.371922133719367e-06, + "loss": 0.1133, + "step": 10009 + }, + { + "epoch": 2.098972530928916, + "grad_norm": 3.2543905763251506, + "learning_rate": 4.370050978566476e-06, + "loss": 0.1209, + "step": 10010 + }, + { + "epoch": 2.0991822184944433, + "grad_norm": 5.4301024990181, + "learning_rate": 4.36818011195824e-06, + "loss": 0.18, + "step": 10011 + }, + { + "epoch": 2.099391906059971, + "grad_norm": 3.4528343859996635, + "learning_rate": 4.366309533990543e-06, + "loss": 0.1319, + "step": 10012 + }, + { + "epoch": 2.099601593625498, + "grad_norm": 3.7322682151174993, + "learning_rate": 4.364439244759248e-06, + "loss": 0.1258, + "step": 10013 + }, + { + "epoch": 2.0998112811910254, + "grad_norm": 4.189320251718574, + "learning_rate": 4.3625692443602204e-06, + "loss": 0.1496, + "step": 10014 + }, + { + "epoch": 2.1000209687565525, + "grad_norm": 3.8661448303702968, + "learning_rate": 4.3606995328892965e-06, + "loss": 0.1065, + "step": 10015 + }, + { + "epoch": 2.10023065632208, + "grad_norm": 5.070207055072522, + "learning_rate": 4.358830110442299e-06, + "loss": 0.1733, + "step": 10016 + }, + { + "epoch": 2.1004403438876076, + "grad_norm": 4.242060838787802, + "learning_rate": 4.356960977115043e-06, + "loss": 0.1394, + "step": 10017 + }, + { + "epoch": 2.1006500314531347, + "grad_norm": 5.0036563898385635, + "learning_rate": 4.355092133003328e-06, + "loss": 0.1445, + "step": 10018 + }, + { + "epoch": 2.100859719018662, + "grad_norm": 4.459394411770401, + "learning_rate": 4.353223578202931e-06, + "loss": 0.1281, + "step": 10019 + }, + { + "epoch": 2.1010694065841897, + "grad_norm": 3.196069591355032, + "learning_rate": 4.351355312809615e-06, + "loss": 0.12, + "step": 10020 + }, + { + "epoch": 2.101279094149717, + "grad_norm": 3.957574425210128, + "learning_rate": 4.349487336919141e-06, + "loss": 0.1489, + "step": 10021 + }, + { + "epoch": 2.1014887817152443, + "grad_norm": 4.139046573621775, + "learning_rate": 4.3476196506272385e-06, + "loss": 0.1412, + "step": 10022 + }, + { + "epoch": 2.101698469280772, + "grad_norm": 3.448373125361909, + "learning_rate": 4.345752254029635e-06, + "loss": 0.1441, + "step": 10023 + }, + { + "epoch": 2.101908156846299, + "grad_norm": 3.9455727134949856, + "learning_rate": 4.343885147222036e-06, + "loss": 0.1689, + "step": 10024 + }, + { + "epoch": 2.1021178444118265, + "grad_norm": 5.735466844701449, + "learning_rate": 4.342018330300129e-06, + "loss": 0.2101, + "step": 10025 + }, + { + "epoch": 2.1023275319773536, + "grad_norm": 5.4001390241609935, + "learning_rate": 4.340151803359596e-06, + "loss": 0.1604, + "step": 10026 + }, + { + "epoch": 2.102537219542881, + "grad_norm": 5.203769880860915, + "learning_rate": 4.338285566496103e-06, + "loss": 0.1766, + "step": 10027 + }, + { + "epoch": 2.1027469071084086, + "grad_norm": 4.069182238778444, + "learning_rate": 4.336419619805294e-06, + "loss": 0.1407, + "step": 10028 + }, + { + "epoch": 2.1029565946739357, + "grad_norm": 5.591594051309537, + "learning_rate": 4.3345539633828e-06, + "loss": 0.2007, + "step": 10029 + }, + { + "epoch": 2.1031662822394632, + "grad_norm": 4.064749381653318, + "learning_rate": 4.332688597324245e-06, + "loss": 0.1432, + "step": 10030 + }, + { + "epoch": 2.1033759698049908, + "grad_norm": 4.5442640731596375, + "learning_rate": 4.330823521725228e-06, + "loss": 0.1735, + "step": 10031 + }, + { + "epoch": 2.103585657370518, + "grad_norm": 10.377040990467885, + "learning_rate": 4.328958736681335e-06, + "loss": 0.1784, + "step": 10032 + }, + { + "epoch": 2.1037953449360454, + "grad_norm": 4.303831628518501, + "learning_rate": 4.327094242288147e-06, + "loss": 0.1731, + "step": 10033 + }, + { + "epoch": 2.1040050325015724, + "grad_norm": 4.787479799134013, + "learning_rate": 4.325230038641214e-06, + "loss": 0.1404, + "step": 10034 + }, + { + "epoch": 2.1042147200671, + "grad_norm": 5.029567796113582, + "learning_rate": 4.323366125836088e-06, + "loss": 0.1774, + "step": 10035 + }, + { + "epoch": 2.1044244076326275, + "grad_norm": 4.642825288977709, + "learning_rate": 4.321502503968294e-06, + "loss": 0.1596, + "step": 10036 + }, + { + "epoch": 2.1046340951981546, + "grad_norm": 4.152967027850551, + "learning_rate": 4.31963917313334e-06, + "loss": 0.1421, + "step": 10037 + }, + { + "epoch": 2.104843782763682, + "grad_norm": 4.192653242644203, + "learning_rate": 4.317776133426732e-06, + "loss": 0.1323, + "step": 10038 + }, + { + "epoch": 2.1050534703292096, + "grad_norm": 3.3844298206748666, + "learning_rate": 4.315913384943955e-06, + "loss": 0.1057, + "step": 10039 + }, + { + "epoch": 2.1052631578947367, + "grad_norm": 4.172212549325048, + "learning_rate": 4.314050927780477e-06, + "loss": 0.1573, + "step": 10040 + }, + { + "epoch": 2.1054728454602643, + "grad_norm": 4.2614245866097695, + "learning_rate": 4.312188762031746e-06, + "loss": 0.1451, + "step": 10041 + }, + { + "epoch": 2.105682533025792, + "grad_norm": 4.721069661333385, + "learning_rate": 4.310326887793211e-06, + "loss": 0.1811, + "step": 10042 + }, + { + "epoch": 2.105892220591319, + "grad_norm": 4.969801811031622, + "learning_rate": 4.308465305160285e-06, + "loss": 0.1261, + "step": 10043 + }, + { + "epoch": 2.1061019081568464, + "grad_norm": 3.223448270226393, + "learning_rate": 4.30660401422839e-06, + "loss": 0.1055, + "step": 10044 + }, + { + "epoch": 2.1063115957223735, + "grad_norm": 4.730753480152726, + "learning_rate": 4.304743015092912e-06, + "loss": 0.1876, + "step": 10045 + }, + { + "epoch": 2.106521283287901, + "grad_norm": 4.301974230073461, + "learning_rate": 4.3028823078492295e-06, + "loss": 0.176, + "step": 10046 + }, + { + "epoch": 2.1067309708534285, + "grad_norm": 5.555576769350366, + "learning_rate": 4.301021892592711e-06, + "loss": 0.172, + "step": 10047 + }, + { + "epoch": 2.1069406584189556, + "grad_norm": 4.747807657954561, + "learning_rate": 4.29916176941871e-06, + "loss": 0.1685, + "step": 10048 + }, + { + "epoch": 2.107150345984483, + "grad_norm": 4.4915366419652845, + "learning_rate": 4.29730193842255e-06, + "loss": 0.155, + "step": 10049 + }, + { + "epoch": 2.1073600335500107, + "grad_norm": 4.678464012342243, + "learning_rate": 4.295442399699557e-06, + "loss": 0.1812, + "step": 10050 + }, + { + "epoch": 2.1075697211155378, + "grad_norm": 3.2853257189271052, + "learning_rate": 4.293583153345036e-06, + "loss": 0.118, + "step": 10051 + }, + { + "epoch": 2.1077794086810653, + "grad_norm": 4.795455820534092, + "learning_rate": 4.291724199454278e-06, + "loss": 0.1766, + "step": 10052 + }, + { + "epoch": 2.1079890962465924, + "grad_norm": 4.687575093402977, + "learning_rate": 4.289865538122552e-06, + "loss": 0.2009, + "step": 10053 + }, + { + "epoch": 2.10819878381212, + "grad_norm": 4.784512134719495, + "learning_rate": 4.288007169445123e-06, + "loss": 0.1826, + "step": 10054 + }, + { + "epoch": 2.1084084713776474, + "grad_norm": 4.619477417222696, + "learning_rate": 4.286149093517231e-06, + "loss": 0.1801, + "step": 10055 + }, + { + "epoch": 2.1086181589431745, + "grad_norm": 3.9323330535721217, + "learning_rate": 4.284291310434111e-06, + "loss": 0.1707, + "step": 10056 + }, + { + "epoch": 2.108827846508702, + "grad_norm": 5.149530038697762, + "learning_rate": 4.2824338202909754e-06, + "loss": 0.1776, + "step": 10057 + }, + { + "epoch": 2.1090375340742296, + "grad_norm": 4.955989889463565, + "learning_rate": 4.280576623183019e-06, + "loss": 0.1552, + "step": 10058 + }, + { + "epoch": 2.1092472216397566, + "grad_norm": 4.591079139635268, + "learning_rate": 4.278719719205431e-06, + "loss": 0.1391, + "step": 10059 + }, + { + "epoch": 2.109456909205284, + "grad_norm": 3.969278193064446, + "learning_rate": 4.276863108453384e-06, + "loss": 0.1348, + "step": 10060 + }, + { + "epoch": 2.1096665967708117, + "grad_norm": 4.602309732614145, + "learning_rate": 4.275006791022028e-06, + "loss": 0.1621, + "step": 10061 + }, + { + "epoch": 2.109876284336339, + "grad_norm": 4.12566193899036, + "learning_rate": 4.273150767006501e-06, + "loss": 0.1738, + "step": 10062 + }, + { + "epoch": 2.1100859719018663, + "grad_norm": 3.4809042918180038, + "learning_rate": 4.271295036501929e-06, + "loss": 0.095, + "step": 10063 + }, + { + "epoch": 2.1102956594673934, + "grad_norm": 4.82748247878098, + "learning_rate": 4.269439599603427e-06, + "loss": 0.1582, + "step": 10064 + }, + { + "epoch": 2.110505347032921, + "grad_norm": 6.328851581587421, + "learning_rate": 4.2675844564060845e-06, + "loss": 0.1776, + "step": 10065 + }, + { + "epoch": 2.1107150345984484, + "grad_norm": 4.463349460662841, + "learning_rate": 4.265729607004977e-06, + "loss": 0.1609, + "step": 10066 + }, + { + "epoch": 2.1109247221639755, + "grad_norm": 3.6917617360512276, + "learning_rate": 4.263875051495176e-06, + "loss": 0.128, + "step": 10067 + }, + { + "epoch": 2.111134409729503, + "grad_norm": 5.4606544466636615, + "learning_rate": 4.262020789971724e-06, + "loss": 0.1832, + "step": 10068 + }, + { + "epoch": 2.1113440972950306, + "grad_norm": 5.008342063081339, + "learning_rate": 4.260166822529662e-06, + "loss": 0.1571, + "step": 10069 + }, + { + "epoch": 2.1115537848605577, + "grad_norm": 3.878436257630092, + "learning_rate": 4.258313149264006e-06, + "loss": 0.1345, + "step": 10070 + }, + { + "epoch": 2.111763472426085, + "grad_norm": 4.419352769981897, + "learning_rate": 4.256459770269755e-06, + "loss": 0.1438, + "step": 10071 + }, + { + "epoch": 2.1119731599916127, + "grad_norm": 5.05863491708066, + "learning_rate": 4.254606685641905e-06, + "loss": 0.1855, + "step": 10072 + }, + { + "epoch": 2.11218284755714, + "grad_norm": 3.827106328373883, + "learning_rate": 4.252753895475432e-06, + "loss": 0.1574, + "step": 10073 + }, + { + "epoch": 2.1123925351226673, + "grad_norm": 4.461140350349457, + "learning_rate": 4.250901399865282e-06, + "loss": 0.1406, + "step": 10074 + }, + { + "epoch": 2.1126022226881944, + "grad_norm": 4.842328730228752, + "learning_rate": 4.2490491989064085e-06, + "loss": 0.1588, + "step": 10075 + }, + { + "epoch": 2.112811910253722, + "grad_norm": 4.132685969345667, + "learning_rate": 4.2471972926937425e-06, + "loss": 0.1462, + "step": 10076 + }, + { + "epoch": 2.1130215978192495, + "grad_norm": 5.006251543968417, + "learning_rate": 4.245345681322191e-06, + "loss": 0.178, + "step": 10077 + }, + { + "epoch": 2.1132312853847766, + "grad_norm": 4.823379335753255, + "learning_rate": 4.243494364886651e-06, + "loss": 0.2049, + "step": 10078 + }, + { + "epoch": 2.113440972950304, + "grad_norm": 3.2864399851430655, + "learning_rate": 4.241643343482013e-06, + "loss": 0.1302, + "step": 10079 + }, + { + "epoch": 2.1136506605158316, + "grad_norm": 4.585328712074158, + "learning_rate": 4.2397926172031366e-06, + "loss": 0.1598, + "step": 10080 + }, + { + "epoch": 2.1138603480813587, + "grad_norm": 4.164435191724604, + "learning_rate": 4.237942186144884e-06, + "loss": 0.1359, + "step": 10081 + }, + { + "epoch": 2.1140700356468862, + "grad_norm": 4.924046823841652, + "learning_rate": 4.236092050402087e-06, + "loss": 0.2007, + "step": 10082 + }, + { + "epoch": 2.1142797232124133, + "grad_norm": 3.935538945533204, + "learning_rate": 4.2342422100695665e-06, + "loss": 0.1543, + "step": 10083 + }, + { + "epoch": 2.114489410777941, + "grad_norm": 3.878665338551989, + "learning_rate": 4.232392665242133e-06, + "loss": 0.1327, + "step": 10084 + }, + { + "epoch": 2.1146990983434684, + "grad_norm": 4.956920865290055, + "learning_rate": 4.230543416014582e-06, + "loss": 0.1683, + "step": 10085 + }, + { + "epoch": 2.1149087859089954, + "grad_norm": 4.278696452348014, + "learning_rate": 4.2286944624816885e-06, + "loss": 0.141, + "step": 10086 + }, + { + "epoch": 2.115118473474523, + "grad_norm": 4.401658686377236, + "learning_rate": 4.226845804738209e-06, + "loss": 0.1528, + "step": 10087 + }, + { + "epoch": 2.1153281610400505, + "grad_norm": 4.196914613907652, + "learning_rate": 4.224997442878898e-06, + "loss": 0.1774, + "step": 10088 + }, + { + "epoch": 2.1155378486055776, + "grad_norm": 3.926638885187505, + "learning_rate": 4.223149376998482e-06, + "loss": 0.1407, + "step": 10089 + }, + { + "epoch": 2.115747536171105, + "grad_norm": 3.96611485764776, + "learning_rate": 4.221301607191684e-06, + "loss": 0.1753, + "step": 10090 + }, + { + "epoch": 2.1159572237366326, + "grad_norm": 4.358617314683359, + "learning_rate": 4.219454133553199e-06, + "loss": 0.1544, + "step": 10091 + }, + { + "epoch": 2.1161669113021597, + "grad_norm": 4.546203519763934, + "learning_rate": 4.217606956177713e-06, + "loss": 0.1647, + "step": 10092 + }, + { + "epoch": 2.1163765988676873, + "grad_norm": 5.119517897537984, + "learning_rate": 4.215760075159903e-06, + "loss": 0.1683, + "step": 10093 + }, + { + "epoch": 2.1165862864332143, + "grad_norm": 3.607397407775089, + "learning_rate": 4.213913490594423e-06, + "loss": 0.1464, + "step": 10094 + }, + { + "epoch": 2.116795973998742, + "grad_norm": 5.405919391867583, + "learning_rate": 4.212067202575907e-06, + "loss": 0.187, + "step": 10095 + }, + { + "epoch": 2.1170056615642694, + "grad_norm": 3.2929630485244346, + "learning_rate": 4.210221211198986e-06, + "loss": 0.1476, + "step": 10096 + }, + { + "epoch": 2.1172153491297965, + "grad_norm": 3.3083251566275638, + "learning_rate": 4.2083755165582726e-06, + "loss": 0.1376, + "step": 10097 + }, + { + "epoch": 2.117425036695324, + "grad_norm": 3.924909992352462, + "learning_rate": 4.20653011874836e-06, + "loss": 0.1685, + "step": 10098 + }, + { + "epoch": 2.1176347242608515, + "grad_norm": 4.089057822028018, + "learning_rate": 4.204685017863824e-06, + "loss": 0.1511, + "step": 10099 + }, + { + "epoch": 2.1178444118263786, + "grad_norm": 4.3237819931346335, + "learning_rate": 4.2028402139992345e-06, + "loss": 0.1598, + "step": 10100 + }, + { + "epoch": 2.118054099391906, + "grad_norm": 4.385749711986025, + "learning_rate": 4.200995707249136e-06, + "loss": 0.1774, + "step": 10101 + }, + { + "epoch": 2.1182637869574332, + "grad_norm": 4.733460921938973, + "learning_rate": 4.199151497708068e-06, + "loss": 0.1938, + "step": 10102 + }, + { + "epoch": 2.1184734745229608, + "grad_norm": 3.685812668143906, + "learning_rate": 4.197307585470547e-06, + "loss": 0.143, + "step": 10103 + }, + { + "epoch": 2.1186831620884883, + "grad_norm": 4.952481969675868, + "learning_rate": 4.195463970631074e-06, + "loss": 0.1575, + "step": 10104 + }, + { + "epoch": 2.1188928496540154, + "grad_norm": 4.63013552708414, + "learning_rate": 4.193620653284138e-06, + "loss": 0.1598, + "step": 10105 + }, + { + "epoch": 2.119102537219543, + "grad_norm": 5.882114104879191, + "learning_rate": 4.19177763352422e-06, + "loss": 0.1788, + "step": 10106 + }, + { + "epoch": 2.1193122247850704, + "grad_norm": 3.960542581436256, + "learning_rate": 4.189934911445771e-06, + "loss": 0.1638, + "step": 10107 + }, + { + "epoch": 2.1195219123505975, + "grad_norm": 4.66233653293019, + "learning_rate": 4.188092487143232e-06, + "loss": 0.1732, + "step": 10108 + }, + { + "epoch": 2.119731599916125, + "grad_norm": 4.375271387495915, + "learning_rate": 4.186250360711035e-06, + "loss": 0.1728, + "step": 10109 + }, + { + "epoch": 2.1199412874816526, + "grad_norm": 4.989013791253462, + "learning_rate": 4.1844085322435924e-06, + "loss": 0.1604, + "step": 10110 + }, + { + "epoch": 2.1201509750471796, + "grad_norm": 2.9117646234746917, + "learning_rate": 4.182567001835295e-06, + "loss": 0.1229, + "step": 10111 + }, + { + "epoch": 2.120360662612707, + "grad_norm": 4.316369253319347, + "learning_rate": 4.180725769580531e-06, + "loss": 0.1794, + "step": 10112 + }, + { + "epoch": 2.1205703501782343, + "grad_norm": 4.67364265328549, + "learning_rate": 4.1788848355736625e-06, + "loss": 0.1846, + "step": 10113 + }, + { + "epoch": 2.120780037743762, + "grad_norm": 3.5326081342207303, + "learning_rate": 4.177044199909045e-06, + "loss": 0.1469, + "step": 10114 + }, + { + "epoch": 2.1209897253092893, + "grad_norm": 6.465095292409778, + "learning_rate": 4.175203862681012e-06, + "loss": 0.215, + "step": 10115 + }, + { + "epoch": 2.1211994128748164, + "grad_norm": 3.773283828322142, + "learning_rate": 4.17336382398388e-06, + "loss": 0.1343, + "step": 10116 + }, + { + "epoch": 2.121409100440344, + "grad_norm": 3.243115020698188, + "learning_rate": 4.171524083911957e-06, + "loss": 0.1119, + "step": 10117 + }, + { + "epoch": 2.1216187880058714, + "grad_norm": 4.545699634466251, + "learning_rate": 4.169684642559538e-06, + "loss": 0.1574, + "step": 10118 + }, + { + "epoch": 2.1218284755713985, + "grad_norm": 4.5269057540890385, + "learning_rate": 4.167845500020894e-06, + "loss": 0.1781, + "step": 10119 + }, + { + "epoch": 2.122038163136926, + "grad_norm": 3.8445812718207533, + "learning_rate": 4.166006656390279e-06, + "loss": 0.1271, + "step": 10120 + }, + { + "epoch": 2.122247850702453, + "grad_norm": 5.203158609751385, + "learning_rate": 4.164168111761942e-06, + "loss": 0.1877, + "step": 10121 + }, + { + "epoch": 2.1224575382679807, + "grad_norm": 3.8834020061099523, + "learning_rate": 4.162329866230114e-06, + "loss": 0.1195, + "step": 10122 + }, + { + "epoch": 2.122667225833508, + "grad_norm": 4.380457757023012, + "learning_rate": 4.1604919198890045e-06, + "loss": 0.1543, + "step": 10123 + }, + { + "epoch": 2.1228769133990353, + "grad_norm": 4.668119681085, + "learning_rate": 4.15865427283281e-06, + "loss": 0.1631, + "step": 10124 + }, + { + "epoch": 2.123086600964563, + "grad_norm": 4.182660033264631, + "learning_rate": 4.156816925155719e-06, + "loss": 0.1416, + "step": 10125 + }, + { + "epoch": 2.1232962885300903, + "grad_norm": 4.902637677481055, + "learning_rate": 4.15497987695189e-06, + "loss": 0.1795, + "step": 10126 + }, + { + "epoch": 2.1235059760956174, + "grad_norm": 3.9500994807432015, + "learning_rate": 4.153143128315486e-06, + "loss": 0.1616, + "step": 10127 + }, + { + "epoch": 2.123715663661145, + "grad_norm": 3.4586483507034287, + "learning_rate": 4.151306679340636e-06, + "loss": 0.1165, + "step": 10128 + }, + { + "epoch": 2.1239253512266725, + "grad_norm": 4.336625512197504, + "learning_rate": 4.149470530121459e-06, + "loss": 0.1723, + "step": 10129 + }, + { + "epoch": 2.1241350387921996, + "grad_norm": 4.558059093789056, + "learning_rate": 4.147634680752064e-06, + "loss": 0.15, + "step": 10130 + }, + { + "epoch": 2.124344726357727, + "grad_norm": 4.177252090244682, + "learning_rate": 4.14579913132655e-06, + "loss": 0.1231, + "step": 10131 + }, + { + "epoch": 2.124554413923254, + "grad_norm": 4.038110633553171, + "learning_rate": 4.143963881938977e-06, + "loss": 0.151, + "step": 10132 + }, + { + "epoch": 2.1247641014887817, + "grad_norm": 3.701671377285011, + "learning_rate": 4.142128932683412e-06, + "loss": 0.1377, + "step": 10133 + }, + { + "epoch": 2.1249737890543092, + "grad_norm": 3.3738755226836123, + "learning_rate": 4.1402942836539015e-06, + "loss": 0.0839, + "step": 10134 + }, + { + "epoch": 2.1251834766198363, + "grad_norm": 4.063641748781456, + "learning_rate": 4.138459934944473e-06, + "loss": 0.1769, + "step": 10135 + }, + { + "epoch": 2.125393164185364, + "grad_norm": 3.931585666586951, + "learning_rate": 4.136625886649136e-06, + "loss": 0.1602, + "step": 10136 + }, + { + "epoch": 2.1256028517508914, + "grad_norm": 4.545788331137572, + "learning_rate": 4.134792138861895e-06, + "loss": 0.2054, + "step": 10137 + }, + { + "epoch": 2.1258125393164184, + "grad_norm": 3.991689931292914, + "learning_rate": 4.132958691676726e-06, + "loss": 0.1393, + "step": 10138 + }, + { + "epoch": 2.126022226881946, + "grad_norm": 4.340126259739262, + "learning_rate": 4.131125545187604e-06, + "loss": 0.1624, + "step": 10139 + }, + { + "epoch": 2.126231914447473, + "grad_norm": 3.78455174407542, + "learning_rate": 4.129292699488477e-06, + "loss": 0.1476, + "step": 10140 + }, + { + "epoch": 2.1264416020130006, + "grad_norm": 4.1655703222570395, + "learning_rate": 4.127460154673276e-06, + "loss": 0.1446, + "step": 10141 + }, + { + "epoch": 2.126651289578528, + "grad_norm": 3.937201620851793, + "learning_rate": 4.125627910835929e-06, + "loss": 0.1443, + "step": 10142 + }, + { + "epoch": 2.126860977144055, + "grad_norm": 4.927761700925233, + "learning_rate": 4.1237959680703435e-06, + "loss": 0.1052, + "step": 10143 + }, + { + "epoch": 2.1270706647095827, + "grad_norm": 4.354194799920581, + "learning_rate": 4.121964326470407e-06, + "loss": 0.1543, + "step": 10144 + }, + { + "epoch": 2.1272803522751103, + "grad_norm": 5.2418347524631415, + "learning_rate": 4.120132986129989e-06, + "loss": 0.1853, + "step": 10145 + }, + { + "epoch": 2.1274900398406373, + "grad_norm": 4.837597365494546, + "learning_rate": 4.1183019471429576e-06, + "loss": 0.198, + "step": 10146 + }, + { + "epoch": 2.127699727406165, + "grad_norm": 3.377443962018991, + "learning_rate": 4.116471209603149e-06, + "loss": 0.1309, + "step": 10147 + }, + { + "epoch": 2.1279094149716924, + "grad_norm": 4.428109567897319, + "learning_rate": 4.1146407736043995e-06, + "loss": 0.1438, + "step": 10148 + }, + { + "epoch": 2.1281191025372195, + "grad_norm": 3.907099874430981, + "learning_rate": 4.112810639240517e-06, + "loss": 0.1663, + "step": 10149 + }, + { + "epoch": 2.128328790102747, + "grad_norm": 4.637491224717953, + "learning_rate": 4.110980806605297e-06, + "loss": 0.1613, + "step": 10150 + }, + { + "epoch": 2.128538477668274, + "grad_norm": 5.351478131527192, + "learning_rate": 4.109151275792525e-06, + "loss": 0.1913, + "step": 10151 + }, + { + "epoch": 2.1287481652338016, + "grad_norm": 5.0326791702639095, + "learning_rate": 4.107322046895974e-06, + "loss": 0.1671, + "step": 10152 + }, + { + "epoch": 2.128957852799329, + "grad_norm": 3.5578423333137064, + "learning_rate": 4.10549312000938e-06, + "loss": 0.1264, + "step": 10153 + }, + { + "epoch": 2.1291675403648562, + "grad_norm": 3.582022682650813, + "learning_rate": 4.103664495226489e-06, + "loss": 0.1536, + "step": 10154 + }, + { + "epoch": 2.1293772279303838, + "grad_norm": 4.257911599169649, + "learning_rate": 4.1018361726410224e-06, + "loss": 0.1477, + "step": 10155 + }, + { + "epoch": 2.1295869154959113, + "grad_norm": 3.9466131342053834, + "learning_rate": 4.100008152346682e-06, + "loss": 0.1515, + "step": 10156 + }, + { + "epoch": 2.1297966030614384, + "grad_norm": 5.447938962424122, + "learning_rate": 4.098180434437154e-06, + "loss": 0.1633, + "step": 10157 + }, + { + "epoch": 2.130006290626966, + "grad_norm": 4.211310278372867, + "learning_rate": 4.096353019006117e-06, + "loss": 0.1506, + "step": 10158 + }, + { + "epoch": 2.130215978192493, + "grad_norm": 3.217935311611867, + "learning_rate": 4.094525906147225e-06, + "loss": 0.1206, + "step": 10159 + }, + { + "epoch": 2.1304256657580205, + "grad_norm": 4.308970804048729, + "learning_rate": 4.092699095954125e-06, + "loss": 0.1325, + "step": 10160 + }, + { + "epoch": 2.130635353323548, + "grad_norm": 5.301004060217339, + "learning_rate": 4.090872588520442e-06, + "loss": 0.158, + "step": 10161 + }, + { + "epoch": 2.130845040889075, + "grad_norm": 4.9455068216367675, + "learning_rate": 4.089046383939784e-06, + "loss": 0.189, + "step": 10162 + }, + { + "epoch": 2.1310547284546026, + "grad_norm": 4.205910351829512, + "learning_rate": 4.087220482305752e-06, + "loss": 0.143, + "step": 10163 + }, + { + "epoch": 2.13126441602013, + "grad_norm": 4.204392619142861, + "learning_rate": 4.085394883711929e-06, + "loss": 0.1732, + "step": 10164 + }, + { + "epoch": 2.1314741035856573, + "grad_norm": 4.78614389960782, + "learning_rate": 4.083569588251876e-06, + "loss": 0.1869, + "step": 10165 + }, + { + "epoch": 2.131683791151185, + "grad_norm": 4.757047615400579, + "learning_rate": 4.0817445960191385e-06, + "loss": 0.1452, + "step": 10166 + }, + { + "epoch": 2.1318934787167123, + "grad_norm": 4.912649891276758, + "learning_rate": 4.079919907107261e-06, + "loss": 0.1227, + "step": 10167 + }, + { + "epoch": 2.1321031662822394, + "grad_norm": 3.887111947164488, + "learning_rate": 4.07809552160975e-06, + "loss": 0.1421, + "step": 10168 + }, + { + "epoch": 2.132312853847767, + "grad_norm": 3.876528291462925, + "learning_rate": 4.07627143962012e-06, + "loss": 0.1552, + "step": 10169 + }, + { + "epoch": 2.132522541413294, + "grad_norm": 4.810092149880191, + "learning_rate": 4.074447661231853e-06, + "loss": 0.173, + "step": 10170 + }, + { + "epoch": 2.1327322289788215, + "grad_norm": 3.8063352203234495, + "learning_rate": 4.0726241865384165e-06, + "loss": 0.1653, + "step": 10171 + }, + { + "epoch": 2.132941916544349, + "grad_norm": 5.490173873852995, + "learning_rate": 4.070801015633274e-06, + "loss": 0.1847, + "step": 10172 + }, + { + "epoch": 2.133151604109876, + "grad_norm": 4.8859261421511375, + "learning_rate": 4.068978148609863e-06, + "loss": 0.1878, + "step": 10173 + }, + { + "epoch": 2.1333612916754037, + "grad_norm": 3.4211518556287617, + "learning_rate": 4.0671555855616065e-06, + "loss": 0.1323, + "step": 10174 + }, + { + "epoch": 2.133570979240931, + "grad_norm": 4.388274629598768, + "learning_rate": 4.065333326581915e-06, + "loss": 0.1571, + "step": 10175 + }, + { + "epoch": 2.1337806668064583, + "grad_norm": 4.757774281818063, + "learning_rate": 4.063511371764188e-06, + "loss": 0.115, + "step": 10176 + }, + { + "epoch": 2.133990354371986, + "grad_norm": 3.5775026591364267, + "learning_rate": 4.0616897212017996e-06, + "loss": 0.127, + "step": 10177 + }, + { + "epoch": 2.134200041937513, + "grad_norm": 4.7226057520241955, + "learning_rate": 4.0598683749881085e-06, + "loss": 0.1322, + "step": 10178 + }, + { + "epoch": 2.1344097295030404, + "grad_norm": 6.709355641660494, + "learning_rate": 4.058047333216469e-06, + "loss": 0.1866, + "step": 10179 + }, + { + "epoch": 2.134619417068568, + "grad_norm": 3.327732615023194, + "learning_rate": 4.056226595980207e-06, + "loss": 0.1068, + "step": 10180 + }, + { + "epoch": 2.134829104634095, + "grad_norm": 4.443679062064012, + "learning_rate": 4.054406163372644e-06, + "loss": 0.1806, + "step": 10181 + }, + { + "epoch": 2.1350387921996226, + "grad_norm": 4.115710777901885, + "learning_rate": 4.052586035487074e-06, + "loss": 0.155, + "step": 10182 + }, + { + "epoch": 2.13524847976515, + "grad_norm": 3.162295644815646, + "learning_rate": 4.050766212416788e-06, + "loss": 0.1166, + "step": 10183 + }, + { + "epoch": 2.135458167330677, + "grad_norm": 3.521417018396423, + "learning_rate": 4.048946694255048e-06, + "loss": 0.1495, + "step": 10184 + }, + { + "epoch": 2.1356678548962047, + "grad_norm": 3.008685069421243, + "learning_rate": 4.047127481095116e-06, + "loss": 0.1208, + "step": 10185 + }, + { + "epoch": 2.1358775424617322, + "grad_norm": 3.9943109226530047, + "learning_rate": 4.045308573030224e-06, + "loss": 0.1571, + "step": 10186 + }, + { + "epoch": 2.1360872300272593, + "grad_norm": 3.4089055212668744, + "learning_rate": 4.0434899701535935e-06, + "loss": 0.1364, + "step": 10187 + }, + { + "epoch": 2.136296917592787, + "grad_norm": 6.292299313163938, + "learning_rate": 4.041671672558432e-06, + "loss": 0.2154, + "step": 10188 + }, + { + "epoch": 2.136506605158314, + "grad_norm": 4.144706815610816, + "learning_rate": 4.03985368033794e-06, + "loss": 0.1283, + "step": 10189 + }, + { + "epoch": 2.1367162927238414, + "grad_norm": 3.355070961173317, + "learning_rate": 4.038035993585275e-06, + "loss": 0.138, + "step": 10190 + }, + { + "epoch": 2.136925980289369, + "grad_norm": 3.7280083949988345, + "learning_rate": 4.0362186123936075e-06, + "loss": 0.1265, + "step": 10191 + }, + { + "epoch": 2.137135667854896, + "grad_norm": 4.056093289675683, + "learning_rate": 4.034401536856083e-06, + "loss": 0.1255, + "step": 10192 + }, + { + "epoch": 2.1373453554204236, + "grad_norm": 4.050370649405686, + "learning_rate": 4.032584767065827e-06, + "loss": 0.1442, + "step": 10193 + }, + { + "epoch": 2.137555042985951, + "grad_norm": 4.681924258824081, + "learning_rate": 4.030768303115947e-06, + "loss": 0.1692, + "step": 10194 + }, + { + "epoch": 2.137764730551478, + "grad_norm": 5.099969483550425, + "learning_rate": 4.028952145099549e-06, + "loss": 0.1699, + "step": 10195 + }, + { + "epoch": 2.1379744181170057, + "grad_norm": 3.577646197429735, + "learning_rate": 4.027136293109705e-06, + "loss": 0.1136, + "step": 10196 + }, + { + "epoch": 2.138184105682533, + "grad_norm": 4.8386777355703385, + "learning_rate": 4.02532074723949e-06, + "loss": 0.1738, + "step": 10197 + }, + { + "epoch": 2.1383937932480603, + "grad_norm": 4.750642685954514, + "learning_rate": 4.023505507581949e-06, + "loss": 0.1286, + "step": 10198 + }, + { + "epoch": 2.138603480813588, + "grad_norm": 5.453403422583522, + "learning_rate": 4.0216905742301125e-06, + "loss": 0.1932, + "step": 10199 + }, + { + "epoch": 2.138813168379115, + "grad_norm": 4.3806140621999905, + "learning_rate": 4.019875947277005e-06, + "loss": 0.1281, + "step": 10200 + }, + { + "epoch": 2.1390228559446425, + "grad_norm": 3.51464371284746, + "learning_rate": 4.018061626815628e-06, + "loss": 0.1315, + "step": 10201 + }, + { + "epoch": 2.13923254351017, + "grad_norm": 4.704765480258965, + "learning_rate": 4.016247612938969e-06, + "loss": 0.1655, + "step": 10202 + }, + { + "epoch": 2.139442231075697, + "grad_norm": 3.710578710753982, + "learning_rate": 4.014433905739994e-06, + "loss": 0.1208, + "step": 10203 + }, + { + "epoch": 2.1396519186412246, + "grad_norm": 2.824973012176952, + "learning_rate": 4.012620505311667e-06, + "loss": 0.1048, + "step": 10204 + }, + { + "epoch": 2.139861606206752, + "grad_norm": 3.304212871752741, + "learning_rate": 4.01080741174692e-06, + "loss": 0.1333, + "step": 10205 + }, + { + "epoch": 2.1400712937722792, + "grad_norm": 4.00001511207721, + "learning_rate": 4.008994625138684e-06, + "loss": 0.1345, + "step": 10206 + }, + { + "epoch": 2.1402809813378068, + "grad_norm": 4.2161891127346784, + "learning_rate": 4.0071821455798655e-06, + "loss": 0.1463, + "step": 10207 + }, + { + "epoch": 2.140490668903334, + "grad_norm": 3.8138949309434116, + "learning_rate": 4.005369973163351e-06, + "loss": 0.1548, + "step": 10208 + }, + { + "epoch": 2.1407003564688614, + "grad_norm": 4.366078323030283, + "learning_rate": 4.0035581079820244e-06, + "loss": 0.1652, + "step": 10209 + }, + { + "epoch": 2.140910044034389, + "grad_norm": 3.7201252453507907, + "learning_rate": 4.00174655012875e-06, + "loss": 0.1329, + "step": 10210 + }, + { + "epoch": 2.141119731599916, + "grad_norm": 4.22883419120911, + "learning_rate": 3.999935299696363e-06, + "loss": 0.177, + "step": 10211 + }, + { + "epoch": 2.1413294191654435, + "grad_norm": 3.474785895417658, + "learning_rate": 3.998124356777697e-06, + "loss": 0.1242, + "step": 10212 + }, + { + "epoch": 2.141539106730971, + "grad_norm": 6.831029604933241, + "learning_rate": 3.996313721465572e-06, + "loss": 0.1633, + "step": 10213 + }, + { + "epoch": 2.141748794296498, + "grad_norm": 5.004839838913437, + "learning_rate": 3.994503393852781e-06, + "loss": 0.1742, + "step": 10214 + }, + { + "epoch": 2.1419584818620256, + "grad_norm": 5.448940202300236, + "learning_rate": 3.992693374032104e-06, + "loss": 0.1694, + "step": 10215 + }, + { + "epoch": 2.1421681694275527, + "grad_norm": 3.97149105611284, + "learning_rate": 3.990883662096314e-06, + "loss": 0.1518, + "step": 10216 + }, + { + "epoch": 2.1423778569930803, + "grad_norm": 4.060765155734436, + "learning_rate": 3.9890742581381545e-06, + "loss": 0.1706, + "step": 10217 + }, + { + "epoch": 2.142587544558608, + "grad_norm": 3.2430490894378527, + "learning_rate": 3.98726516225037e-06, + "loss": 0.1362, + "step": 10218 + }, + { + "epoch": 2.142797232124135, + "grad_norm": 4.474665573475698, + "learning_rate": 3.9854563745256734e-06, + "loss": 0.1348, + "step": 10219 + }, + { + "epoch": 2.1430069196896624, + "grad_norm": 3.6961830178120034, + "learning_rate": 3.983647895056766e-06, + "loss": 0.1207, + "step": 10220 + }, + { + "epoch": 2.14321660725519, + "grad_norm": 4.937679099466792, + "learning_rate": 3.981839723936338e-06, + "loss": 0.1862, + "step": 10221 + }, + { + "epoch": 2.143426294820717, + "grad_norm": 3.690501765546378, + "learning_rate": 3.980031861257066e-06, + "loss": 0.128, + "step": 10222 + }, + { + "epoch": 2.1436359823862445, + "grad_norm": 4.3130680932160095, + "learning_rate": 3.978224307111603e-06, + "loss": 0.1468, + "step": 10223 + }, + { + "epoch": 2.143845669951772, + "grad_norm": 4.945284649204034, + "learning_rate": 3.976417061592585e-06, + "loss": 0.1742, + "step": 10224 + }, + { + "epoch": 2.144055357517299, + "grad_norm": 3.727374764555077, + "learning_rate": 3.974610124792642e-06, + "loss": 0.1529, + "step": 10225 + }, + { + "epoch": 2.1442650450828267, + "grad_norm": 4.5375334455905065, + "learning_rate": 3.972803496804378e-06, + "loss": 0.1637, + "step": 10226 + }, + { + "epoch": 2.1444747326483538, + "grad_norm": 3.709471930166969, + "learning_rate": 3.970997177720392e-06, + "loss": 0.141, + "step": 10227 + }, + { + "epoch": 2.1446844202138813, + "grad_norm": 4.079453304295589, + "learning_rate": 3.969191167633257e-06, + "loss": 0.1257, + "step": 10228 + }, + { + "epoch": 2.144894107779409, + "grad_norm": 4.544578940941523, + "learning_rate": 3.967385466635531e-06, + "loss": 0.1664, + "step": 10229 + }, + { + "epoch": 2.145103795344936, + "grad_norm": 4.439339712354099, + "learning_rate": 3.9655800748197645e-06, + "loss": 0.1424, + "step": 10230 + }, + { + "epoch": 2.1453134829104634, + "grad_norm": 3.951014624152122, + "learning_rate": 3.963774992278491e-06, + "loss": 0.1577, + "step": 10231 + }, + { + "epoch": 2.145523170475991, + "grad_norm": 3.303602404309969, + "learning_rate": 3.961970219104212e-06, + "loss": 0.12, + "step": 10232 + }, + { + "epoch": 2.145732858041518, + "grad_norm": 3.541266713995233, + "learning_rate": 3.960165755389431e-06, + "loss": 0.144, + "step": 10233 + }, + { + "epoch": 2.1459425456070456, + "grad_norm": 3.6086508076778285, + "learning_rate": 3.958361601226635e-06, + "loss": 0.146, + "step": 10234 + }, + { + "epoch": 2.1461522331725726, + "grad_norm": 3.071972091181923, + "learning_rate": 3.956557756708285e-06, + "loss": 0.1234, + "step": 10235 + }, + { + "epoch": 2.1463619207381, + "grad_norm": 5.557180803789155, + "learning_rate": 3.954754221926827e-06, + "loss": 0.2322, + "step": 10236 + }, + { + "epoch": 2.1465716083036277, + "grad_norm": 4.645234789793375, + "learning_rate": 3.952950996974706e-06, + "loss": 0.1665, + "step": 10237 + }, + { + "epoch": 2.146781295869155, + "grad_norm": 4.644390479842836, + "learning_rate": 3.9511480819443295e-06, + "loss": 0.1568, + "step": 10238 + }, + { + "epoch": 2.1469909834346823, + "grad_norm": 3.844798241102887, + "learning_rate": 3.949345476928109e-06, + "loss": 0.1726, + "step": 10239 + }, + { + "epoch": 2.14720067100021, + "grad_norm": 4.0235761240114645, + "learning_rate": 3.947543182018427e-06, + "loss": 0.1423, + "step": 10240 + }, + { + "epoch": 2.147410358565737, + "grad_norm": 5.305409190296219, + "learning_rate": 3.945741197307651e-06, + "loss": 0.1742, + "step": 10241 + }, + { + "epoch": 2.1476200461312644, + "grad_norm": 3.12150027658663, + "learning_rate": 3.943939522888138e-06, + "loss": 0.1212, + "step": 10242 + }, + { + "epoch": 2.147829733696792, + "grad_norm": 4.4521856609745525, + "learning_rate": 3.942138158852233e-06, + "loss": 0.1695, + "step": 10243 + }, + { + "epoch": 2.148039421262319, + "grad_norm": 4.350284112977845, + "learning_rate": 3.940337105292255e-06, + "loss": 0.17, + "step": 10244 + }, + { + "epoch": 2.1482491088278466, + "grad_norm": 3.623826233435792, + "learning_rate": 3.938536362300505e-06, + "loss": 0.1371, + "step": 10245 + }, + { + "epoch": 2.1484587963933737, + "grad_norm": 4.881080766673594, + "learning_rate": 3.936735929969281e-06, + "loss": 0.1974, + "step": 10246 + }, + { + "epoch": 2.148668483958901, + "grad_norm": 5.026858077527963, + "learning_rate": 3.93493580839086e-06, + "loss": 0.1772, + "step": 10247 + }, + { + "epoch": 2.1488781715244287, + "grad_norm": 2.7366228822996934, + "learning_rate": 3.933135997657497e-06, + "loss": 0.0839, + "step": 10248 + }, + { + "epoch": 2.149087859089956, + "grad_norm": 4.393154899593718, + "learning_rate": 3.931336497861433e-06, + "loss": 0.1595, + "step": 10249 + }, + { + "epoch": 2.1492975466554833, + "grad_norm": 3.771485766371555, + "learning_rate": 3.929537309094904e-06, + "loss": 0.1356, + "step": 10250 + }, + { + "epoch": 2.149507234221011, + "grad_norm": 4.106900801554359, + "learning_rate": 3.927738431450116e-06, + "loss": 0.1572, + "step": 10251 + }, + { + "epoch": 2.149716921786538, + "grad_norm": 4.268596240966985, + "learning_rate": 3.925939865019262e-06, + "loss": 0.1492, + "step": 10252 + }, + { + "epoch": 2.1499266093520655, + "grad_norm": 4.860702241878373, + "learning_rate": 3.924141609894528e-06, + "loss": 0.1902, + "step": 10253 + }, + { + "epoch": 2.1501362969175926, + "grad_norm": 4.520183363697384, + "learning_rate": 3.92234366616807e-06, + "loss": 0.1666, + "step": 10254 + }, + { + "epoch": 2.15034598448312, + "grad_norm": 3.8261630062810554, + "learning_rate": 3.920546033932044e-06, + "loss": 0.1639, + "step": 10255 + }, + { + "epoch": 2.1505556720486476, + "grad_norm": 3.8025096376025047, + "learning_rate": 3.918748713278579e-06, + "loss": 0.1308, + "step": 10256 + }, + { + "epoch": 2.1507653596141747, + "grad_norm": 3.368423962803322, + "learning_rate": 3.916951704299786e-06, + "loss": 0.145, + "step": 10257 + }, + { + "epoch": 2.1509750471797022, + "grad_norm": 3.604569715272275, + "learning_rate": 3.91515500708777e-06, + "loss": 0.1484, + "step": 10258 + }, + { + "epoch": 2.1511847347452298, + "grad_norm": 4.399770835695097, + "learning_rate": 3.913358621734615e-06, + "loss": 0.1305, + "step": 10259 + }, + { + "epoch": 2.151394422310757, + "grad_norm": 3.866625956503527, + "learning_rate": 3.9115625483323885e-06, + "loss": 0.1026, + "step": 10260 + }, + { + "epoch": 2.1516041098762844, + "grad_norm": 4.513341007626147, + "learning_rate": 3.909766786973137e-06, + "loss": 0.158, + "step": 10261 + }, + { + "epoch": 2.151813797441812, + "grad_norm": 4.601239239501308, + "learning_rate": 3.907971337748904e-06, + "loss": 0.158, + "step": 10262 + }, + { + "epoch": 2.152023485007339, + "grad_norm": 3.4288479114325234, + "learning_rate": 3.9061762007517045e-06, + "loss": 0.1248, + "step": 10263 + }, + { + "epoch": 2.1522331725728665, + "grad_norm": 4.668719233752378, + "learning_rate": 3.904381376073546e-06, + "loss": 0.1449, + "step": 10264 + }, + { + "epoch": 2.1524428601383936, + "grad_norm": 3.8480362376539543, + "learning_rate": 3.9025868638064135e-06, + "loss": 0.1363, + "step": 10265 + }, + { + "epoch": 2.152652547703921, + "grad_norm": 4.643328068394935, + "learning_rate": 3.900792664042278e-06, + "loss": 0.1526, + "step": 10266 + }, + { + "epoch": 2.1528622352694486, + "grad_norm": 3.577727022641396, + "learning_rate": 3.898998776873095e-06, + "loss": 0.1487, + "step": 10267 + }, + { + "epoch": 2.1530719228349757, + "grad_norm": 5.218745700254782, + "learning_rate": 3.897205202390815e-06, + "loss": 0.1854, + "step": 10268 + }, + { + "epoch": 2.1532816104005033, + "grad_norm": 3.925820989668583, + "learning_rate": 3.895411940687345e-06, + "loss": 0.1617, + "step": 10269 + }, + { + "epoch": 2.153491297966031, + "grad_norm": 4.708899920666233, + "learning_rate": 3.893618991854602e-06, + "loss": 0.181, + "step": 10270 + }, + { + "epoch": 2.153700985531558, + "grad_norm": 4.994359743371025, + "learning_rate": 3.891826355984479e-06, + "loss": 0.1592, + "step": 10271 + }, + { + "epoch": 2.1539106730970854, + "grad_norm": 4.9125148361272615, + "learning_rate": 3.89003403316885e-06, + "loss": 0.2269, + "step": 10272 + }, + { + "epoch": 2.1541203606626125, + "grad_norm": 3.7447386156147924, + "learning_rate": 3.88824202349957e-06, + "loss": 0.1523, + "step": 10273 + }, + { + "epoch": 2.15433004822814, + "grad_norm": 4.1215467607407845, + "learning_rate": 3.886450327068491e-06, + "loss": 0.1349, + "step": 10274 + }, + { + "epoch": 2.1545397357936675, + "grad_norm": 3.3658178408843935, + "learning_rate": 3.884658943967432e-06, + "loss": 0.1401, + "step": 10275 + }, + { + "epoch": 2.1547494233591946, + "grad_norm": 4.544619638715648, + "learning_rate": 3.882867874288212e-06, + "loss": 0.1624, + "step": 10276 + }, + { + "epoch": 2.154959110924722, + "grad_norm": 5.357227598319061, + "learning_rate": 3.8810771181226235e-06, + "loss": 0.1793, + "step": 10277 + }, + { + "epoch": 2.1551687984902497, + "grad_norm": 3.6002854624584697, + "learning_rate": 3.879286675562441e-06, + "loss": 0.143, + "step": 10278 + }, + { + "epoch": 2.1553784860557768, + "grad_norm": 3.9447058019985612, + "learning_rate": 3.877496546699433e-06, + "loss": 0.1259, + "step": 10279 + }, + { + "epoch": 2.1555881736213043, + "grad_norm": 5.090011686312419, + "learning_rate": 3.875706731625349e-06, + "loss": 0.193, + "step": 10280 + }, + { + "epoch": 2.155797861186832, + "grad_norm": 4.0802611252451655, + "learning_rate": 3.8739172304319185e-06, + "loss": 0.1469, + "step": 10281 + }, + { + "epoch": 2.156007548752359, + "grad_norm": 6.063055375557348, + "learning_rate": 3.87212804321085e-06, + "loss": 0.1772, + "step": 10282 + }, + { + "epoch": 2.1562172363178864, + "grad_norm": 4.621646894552943, + "learning_rate": 3.870339170053851e-06, + "loss": 0.1436, + "step": 10283 + }, + { + "epoch": 2.156426923883414, + "grad_norm": 4.3106142211867855, + "learning_rate": 3.868550611052598e-06, + "loss": 0.1449, + "step": 10284 + }, + { + "epoch": 2.156636611448941, + "grad_norm": 4.019818729487891, + "learning_rate": 3.8667623662987655e-06, + "loss": 0.1374, + "step": 10285 + }, + { + "epoch": 2.1568462990144686, + "grad_norm": 4.967431408251424, + "learning_rate": 3.864974435883997e-06, + "loss": 0.1427, + "step": 10286 + }, + { + "epoch": 2.1570559865799956, + "grad_norm": 3.894330024542563, + "learning_rate": 3.863186819899927e-06, + "loss": 0.158, + "step": 10287 + }, + { + "epoch": 2.157265674145523, + "grad_norm": 4.3751328910075875, + "learning_rate": 3.861399518438176e-06, + "loss": 0.1656, + "step": 10288 + }, + { + "epoch": 2.1574753617110507, + "grad_norm": 3.5823785221965263, + "learning_rate": 3.859612531590353e-06, + "loss": 0.1329, + "step": 10289 + }, + { + "epoch": 2.157685049276578, + "grad_norm": 3.889814684066467, + "learning_rate": 3.857825859448031e-06, + "loss": 0.1489, + "step": 10290 + }, + { + "epoch": 2.1578947368421053, + "grad_norm": 4.9848592709399915, + "learning_rate": 3.856039502102788e-06, + "loss": 0.2092, + "step": 10291 + }, + { + "epoch": 2.1581044244076324, + "grad_norm": 3.7051199753307573, + "learning_rate": 3.854253459646179e-06, + "loss": 0.1072, + "step": 10292 + }, + { + "epoch": 2.15831411197316, + "grad_norm": 3.321550239048246, + "learning_rate": 3.852467732169739e-06, + "loss": 0.1103, + "step": 10293 + }, + { + "epoch": 2.1585237995386874, + "grad_norm": 4.333261889831508, + "learning_rate": 3.850682319764988e-06, + "loss": 0.1382, + "step": 10294 + }, + { + "epoch": 2.1587334871042145, + "grad_norm": 4.9737240282806345, + "learning_rate": 3.848897222523436e-06, + "loss": 0.176, + "step": 10295 + }, + { + "epoch": 2.158943174669742, + "grad_norm": 5.8139843942408485, + "learning_rate": 3.847112440536567e-06, + "loss": 0.1637, + "step": 10296 + }, + { + "epoch": 2.1591528622352696, + "grad_norm": 4.893167442342231, + "learning_rate": 3.845327973895859e-06, + "loss": 0.1298, + "step": 10297 + }, + { + "epoch": 2.1593625498007967, + "grad_norm": 4.183142974799366, + "learning_rate": 3.843543822692769e-06, + "loss": 0.1383, + "step": 10298 + }, + { + "epoch": 2.159572237366324, + "grad_norm": 3.5153568968269795, + "learning_rate": 3.841759987018731e-06, + "loss": 0.1328, + "step": 10299 + }, + { + "epoch": 2.1597819249318517, + "grad_norm": 4.3158992023968485, + "learning_rate": 3.839976466965176e-06, + "loss": 0.1861, + "step": 10300 + }, + { + "epoch": 2.159991612497379, + "grad_norm": 3.869353492394298, + "learning_rate": 3.838193262623514e-06, + "loss": 0.1166, + "step": 10301 + }, + { + "epoch": 2.1602013000629063, + "grad_norm": 3.833974048898893, + "learning_rate": 3.836410374085134e-06, + "loss": 0.1196, + "step": 10302 + }, + { + "epoch": 2.160410987628434, + "grad_norm": 4.710967247269625, + "learning_rate": 3.834627801441409e-06, + "loss": 0.1544, + "step": 10303 + }, + { + "epoch": 2.160620675193961, + "grad_norm": 4.500952119278785, + "learning_rate": 3.832845544783703e-06, + "loss": 0.1333, + "step": 10304 + }, + { + "epoch": 2.1608303627594885, + "grad_norm": 3.540504490872378, + "learning_rate": 3.831063604203363e-06, + "loss": 0.1148, + "step": 10305 + }, + { + "epoch": 2.1610400503250156, + "grad_norm": 4.065927088944747, + "learning_rate": 3.829281979791711e-06, + "loss": 0.1458, + "step": 10306 + }, + { + "epoch": 2.161249737890543, + "grad_norm": 3.5827022833170776, + "learning_rate": 3.8275006716400585e-06, + "loss": 0.1381, + "step": 10307 + }, + { + "epoch": 2.1614594254560706, + "grad_norm": 3.7723839104165453, + "learning_rate": 3.825719679839706e-06, + "loss": 0.1517, + "step": 10308 + }, + { + "epoch": 2.1616691130215977, + "grad_norm": 3.8250447005732915, + "learning_rate": 3.823939004481923e-06, + "loss": 0.1432, + "step": 10309 + }, + { + "epoch": 2.1618788005871252, + "grad_norm": 3.9967357353104327, + "learning_rate": 3.822158645657983e-06, + "loss": 0.1532, + "step": 10310 + }, + { + "epoch": 2.1620884881526528, + "grad_norm": 3.8634173472455173, + "learning_rate": 3.820378603459127e-06, + "loss": 0.1628, + "step": 10311 + }, + { + "epoch": 2.16229817571818, + "grad_norm": 3.323198722801173, + "learning_rate": 3.818598877976583e-06, + "loss": 0.111, + "step": 10312 + }, + { + "epoch": 2.1625078632837074, + "grad_norm": 3.580256053677794, + "learning_rate": 3.816819469301569e-06, + "loss": 0.1206, + "step": 10313 + }, + { + "epoch": 2.1627175508492344, + "grad_norm": 3.63978966824555, + "learning_rate": 3.815040377525283e-06, + "loss": 0.1163, + "step": 10314 + }, + { + "epoch": 2.162927238414762, + "grad_norm": 4.8457774877704285, + "learning_rate": 3.8132616027389e-06, + "loss": 0.1808, + "step": 10315 + }, + { + "epoch": 2.1631369259802895, + "grad_norm": 4.911369531649282, + "learning_rate": 3.811483145033592e-06, + "loss": 0.1801, + "step": 10316 + }, + { + "epoch": 2.1633466135458166, + "grad_norm": 4.1665466674770535, + "learning_rate": 3.809705004500509e-06, + "loss": 0.1739, + "step": 10317 + }, + { + "epoch": 2.163556301111344, + "grad_norm": 4.1553721080007, + "learning_rate": 3.8079271812307807e-06, + "loss": 0.135, + "step": 10318 + }, + { + "epoch": 2.1637659886768716, + "grad_norm": 6.522481596317578, + "learning_rate": 3.8061496753155204e-06, + "loss": 0.1679, + "step": 10319 + }, + { + "epoch": 2.1639756762423987, + "grad_norm": 4.0296596464567855, + "learning_rate": 3.8043724868458364e-06, + "loss": 0.1581, + "step": 10320 + }, + { + "epoch": 2.1641853638079263, + "grad_norm": 2.910459594629989, + "learning_rate": 3.8025956159128043e-06, + "loss": 0.119, + "step": 10321 + }, + { + "epoch": 2.164395051373454, + "grad_norm": 4.526608802515305, + "learning_rate": 3.800819062607498e-06, + "loss": 0.1678, + "step": 10322 + }, + { + "epoch": 2.164604738938981, + "grad_norm": 3.4757800200782345, + "learning_rate": 3.799042827020968e-06, + "loss": 0.1369, + "step": 10323 + }, + { + "epoch": 2.1648144265045084, + "grad_norm": 5.003717586994313, + "learning_rate": 3.797266909244245e-06, + "loss": 0.1943, + "step": 10324 + }, + { + "epoch": 2.1650241140700355, + "grad_norm": 4.417448058040557, + "learning_rate": 3.7954913093683497e-06, + "loss": 0.1388, + "step": 10325 + }, + { + "epoch": 2.165233801635563, + "grad_norm": 4.163842074623893, + "learning_rate": 3.79371602748429e-06, + "loss": 0.1761, + "step": 10326 + }, + { + "epoch": 2.1654434892010905, + "grad_norm": 4.955661949231234, + "learning_rate": 3.791941063683048e-06, + "loss": 0.1589, + "step": 10327 + }, + { + "epoch": 2.1656531767666176, + "grad_norm": 3.3814612606396084, + "learning_rate": 3.790166418055591e-06, + "loss": 0.1166, + "step": 10328 + }, + { + "epoch": 2.165862864332145, + "grad_norm": 3.3215147568184524, + "learning_rate": 3.7883920906928784e-06, + "loss": 0.1182, + "step": 10329 + }, + { + "epoch": 2.1660725518976727, + "grad_norm": 4.115026731281144, + "learning_rate": 3.7866180816858444e-06, + "loss": 0.1514, + "step": 10330 + }, + { + "epoch": 2.1662822394631998, + "grad_norm": 3.9539540771018995, + "learning_rate": 3.7848443911254074e-06, + "loss": 0.1438, + "step": 10331 + }, + { + "epoch": 2.1664919270287273, + "grad_norm": 3.944924922828241, + "learning_rate": 3.7830710191024777e-06, + "loss": 0.1426, + "step": 10332 + }, + { + "epoch": 2.1667016145942544, + "grad_norm": 5.219824693326552, + "learning_rate": 3.7812979657079385e-06, + "loss": 0.1548, + "step": 10333 + }, + { + "epoch": 2.166911302159782, + "grad_norm": 6.686378470468252, + "learning_rate": 3.779525231032667e-06, + "loss": 0.2362, + "step": 10334 + }, + { + "epoch": 2.1671209897253094, + "grad_norm": 4.41325012428354, + "learning_rate": 3.777752815167517e-06, + "loss": 0.1304, + "step": 10335 + }, + { + "epoch": 2.1673306772908365, + "grad_norm": 4.666946070962675, + "learning_rate": 3.7759807182033236e-06, + "loss": 0.1679, + "step": 10336 + }, + { + "epoch": 2.167540364856364, + "grad_norm": 4.61871361230585, + "learning_rate": 3.7742089402309122e-06, + "loss": 0.1526, + "step": 10337 + }, + { + "epoch": 2.1677500524218916, + "grad_norm": 5.5695614505724365, + "learning_rate": 3.772437481341096e-06, + "loss": 0.1771, + "step": 10338 + }, + { + "epoch": 2.1679597399874186, + "grad_norm": 3.8917628856711217, + "learning_rate": 3.770666341624659e-06, + "loss": 0.1568, + "step": 10339 + }, + { + "epoch": 2.168169427552946, + "grad_norm": 3.9307803032551276, + "learning_rate": 3.768895521172373e-06, + "loss": 0.1312, + "step": 10340 + }, + { + "epoch": 2.1683791151184737, + "grad_norm": 4.004573833706044, + "learning_rate": 3.767125020075002e-06, + "loss": 0.1644, + "step": 10341 + }, + { + "epoch": 2.168588802684001, + "grad_norm": 4.852865509553495, + "learning_rate": 3.765354838423281e-06, + "loss": 0.1489, + "step": 10342 + }, + { + "epoch": 2.1687984902495283, + "grad_norm": 4.266909237431339, + "learning_rate": 3.7635849763079413e-06, + "loss": 0.1487, + "step": 10343 + }, + { + "epoch": 2.1690081778150554, + "grad_norm": 3.600150534923006, + "learning_rate": 3.761815433819688e-06, + "loss": 0.1381, + "step": 10344 + }, + { + "epoch": 2.169217865380583, + "grad_norm": 3.8982235636233513, + "learning_rate": 3.7600462110492097e-06, + "loss": 0.1323, + "step": 10345 + }, + { + "epoch": 2.1694275529461104, + "grad_norm": 4.332766841079706, + "learning_rate": 3.758277308087186e-06, + "loss": 0.13, + "step": 10346 + }, + { + "epoch": 2.1696372405116375, + "grad_norm": 4.95122904396534, + "learning_rate": 3.7565087250242825e-06, + "loss": 0.1457, + "step": 10347 + }, + { + "epoch": 2.169846928077165, + "grad_norm": 3.9682599464666954, + "learning_rate": 3.7547404619511294e-06, + "loss": 0.1554, + "step": 10348 + }, + { + "epoch": 2.1700566156426926, + "grad_norm": 4.095760828256584, + "learning_rate": 3.752972518958359e-06, + "loss": 0.148, + "step": 10349 + }, + { + "epoch": 2.1702663032082197, + "grad_norm": 3.5423264474596796, + "learning_rate": 3.7512048961365855e-06, + "loss": 0.134, + "step": 10350 + }, + { + "epoch": 2.170475990773747, + "grad_norm": 4.731927735995504, + "learning_rate": 3.7494375935763995e-06, + "loss": 0.1553, + "step": 10351 + }, + { + "epoch": 2.1706856783392743, + "grad_norm": 4.617120888454631, + "learning_rate": 3.7476706113683737e-06, + "loss": 0.181, + "step": 10352 + }, + { + "epoch": 2.170895365904802, + "grad_norm": 3.3312324097953536, + "learning_rate": 3.745903949603077e-06, + "loss": 0.1435, + "step": 10353 + }, + { + "epoch": 2.1711050534703293, + "grad_norm": 4.191360713803817, + "learning_rate": 3.7441376083710457e-06, + "loss": 0.1841, + "step": 10354 + }, + { + "epoch": 2.1713147410358564, + "grad_norm": 3.981329947814955, + "learning_rate": 3.7423715877628166e-06, + "loss": 0.1473, + "step": 10355 + }, + { + "epoch": 2.171524428601384, + "grad_norm": 3.9653747961666825, + "learning_rate": 3.7406058878688967e-06, + "loss": 0.1599, + "step": 10356 + }, + { + "epoch": 2.1717341161669115, + "grad_norm": 5.672372827672104, + "learning_rate": 3.7388405087797786e-06, + "loss": 0.19, + "step": 10357 + }, + { + "epoch": 2.1719438037324386, + "grad_norm": 3.163269765690655, + "learning_rate": 3.737075450585943e-06, + "loss": 0.1036, + "step": 10358 + }, + { + "epoch": 2.172153491297966, + "grad_norm": 4.604462060597182, + "learning_rate": 3.7353107133778566e-06, + "loss": 0.1403, + "step": 10359 + }, + { + "epoch": 2.1723631788634936, + "grad_norm": 3.697339097810392, + "learning_rate": 3.7335462972459623e-06, + "loss": 0.1517, + "step": 10360 + }, + { + "epoch": 2.1725728664290207, + "grad_norm": 4.660279528361429, + "learning_rate": 3.7317822022806847e-06, + "loss": 0.1895, + "step": 10361 + }, + { + "epoch": 2.1727825539945482, + "grad_norm": 3.6950173092489185, + "learning_rate": 3.730018428572445e-06, + "loss": 0.1476, + "step": 10362 + }, + { + "epoch": 2.1729922415600753, + "grad_norm": 5.608187412786769, + "learning_rate": 3.7282549762116317e-06, + "loss": 0.1959, + "step": 10363 + }, + { + "epoch": 2.173201929125603, + "grad_norm": 3.98769691249305, + "learning_rate": 3.726491845288632e-06, + "loss": 0.1447, + "step": 10364 + }, + { + "epoch": 2.1734116166911304, + "grad_norm": 3.950655426605788, + "learning_rate": 3.7247290358938036e-06, + "loss": 0.1522, + "step": 10365 + }, + { + "epoch": 2.1736213042566574, + "grad_norm": 4.6251148494752785, + "learning_rate": 3.7229665481174983e-06, + "loss": 0.1423, + "step": 10366 + }, + { + "epoch": 2.173830991822185, + "grad_norm": 3.7487452702668933, + "learning_rate": 3.721204382050042e-06, + "loss": 0.1349, + "step": 10367 + }, + { + "epoch": 2.1740406793877125, + "grad_norm": 3.7056118645306944, + "learning_rate": 3.7194425377817533e-06, + "loss": 0.134, + "step": 10368 + }, + { + "epoch": 2.1742503669532396, + "grad_norm": 4.557906263643729, + "learning_rate": 3.7176810154029288e-06, + "loss": 0.1685, + "step": 10369 + }, + { + "epoch": 2.174460054518767, + "grad_norm": 3.7236536372072893, + "learning_rate": 3.7159198150038445e-06, + "loss": 0.1173, + "step": 10370 + }, + { + "epoch": 2.174669742084294, + "grad_norm": 3.977368730173153, + "learning_rate": 3.714158936674769e-06, + "loss": 0.1521, + "step": 10371 + }, + { + "epoch": 2.1748794296498217, + "grad_norm": 3.728554326916154, + "learning_rate": 3.712398380505956e-06, + "loss": 0.1293, + "step": 10372 + }, + { + "epoch": 2.1750891172153493, + "grad_norm": 4.019206693139448, + "learning_rate": 3.710638146587625e-06, + "loss": 0.1657, + "step": 10373 + }, + { + "epoch": 2.1752988047808763, + "grad_norm": 3.291468484294927, + "learning_rate": 3.7088782350099985e-06, + "loss": 0.1265, + "step": 10374 + }, + { + "epoch": 2.175508492346404, + "grad_norm": 4.509319147433582, + "learning_rate": 3.707118645863276e-06, + "loss": 0.1382, + "step": 10375 + }, + { + "epoch": 2.1757181799119314, + "grad_norm": 4.012120337987077, + "learning_rate": 3.7053593792376375e-06, + "loss": 0.1406, + "step": 10376 + }, + { + "epoch": 2.1759278674774585, + "grad_norm": 4.246031998919653, + "learning_rate": 3.7036004352232457e-06, + "loss": 0.1396, + "step": 10377 + }, + { + "epoch": 2.176137555042986, + "grad_norm": 4.757468103740484, + "learning_rate": 3.7018418139102553e-06, + "loss": 0.1388, + "step": 10378 + }, + { + "epoch": 2.1763472426085135, + "grad_norm": 4.038885202611624, + "learning_rate": 3.7000835153887915e-06, + "loss": 0.1559, + "step": 10379 + }, + { + "epoch": 2.1765569301740406, + "grad_norm": 4.803492447664633, + "learning_rate": 3.698325539748978e-06, + "loss": 0.1415, + "step": 10380 + }, + { + "epoch": 2.176766617739568, + "grad_norm": 5.506047016950488, + "learning_rate": 3.69656788708091e-06, + "loss": 0.191, + "step": 10381 + }, + { + "epoch": 2.1769763053050952, + "grad_norm": 4.214636155001907, + "learning_rate": 3.6948105574746674e-06, + "loss": 0.1758, + "step": 10382 + }, + { + "epoch": 2.1771859928706228, + "grad_norm": 4.113497776597873, + "learning_rate": 3.693053551020319e-06, + "loss": 0.1314, + "step": 10383 + }, + { + "epoch": 2.1773956804361503, + "grad_norm": 5.597044272255223, + "learning_rate": 3.69129686780792e-06, + "loss": 0.228, + "step": 10384 + }, + { + "epoch": 2.1776053680016774, + "grad_norm": 4.7517241319113985, + "learning_rate": 3.689540507927497e-06, + "loss": 0.1742, + "step": 10385 + }, + { + "epoch": 2.177815055567205, + "grad_norm": 3.7885796503509157, + "learning_rate": 3.6877844714690646e-06, + "loss": 0.1297, + "step": 10386 + }, + { + "epoch": 2.1780247431327324, + "grad_norm": 4.771173721776659, + "learning_rate": 3.6860287585226294e-06, + "loss": 0.1831, + "step": 10387 + }, + { + "epoch": 2.1782344306982595, + "grad_norm": 4.434410180109572, + "learning_rate": 3.6842733691781686e-06, + "loss": 0.1447, + "step": 10388 + }, + { + "epoch": 2.178444118263787, + "grad_norm": 3.9918866551573475, + "learning_rate": 3.682518303525654e-06, + "loss": 0.1208, + "step": 10389 + }, + { + "epoch": 2.178653805829314, + "grad_norm": 5.57443055857864, + "learning_rate": 3.680763561655034e-06, + "loss": 0.2128, + "step": 10390 + }, + { + "epoch": 2.1788634933948416, + "grad_norm": 3.9806744170939776, + "learning_rate": 3.679009143656238e-06, + "loss": 0.1491, + "step": 10391 + }, + { + "epoch": 2.179073180960369, + "grad_norm": 3.8824567439935964, + "learning_rate": 3.6772550496191906e-06, + "loss": 0.1511, + "step": 10392 + }, + { + "epoch": 2.1792828685258963, + "grad_norm": 4.97616985103004, + "learning_rate": 3.6755012796337885e-06, + "loss": 0.2004, + "step": 10393 + }, + { + "epoch": 2.179492556091424, + "grad_norm": 5.217769227177019, + "learning_rate": 3.67374783378991e-06, + "loss": 0.1823, + "step": 10394 + }, + { + "epoch": 2.1797022436569513, + "grad_norm": 4.620773382319504, + "learning_rate": 3.671994712177429e-06, + "loss": 0.1647, + "step": 10395 + }, + { + "epoch": 2.1799119312224784, + "grad_norm": 4.562600439559105, + "learning_rate": 3.6702419148861966e-06, + "loss": 0.173, + "step": 10396 + }, + { + "epoch": 2.180121618788006, + "grad_norm": 3.4777209003527654, + "learning_rate": 3.668489442006046e-06, + "loss": 0.1294, + "step": 10397 + }, + { + "epoch": 2.1803313063535334, + "grad_norm": 4.433330056566309, + "learning_rate": 3.6667372936267888e-06, + "loss": 0.1412, + "step": 10398 + }, + { + "epoch": 2.1805409939190605, + "grad_norm": 4.073083087577789, + "learning_rate": 3.6649854698382325e-06, + "loss": 0.1654, + "step": 10399 + }, + { + "epoch": 2.180750681484588, + "grad_norm": 4.66692724539925, + "learning_rate": 3.6632339707301557e-06, + "loss": 0.1535, + "step": 10400 + }, + { + "epoch": 2.180960369050115, + "grad_norm": 4.700690783896078, + "learning_rate": 3.6614827963923315e-06, + "loss": 0.1802, + "step": 10401 + }, + { + "epoch": 2.1811700566156427, + "grad_norm": 4.1870644202961085, + "learning_rate": 3.6597319469145075e-06, + "loss": 0.1553, + "step": 10402 + }, + { + "epoch": 2.18137974418117, + "grad_norm": 3.196730751605853, + "learning_rate": 3.6579814223864153e-06, + "loss": 0.1029, + "step": 10403 + }, + { + "epoch": 2.1815894317466973, + "grad_norm": 4.148012508562863, + "learning_rate": 3.656231222897774e-06, + "loss": 0.1324, + "step": 10404 + }, + { + "epoch": 2.181799119312225, + "grad_norm": 5.71753239469698, + "learning_rate": 3.654481348538289e-06, + "loss": 0.1787, + "step": 10405 + }, + { + "epoch": 2.1820088068777523, + "grad_norm": 5.034997349278202, + "learning_rate": 3.6527317993976397e-06, + "loss": 0.1643, + "step": 10406 + }, + { + "epoch": 2.1822184944432794, + "grad_norm": 4.509004937635389, + "learning_rate": 3.6509825755654926e-06, + "loss": 0.1372, + "step": 10407 + }, + { + "epoch": 2.182428182008807, + "grad_norm": 4.805080597776622, + "learning_rate": 3.649233677131503e-06, + "loss": 0.1609, + "step": 10408 + }, + { + "epoch": 2.182637869574334, + "grad_norm": 3.7318194085011305, + "learning_rate": 3.647485104185302e-06, + "loss": 0.1396, + "step": 10409 + }, + { + "epoch": 2.1828475571398616, + "grad_norm": 4.289761972737748, + "learning_rate": 3.645736856816504e-06, + "loss": 0.1703, + "step": 10410 + }, + { + "epoch": 2.183057244705389, + "grad_norm": 3.474863217855607, + "learning_rate": 3.6439889351147174e-06, + "loss": 0.092, + "step": 10411 + }, + { + "epoch": 2.183266932270916, + "grad_norm": 4.602450605675988, + "learning_rate": 3.6422413391695165e-06, + "loss": 0.1704, + "step": 10412 + }, + { + "epoch": 2.1834766198364437, + "grad_norm": 4.936071297251656, + "learning_rate": 3.640494069070479e-06, + "loss": 0.1557, + "step": 10413 + }, + { + "epoch": 2.1836863074019712, + "grad_norm": 3.226485422913305, + "learning_rate": 3.6387471249071505e-06, + "loss": 0.1169, + "step": 10414 + }, + { + "epoch": 2.1838959949674983, + "grad_norm": 4.978268739947345, + "learning_rate": 3.6370005067690606e-06, + "loss": 0.1645, + "step": 10415 + }, + { + "epoch": 2.184105682533026, + "grad_norm": 3.9998159338279216, + "learning_rate": 3.6352542147457302e-06, + "loss": 0.1241, + "step": 10416 + }, + { + "epoch": 2.1843153700985534, + "grad_norm": 4.0094225157654195, + "learning_rate": 3.633508248926665e-06, + "loss": 0.1462, + "step": 10417 + }, + { + "epoch": 2.1845250576640804, + "grad_norm": 4.04747342666115, + "learning_rate": 3.6317626094013436e-06, + "loss": 0.13, + "step": 10418 + }, + { + "epoch": 2.184734745229608, + "grad_norm": 3.787128988596868, + "learning_rate": 3.6300172962592307e-06, + "loss": 0.114, + "step": 10419 + }, + { + "epoch": 2.184944432795135, + "grad_norm": 4.966833053551608, + "learning_rate": 3.628272309589783e-06, + "loss": 0.1617, + "step": 10420 + }, + { + "epoch": 2.1851541203606626, + "grad_norm": 4.400481545456757, + "learning_rate": 3.6265276494824265e-06, + "loss": 0.1531, + "step": 10421 + }, + { + "epoch": 2.18536380792619, + "grad_norm": 3.709960444847731, + "learning_rate": 3.6247833160265867e-06, + "loss": 0.1316, + "step": 10422 + }, + { + "epoch": 2.185573495491717, + "grad_norm": 4.313501832960126, + "learning_rate": 3.6230393093116578e-06, + "loss": 0.1478, + "step": 10423 + }, + { + "epoch": 2.1857831830572447, + "grad_norm": 3.596884656255381, + "learning_rate": 3.621295629427022e-06, + "loss": 0.1414, + "step": 10424 + }, + { + "epoch": 2.1859928706227723, + "grad_norm": 4.74164049459647, + "learning_rate": 3.619552276462048e-06, + "loss": 0.169, + "step": 10425 + }, + { + "epoch": 2.1862025581882993, + "grad_norm": 4.803127498930869, + "learning_rate": 3.61780925050609e-06, + "loss": 0.1552, + "step": 10426 + }, + { + "epoch": 2.186412245753827, + "grad_norm": 5.133960379966977, + "learning_rate": 3.616066551648478e-06, + "loss": 0.1366, + "step": 10427 + }, + { + "epoch": 2.186621933319354, + "grad_norm": 4.637919501072803, + "learning_rate": 3.6143241799785234e-06, + "loss": 0.1734, + "step": 10428 + }, + { + "epoch": 2.1868316208848815, + "grad_norm": 4.177126832482457, + "learning_rate": 3.6125821355855293e-06, + "loss": 0.1318, + "step": 10429 + }, + { + "epoch": 2.187041308450409, + "grad_norm": 4.603232760229832, + "learning_rate": 3.6108404185587877e-06, + "loss": 0.163, + "step": 10430 + }, + { + "epoch": 2.187250996015936, + "grad_norm": 4.582153423055365, + "learning_rate": 3.609099028987547e-06, + "loss": 0.194, + "step": 10431 + }, + { + "epoch": 2.1874606835814636, + "grad_norm": 5.33181598546689, + "learning_rate": 3.607357966961067e-06, + "loss": 0.1913, + "step": 10432 + }, + { + "epoch": 2.187670371146991, + "grad_norm": 4.005593598400876, + "learning_rate": 3.6056172325685824e-06, + "loss": 0.1377, + "step": 10433 + }, + { + "epoch": 2.1878800587125182, + "grad_norm": 4.264745274089219, + "learning_rate": 3.603876825899304e-06, + "loss": 0.1398, + "step": 10434 + }, + { + "epoch": 2.1880897462780458, + "grad_norm": 3.4417428758340374, + "learning_rate": 3.6021367470424294e-06, + "loss": 0.1297, + "step": 10435 + }, + { + "epoch": 2.1882994338435733, + "grad_norm": 4.998102547897121, + "learning_rate": 3.6003969960871453e-06, + "loss": 0.1977, + "step": 10436 + }, + { + "epoch": 2.1885091214091004, + "grad_norm": 4.288684259171469, + "learning_rate": 3.5986575731226124e-06, + "loss": 0.136, + "step": 10437 + }, + { + "epoch": 2.188718808974628, + "grad_norm": 4.2312655452545895, + "learning_rate": 3.596918478237985e-06, + "loss": 0.1736, + "step": 10438 + }, + { + "epoch": 2.188928496540155, + "grad_norm": 3.9962452693592545, + "learning_rate": 3.5951797115223895e-06, + "loss": 0.1334, + "step": 10439 + }, + { + "epoch": 2.1891381841056825, + "grad_norm": 4.682006963641091, + "learning_rate": 3.59344127306494e-06, + "loss": 0.2043, + "step": 10440 + }, + { + "epoch": 2.18934787167121, + "grad_norm": 4.4338111249637695, + "learning_rate": 3.591703162954737e-06, + "loss": 0.1643, + "step": 10441 + }, + { + "epoch": 2.189557559236737, + "grad_norm": 3.6987520607547832, + "learning_rate": 3.5899653812808656e-06, + "loss": 0.1392, + "step": 10442 + }, + { + "epoch": 2.1897672468022646, + "grad_norm": 3.858218074185271, + "learning_rate": 3.588227928132385e-06, + "loss": 0.1162, + "step": 10443 + }, + { + "epoch": 2.189976934367792, + "grad_norm": 4.567917550348727, + "learning_rate": 3.5864908035983416e-06, + "loss": 0.1598, + "step": 10444 + }, + { + "epoch": 2.1901866219333193, + "grad_norm": 4.224606129728385, + "learning_rate": 3.5847540077677702e-06, + "loss": 0.1675, + "step": 10445 + }, + { + "epoch": 2.190396309498847, + "grad_norm": 4.072874673174045, + "learning_rate": 3.58301754072968e-06, + "loss": 0.1584, + "step": 10446 + }, + { + "epoch": 2.190605997064374, + "grad_norm": 4.152993738299188, + "learning_rate": 3.5812814025730724e-06, + "loss": 0.1491, + "step": 10447 + }, + { + "epoch": 2.1908156846299014, + "grad_norm": 3.3215017015813215, + "learning_rate": 3.579545593386927e-06, + "loss": 0.1087, + "step": 10448 + }, + { + "epoch": 2.191025372195429, + "grad_norm": 4.906089861345951, + "learning_rate": 3.5778101132602026e-06, + "loss": 0.174, + "step": 10449 + }, + { + "epoch": 2.191235059760956, + "grad_norm": 3.5663943431525964, + "learning_rate": 3.576074962281848e-06, + "loss": 0.148, + "step": 10450 + }, + { + "epoch": 2.1914447473264835, + "grad_norm": 4.099145239644733, + "learning_rate": 3.5743401405407998e-06, + "loss": 0.1213, + "step": 10451 + }, + { + "epoch": 2.191654434892011, + "grad_norm": 3.9316863565679934, + "learning_rate": 3.5726056481259572e-06, + "loss": 0.1528, + "step": 10452 + }, + { + "epoch": 2.191864122457538, + "grad_norm": 3.4746753669339827, + "learning_rate": 3.5708714851262238e-06, + "loss": 0.1167, + "step": 10453 + }, + { + "epoch": 2.1920738100230657, + "grad_norm": 5.261791192110429, + "learning_rate": 3.569137651630481e-06, + "loss": 0.1855, + "step": 10454 + }, + { + "epoch": 2.192283497588593, + "grad_norm": 4.6948533865131195, + "learning_rate": 3.5674041477275867e-06, + "loss": 0.1934, + "step": 10455 + }, + { + "epoch": 2.1924931851541203, + "grad_norm": 4.354187519282696, + "learning_rate": 3.565670973506383e-06, + "loss": 0.175, + "step": 10456 + }, + { + "epoch": 2.192702872719648, + "grad_norm": 5.609629806891594, + "learning_rate": 3.563938129055705e-06, + "loss": 0.1993, + "step": 10457 + }, + { + "epoch": 2.192912560285175, + "grad_norm": 3.5434957148393367, + "learning_rate": 3.562205614464358e-06, + "loss": 0.1307, + "step": 10458 + }, + { + "epoch": 2.1931222478507024, + "grad_norm": 5.3123534010547, + "learning_rate": 3.560473429821142e-06, + "loss": 0.1777, + "step": 10459 + }, + { + "epoch": 2.19333193541623, + "grad_norm": 5.020795273170776, + "learning_rate": 3.5587415752148326e-06, + "loss": 0.1552, + "step": 10460 + }, + { + "epoch": 2.193541622981757, + "grad_norm": 3.920912824297932, + "learning_rate": 3.5570100507341843e-06, + "loss": 0.1249, + "step": 10461 + }, + { + "epoch": 2.1937513105472846, + "grad_norm": 3.69291304408508, + "learning_rate": 3.5552788564679464e-06, + "loss": 0.1523, + "step": 10462 + }, + { + "epoch": 2.193960998112812, + "grad_norm": 4.081209041070821, + "learning_rate": 3.5535479925048477e-06, + "loss": 0.1214, + "step": 10463 + }, + { + "epoch": 2.194170685678339, + "grad_norm": 3.7475962413922206, + "learning_rate": 3.5518174589335964e-06, + "loss": 0.1356, + "step": 10464 + }, + { + "epoch": 2.1943803732438667, + "grad_norm": 4.085818118016188, + "learning_rate": 3.55008725584288e-06, + "loss": 0.1325, + "step": 10465 + }, + { + "epoch": 2.194590060809394, + "grad_norm": 4.152270145068927, + "learning_rate": 3.5483573833213815e-06, + "loss": 0.135, + "step": 10466 + }, + { + "epoch": 2.1947997483749213, + "grad_norm": 3.658197170513417, + "learning_rate": 3.546627841457755e-06, + "loss": 0.1187, + "step": 10467 + }, + { + "epoch": 2.195009435940449, + "grad_norm": 4.107237872048693, + "learning_rate": 3.544898630340646e-06, + "loss": 0.1411, + "step": 10468 + }, + { + "epoch": 2.195219123505976, + "grad_norm": 5.768380735941217, + "learning_rate": 3.5431697500586795e-06, + "loss": 0.1677, + "step": 10469 + }, + { + "epoch": 2.1954288110715034, + "grad_norm": 4.2366587294187665, + "learning_rate": 3.5414412007004573e-06, + "loss": 0.1625, + "step": 10470 + }, + { + "epoch": 2.195638498637031, + "grad_norm": 4.4287366883140145, + "learning_rate": 3.5397129823545795e-06, + "loss": 0.1659, + "step": 10471 + }, + { + "epoch": 2.195848186202558, + "grad_norm": 3.8956753743029675, + "learning_rate": 3.5379850951096163e-06, + "loss": 0.1204, + "step": 10472 + }, + { + "epoch": 2.1960578737680856, + "grad_norm": 3.4117243215979696, + "learning_rate": 3.536257539054121e-06, + "loss": 0.1294, + "step": 10473 + }, + { + "epoch": 2.196267561333613, + "grad_norm": 3.7019824450155827, + "learning_rate": 3.5345303142766387e-06, + "loss": 0.1248, + "step": 10474 + }, + { + "epoch": 2.19647724889914, + "grad_norm": 3.9654650854328795, + "learning_rate": 3.5328034208656935e-06, + "loss": 0.1352, + "step": 10475 + }, + { + "epoch": 2.1966869364646677, + "grad_norm": 3.6986469903974233, + "learning_rate": 3.531076858909791e-06, + "loss": 0.1339, + "step": 10476 + }, + { + "epoch": 2.196896624030195, + "grad_norm": 4.972096715665888, + "learning_rate": 3.5293506284974155e-06, + "loss": 0.162, + "step": 10477 + }, + { + "epoch": 2.1971063115957223, + "grad_norm": 5.667536979690002, + "learning_rate": 3.5276247297170453e-06, + "loss": 0.1481, + "step": 10478 + }, + { + "epoch": 2.19731599916125, + "grad_norm": 3.84387988238395, + "learning_rate": 3.5258991626571317e-06, + "loss": 0.1137, + "step": 10479 + }, + { + "epoch": 2.197525686726777, + "grad_norm": 3.760782493334706, + "learning_rate": 3.524173927406117e-06, + "loss": 0.1674, + "step": 10480 + }, + { + "epoch": 2.1977353742923045, + "grad_norm": 5.048411022274463, + "learning_rate": 3.522449024052421e-06, + "loss": 0.1527, + "step": 10481 + }, + { + "epoch": 2.197945061857832, + "grad_norm": 6.228036933870606, + "learning_rate": 3.520724452684443e-06, + "loss": 0.184, + "step": 10482 + }, + { + "epoch": 2.198154749423359, + "grad_norm": 3.9885158448806446, + "learning_rate": 3.5190002133905753e-06, + "loss": 0.1721, + "step": 10483 + }, + { + "epoch": 2.1983644369888866, + "grad_norm": 4.66226562534614, + "learning_rate": 3.517276306259191e-06, + "loss": 0.1672, + "step": 10484 + }, + { + "epoch": 2.1985741245544137, + "grad_norm": 4.835441321052882, + "learning_rate": 3.51555273137864e-06, + "loss": 0.1445, + "step": 10485 + }, + { + "epoch": 2.1987838121199412, + "grad_norm": 4.4400048666185805, + "learning_rate": 3.513829488837254e-06, + "loss": 0.1234, + "step": 10486 + }, + { + "epoch": 2.1989934996854688, + "grad_norm": 4.70895698180235, + "learning_rate": 3.5121065787233566e-06, + "loss": 0.1962, + "step": 10487 + }, + { + "epoch": 2.199203187250996, + "grad_norm": 4.813402709251195, + "learning_rate": 3.5103840011252556e-06, + "loss": 0.1565, + "step": 10488 + }, + { + "epoch": 2.1994128748165234, + "grad_norm": 3.775074632537291, + "learning_rate": 3.508661756131224e-06, + "loss": 0.1229, + "step": 10489 + }, + { + "epoch": 2.199622562382051, + "grad_norm": 5.436575849762496, + "learning_rate": 3.5069398438295367e-06, + "loss": 0.173, + "step": 10490 + }, + { + "epoch": 2.199832249947578, + "grad_norm": 5.725163862744351, + "learning_rate": 3.5052182643084476e-06, + "loss": 0.1578, + "step": 10491 + }, + { + "epoch": 2.2000419375131055, + "grad_norm": 3.6001541904034267, + "learning_rate": 3.5034970176561868e-06, + "loss": 0.1392, + "step": 10492 + }, + { + "epoch": 2.200251625078633, + "grad_norm": 4.308541146283223, + "learning_rate": 3.5017761039609677e-06, + "loss": 0.1353, + "step": 10493 + }, + { + "epoch": 2.20046131264416, + "grad_norm": 4.4183231738813715, + "learning_rate": 3.500055523310997e-06, + "loss": 0.1411, + "step": 10494 + }, + { + "epoch": 2.2006710002096876, + "grad_norm": 4.93957161864954, + "learning_rate": 3.498335275794451e-06, + "loss": 0.1704, + "step": 10495 + }, + { + "epoch": 2.2008806877752147, + "grad_norm": 4.817211243280632, + "learning_rate": 3.4966153614995024e-06, + "loss": 0.1933, + "step": 10496 + }, + { + "epoch": 2.2010903753407423, + "grad_norm": 6.067319472415323, + "learning_rate": 3.4948957805142957e-06, + "loss": 0.2036, + "step": 10497 + }, + { + "epoch": 2.20130006290627, + "grad_norm": 4.360478969711781, + "learning_rate": 3.4931765329269605e-06, + "loss": 0.1713, + "step": 10498 + }, + { + "epoch": 2.201509750471797, + "grad_norm": 3.5080431833645584, + "learning_rate": 3.491457618825612e-06, + "loss": 0.1345, + "step": 10499 + }, + { + "epoch": 2.2017194380373244, + "grad_norm": 4.339862703699154, + "learning_rate": 3.4897390382983544e-06, + "loss": 0.1132, + "step": 10500 + }, + { + "epoch": 2.201929125602852, + "grad_norm": 3.009003361056837, + "learning_rate": 3.4880207914332608e-06, + "loss": 0.1429, + "step": 10501 + }, + { + "epoch": 2.202138813168379, + "grad_norm": 4.898990018155972, + "learning_rate": 3.486302878318393e-06, + "loss": 0.1325, + "step": 10502 + }, + { + "epoch": 2.2023485007339065, + "grad_norm": 4.820679789728473, + "learning_rate": 3.484585299041804e-06, + "loss": 0.2099, + "step": 10503 + }, + { + "epoch": 2.2025581882994336, + "grad_norm": 5.059341046650569, + "learning_rate": 3.482868053691515e-06, + "loss": 0.1806, + "step": 10504 + }, + { + "epoch": 2.202767875864961, + "grad_norm": 5.234688695094377, + "learning_rate": 3.4811511423555457e-06, + "loss": 0.203, + "step": 10505 + }, + { + "epoch": 2.2029775634304887, + "grad_norm": 4.168740797287372, + "learning_rate": 3.4794345651218864e-06, + "loss": 0.1594, + "step": 10506 + }, + { + "epoch": 2.2031872509960158, + "grad_norm": 4.482010581068789, + "learning_rate": 3.4777183220785106e-06, + "loss": 0.1553, + "step": 10507 + }, + { + "epoch": 2.2033969385615433, + "grad_norm": 3.528291660976985, + "learning_rate": 3.476002413313384e-06, + "loss": 0.1626, + "step": 10508 + }, + { + "epoch": 2.203606626127071, + "grad_norm": 4.711893691774915, + "learning_rate": 3.4742868389144557e-06, + "loss": 0.1297, + "step": 10509 + }, + { + "epoch": 2.203816313692598, + "grad_norm": 4.735618789096599, + "learning_rate": 3.4725715989696385e-06, + "loss": 0.1687, + "step": 10510 + }, + { + "epoch": 2.2040260012581254, + "grad_norm": 3.0256162683806016, + "learning_rate": 3.4708566935668484e-06, + "loss": 0.087, + "step": 10511 + }, + { + "epoch": 2.204235688823653, + "grad_norm": 4.038067414726757, + "learning_rate": 3.4691421227939804e-06, + "loss": 0.1382, + "step": 10512 + }, + { + "epoch": 2.20444537638918, + "grad_norm": 2.663010938894181, + "learning_rate": 3.4674278867389043e-06, + "loss": 0.1009, + "step": 10513 + }, + { + "epoch": 2.2046550639547076, + "grad_norm": 3.522530704946384, + "learning_rate": 3.4657139854894774e-06, + "loss": 0.0998, + "step": 10514 + }, + { + "epoch": 2.2048647515202346, + "grad_norm": 4.30843232419637, + "learning_rate": 3.464000419133545e-06, + "loss": 0.1799, + "step": 10515 + }, + { + "epoch": 2.205074439085762, + "grad_norm": 4.666999334763687, + "learning_rate": 3.462287187758924e-06, + "loss": 0.1556, + "step": 10516 + }, + { + "epoch": 2.2052841266512897, + "grad_norm": 4.320642539329774, + "learning_rate": 3.460574291453427e-06, + "loss": 0.1501, + "step": 10517 + }, + { + "epoch": 2.205493814216817, + "grad_norm": 3.197882142366623, + "learning_rate": 3.458861730304839e-06, + "loss": 0.103, + "step": 10518 + }, + { + "epoch": 2.2057035017823443, + "grad_norm": 3.697010250738273, + "learning_rate": 3.4571495044009294e-06, + "loss": 0.0923, + "step": 10519 + }, + { + "epoch": 2.205913189347872, + "grad_norm": 3.679497328301629, + "learning_rate": 3.4554376138294553e-06, + "loss": 0.1173, + "step": 10520 + }, + { + "epoch": 2.206122876913399, + "grad_norm": 6.508857305302051, + "learning_rate": 3.4537260586781587e-06, + "loss": 0.2256, + "step": 10521 + }, + { + "epoch": 2.2063325644789265, + "grad_norm": 3.863817772189926, + "learning_rate": 3.452014839034754e-06, + "loss": 0.1424, + "step": 10522 + }, + { + "epoch": 2.2065422520444535, + "grad_norm": 3.972804207892561, + "learning_rate": 3.4503039549869433e-06, + "loss": 0.1384, + "step": 10523 + }, + { + "epoch": 2.206751939609981, + "grad_norm": 4.614506396101835, + "learning_rate": 3.448593406622419e-06, + "loss": 0.1612, + "step": 10524 + }, + { + "epoch": 2.2069616271755086, + "grad_norm": 3.161308821349727, + "learning_rate": 3.4468831940288404e-06, + "loss": 0.116, + "step": 10525 + }, + { + "epoch": 2.2071713147410357, + "grad_norm": 3.1160466555212025, + "learning_rate": 3.4451733172938685e-06, + "loss": 0.1487, + "step": 10526 + }, + { + "epoch": 2.207381002306563, + "grad_norm": 5.038137294236161, + "learning_rate": 3.443463776505133e-06, + "loss": 0.1304, + "step": 10527 + }, + { + "epoch": 2.2075906898720907, + "grad_norm": 4.611214281399228, + "learning_rate": 3.4417545717502466e-06, + "loss": 0.1523, + "step": 10528 + }, + { + "epoch": 2.207800377437618, + "grad_norm": 4.177045506310439, + "learning_rate": 3.4400457031168134e-06, + "loss": 0.1525, + "step": 10529 + }, + { + "epoch": 2.2080100650031453, + "grad_norm": 5.7185468719057955, + "learning_rate": 3.4383371706924216e-06, + "loss": 0.1971, + "step": 10530 + }, + { + "epoch": 2.208219752568673, + "grad_norm": 4.612471219896324, + "learning_rate": 3.436628974564625e-06, + "loss": 0.1731, + "step": 10531 + }, + { + "epoch": 2.2084294401342, + "grad_norm": 4.713898468585529, + "learning_rate": 3.434921114820976e-06, + "loss": 0.1857, + "step": 10532 + }, + { + "epoch": 2.2086391276997275, + "grad_norm": 4.558450999390657, + "learning_rate": 3.4332135915490106e-06, + "loss": 0.1291, + "step": 10533 + }, + { + "epoch": 2.208848815265255, + "grad_norm": 4.1623740809072025, + "learning_rate": 3.431506404836238e-06, + "loss": 0.1525, + "step": 10534 + }, + { + "epoch": 2.209058502830782, + "grad_norm": 4.698715251956241, + "learning_rate": 3.4297995547701504e-06, + "loss": 0.166, + "step": 10535 + }, + { + "epoch": 2.2092681903963096, + "grad_norm": 3.3048108691664315, + "learning_rate": 3.4280930414382353e-06, + "loss": 0.12, + "step": 10536 + }, + { + "epoch": 2.2094778779618367, + "grad_norm": 4.77923003217397, + "learning_rate": 3.4263868649279463e-06, + "loss": 0.1591, + "step": 10537 + }, + { + "epoch": 2.2096875655273642, + "grad_norm": 3.681016843730481, + "learning_rate": 3.424681025326735e-06, + "loss": 0.1479, + "step": 10538 + }, + { + "epoch": 2.2098972530928918, + "grad_norm": 5.848102685205805, + "learning_rate": 3.422975522722026e-06, + "loss": 0.1527, + "step": 10539 + }, + { + "epoch": 2.210106940658419, + "grad_norm": 4.496715792620476, + "learning_rate": 3.421270357201225e-06, + "loss": 0.1321, + "step": 10540 + }, + { + "epoch": 2.2103166282239464, + "grad_norm": 3.9315476034637364, + "learning_rate": 3.4195655288517295e-06, + "loss": 0.1293, + "step": 10541 + }, + { + "epoch": 2.2105263157894735, + "grad_norm": 6.1926348285561055, + "learning_rate": 3.4178610377609155e-06, + "loss": 0.1196, + "step": 10542 + }, + { + "epoch": 2.210736003355001, + "grad_norm": 4.3186473056001375, + "learning_rate": 3.416156884016141e-06, + "loss": 0.1275, + "step": 10543 + }, + { + "epoch": 2.2109456909205285, + "grad_norm": 4.200387211734999, + "learning_rate": 3.4144530677047418e-06, + "loss": 0.1897, + "step": 10544 + }, + { + "epoch": 2.2111553784860556, + "grad_norm": 4.444594369753747, + "learning_rate": 3.412749588914047e-06, + "loss": 0.1418, + "step": 10545 + }, + { + "epoch": 2.211365066051583, + "grad_norm": 6.035932093197452, + "learning_rate": 3.411046447731359e-06, + "loss": 0.1671, + "step": 10546 + }, + { + "epoch": 2.2115747536171106, + "grad_norm": 4.597523340963421, + "learning_rate": 3.4093436442439697e-06, + "loss": 0.1614, + "step": 10547 + }, + { + "epoch": 2.2117844411826377, + "grad_norm": 3.913425189278523, + "learning_rate": 3.407641178539148e-06, + "loss": 0.1421, + "step": 10548 + }, + { + "epoch": 2.2119941287481653, + "grad_norm": 4.0555145765201805, + "learning_rate": 3.4059390507041523e-06, + "loss": 0.1326, + "step": 10549 + }, + { + "epoch": 2.212203816313693, + "grad_norm": 3.251623869958489, + "learning_rate": 3.4042372608262175e-06, + "loss": 0.1155, + "step": 10550 + }, + { + "epoch": 2.21241350387922, + "grad_norm": 4.3015527008423415, + "learning_rate": 3.40253580899256e-06, + "loss": 0.1508, + "step": 10551 + }, + { + "epoch": 2.2126231914447474, + "grad_norm": 4.599544889560929, + "learning_rate": 3.400834695290387e-06, + "loss": 0.1498, + "step": 10552 + }, + { + "epoch": 2.212832879010275, + "grad_norm": 3.98689611312043, + "learning_rate": 3.3991339198068795e-06, + "loss": 0.1213, + "step": 10553 + }, + { + "epoch": 2.213042566575802, + "grad_norm": 3.44408713423122, + "learning_rate": 3.39743348262921e-06, + "loss": 0.1105, + "step": 10554 + }, + { + "epoch": 2.2132522541413295, + "grad_norm": 4.2247241265191855, + "learning_rate": 3.395733383844526e-06, + "loss": 0.1448, + "step": 10555 + }, + { + "epoch": 2.2134619417068566, + "grad_norm": 4.779136395069259, + "learning_rate": 3.3940336235399574e-06, + "loss": 0.1761, + "step": 10556 + }, + { + "epoch": 2.213671629272384, + "grad_norm": 5.4207493341137685, + "learning_rate": 3.392334201802623e-06, + "loss": 0.2026, + "step": 10557 + }, + { + "epoch": 2.2138813168379117, + "grad_norm": 4.301414001442894, + "learning_rate": 3.390635118719625e-06, + "loss": 0.1576, + "step": 10558 + }, + { + "epoch": 2.2140910044034388, + "grad_norm": 3.338157722780891, + "learning_rate": 3.38893637437804e-06, + "loss": 0.1338, + "step": 10559 + }, + { + "epoch": 2.2143006919689663, + "grad_norm": 5.022263332220146, + "learning_rate": 3.387237968864929e-06, + "loss": 0.1937, + "step": 10560 + }, + { + "epoch": 2.214510379534494, + "grad_norm": 3.1090608944638554, + "learning_rate": 3.385539902267344e-06, + "loss": 0.1124, + "step": 10561 + }, + { + "epoch": 2.214720067100021, + "grad_norm": 5.215638916598176, + "learning_rate": 3.3838421746723083e-06, + "loss": 0.1619, + "step": 10562 + }, + { + "epoch": 2.2149297546655484, + "grad_norm": 4.31591027115965, + "learning_rate": 3.38214478616684e-06, + "loss": 0.1137, + "step": 10563 + }, + { + "epoch": 2.2151394422310755, + "grad_norm": 4.076780061126983, + "learning_rate": 3.3804477368379297e-06, + "loss": 0.1315, + "step": 10564 + }, + { + "epoch": 2.215349129796603, + "grad_norm": 3.4126418736210766, + "learning_rate": 3.3787510267725497e-06, + "loss": 0.1012, + "step": 10565 + }, + { + "epoch": 2.2155588173621306, + "grad_norm": 3.8985707238994496, + "learning_rate": 3.3770546560576644e-06, + "loss": 0.135, + "step": 10566 + }, + { + "epoch": 2.2157685049276576, + "grad_norm": 3.7325222353963996, + "learning_rate": 3.3753586247802204e-06, + "loss": 0.1133, + "step": 10567 + }, + { + "epoch": 2.215978192493185, + "grad_norm": 3.578163677739487, + "learning_rate": 3.373662933027131e-06, + "loss": 0.1234, + "step": 10568 + }, + { + "epoch": 2.2161878800587127, + "grad_norm": 4.0290058944375104, + "learning_rate": 3.3719675808853093e-06, + "loss": 0.1889, + "step": 10569 + }, + { + "epoch": 2.21639756762424, + "grad_norm": 4.057417792827492, + "learning_rate": 3.3702725684416482e-06, + "loss": 0.1107, + "step": 10570 + }, + { + "epoch": 2.2166072551897673, + "grad_norm": 4.947652640492737, + "learning_rate": 3.368577895783016e-06, + "loss": 0.1604, + "step": 10571 + }, + { + "epoch": 2.216816942755295, + "grad_norm": 3.4559309237423026, + "learning_rate": 3.3668835629962647e-06, + "loss": 0.1413, + "step": 10572 + }, + { + "epoch": 2.217026630320822, + "grad_norm": 3.5726464751194538, + "learning_rate": 3.365189570168239e-06, + "loss": 0.1414, + "step": 10573 + }, + { + "epoch": 2.2172363178863495, + "grad_norm": 4.581596600241903, + "learning_rate": 3.3634959173857528e-06, + "loss": 0.1562, + "step": 10574 + }, + { + "epoch": 2.2174460054518765, + "grad_norm": 3.0616292761040875, + "learning_rate": 3.3618026047356132e-06, + "loss": 0.0923, + "step": 10575 + }, + { + "epoch": 2.217655693017404, + "grad_norm": 4.443749971723422, + "learning_rate": 3.3601096323046045e-06, + "loss": 0.1367, + "step": 10576 + }, + { + "epoch": 2.2178653805829316, + "grad_norm": 3.7921328324508163, + "learning_rate": 3.3584170001794902e-06, + "loss": 0.1488, + "step": 10577 + }, + { + "epoch": 2.2180750681484587, + "grad_norm": 4.186439901959596, + "learning_rate": 3.356724708447023e-06, + "loss": 0.1617, + "step": 10578 + }, + { + "epoch": 2.218284755713986, + "grad_norm": 4.32495318969111, + "learning_rate": 3.355032757193941e-06, + "loss": 0.1233, + "step": 10579 + }, + { + "epoch": 2.2184944432795137, + "grad_norm": 3.4817229891839694, + "learning_rate": 3.3533411465069544e-06, + "loss": 0.1395, + "step": 10580 + }, + { + "epoch": 2.218704130845041, + "grad_norm": 3.4824599062329344, + "learning_rate": 3.351649876472759e-06, + "loss": 0.1291, + "step": 10581 + }, + { + "epoch": 2.2189138184105683, + "grad_norm": 5.23394493837928, + "learning_rate": 3.3499589471780425e-06, + "loss": 0.1862, + "step": 10582 + }, + { + "epoch": 2.2191235059760954, + "grad_norm": 4.931904266411021, + "learning_rate": 3.34826835870946e-06, + "loss": 0.1384, + "step": 10583 + }, + { + "epoch": 2.219333193541623, + "grad_norm": 4.8460089066395255, + "learning_rate": 3.3465781111536634e-06, + "loss": 0.1339, + "step": 10584 + }, + { + "epoch": 2.2195428811071505, + "grad_norm": 5.723602856536666, + "learning_rate": 3.3448882045972797e-06, + "loss": 0.1603, + "step": 10585 + }, + { + "epoch": 2.2197525686726776, + "grad_norm": 4.319926770022346, + "learning_rate": 3.3431986391269146e-06, + "loss": 0.1416, + "step": 10586 + }, + { + "epoch": 2.219962256238205, + "grad_norm": 4.739797232011968, + "learning_rate": 3.3415094148291637e-06, + "loss": 0.1243, + "step": 10587 + }, + { + "epoch": 2.2201719438037326, + "grad_norm": 4.957718617779369, + "learning_rate": 3.339820531790612e-06, + "loss": 0.1232, + "step": 10588 + }, + { + "epoch": 2.2203816313692597, + "grad_norm": 4.106930493258497, + "learning_rate": 3.3381319900978027e-06, + "loss": 0.1512, + "step": 10589 + }, + { + "epoch": 2.2205913189347872, + "grad_norm": 4.673163328662491, + "learning_rate": 3.3364437898372826e-06, + "loss": 0.1529, + "step": 10590 + }, + { + "epoch": 2.2208010065003148, + "grad_norm": 3.558039461232673, + "learning_rate": 3.334755931095579e-06, + "loss": 0.1312, + "step": 10591 + }, + { + "epoch": 2.221010694065842, + "grad_norm": 3.597685987223306, + "learning_rate": 3.333068413959195e-06, + "loss": 0.1353, + "step": 10592 + }, + { + "epoch": 2.2212203816313694, + "grad_norm": 3.7332485875114365, + "learning_rate": 3.331381238514614e-06, + "loss": 0.1203, + "step": 10593 + }, + { + "epoch": 2.2214300691968965, + "grad_norm": 2.9904384411037235, + "learning_rate": 3.329694404848315e-06, + "loss": 0.1238, + "step": 10594 + }, + { + "epoch": 2.221639756762424, + "grad_norm": 4.711471606219611, + "learning_rate": 3.3280079130467426e-06, + "loss": 0.1755, + "step": 10595 + }, + { + "epoch": 2.2218494443279515, + "grad_norm": 3.8994048355169157, + "learning_rate": 3.3263217631963398e-06, + "loss": 0.1324, + "step": 10596 + }, + { + "epoch": 2.2220591318934786, + "grad_norm": 3.235770309683167, + "learning_rate": 3.3246359553835217e-06, + "loss": 0.1164, + "step": 10597 + }, + { + "epoch": 2.222268819459006, + "grad_norm": 4.662696884062969, + "learning_rate": 3.3229504896946852e-06, + "loss": 0.1802, + "step": 10598 + }, + { + "epoch": 2.2224785070245336, + "grad_norm": 5.909160384507709, + "learning_rate": 3.3212653662162177e-06, + "loss": 0.2086, + "step": 10599 + }, + { + "epoch": 2.2226881945900607, + "grad_norm": 4.3328417253334015, + "learning_rate": 3.3195805850344862e-06, + "loss": 0.1693, + "step": 10600 + }, + { + "epoch": 2.2228978821555883, + "grad_norm": 3.4605092190390185, + "learning_rate": 3.317896146235837e-06, + "loss": 0.1323, + "step": 10601 + }, + { + "epoch": 2.2231075697211153, + "grad_norm": 5.048474557143465, + "learning_rate": 3.3162120499065964e-06, + "loss": 0.1673, + "step": 10602 + }, + { + "epoch": 2.223317257286643, + "grad_norm": 3.977703827457185, + "learning_rate": 3.3145282961330836e-06, + "loss": 0.1661, + "step": 10603 + }, + { + "epoch": 2.2235269448521704, + "grad_norm": 4.570787231943582, + "learning_rate": 3.312844885001588e-06, + "loss": 0.1128, + "step": 10604 + }, + { + "epoch": 2.2237366324176975, + "grad_norm": 3.4518611494357483, + "learning_rate": 3.3111618165983938e-06, + "loss": 0.1073, + "step": 10605 + }, + { + "epoch": 2.223946319983225, + "grad_norm": 4.767025805207989, + "learning_rate": 3.3094790910097573e-06, + "loss": 0.1815, + "step": 10606 + }, + { + "epoch": 2.2241560075487525, + "grad_norm": 4.5679070930550765, + "learning_rate": 3.3077967083219185e-06, + "loss": 0.1179, + "step": 10607 + }, + { + "epoch": 2.2243656951142796, + "grad_norm": 4.9873069294838, + "learning_rate": 3.3061146686211054e-06, + "loss": 0.1342, + "step": 10608 + }, + { + "epoch": 2.224575382679807, + "grad_norm": 3.4742804014925115, + "learning_rate": 3.3044329719935285e-06, + "loss": 0.1173, + "step": 10609 + }, + { + "epoch": 2.2247850702453347, + "grad_norm": 4.359742076413879, + "learning_rate": 3.302751618525375e-06, + "loss": 0.1705, + "step": 10610 + }, + { + "epoch": 2.2249947578108618, + "grad_norm": 4.19551151130385, + "learning_rate": 3.301070608302813e-06, + "loss": 0.1388, + "step": 10611 + }, + { + "epoch": 2.2252044453763893, + "grad_norm": 3.995780906942006, + "learning_rate": 3.2993899414120044e-06, + "loss": 0.1428, + "step": 10612 + }, + { + "epoch": 2.2254141329419164, + "grad_norm": 4.6427190042705035, + "learning_rate": 3.2977096179390832e-06, + "loss": 0.1374, + "step": 10613 + }, + { + "epoch": 2.225623820507444, + "grad_norm": 4.706071616326585, + "learning_rate": 3.2960296379701662e-06, + "loss": 0.1438, + "step": 10614 + }, + { + "epoch": 2.2258335080729714, + "grad_norm": 3.0932886108939606, + "learning_rate": 3.2943500015913564e-06, + "loss": 0.1142, + "step": 10615 + }, + { + "epoch": 2.2260431956384985, + "grad_norm": 4.460160031520024, + "learning_rate": 3.2926707088887433e-06, + "loss": 0.1673, + "step": 10616 + }, + { + "epoch": 2.226252883204026, + "grad_norm": 3.8306100276362125, + "learning_rate": 3.29099175994839e-06, + "loss": 0.136, + "step": 10617 + }, + { + "epoch": 2.2264625707695536, + "grad_norm": 4.628461826191929, + "learning_rate": 3.2893131548563406e-06, + "loss": 0.1868, + "step": 10618 + }, + { + "epoch": 2.2266722583350806, + "grad_norm": 4.758077852478928, + "learning_rate": 3.2876348936986357e-06, + "loss": 0.1503, + "step": 10619 + }, + { + "epoch": 2.226881945900608, + "grad_norm": 3.6317435033830887, + "learning_rate": 3.2859569765612798e-06, + "loss": 0.1525, + "step": 10620 + }, + { + "epoch": 2.2270916334661353, + "grad_norm": 4.278099658006646, + "learning_rate": 3.284279403530278e-06, + "loss": 0.1495, + "step": 10621 + }, + { + "epoch": 2.227301321031663, + "grad_norm": 4.312772193620942, + "learning_rate": 3.2826021746916047e-06, + "loss": 0.1593, + "step": 10622 + }, + { + "epoch": 2.2275110085971903, + "grad_norm": 4.6337431699371505, + "learning_rate": 3.280925290131217e-06, + "loss": 0.1677, + "step": 10623 + }, + { + "epoch": 2.2277206961627174, + "grad_norm": 3.6948679268208875, + "learning_rate": 3.279248749935062e-06, + "loss": 0.1423, + "step": 10624 + }, + { + "epoch": 2.227930383728245, + "grad_norm": 4.177971836722947, + "learning_rate": 3.277572554189069e-06, + "loss": 0.1435, + "step": 10625 + }, + { + "epoch": 2.2281400712937725, + "grad_norm": 4.301396948760756, + "learning_rate": 3.2758967029791412e-06, + "loss": 0.1412, + "step": 10626 + }, + { + "epoch": 2.2283497588592995, + "grad_norm": 3.4752779585961453, + "learning_rate": 3.2742211963911676e-06, + "loss": 0.1206, + "step": 10627 + }, + { + "epoch": 2.228559446424827, + "grad_norm": 5.306285696757676, + "learning_rate": 3.272546034511025e-06, + "loss": 0.1862, + "step": 10628 + }, + { + "epoch": 2.2287691339903546, + "grad_norm": 4.743314596531981, + "learning_rate": 3.270871217424567e-06, + "loss": 0.1337, + "step": 10629 + }, + { + "epoch": 2.2289788215558817, + "grad_norm": 3.4118717446034785, + "learning_rate": 3.2691967452176264e-06, + "loss": 0.1165, + "step": 10630 + }, + { + "epoch": 2.229188509121409, + "grad_norm": 4.1969270153552145, + "learning_rate": 3.26752261797603e-06, + "loss": 0.1285, + "step": 10631 + }, + { + "epoch": 2.2293981966869363, + "grad_norm": 5.0494130550202465, + "learning_rate": 3.2658488357855734e-06, + "loss": 0.1964, + "step": 10632 + }, + { + "epoch": 2.229607884252464, + "grad_norm": 4.371426353209292, + "learning_rate": 3.264175398732047e-06, + "loss": 0.1615, + "step": 10633 + }, + { + "epoch": 2.2298175718179913, + "grad_norm": 6.1462048238815745, + "learning_rate": 3.2625023069012128e-06, + "loss": 0.1795, + "step": 10634 + }, + { + "epoch": 2.2300272593835184, + "grad_norm": 4.283301703261908, + "learning_rate": 3.2608295603788197e-06, + "loss": 0.1465, + "step": 10635 + }, + { + "epoch": 2.230236946949046, + "grad_norm": 3.879640179953613, + "learning_rate": 3.2591571592505987e-06, + "loss": 0.121, + "step": 10636 + }, + { + "epoch": 2.2304466345145735, + "grad_norm": 4.589213017893935, + "learning_rate": 3.2574851036022692e-06, + "loss": 0.1549, + "step": 10637 + }, + { + "epoch": 2.2306563220801006, + "grad_norm": 3.506850279898489, + "learning_rate": 3.2558133935195213e-06, + "loss": 0.1075, + "step": 10638 + }, + { + "epoch": 2.230866009645628, + "grad_norm": 4.528415173616114, + "learning_rate": 3.254142029088031e-06, + "loss": 0.1488, + "step": 10639 + }, + { + "epoch": 2.231075697211155, + "grad_norm": 4.022609724380287, + "learning_rate": 3.2524710103934653e-06, + "loss": 0.1401, + "step": 10640 + }, + { + "epoch": 2.2312853847766827, + "grad_norm": 6.533516023039647, + "learning_rate": 3.2508003375214603e-06, + "loss": 0.2036, + "step": 10641 + }, + { + "epoch": 2.2314950723422102, + "grad_norm": 5.131158955943327, + "learning_rate": 3.2491300105576473e-06, + "loss": 0.1578, + "step": 10642 + }, + { + "epoch": 2.2317047599077373, + "grad_norm": 5.013241872507247, + "learning_rate": 3.2474600295876304e-06, + "loss": 0.1583, + "step": 10643 + }, + { + "epoch": 2.231914447473265, + "grad_norm": 5.7839247309075725, + "learning_rate": 3.2457903946969938e-06, + "loss": 0.1961, + "step": 10644 + }, + { + "epoch": 2.2321241350387924, + "grad_norm": 4.152706619445699, + "learning_rate": 3.2441211059713153e-06, + "loss": 0.1544, + "step": 10645 + }, + { + "epoch": 2.2323338226043195, + "grad_norm": 5.0144134919604015, + "learning_rate": 3.2424521634961535e-06, + "loss": 0.1394, + "step": 10646 + }, + { + "epoch": 2.232543510169847, + "grad_norm": 4.2678485023080945, + "learning_rate": 3.240783567357033e-06, + "loss": 0.1653, + "step": 10647 + }, + { + "epoch": 2.2327531977353745, + "grad_norm": 4.379278766162334, + "learning_rate": 3.2391153176394773e-06, + "loss": 0.1685, + "step": 10648 + }, + { + "epoch": 2.2329628853009016, + "grad_norm": 4.962994492594998, + "learning_rate": 3.2374474144289912e-06, + "loss": 0.1755, + "step": 10649 + }, + { + "epoch": 2.233172572866429, + "grad_norm": 3.531747805190829, + "learning_rate": 3.2357798578110545e-06, + "loss": 0.0953, + "step": 10650 + }, + { + "epoch": 2.233382260431956, + "grad_norm": 3.689540673813028, + "learning_rate": 3.2341126478711284e-06, + "loss": 0.1353, + "step": 10651 + }, + { + "epoch": 2.2335919479974837, + "grad_norm": 4.279537920929235, + "learning_rate": 3.2324457846946668e-06, + "loss": 0.1679, + "step": 10652 + }, + { + "epoch": 2.2338016355630113, + "grad_norm": 4.114438663815979, + "learning_rate": 3.2307792683670937e-06, + "loss": 0.1606, + "step": 10653 + }, + { + "epoch": 2.2340113231285383, + "grad_norm": 4.056885208687246, + "learning_rate": 3.229113098973826e-06, + "loss": 0.1572, + "step": 10654 + }, + { + "epoch": 2.234221010694066, + "grad_norm": 4.116459934672251, + "learning_rate": 3.2274472766002573e-06, + "loss": 0.1246, + "step": 10655 + }, + { + "epoch": 2.2344306982595934, + "grad_norm": 4.718193127325923, + "learning_rate": 3.225781801331758e-06, + "loss": 0.1309, + "step": 10656 + }, + { + "epoch": 2.2346403858251205, + "grad_norm": 3.687084362985051, + "learning_rate": 3.2241166732536898e-06, + "loss": 0.1391, + "step": 10657 + }, + { + "epoch": 2.234850073390648, + "grad_norm": 6.079697287451632, + "learning_rate": 3.2224518924513983e-06, + "loss": 0.1869, + "step": 10658 + }, + { + "epoch": 2.235059760956175, + "grad_norm": 3.749754526130742, + "learning_rate": 3.220787459010203e-06, + "loss": 0.1584, + "step": 10659 + }, + { + "epoch": 2.2352694485217026, + "grad_norm": 3.726670840675818, + "learning_rate": 3.2191233730154035e-06, + "loss": 0.1347, + "step": 10660 + }, + { + "epoch": 2.23547913608723, + "grad_norm": 3.8463145648327997, + "learning_rate": 3.217459634552297e-06, + "loss": 0.1429, + "step": 10661 + }, + { + "epoch": 2.2356888236527572, + "grad_norm": 3.6735594308711645, + "learning_rate": 3.215796243706143e-06, + "loss": 0.133, + "step": 10662 + }, + { + "epoch": 2.2358985112182848, + "grad_norm": 5.159506808179413, + "learning_rate": 3.214133200562203e-06, + "loss": 0.1444, + "step": 10663 + }, + { + "epoch": 2.2361081987838123, + "grad_norm": 4.251251132308287, + "learning_rate": 3.2124705052057047e-06, + "loss": 0.1237, + "step": 10664 + }, + { + "epoch": 2.2363178863493394, + "grad_norm": 4.128294289381434, + "learning_rate": 3.210808157721863e-06, + "loss": 0.1304, + "step": 10665 + }, + { + "epoch": 2.236527573914867, + "grad_norm": 3.9921345576603606, + "learning_rate": 3.209146158195877e-06, + "loss": 0.1681, + "step": 10666 + }, + { + "epoch": 2.2367372614803944, + "grad_norm": 3.380505274633623, + "learning_rate": 3.2074845067129367e-06, + "loss": 0.1082, + "step": 10667 + }, + { + "epoch": 2.2369469490459215, + "grad_norm": 5.5600830677914574, + "learning_rate": 3.2058232033581893e-06, + "loss": 0.1861, + "step": 10668 + }, + { + "epoch": 2.237156636611449, + "grad_norm": 3.9523289763903477, + "learning_rate": 3.2041622482167867e-06, + "loss": 0.1567, + "step": 10669 + }, + { + "epoch": 2.237366324176976, + "grad_norm": 3.765372948501272, + "learning_rate": 3.202501641373855e-06, + "loss": 0.1644, + "step": 10670 + }, + { + "epoch": 2.2375760117425036, + "grad_norm": 4.350471809551431, + "learning_rate": 3.2008413829145113e-06, + "loss": 0.1074, + "step": 10671 + }, + { + "epoch": 2.237785699308031, + "grad_norm": 4.640865153442866, + "learning_rate": 3.1991814729238325e-06, + "loss": 0.1933, + "step": 10672 + }, + { + "epoch": 2.2379953868735583, + "grad_norm": 3.9901887373618057, + "learning_rate": 3.1975219114868994e-06, + "loss": 0.1542, + "step": 10673 + }, + { + "epoch": 2.238205074439086, + "grad_norm": 3.4396244096590936, + "learning_rate": 3.195862698688769e-06, + "loss": 0.131, + "step": 10674 + }, + { + "epoch": 2.2384147620046133, + "grad_norm": 3.939392748844276, + "learning_rate": 3.1942038346144766e-06, + "loss": 0.167, + "step": 10675 + }, + { + "epoch": 2.2386244495701404, + "grad_norm": 4.83238479485785, + "learning_rate": 3.1925453193490387e-06, + "loss": 0.1666, + "step": 10676 + }, + { + "epoch": 2.238834137135668, + "grad_norm": 4.3705087586477855, + "learning_rate": 3.1908871529774623e-06, + "loss": 0.1569, + "step": 10677 + }, + { + "epoch": 2.239043824701195, + "grad_norm": 3.4111072186848514, + "learning_rate": 3.189229335584727e-06, + "loss": 0.1538, + "step": 10678 + }, + { + "epoch": 2.2392535122667225, + "grad_norm": 3.298364960160477, + "learning_rate": 3.187571867255803e-06, + "loss": 0.1087, + "step": 10679 + }, + { + "epoch": 2.23946319983225, + "grad_norm": 4.038776106475902, + "learning_rate": 3.185914748075638e-06, + "loss": 0.1476, + "step": 10680 + }, + { + "epoch": 2.239672887397777, + "grad_norm": 3.9246584661438995, + "learning_rate": 3.184257978129156e-06, + "loss": 0.1407, + "step": 10681 + }, + { + "epoch": 2.2398825749633047, + "grad_norm": 3.7005412097884887, + "learning_rate": 3.182601557501275e-06, + "loss": 0.0989, + "step": 10682 + }, + { + "epoch": 2.240092262528832, + "grad_norm": 4.557648386050514, + "learning_rate": 3.1809454862768908e-06, + "loss": 0.1641, + "step": 10683 + }, + { + "epoch": 2.2403019500943593, + "grad_norm": 3.5916031962341264, + "learning_rate": 3.1792897645408783e-06, + "loss": 0.1467, + "step": 10684 + }, + { + "epoch": 2.240511637659887, + "grad_norm": 3.225484030173698, + "learning_rate": 3.1776343923780915e-06, + "loss": 0.1346, + "step": 10685 + }, + { + "epoch": 2.2407213252254143, + "grad_norm": 3.618816861698131, + "learning_rate": 3.1759793698733776e-06, + "loss": 0.119, + "step": 10686 + }, + { + "epoch": 2.2409310127909414, + "grad_norm": 4.201344782604311, + "learning_rate": 3.1743246971115538e-06, + "loss": 0.1371, + "step": 10687 + }, + { + "epoch": 2.241140700356469, + "grad_norm": 4.535607229484524, + "learning_rate": 3.172670374177431e-06, + "loss": 0.1626, + "step": 10688 + }, + { + "epoch": 2.241350387921996, + "grad_norm": 4.2331814957837315, + "learning_rate": 3.171016401155793e-06, + "loss": 0.1611, + "step": 10689 + }, + { + "epoch": 2.2415600754875236, + "grad_norm": 3.5263875962958005, + "learning_rate": 3.169362778131406e-06, + "loss": 0.1291, + "step": 10690 + }, + { + "epoch": 2.241769763053051, + "grad_norm": 4.516105193828841, + "learning_rate": 3.167709505189027e-06, + "loss": 0.1552, + "step": 10691 + }, + { + "epoch": 2.241979450618578, + "grad_norm": 3.6479127601514203, + "learning_rate": 3.166056582413385e-06, + "loss": 0.1224, + "step": 10692 + }, + { + "epoch": 2.2421891381841057, + "grad_norm": 4.652470713467641, + "learning_rate": 3.1644040098891938e-06, + "loss": 0.1397, + "step": 10693 + }, + { + "epoch": 2.2423988257496332, + "grad_norm": 4.04336450339858, + "learning_rate": 3.1627517877011516e-06, + "loss": 0.1552, + "step": 10694 + }, + { + "epoch": 2.2426085133151603, + "grad_norm": 4.304574538521102, + "learning_rate": 3.1610999159339427e-06, + "loss": 0.143, + "step": 10695 + }, + { + "epoch": 2.242818200880688, + "grad_norm": 4.5750593121446155, + "learning_rate": 3.1594483946722254e-06, + "loss": 0.1919, + "step": 10696 + }, + { + "epoch": 2.243027888446215, + "grad_norm": 3.8149046175335286, + "learning_rate": 3.1577972240006384e-06, + "loss": 0.1427, + "step": 10697 + }, + { + "epoch": 2.2432375760117425, + "grad_norm": 3.4402778063016735, + "learning_rate": 3.1561464040038137e-06, + "loss": 0.1197, + "step": 10698 + }, + { + "epoch": 2.24344726357727, + "grad_norm": 5.4428125420926445, + "learning_rate": 3.154495934766353e-06, + "loss": 0.1883, + "step": 10699 + }, + { + "epoch": 2.243656951142797, + "grad_norm": 4.041442735823655, + "learning_rate": 3.1528458163728515e-06, + "loss": 0.1231, + "step": 10700 + }, + { + "epoch": 2.2438666387083246, + "grad_norm": 4.598540097774539, + "learning_rate": 3.1511960489078775e-06, + "loss": 0.1979, + "step": 10701 + }, + { + "epoch": 2.244076326273852, + "grad_norm": 3.6264224777376746, + "learning_rate": 3.149546632455982e-06, + "loss": 0.1357, + "step": 10702 + }, + { + "epoch": 2.244286013839379, + "grad_norm": 4.407547328787699, + "learning_rate": 3.147897567101703e-06, + "loss": 0.1559, + "step": 10703 + }, + { + "epoch": 2.2444957014049067, + "grad_norm": 3.905412825519575, + "learning_rate": 3.146248852929561e-06, + "loss": 0.1309, + "step": 10704 + }, + { + "epoch": 2.2447053889704343, + "grad_norm": 4.681106883892964, + "learning_rate": 3.1446004900240534e-06, + "loss": 0.112, + "step": 10705 + }, + { + "epoch": 2.2449150765359613, + "grad_norm": 2.833973808978711, + "learning_rate": 3.1429524784696565e-06, + "loss": 0.1246, + "step": 10706 + }, + { + "epoch": 2.245124764101489, + "grad_norm": 3.9934521939892926, + "learning_rate": 3.141304818350841e-06, + "loss": 0.1485, + "step": 10707 + }, + { + "epoch": 2.245334451667016, + "grad_norm": 4.741158715735433, + "learning_rate": 3.1396575097520498e-06, + "loss": 0.1588, + "step": 10708 + }, + { + "epoch": 2.2455441392325435, + "grad_norm": 3.5752646485221615, + "learning_rate": 3.1380105527577074e-06, + "loss": 0.1109, + "step": 10709 + }, + { + "epoch": 2.245753826798071, + "grad_norm": 5.028553363506122, + "learning_rate": 3.136363947452228e-06, + "loss": 0.1605, + "step": 10710 + }, + { + "epoch": 2.245963514363598, + "grad_norm": 4.104847486031501, + "learning_rate": 3.1347176939199984e-06, + "loss": 0.1464, + "step": 10711 + }, + { + "epoch": 2.2461732019291256, + "grad_norm": 3.374350047140469, + "learning_rate": 3.1330717922453967e-06, + "loss": 0.1121, + "step": 10712 + }, + { + "epoch": 2.246382889494653, + "grad_norm": 3.392144764759944, + "learning_rate": 3.131426242512775e-06, + "loss": 0.1469, + "step": 10713 + }, + { + "epoch": 2.2465925770601802, + "grad_norm": 3.2207502618443646, + "learning_rate": 3.1297810448064693e-06, + "loss": 0.113, + "step": 10714 + }, + { + "epoch": 2.2468022646257078, + "grad_norm": 5.405880838001209, + "learning_rate": 3.1281361992108005e-06, + "loss": 0.2209, + "step": 10715 + }, + { + "epoch": 2.247011952191235, + "grad_norm": 4.680217965403063, + "learning_rate": 3.126491705810073e-06, + "loss": 0.168, + "step": 10716 + }, + { + "epoch": 2.2472216397567624, + "grad_norm": 4.6834422657508, + "learning_rate": 3.1248475646885678e-06, + "loss": 0.1571, + "step": 10717 + }, + { + "epoch": 2.24743132732229, + "grad_norm": 5.046100292557407, + "learning_rate": 3.1232037759305456e-06, + "loss": 0.2107, + "step": 10718 + }, + { + "epoch": 2.247641014887817, + "grad_norm": 4.864552236797289, + "learning_rate": 3.1215603396202608e-06, + "loss": 0.1462, + "step": 10719 + }, + { + "epoch": 2.2478507024533445, + "grad_norm": 7.729762780229888, + "learning_rate": 3.1199172558419342e-06, + "loss": 0.1795, + "step": 10720 + }, + { + "epoch": 2.248060390018872, + "grad_norm": 5.103076269261817, + "learning_rate": 3.1182745246797852e-06, + "loss": 0.1758, + "step": 10721 + }, + { + "epoch": 2.248270077584399, + "grad_norm": 3.3815289499040335, + "learning_rate": 3.116632146218003e-06, + "loss": 0.1219, + "step": 10722 + }, + { + "epoch": 2.2484797651499266, + "grad_norm": 4.730575898742255, + "learning_rate": 3.1149901205407572e-06, + "loss": 0.159, + "step": 10723 + }, + { + "epoch": 2.248689452715454, + "grad_norm": 3.56039129046489, + "learning_rate": 3.113348447732211e-06, + "loss": 0.1468, + "step": 10724 + }, + { + "epoch": 2.2488991402809813, + "grad_norm": 3.4961659137641083, + "learning_rate": 3.1117071278765066e-06, + "loss": 0.1414, + "step": 10725 + }, + { + "epoch": 2.249108827846509, + "grad_norm": 5.067809440016506, + "learning_rate": 3.110066161057752e-06, + "loss": 0.1593, + "step": 10726 + }, + { + "epoch": 2.249318515412036, + "grad_norm": 4.653335047595459, + "learning_rate": 3.108425547360058e-06, + "loss": 0.1496, + "step": 10727 + }, + { + "epoch": 2.2495282029775634, + "grad_norm": 5.7321469455472, + "learning_rate": 3.10678528686751e-06, + "loss": 0.1499, + "step": 10728 + }, + { + "epoch": 2.249737890543091, + "grad_norm": 4.89149136960824, + "learning_rate": 3.1051453796641718e-06, + "loss": 0.1487, + "step": 10729 + }, + { + "epoch": 2.249947578108618, + "grad_norm": 5.173151876393134, + "learning_rate": 3.1035058258340876e-06, + "loss": 0.1947, + "step": 10730 + }, + { + "epoch": 2.2501572656741455, + "grad_norm": 4.387382183707479, + "learning_rate": 3.1018666254612927e-06, + "loss": 0.1841, + "step": 10731 + }, + { + "epoch": 2.250366953239673, + "grad_norm": 4.4664996662933385, + "learning_rate": 3.1002277786297996e-06, + "loss": 0.1711, + "step": 10732 + }, + { + "epoch": 2.2505766408052, + "grad_norm": 4.198477045803032, + "learning_rate": 3.0985892854235998e-06, + "loss": 0.145, + "step": 10733 + }, + { + "epoch": 2.2507863283707277, + "grad_norm": 5.16231479788112, + "learning_rate": 3.0969511459266666e-06, + "loss": 0.1623, + "step": 10734 + }, + { + "epoch": 2.2509960159362548, + "grad_norm": 4.920457625285519, + "learning_rate": 3.0953133602229624e-06, + "loss": 0.1661, + "step": 10735 + }, + { + "epoch": 2.2512057035017823, + "grad_norm": 4.352948071565112, + "learning_rate": 3.0936759283964204e-06, + "loss": 0.118, + "step": 10736 + }, + { + "epoch": 2.25141539106731, + "grad_norm": 3.661842826933915, + "learning_rate": 3.0920388505309697e-06, + "loss": 0.143, + "step": 10737 + }, + { + "epoch": 2.251625078632837, + "grad_norm": 3.448088734755034, + "learning_rate": 3.090402126710509e-06, + "loss": 0.122, + "step": 10738 + }, + { + "epoch": 2.2518347661983644, + "grad_norm": 5.049641111539313, + "learning_rate": 3.08876575701892e-06, + "loss": 0.1613, + "step": 10739 + }, + { + "epoch": 2.252044453763892, + "grad_norm": 5.158595952181279, + "learning_rate": 3.0871297415400727e-06, + "loss": 0.183, + "step": 10740 + }, + { + "epoch": 2.252254141329419, + "grad_norm": 4.959679163330671, + "learning_rate": 3.0854940803578194e-06, + "loss": 0.1631, + "step": 10741 + }, + { + "epoch": 2.2524638288949466, + "grad_norm": 3.9956075802820417, + "learning_rate": 3.0838587735559866e-06, + "loss": 0.1463, + "step": 10742 + }, + { + "epoch": 2.252673516460474, + "grad_norm": 4.504463682186625, + "learning_rate": 3.082223821218384e-06, + "loss": 0.1783, + "step": 10743 + }, + { + "epoch": 2.252883204026001, + "grad_norm": 4.877590025377896, + "learning_rate": 3.0805892234288127e-06, + "loss": 0.1578, + "step": 10744 + }, + { + "epoch": 2.2530928915915287, + "grad_norm": 4.2558157596042605, + "learning_rate": 3.0789549802710406e-06, + "loss": 0.1532, + "step": 10745 + }, + { + "epoch": 2.2533025791570562, + "grad_norm": 3.861508836868627, + "learning_rate": 3.0773210918288333e-06, + "loss": 0.1202, + "step": 10746 + }, + { + "epoch": 2.2535122667225833, + "grad_norm": 4.816384688056086, + "learning_rate": 3.0756875581859267e-06, + "loss": 0.1544, + "step": 10747 + }, + { + "epoch": 2.253721954288111, + "grad_norm": 4.612391747471403, + "learning_rate": 3.0740543794260403e-06, + "loss": 0.1665, + "step": 10748 + }, + { + "epoch": 2.253931641853638, + "grad_norm": 3.787564757170909, + "learning_rate": 3.0724215556328784e-06, + "loss": 0.128, + "step": 10749 + }, + { + "epoch": 2.2541413294191655, + "grad_norm": 5.353320700650199, + "learning_rate": 3.0707890868901337e-06, + "loss": 0.1346, + "step": 10750 + }, + { + "epoch": 2.254351016984693, + "grad_norm": 4.311020272899277, + "learning_rate": 3.0691569732814596e-06, + "loss": 0.1513, + "step": 10751 + }, + { + "epoch": 2.25456070455022, + "grad_norm": 4.165035291851166, + "learning_rate": 3.067525214890512e-06, + "loss": 0.1436, + "step": 10752 + }, + { + "epoch": 2.2547703921157476, + "grad_norm": 5.877515509216618, + "learning_rate": 3.065893811800924e-06, + "loss": 0.1681, + "step": 10753 + }, + { + "epoch": 2.2549800796812747, + "grad_norm": 3.8060323708536203, + "learning_rate": 3.064262764096304e-06, + "loss": 0.1377, + "step": 10754 + }, + { + "epoch": 2.255189767246802, + "grad_norm": 4.768631616737858, + "learning_rate": 3.0626320718602442e-06, + "loss": 0.1518, + "step": 10755 + }, + { + "epoch": 2.2553994548123297, + "grad_norm": 4.540114404197204, + "learning_rate": 3.0610017351763265e-06, + "loss": 0.1389, + "step": 10756 + }, + { + "epoch": 2.255609142377857, + "grad_norm": 5.28016184884163, + "learning_rate": 3.059371754128102e-06, + "loss": 0.1758, + "step": 10757 + }, + { + "epoch": 2.2558188299433843, + "grad_norm": 4.282707146935706, + "learning_rate": 3.057742128799115e-06, + "loss": 0.1199, + "step": 10758 + }, + { + "epoch": 2.256028517508912, + "grad_norm": 3.502794799874295, + "learning_rate": 3.056112859272886e-06, + "loss": 0.1322, + "step": 10759 + }, + { + "epoch": 2.256238205074439, + "grad_norm": 3.8887861468310607, + "learning_rate": 3.054483945632912e-06, + "loss": 0.1267, + "step": 10760 + }, + { + "epoch": 2.2564478926399665, + "grad_norm": 4.286909632059399, + "learning_rate": 3.052855387962683e-06, + "loss": 0.1456, + "step": 10761 + }, + { + "epoch": 2.256657580205494, + "grad_norm": 4.405800026844709, + "learning_rate": 3.0512271863456676e-06, + "loss": 0.1522, + "step": 10762 + }, + { + "epoch": 2.256867267771021, + "grad_norm": 4.02597406666974, + "learning_rate": 3.049599340865311e-06, + "loss": 0.1418, + "step": 10763 + }, + { + "epoch": 2.2570769553365486, + "grad_norm": 4.082163738537335, + "learning_rate": 3.0479718516050406e-06, + "loss": 0.1658, + "step": 10764 + }, + { + "epoch": 2.257286642902076, + "grad_norm": 4.852521962642193, + "learning_rate": 3.046344718648272e-06, + "loss": 0.1808, + "step": 10765 + }, + { + "epoch": 2.2574963304676032, + "grad_norm": 3.609655604986204, + "learning_rate": 3.0447179420783945e-06, + "loss": 0.1043, + "step": 10766 + }, + { + "epoch": 2.2577060180331308, + "grad_norm": 3.376504162623587, + "learning_rate": 3.043091521978789e-06, + "loss": 0.1037, + "step": 10767 + }, + { + "epoch": 2.257915705598658, + "grad_norm": 4.18927843690016, + "learning_rate": 3.0414654584328096e-06, + "loss": 0.1472, + "step": 10768 + }, + { + "epoch": 2.2581253931641854, + "grad_norm": 4.142521652010309, + "learning_rate": 3.0398397515237908e-06, + "loss": 0.1471, + "step": 10769 + }, + { + "epoch": 2.258335080729713, + "grad_norm": 3.760157250159119, + "learning_rate": 3.03821440133506e-06, + "loss": 0.0864, + "step": 10770 + }, + { + "epoch": 2.25854476829524, + "grad_norm": 3.5359133254850246, + "learning_rate": 3.0365894079499158e-06, + "loss": 0.0938, + "step": 10771 + }, + { + "epoch": 2.2587544558607675, + "grad_norm": 4.259580070305667, + "learning_rate": 3.0349647714516396e-06, + "loss": 0.1704, + "step": 10772 + }, + { + "epoch": 2.2589641434262946, + "grad_norm": 3.770184737223212, + "learning_rate": 3.033340491923498e-06, + "loss": 0.1345, + "step": 10773 + }, + { + "epoch": 2.259173830991822, + "grad_norm": 3.8208175270030473, + "learning_rate": 3.0317165694487415e-06, + "loss": 0.1216, + "step": 10774 + }, + { + "epoch": 2.2593835185573496, + "grad_norm": 3.561508016419555, + "learning_rate": 3.030093004110598e-06, + "loss": 0.124, + "step": 10775 + }, + { + "epoch": 2.2595932061228767, + "grad_norm": 3.9073214899595934, + "learning_rate": 3.028469795992274e-06, + "loss": 0.1443, + "step": 10776 + }, + { + "epoch": 2.2598028936884043, + "grad_norm": 3.5160419215571017, + "learning_rate": 3.0268469451769654e-06, + "loss": 0.123, + "step": 10777 + }, + { + "epoch": 2.260012581253932, + "grad_norm": 5.5583007611969855, + "learning_rate": 3.0252244517478447e-06, + "loss": 0.1534, + "step": 10778 + }, + { + "epoch": 2.260222268819459, + "grad_norm": 3.2243622002638257, + "learning_rate": 3.023602315788069e-06, + "loss": 0.1185, + "step": 10779 + }, + { + "epoch": 2.2604319563849864, + "grad_norm": 3.5558776321170824, + "learning_rate": 3.0219805373807752e-06, + "loss": 0.1256, + "step": 10780 + }, + { + "epoch": 2.260641643950514, + "grad_norm": 5.852776329887172, + "learning_rate": 3.0203591166090796e-06, + "loss": 0.1665, + "step": 10781 + }, + { + "epoch": 2.260851331516041, + "grad_norm": 3.9080746374356083, + "learning_rate": 3.0187380535560838e-06, + "loss": 0.1147, + "step": 10782 + }, + { + "epoch": 2.2610610190815685, + "grad_norm": 3.7383703280491725, + "learning_rate": 3.0171173483048744e-06, + "loss": 0.1369, + "step": 10783 + }, + { + "epoch": 2.261270706647096, + "grad_norm": 4.2241462794518645, + "learning_rate": 3.015497000938512e-06, + "loss": 0.1618, + "step": 10784 + }, + { + "epoch": 2.261480394212623, + "grad_norm": 4.386034082502578, + "learning_rate": 3.0138770115400396e-06, + "loss": 0.1556, + "step": 10785 + }, + { + "epoch": 2.2616900817781507, + "grad_norm": 4.389248930771413, + "learning_rate": 3.0122573801924903e-06, + "loss": 0.1648, + "step": 10786 + }, + { + "epoch": 2.2618997693436778, + "grad_norm": 4.762619257881491, + "learning_rate": 3.0106381069788694e-06, + "loss": 0.1467, + "step": 10787 + }, + { + "epoch": 2.2621094569092053, + "grad_norm": 6.351787661682319, + "learning_rate": 3.009019191982164e-06, + "loss": 0.1403, + "step": 10788 + }, + { + "epoch": 2.262319144474733, + "grad_norm": 5.135216116777504, + "learning_rate": 3.0074006352853557e-06, + "loss": 0.1674, + "step": 10789 + }, + { + "epoch": 2.26252883204026, + "grad_norm": 4.132108517975747, + "learning_rate": 3.0057824369713883e-06, + "loss": 0.1427, + "step": 10790 + }, + { + "epoch": 2.2627385196057874, + "grad_norm": 3.653071605095956, + "learning_rate": 3.004164597123205e-06, + "loss": 0.1273, + "step": 10791 + }, + { + "epoch": 2.2629482071713145, + "grad_norm": 4.958798621005177, + "learning_rate": 3.0025471158237162e-06, + "loss": 0.1599, + "step": 10792 + }, + { + "epoch": 2.263157894736842, + "grad_norm": 3.711438651627813, + "learning_rate": 3.000929993155829e-06, + "loss": 0.1395, + "step": 10793 + }, + { + "epoch": 2.2633675823023696, + "grad_norm": 4.956851699239931, + "learning_rate": 2.999313229202414e-06, + "loss": 0.1815, + "step": 10794 + }, + { + "epoch": 2.2635772698678966, + "grad_norm": 3.4254172346776164, + "learning_rate": 2.997696824046342e-06, + "loss": 0.1106, + "step": 10795 + }, + { + "epoch": 2.263786957433424, + "grad_norm": 4.063648494813445, + "learning_rate": 2.9960807777704527e-06, + "loss": 0.1296, + "step": 10796 + }, + { + "epoch": 2.2639966449989517, + "grad_norm": 3.6339727701038536, + "learning_rate": 2.9944650904575677e-06, + "loss": 0.1331, + "step": 10797 + }, + { + "epoch": 2.264206332564479, + "grad_norm": 4.242863508473164, + "learning_rate": 2.992849762190497e-06, + "loss": 0.1249, + "step": 10798 + }, + { + "epoch": 2.2644160201300063, + "grad_norm": 4.119778938292382, + "learning_rate": 2.9912347930520334e-06, + "loss": 0.1026, + "step": 10799 + }, + { + "epoch": 2.264625707695534, + "grad_norm": 4.784281318046052, + "learning_rate": 2.9896201831249426e-06, + "loss": 0.177, + "step": 10800 + }, + { + "epoch": 2.264835395261061, + "grad_norm": 3.519635691003411, + "learning_rate": 2.988005932491973e-06, + "loss": 0.1316, + "step": 10801 + }, + { + "epoch": 2.2650450828265885, + "grad_norm": 3.903676600011058, + "learning_rate": 2.9863920412358637e-06, + "loss": 0.1623, + "step": 10802 + }, + { + "epoch": 2.265254770392116, + "grad_norm": 3.934792121922775, + "learning_rate": 2.9847785094393243e-06, + "loss": 0.1668, + "step": 10803 + }, + { + "epoch": 2.265464457957643, + "grad_norm": 3.5384464209255544, + "learning_rate": 2.983165337185057e-06, + "loss": 0.116, + "step": 10804 + }, + { + "epoch": 2.2656741455231706, + "grad_norm": 3.220997227853989, + "learning_rate": 2.9815525245557355e-06, + "loss": 0.1003, + "step": 10805 + }, + { + "epoch": 2.2658838330886977, + "grad_norm": 4.010711161500659, + "learning_rate": 2.9799400716340167e-06, + "loss": 0.1486, + "step": 10806 + }, + { + "epoch": 2.266093520654225, + "grad_norm": 3.724318805681504, + "learning_rate": 2.9783279785025455e-06, + "loss": 0.1271, + "step": 10807 + }, + { + "epoch": 2.2663032082197527, + "grad_norm": 5.9147519830319855, + "learning_rate": 2.976716245243949e-06, + "loss": 0.1656, + "step": 10808 + }, + { + "epoch": 2.26651289578528, + "grad_norm": 3.625110983113078, + "learning_rate": 2.9751048719408216e-06, + "loss": 0.1165, + "step": 10809 + }, + { + "epoch": 2.2667225833508073, + "grad_norm": 4.343066883595352, + "learning_rate": 2.9734938586757522e-06, + "loss": 0.1723, + "step": 10810 + }, + { + "epoch": 2.2669322709163344, + "grad_norm": 4.519290255556346, + "learning_rate": 2.971883205531313e-06, + "loss": 0.1806, + "step": 10811 + }, + { + "epoch": 2.267141958481862, + "grad_norm": 3.992865376113029, + "learning_rate": 2.9702729125900497e-06, + "loss": 0.1434, + "step": 10812 + }, + { + "epoch": 2.2673516460473895, + "grad_norm": 5.190167415771511, + "learning_rate": 2.968662979934488e-06, + "loss": 0.1837, + "step": 10813 + }, + { + "epoch": 2.2675613336129166, + "grad_norm": 3.680736822779035, + "learning_rate": 2.967053407647148e-06, + "loss": 0.1349, + "step": 10814 + }, + { + "epoch": 2.267771021178444, + "grad_norm": 4.575894058145139, + "learning_rate": 2.965444195810515e-06, + "loss": 0.1569, + "step": 10815 + }, + { + "epoch": 2.2679807087439716, + "grad_norm": 3.6588259245477124, + "learning_rate": 2.963835344507071e-06, + "loss": 0.1338, + "step": 10816 + }, + { + "epoch": 2.2681903963094987, + "grad_norm": 4.454585144055542, + "learning_rate": 2.96222685381927e-06, + "loss": 0.1416, + "step": 10817 + }, + { + "epoch": 2.2684000838750262, + "grad_norm": 4.9222067865871235, + "learning_rate": 2.9606187238295446e-06, + "loss": 0.1704, + "step": 10818 + }, + { + "epoch": 2.2686097714405538, + "grad_norm": 5.878330757799028, + "learning_rate": 2.9590109546203196e-06, + "loss": 0.1969, + "step": 10819 + }, + { + "epoch": 2.268819459006081, + "grad_norm": 4.374341316323395, + "learning_rate": 2.957403546273998e-06, + "loss": 0.1374, + "step": 10820 + }, + { + "epoch": 2.2690291465716084, + "grad_norm": 3.616805945137954, + "learning_rate": 2.9557964988729606e-06, + "loss": 0.1512, + "step": 10821 + }, + { + "epoch": 2.269238834137136, + "grad_norm": 4.158588774981077, + "learning_rate": 2.954189812499565e-06, + "loss": 0.132, + "step": 10822 + }, + { + "epoch": 2.269448521702663, + "grad_norm": 3.88399857024793, + "learning_rate": 2.952583487236167e-06, + "loss": 0.1399, + "step": 10823 + }, + { + "epoch": 2.2696582092681905, + "grad_norm": 4.035728487294192, + "learning_rate": 2.950977523165084e-06, + "loss": 0.1456, + "step": 10824 + }, + { + "epoch": 2.2698678968337176, + "grad_norm": 4.303443582495658, + "learning_rate": 2.949371920368632e-06, + "loss": 0.1035, + "step": 10825 + }, + { + "epoch": 2.270077584399245, + "grad_norm": 5.761793592742784, + "learning_rate": 2.947766678929098e-06, + "loss": 0.2102, + "step": 10826 + }, + { + "epoch": 2.2702872719647726, + "grad_norm": 3.7182545241688922, + "learning_rate": 2.9461617989287504e-06, + "loss": 0.13, + "step": 10827 + }, + { + "epoch": 2.2704969595302997, + "grad_norm": 4.19171275587322, + "learning_rate": 2.9445572804498446e-06, + "loss": 0.1367, + "step": 10828 + }, + { + "epoch": 2.2707066470958273, + "grad_norm": 3.848244528166398, + "learning_rate": 2.942953123574621e-06, + "loss": 0.1292, + "step": 10829 + }, + { + "epoch": 2.2709163346613543, + "grad_norm": 4.2612739434881615, + "learning_rate": 2.9413493283852847e-06, + "loss": 0.1423, + "step": 10830 + }, + { + "epoch": 2.271126022226882, + "grad_norm": 3.998077519648316, + "learning_rate": 2.939745894964037e-06, + "loss": 0.1322, + "step": 10831 + }, + { + "epoch": 2.2713357097924094, + "grad_norm": 3.9673085609332546, + "learning_rate": 2.938142823393061e-06, + "loss": 0.1237, + "step": 10832 + }, + { + "epoch": 2.2715453973579365, + "grad_norm": 4.70776790576569, + "learning_rate": 2.9365401137545123e-06, + "loss": 0.1692, + "step": 10833 + }, + { + "epoch": 2.271755084923464, + "grad_norm": 4.793834441964729, + "learning_rate": 2.9349377661305325e-06, + "loss": 0.1757, + "step": 10834 + }, + { + "epoch": 2.2719647724889915, + "grad_norm": 4.66836977197758, + "learning_rate": 2.933335780603248e-06, + "loss": 0.187, + "step": 10835 + }, + { + "epoch": 2.2721744600545186, + "grad_norm": 4.285000965839606, + "learning_rate": 2.9317341572547574e-06, + "loss": 0.1619, + "step": 10836 + }, + { + "epoch": 2.272384147620046, + "grad_norm": 5.882703803553985, + "learning_rate": 2.9301328961671548e-06, + "loss": 0.1673, + "step": 10837 + }, + { + "epoch": 2.2725938351855737, + "grad_norm": 4.07008667667668, + "learning_rate": 2.9285319974225033e-06, + "loss": 0.1508, + "step": 10838 + }, + { + "epoch": 2.2728035227511008, + "grad_norm": 3.184363690106782, + "learning_rate": 2.926931461102849e-06, + "loss": 0.1246, + "step": 10839 + }, + { + "epoch": 2.2730132103166283, + "grad_norm": 4.259076507754931, + "learning_rate": 2.925331287290224e-06, + "loss": 0.1315, + "step": 10840 + }, + { + "epoch": 2.273222897882156, + "grad_norm": 4.39751364669792, + "learning_rate": 2.9237314760666436e-06, + "loss": 0.1525, + "step": 10841 + }, + { + "epoch": 2.273432585447683, + "grad_norm": 4.845012199337564, + "learning_rate": 2.9221320275140996e-06, + "loss": 0.1809, + "step": 10842 + }, + { + "epoch": 2.2736422730132104, + "grad_norm": 3.873528490021121, + "learning_rate": 2.920532941714561e-06, + "loss": 0.133, + "step": 10843 + }, + { + "epoch": 2.2738519605787375, + "grad_norm": 4.602904997656741, + "learning_rate": 2.9189342187499913e-06, + "loss": 0.19, + "step": 10844 + }, + { + "epoch": 2.274061648144265, + "grad_norm": 4.656508198162215, + "learning_rate": 2.9173358587023215e-06, + "loss": 0.1515, + "step": 10845 + }, + { + "epoch": 2.2742713357097926, + "grad_norm": 3.511222136721952, + "learning_rate": 2.915737861653476e-06, + "loss": 0.1267, + "step": 10846 + }, + { + "epoch": 2.2744810232753196, + "grad_norm": 4.887492108860774, + "learning_rate": 2.9141402276853527e-06, + "loss": 0.1108, + "step": 10847 + }, + { + "epoch": 2.274690710840847, + "grad_norm": 4.107942201418428, + "learning_rate": 2.912542956879828e-06, + "loss": 0.1612, + "step": 10848 + }, + { + "epoch": 2.2749003984063743, + "grad_norm": 4.382246120681435, + "learning_rate": 2.9109460493187715e-06, + "loss": 0.1592, + "step": 10849 + }, + { + "epoch": 2.275110085971902, + "grad_norm": 4.689642117714378, + "learning_rate": 2.909349505084029e-06, + "loss": 0.1746, + "step": 10850 + }, + { + "epoch": 2.2753197735374293, + "grad_norm": 3.531439313650851, + "learning_rate": 2.9077533242574184e-06, + "loss": 0.1405, + "step": 10851 + }, + { + "epoch": 2.2755294611029564, + "grad_norm": 3.7598197933109607, + "learning_rate": 2.9061575069207494e-06, + "loss": 0.1325, + "step": 10852 + }, + { + "epoch": 2.275739148668484, + "grad_norm": 4.357492588350868, + "learning_rate": 2.904562053155816e-06, + "loss": 0.1399, + "step": 10853 + }, + { + "epoch": 2.2759488362340115, + "grad_norm": 3.959555353262392, + "learning_rate": 2.9029669630443834e-06, + "loss": 0.1622, + "step": 10854 + }, + { + "epoch": 2.2761585237995385, + "grad_norm": 3.408510224420956, + "learning_rate": 2.9013722366681997e-06, + "loss": 0.1034, + "step": 10855 + }, + { + "epoch": 2.276368211365066, + "grad_norm": 4.545920661949794, + "learning_rate": 2.899777874109001e-06, + "loss": 0.1474, + "step": 10856 + }, + { + "epoch": 2.2765778989305936, + "grad_norm": 3.6360900411770922, + "learning_rate": 2.8981838754485036e-06, + "loss": 0.143, + "step": 10857 + }, + { + "epoch": 2.2767875864961207, + "grad_norm": 4.126495728531195, + "learning_rate": 2.896590240768401e-06, + "loss": 0.1315, + "step": 10858 + }, + { + "epoch": 2.276997274061648, + "grad_norm": 6.605061217475966, + "learning_rate": 2.8949969701503653e-06, + "loss": 0.1436, + "step": 10859 + }, + { + "epoch": 2.2772069616271757, + "grad_norm": 4.371259412511121, + "learning_rate": 2.8934040636760608e-06, + "loss": 0.1517, + "step": 10860 + }, + { + "epoch": 2.277416649192703, + "grad_norm": 4.684086543430655, + "learning_rate": 2.891811521427121e-06, + "loss": 0.1738, + "step": 10861 + }, + { + "epoch": 2.2776263367582303, + "grad_norm": 4.905707287726102, + "learning_rate": 2.8902193434851723e-06, + "loss": 0.1439, + "step": 10862 + }, + { + "epoch": 2.2778360243237574, + "grad_norm": 3.8266854917527926, + "learning_rate": 2.8886275299318133e-06, + "loss": 0.1527, + "step": 10863 + }, + { + "epoch": 2.278045711889285, + "grad_norm": 3.872439496742937, + "learning_rate": 2.887036080848624e-06, + "loss": 0.1578, + "step": 10864 + }, + { + "epoch": 2.2782553994548125, + "grad_norm": 3.944525539759928, + "learning_rate": 2.8854449963171737e-06, + "loss": 0.1522, + "step": 10865 + }, + { + "epoch": 2.2784650870203396, + "grad_norm": 5.640306986055992, + "learning_rate": 2.8838542764190113e-06, + "loss": 0.2121, + "step": 10866 + }, + { + "epoch": 2.278674774585867, + "grad_norm": 4.2313843551486725, + "learning_rate": 2.882263921235655e-06, + "loss": 0.136, + "step": 10867 + }, + { + "epoch": 2.278884462151394, + "grad_norm": 3.734333618186167, + "learning_rate": 2.8806739308486164e-06, + "loss": 0.1337, + "step": 10868 + }, + { + "epoch": 2.2790941497169217, + "grad_norm": 4.5720340818010685, + "learning_rate": 2.8790843053393915e-06, + "loss": 0.1497, + "step": 10869 + }, + { + "epoch": 2.2793038372824492, + "grad_norm": 5.102434722344642, + "learning_rate": 2.8774950447894455e-06, + "loss": 0.196, + "step": 10870 + }, + { + "epoch": 2.2795135248479763, + "grad_norm": 3.4647175461219244, + "learning_rate": 2.8759061492802286e-06, + "loss": 0.1395, + "step": 10871 + }, + { + "epoch": 2.279723212413504, + "grad_norm": 3.8137786472061084, + "learning_rate": 2.874317618893181e-06, + "loss": 0.1335, + "step": 10872 + }, + { + "epoch": 2.2799328999790314, + "grad_norm": 3.0033002310764525, + "learning_rate": 2.8727294537097107e-06, + "loss": 0.1053, + "step": 10873 + }, + { + "epoch": 2.2801425875445585, + "grad_norm": 3.4224330276172394, + "learning_rate": 2.8711416538112214e-06, + "loss": 0.1314, + "step": 10874 + }, + { + "epoch": 2.280352275110086, + "grad_norm": 4.411927999008102, + "learning_rate": 2.869554219279086e-06, + "loss": 0.1156, + "step": 10875 + }, + { + "epoch": 2.2805619626756135, + "grad_norm": 4.815637932898665, + "learning_rate": 2.8679671501946617e-06, + "loss": 0.1665, + "step": 10876 + }, + { + "epoch": 2.2807716502411406, + "grad_norm": 4.219286958201543, + "learning_rate": 2.86638044663929e-06, + "loss": 0.1426, + "step": 10877 + }, + { + "epoch": 2.280981337806668, + "grad_norm": 3.7153800293720622, + "learning_rate": 2.8647941086942967e-06, + "loss": 0.1346, + "step": 10878 + }, + { + "epoch": 2.2811910253721956, + "grad_norm": 4.4891993870687195, + "learning_rate": 2.8632081364409804e-06, + "loss": 0.1612, + "step": 10879 + }, + { + "epoch": 2.2814007129377227, + "grad_norm": 5.333259823322492, + "learning_rate": 2.8616225299606228e-06, + "loss": 0.1905, + "step": 10880 + }, + { + "epoch": 2.2816104005032503, + "grad_norm": 3.892640192888336, + "learning_rate": 2.860037289334494e-06, + "loss": 0.1081, + "step": 10881 + }, + { + "epoch": 2.2818200880687773, + "grad_norm": 3.121292950721039, + "learning_rate": 2.858452414643834e-06, + "loss": 0.1059, + "step": 10882 + }, + { + "epoch": 2.282029775634305, + "grad_norm": 4.3195703498059395, + "learning_rate": 2.856867905969879e-06, + "loss": 0.1483, + "step": 10883 + }, + { + "epoch": 2.2822394631998324, + "grad_norm": 4.501755302741633, + "learning_rate": 2.8552837633938314e-06, + "loss": 0.1624, + "step": 10884 + }, + { + "epoch": 2.2824491507653595, + "grad_norm": 3.8154818098268546, + "learning_rate": 2.8536999869968807e-06, + "loss": 0.1548, + "step": 10885 + }, + { + "epoch": 2.282658838330887, + "grad_norm": 3.99808293556011, + "learning_rate": 2.852116576860199e-06, + "loss": 0.1541, + "step": 10886 + }, + { + "epoch": 2.2828685258964145, + "grad_norm": 4.1879128730303234, + "learning_rate": 2.8505335330649476e-06, + "loss": 0.1562, + "step": 10887 + }, + { + "epoch": 2.2830782134619416, + "grad_norm": 5.431519078517125, + "learning_rate": 2.8489508556922464e-06, + "loss": 0.1866, + "step": 10888 + }, + { + "epoch": 2.283287901027469, + "grad_norm": 4.785016465188952, + "learning_rate": 2.847368544823216e-06, + "loss": 0.1786, + "step": 10889 + }, + { + "epoch": 2.2834975885929962, + "grad_norm": 3.9230401757289544, + "learning_rate": 2.845786600538957e-06, + "loss": 0.1495, + "step": 10890 + }, + { + "epoch": 2.2837072761585238, + "grad_norm": 4.843360176089437, + "learning_rate": 2.844205022920542e-06, + "loss": 0.1813, + "step": 10891 + }, + { + "epoch": 2.2839169637240513, + "grad_norm": 5.499741147700227, + "learning_rate": 2.842623812049029e-06, + "loss": 0.1788, + "step": 10892 + }, + { + "epoch": 2.2841266512895784, + "grad_norm": 4.780227833531306, + "learning_rate": 2.8410429680054617e-06, + "loss": 0.1541, + "step": 10893 + }, + { + "epoch": 2.284336338855106, + "grad_norm": 5.436887340210706, + "learning_rate": 2.8394624908708568e-06, + "loss": 0.1522, + "step": 10894 + }, + { + "epoch": 2.2845460264206334, + "grad_norm": 3.6591589581961452, + "learning_rate": 2.8378823807262202e-06, + "loss": 0.149, + "step": 10895 + }, + { + "epoch": 2.2847557139861605, + "grad_norm": 3.4198537755217413, + "learning_rate": 2.8363026376525347e-06, + "loss": 0.1154, + "step": 10896 + }, + { + "epoch": 2.284965401551688, + "grad_norm": 4.329877872852249, + "learning_rate": 2.834723261730761e-06, + "loss": 0.1568, + "step": 10897 + }, + { + "epoch": 2.2851750891172156, + "grad_norm": 4.851258982694854, + "learning_rate": 2.8331442530418463e-06, + "loss": 0.1631, + "step": 10898 + }, + { + "epoch": 2.2853847766827426, + "grad_norm": 3.551350542815262, + "learning_rate": 2.8315656116667234e-06, + "loss": 0.1177, + "step": 10899 + }, + { + "epoch": 2.28559446424827, + "grad_norm": 3.911238506998745, + "learning_rate": 2.829987337686295e-06, + "loss": 0.152, + "step": 10900 + }, + { + "epoch": 2.2858041518137973, + "grad_norm": 3.2436121767719146, + "learning_rate": 2.828409431181448e-06, + "loss": 0.1052, + "step": 10901 + }, + { + "epoch": 2.286013839379325, + "grad_norm": 5.065058671621125, + "learning_rate": 2.8268318922330597e-06, + "loss": 0.1328, + "step": 10902 + }, + { + "epoch": 2.2862235269448523, + "grad_norm": 3.688865624184741, + "learning_rate": 2.825254720921975e-06, + "loss": 0.1356, + "step": 10903 + }, + { + "epoch": 2.2864332145103794, + "grad_norm": 4.472086900387992, + "learning_rate": 2.823677917329033e-06, + "loss": 0.114, + "step": 10904 + }, + { + "epoch": 2.286642902075907, + "grad_norm": 4.257352103596923, + "learning_rate": 2.8221014815350447e-06, + "loss": 0.1657, + "step": 10905 + }, + { + "epoch": 2.2868525896414345, + "grad_norm": 4.890087067822891, + "learning_rate": 2.8205254136208016e-06, + "loss": 0.129, + "step": 10906 + }, + { + "epoch": 2.2870622772069615, + "grad_norm": 4.336838988182833, + "learning_rate": 2.818949713667083e-06, + "loss": 0.1662, + "step": 10907 + }, + { + "epoch": 2.287271964772489, + "grad_norm": 3.924415873564747, + "learning_rate": 2.817374381754653e-06, + "loss": 0.136, + "step": 10908 + }, + { + "epoch": 2.287481652338016, + "grad_norm": 4.327641954432818, + "learning_rate": 2.815799417964238e-06, + "loss": 0.1365, + "step": 10909 + }, + { + "epoch": 2.2876913399035437, + "grad_norm": 5.480002568661021, + "learning_rate": 2.814224822376562e-06, + "loss": 0.2157, + "step": 10910 + }, + { + "epoch": 2.287901027469071, + "grad_norm": 4.577886241266605, + "learning_rate": 2.8126505950723305e-06, + "loss": 0.1727, + "step": 10911 + }, + { + "epoch": 2.2881107150345983, + "grad_norm": 4.649738268214217, + "learning_rate": 2.811076736132221e-06, + "loss": 0.1809, + "step": 10912 + }, + { + "epoch": 2.288320402600126, + "grad_norm": 4.032422574647493, + "learning_rate": 2.8095032456368953e-06, + "loss": 0.1156, + "step": 10913 + }, + { + "epoch": 2.2885300901656533, + "grad_norm": 4.929841683847431, + "learning_rate": 2.8079301236670032e-06, + "loss": 0.1333, + "step": 10914 + }, + { + "epoch": 2.2887397777311804, + "grad_norm": 4.534548000746281, + "learning_rate": 2.806357370303162e-06, + "loss": 0.186, + "step": 10915 + }, + { + "epoch": 2.288949465296708, + "grad_norm": 4.751398053541186, + "learning_rate": 2.8047849856259856e-06, + "loss": 0.1805, + "step": 10916 + }, + { + "epoch": 2.2891591528622355, + "grad_norm": 4.249496927293688, + "learning_rate": 2.803212969716055e-06, + "loss": 0.1117, + "step": 10917 + }, + { + "epoch": 2.2893688404277626, + "grad_norm": 4.1953376578574755, + "learning_rate": 2.801641322653945e-06, + "loss": 0.1538, + "step": 10918 + }, + { + "epoch": 2.28957852799329, + "grad_norm": 3.8180116843929297, + "learning_rate": 2.8000700445202e-06, + "loss": 0.1256, + "step": 10919 + }, + { + "epoch": 2.289788215558817, + "grad_norm": 4.61342975553467, + "learning_rate": 2.798499135395355e-06, + "loss": 0.1548, + "step": 10920 + }, + { + "epoch": 2.2899979031243447, + "grad_norm": 5.468465751174898, + "learning_rate": 2.7969285953599203e-06, + "loss": 0.1234, + "step": 10921 + }, + { + "epoch": 2.2902075906898722, + "grad_norm": 4.820803288641013, + "learning_rate": 2.7953584244943842e-06, + "loss": 0.1697, + "step": 10922 + }, + { + "epoch": 2.2904172782553993, + "grad_norm": 5.563363845909404, + "learning_rate": 2.7937886228792256e-06, + "loss": 0.193, + "step": 10923 + }, + { + "epoch": 2.290626965820927, + "grad_norm": 5.1031562862616875, + "learning_rate": 2.7922191905949e-06, + "loss": 0.1628, + "step": 10924 + }, + { + "epoch": 2.2908366533864544, + "grad_norm": 4.684495202838718, + "learning_rate": 2.7906501277218447e-06, + "loss": 0.1275, + "step": 10925 + }, + { + "epoch": 2.2910463409519815, + "grad_norm": 4.09148039356376, + "learning_rate": 2.7890814343404694e-06, + "loss": 0.1323, + "step": 10926 + }, + { + "epoch": 2.291256028517509, + "grad_norm": 3.6203582418177676, + "learning_rate": 2.78751311053118e-06, + "loss": 0.1215, + "step": 10927 + }, + { + "epoch": 2.291465716083036, + "grad_norm": 3.776714307517992, + "learning_rate": 2.7859451563743523e-06, + "loss": 0.1423, + "step": 10928 + }, + { + "epoch": 2.2916754036485636, + "grad_norm": 5.028675429087357, + "learning_rate": 2.7843775719503486e-06, + "loss": 0.1366, + "step": 10929 + }, + { + "epoch": 2.291885091214091, + "grad_norm": 4.713518592126404, + "learning_rate": 2.7828103573395103e-06, + "loss": 0.145, + "step": 10930 + }, + { + "epoch": 2.292094778779618, + "grad_norm": 5.908249613996727, + "learning_rate": 2.781243512622155e-06, + "loss": 0.1478, + "step": 10931 + }, + { + "epoch": 2.2923044663451457, + "grad_norm": 4.501431416301457, + "learning_rate": 2.7796770378785933e-06, + "loss": 0.1512, + "step": 10932 + }, + { + "epoch": 2.2925141539106733, + "grad_norm": 5.3048531679526185, + "learning_rate": 2.7781109331891063e-06, + "loss": 0.1785, + "step": 10933 + }, + { + "epoch": 2.2927238414762003, + "grad_norm": 3.916355861595163, + "learning_rate": 2.7765451986339566e-06, + "loss": 0.125, + "step": 10934 + }, + { + "epoch": 2.292933529041728, + "grad_norm": 4.832964778357069, + "learning_rate": 2.7749798342933944e-06, + "loss": 0.1641, + "step": 10935 + }, + { + "epoch": 2.2931432166072554, + "grad_norm": 4.014210604039888, + "learning_rate": 2.773414840247649e-06, + "loss": 0.1588, + "step": 10936 + }, + { + "epoch": 2.2933529041727825, + "grad_norm": 5.217905132852857, + "learning_rate": 2.771850216576928e-06, + "loss": 0.1872, + "step": 10937 + }, + { + "epoch": 2.29356259173831, + "grad_norm": 4.599792998793363, + "learning_rate": 2.7702859633614154e-06, + "loss": 0.137, + "step": 10938 + }, + { + "epoch": 2.293772279303837, + "grad_norm": 3.833527536084704, + "learning_rate": 2.768722080681291e-06, + "loss": 0.1406, + "step": 10939 + }, + { + "epoch": 2.2939819668693646, + "grad_norm": 5.32027297733488, + "learning_rate": 2.767158568616698e-06, + "loss": 0.186, + "step": 10940 + }, + { + "epoch": 2.294191654434892, + "grad_norm": 3.7333278416088924, + "learning_rate": 2.765595427247777e-06, + "loss": 0.1484, + "step": 10941 + }, + { + "epoch": 2.2944013420004192, + "grad_norm": 3.5176573271309834, + "learning_rate": 2.764032656654637e-06, + "loss": 0.1137, + "step": 10942 + }, + { + "epoch": 2.2946110295659468, + "grad_norm": 4.527064273965617, + "learning_rate": 2.7624702569173713e-06, + "loss": 0.1428, + "step": 10943 + }, + { + "epoch": 2.2948207171314743, + "grad_norm": 4.9475130730125425, + "learning_rate": 2.760908228116057e-06, + "loss": 0.1588, + "step": 10944 + }, + { + "epoch": 2.2950304046970014, + "grad_norm": 5.048553808441142, + "learning_rate": 2.7593465703307555e-06, + "loss": 0.182, + "step": 10945 + }, + { + "epoch": 2.295240092262529, + "grad_norm": 4.486613381633271, + "learning_rate": 2.7577852836414997e-06, + "loss": 0.1371, + "step": 10946 + }, + { + "epoch": 2.295449779828056, + "grad_norm": 4.230508060924311, + "learning_rate": 2.7562243681283063e-06, + "loss": 0.1551, + "step": 10947 + }, + { + "epoch": 2.2956594673935835, + "grad_norm": 3.657073109136003, + "learning_rate": 2.754663823871181e-06, + "loss": 0.1379, + "step": 10948 + }, + { + "epoch": 2.295869154959111, + "grad_norm": 5.476866868740158, + "learning_rate": 2.7531036509501006e-06, + "loss": 0.1672, + "step": 10949 + }, + { + "epoch": 2.296078842524638, + "grad_norm": 3.747436450620385, + "learning_rate": 2.7515438494450243e-06, + "loss": 0.1358, + "step": 10950 + }, + { + "epoch": 2.2962885300901656, + "grad_norm": 4.158123442037323, + "learning_rate": 2.7499844194359016e-06, + "loss": 0.1626, + "step": 10951 + }, + { + "epoch": 2.296498217655693, + "grad_norm": 5.662663579374337, + "learning_rate": 2.748425361002648e-06, + "loss": 0.1595, + "step": 10952 + }, + { + "epoch": 2.2967079052212203, + "grad_norm": 3.820839838919107, + "learning_rate": 2.7468666742251747e-06, + "loss": 0.1651, + "step": 10953 + }, + { + "epoch": 2.296917592786748, + "grad_norm": 3.9349774907179036, + "learning_rate": 2.745308359183364e-06, + "loss": 0.1215, + "step": 10954 + }, + { + "epoch": 2.2971272803522753, + "grad_norm": 4.610434473242038, + "learning_rate": 2.7437504159570804e-06, + "loss": 0.1553, + "step": 10955 + }, + { + "epoch": 2.2973369679178024, + "grad_norm": 2.6658628917792524, + "learning_rate": 2.7421928446261726e-06, + "loss": 0.0835, + "step": 10956 + }, + { + "epoch": 2.29754665548333, + "grad_norm": 4.406227439795561, + "learning_rate": 2.740635645270473e-06, + "loss": 0.1381, + "step": 10957 + }, + { + "epoch": 2.2977563430488575, + "grad_norm": 4.180955645660982, + "learning_rate": 2.739078817969787e-06, + "loss": 0.1385, + "step": 10958 + }, + { + "epoch": 2.2979660306143845, + "grad_norm": 4.006099900622913, + "learning_rate": 2.7375223628039025e-06, + "loss": 0.1271, + "step": 10959 + }, + { + "epoch": 2.298175718179912, + "grad_norm": 4.392571319093092, + "learning_rate": 2.7359662798525953e-06, + "loss": 0.1356, + "step": 10960 + }, + { + "epoch": 2.298385405745439, + "grad_norm": 3.636146971920943, + "learning_rate": 2.7344105691956122e-06, + "loss": 0.1437, + "step": 10961 + }, + { + "epoch": 2.2985950933109667, + "grad_norm": 5.0847305524419, + "learning_rate": 2.7328552309126934e-06, + "loss": 0.1868, + "step": 10962 + }, + { + "epoch": 2.298804780876494, + "grad_norm": 3.701476413708825, + "learning_rate": 2.7313002650835463e-06, + "loss": 0.1408, + "step": 10963 + }, + { + "epoch": 2.2990144684420213, + "grad_norm": 4.901698596266978, + "learning_rate": 2.7297456717878656e-06, + "loss": 0.1496, + "step": 10964 + }, + { + "epoch": 2.299224156007549, + "grad_norm": 5.314994755671248, + "learning_rate": 2.7281914511053285e-06, + "loss": 0.1327, + "step": 10965 + }, + { + "epoch": 2.299433843573076, + "grad_norm": 4.182942187325313, + "learning_rate": 2.7266376031155984e-06, + "loss": 0.1471, + "step": 10966 + }, + { + "epoch": 2.2996435311386034, + "grad_norm": 3.817063763514465, + "learning_rate": 2.725084127898299e-06, + "loss": 0.1104, + "step": 10967 + }, + { + "epoch": 2.299853218704131, + "grad_norm": 4.3267705169489155, + "learning_rate": 2.723531025533056e-06, + "loss": 0.1567, + "step": 10968 + }, + { + "epoch": 2.300062906269658, + "grad_norm": 5.256568771421014, + "learning_rate": 2.721978296099472e-06, + "loss": 0.1567, + "step": 10969 + }, + { + "epoch": 2.3002725938351856, + "grad_norm": 4.317323868510292, + "learning_rate": 2.7204259396771236e-06, + "loss": 0.1539, + "step": 10970 + }, + { + "epoch": 2.300482281400713, + "grad_norm": 3.7762725267391244, + "learning_rate": 2.7188739563455677e-06, + "loss": 0.1661, + "step": 10971 + }, + { + "epoch": 2.30069196896624, + "grad_norm": 3.96655206650356, + "learning_rate": 2.7173223461843535e-06, + "loss": 0.1327, + "step": 10972 + }, + { + "epoch": 2.3009016565317677, + "grad_norm": 4.938880370582554, + "learning_rate": 2.715771109272998e-06, + "loss": 0.1603, + "step": 10973 + }, + { + "epoch": 2.3011113440972952, + "grad_norm": 3.745290601015423, + "learning_rate": 2.7142202456910085e-06, + "loss": 0.1161, + "step": 10974 + }, + { + "epoch": 2.3013210316628223, + "grad_norm": 4.829599929006866, + "learning_rate": 2.7126697555178683e-06, + "loss": 0.1793, + "step": 10975 + }, + { + "epoch": 2.30153071922835, + "grad_norm": 4.236044629448255, + "learning_rate": 2.7111196388330396e-06, + "loss": 0.132, + "step": 10976 + }, + { + "epoch": 2.3017404067938774, + "grad_norm": 4.993906754488048, + "learning_rate": 2.709569895715972e-06, + "loss": 0.154, + "step": 10977 + }, + { + "epoch": 2.3019500943594045, + "grad_norm": 3.6426238826710593, + "learning_rate": 2.7080205262460934e-06, + "loss": 0.1378, + "step": 10978 + }, + { + "epoch": 2.302159781924932, + "grad_norm": 3.8866238907160073, + "learning_rate": 2.706471530502811e-06, + "loss": 0.1437, + "step": 10979 + }, + { + "epoch": 2.302369469490459, + "grad_norm": 5.205591909304769, + "learning_rate": 2.7049229085655103e-06, + "loss": 0.1572, + "step": 10980 + }, + { + "epoch": 2.3025791570559866, + "grad_norm": 3.361814001701675, + "learning_rate": 2.7033746605135627e-06, + "loss": 0.1192, + "step": 10981 + }, + { + "epoch": 2.302788844621514, + "grad_norm": 3.8730219995543362, + "learning_rate": 2.7018267864263215e-06, + "loss": 0.158, + "step": 10982 + }, + { + "epoch": 2.302998532187041, + "grad_norm": 4.309931794256158, + "learning_rate": 2.7002792863831162e-06, + "loss": 0.162, + "step": 10983 + }, + { + "epoch": 2.3032082197525687, + "grad_norm": 3.2208454119156444, + "learning_rate": 2.6987321604632545e-06, + "loss": 0.1356, + "step": 10984 + }, + { + "epoch": 2.303417907318096, + "grad_norm": 4.452595458522969, + "learning_rate": 2.6971854087460368e-06, + "loss": 0.1527, + "step": 10985 + }, + { + "epoch": 2.3036275948836233, + "grad_norm": 3.173403539647856, + "learning_rate": 2.6956390313107293e-06, + "loss": 0.0895, + "step": 10986 + }, + { + "epoch": 2.303837282449151, + "grad_norm": 3.537001705979659, + "learning_rate": 2.694093028236594e-06, + "loss": 0.1237, + "step": 10987 + }, + { + "epoch": 2.304046970014678, + "grad_norm": 4.038299730515624, + "learning_rate": 2.692547399602862e-06, + "loss": 0.1532, + "step": 10988 + }, + { + "epoch": 2.3042566575802055, + "grad_norm": 4.397544426520461, + "learning_rate": 2.691002145488747e-06, + "loss": 0.1726, + "step": 10989 + }, + { + "epoch": 2.304466345145733, + "grad_norm": 3.750266184027141, + "learning_rate": 2.6894572659734485e-06, + "loss": 0.1446, + "step": 10990 + }, + { + "epoch": 2.30467603271126, + "grad_norm": 3.485296211674269, + "learning_rate": 2.687912761136151e-06, + "loss": 0.1179, + "step": 10991 + }, + { + "epoch": 2.3048857202767876, + "grad_norm": 3.920530570125744, + "learning_rate": 2.686368631056001e-06, + "loss": 0.1492, + "step": 10992 + }, + { + "epoch": 2.305095407842315, + "grad_norm": 3.3642528232155446, + "learning_rate": 2.684824875812143e-06, + "loss": 0.0898, + "step": 10993 + }, + { + "epoch": 2.3053050954078422, + "grad_norm": 5.588708162419901, + "learning_rate": 2.6832814954837004e-06, + "loss": 0.187, + "step": 10994 + }, + { + "epoch": 2.3055147829733698, + "grad_norm": 3.4597144806253626, + "learning_rate": 2.6817384901497713e-06, + "loss": 0.134, + "step": 10995 + }, + { + "epoch": 2.3057244705388973, + "grad_norm": 4.627972614820141, + "learning_rate": 2.680195859889435e-06, + "loss": 0.1528, + "step": 10996 + }, + { + "epoch": 2.3059341581044244, + "grad_norm": 4.2301864415351975, + "learning_rate": 2.678653604781759e-06, + "loss": 0.1558, + "step": 10997 + }, + { + "epoch": 2.306143845669952, + "grad_norm": 4.677111590584339, + "learning_rate": 2.6771117249057822e-06, + "loss": 0.1444, + "step": 10998 + }, + { + "epoch": 2.306353533235479, + "grad_norm": 5.284685089921044, + "learning_rate": 2.675570220340532e-06, + "loss": 0.1605, + "step": 10999 + }, + { + "epoch": 2.3065632208010065, + "grad_norm": 7.46246771698627, + "learning_rate": 2.6740290911650125e-06, + "loss": 0.2058, + "step": 11000 + }, + { + "epoch": 2.306772908366534, + "grad_norm": 3.2256380773378823, + "learning_rate": 2.6724883374582044e-06, + "loss": 0.1147, + "step": 11001 + }, + { + "epoch": 2.306982595932061, + "grad_norm": 3.7999741504952276, + "learning_rate": 2.670947959299078e-06, + "loss": 0.1344, + "step": 11002 + }, + { + "epoch": 2.3071922834975886, + "grad_norm": 3.4418588787445183, + "learning_rate": 2.669407956766584e-06, + "loss": 0.128, + "step": 11003 + }, + { + "epoch": 2.3074019710631157, + "grad_norm": 4.012557444196535, + "learning_rate": 2.6678683299396455e-06, + "loss": 0.1492, + "step": 11004 + }, + { + "epoch": 2.3076116586286433, + "grad_norm": 4.008298727172779, + "learning_rate": 2.6663290788971705e-06, + "loss": 0.1357, + "step": 11005 + }, + { + "epoch": 2.307821346194171, + "grad_norm": 5.146556463587217, + "learning_rate": 2.664790203718052e-06, + "loss": 0.1559, + "step": 11006 + }, + { + "epoch": 2.308031033759698, + "grad_norm": 3.6543128631543795, + "learning_rate": 2.6632517044811546e-06, + "loss": 0.157, + "step": 11007 + }, + { + "epoch": 2.3082407213252254, + "grad_norm": 4.046779356658434, + "learning_rate": 2.661713581265336e-06, + "loss": 0.1323, + "step": 11008 + }, + { + "epoch": 2.308450408890753, + "grad_norm": 4.505874717886358, + "learning_rate": 2.660175834149423e-06, + "loss": 0.1561, + "step": 11009 + }, + { + "epoch": 2.30866009645628, + "grad_norm": 4.396419738199487, + "learning_rate": 2.6586384632122264e-06, + "loss": 0.1034, + "step": 11010 + }, + { + "epoch": 2.3088697840218075, + "grad_norm": 4.683882187034118, + "learning_rate": 2.6571014685325446e-06, + "loss": 0.1405, + "step": 11011 + }, + { + "epoch": 2.309079471587335, + "grad_norm": 4.701119434506378, + "learning_rate": 2.6555648501891473e-06, + "loss": 0.1718, + "step": 11012 + }, + { + "epoch": 2.309289159152862, + "grad_norm": 3.2700157587019087, + "learning_rate": 2.654028608260788e-06, + "loss": 0.121, + "step": 11013 + }, + { + "epoch": 2.3094988467183897, + "grad_norm": 3.900264992861094, + "learning_rate": 2.6524927428262024e-06, + "loss": 0.1355, + "step": 11014 + }, + { + "epoch": 2.309708534283917, + "grad_norm": 4.223028631270602, + "learning_rate": 2.6509572539641115e-06, + "loss": 0.1363, + "step": 11015 + }, + { + "epoch": 2.3099182218494443, + "grad_norm": 5.25396807609399, + "learning_rate": 2.6494221417532064e-06, + "loss": 0.1935, + "step": 11016 + }, + { + "epoch": 2.310127909414972, + "grad_norm": 4.759936659032829, + "learning_rate": 2.6478874062721626e-06, + "loss": 0.1524, + "step": 11017 + }, + { + "epoch": 2.310337596980499, + "grad_norm": 5.257778133248842, + "learning_rate": 2.6463530475996445e-06, + "loss": 0.1613, + "step": 11018 + }, + { + "epoch": 2.3105472845460264, + "grad_norm": 3.673564899557511, + "learning_rate": 2.6448190658142838e-06, + "loss": 0.1091, + "step": 11019 + }, + { + "epoch": 2.310756972111554, + "grad_norm": 4.246249811284005, + "learning_rate": 2.6432854609947055e-06, + "loss": 0.1723, + "step": 11020 + }, + { + "epoch": 2.310966659677081, + "grad_norm": 3.662389159895316, + "learning_rate": 2.6417522332195066e-06, + "loss": 0.1432, + "step": 11021 + }, + { + "epoch": 2.3111763472426086, + "grad_norm": 5.169935543833419, + "learning_rate": 2.6402193825672664e-06, + "loss": 0.1807, + "step": 11022 + }, + { + "epoch": 2.3113860348081356, + "grad_norm": 4.048917475315759, + "learning_rate": 2.6386869091165467e-06, + "loss": 0.1464, + "step": 11023 + }, + { + "epoch": 2.311595722373663, + "grad_norm": 4.236512843488925, + "learning_rate": 2.6371548129458935e-06, + "loss": 0.1611, + "step": 11024 + }, + { + "epoch": 2.3118054099391907, + "grad_norm": 4.644487607364284, + "learning_rate": 2.6356230941338256e-06, + "loss": 0.1558, + "step": 11025 + }, + { + "epoch": 2.312015097504718, + "grad_norm": 4.922419772766671, + "learning_rate": 2.634091752758845e-06, + "loss": 0.1656, + "step": 11026 + }, + { + "epoch": 2.3122247850702453, + "grad_norm": 4.846652978973002, + "learning_rate": 2.632560788899441e-06, + "loss": 0.1654, + "step": 11027 + }, + { + "epoch": 2.312434472635773, + "grad_norm": 4.270031500418427, + "learning_rate": 2.631030202634074e-06, + "loss": 0.16, + "step": 11028 + }, + { + "epoch": 2.3126441602013, + "grad_norm": 3.817901552661584, + "learning_rate": 2.629499994041187e-06, + "loss": 0.1314, + "step": 11029 + }, + { + "epoch": 2.3128538477668275, + "grad_norm": 4.346036836048172, + "learning_rate": 2.6279701631992115e-06, + "loss": 0.158, + "step": 11030 + }, + { + "epoch": 2.313063535332355, + "grad_norm": 3.9678589778545326, + "learning_rate": 2.6264407101865485e-06, + "loss": 0.1803, + "step": 11031 + }, + { + "epoch": 2.313273222897882, + "grad_norm": 5.078822317021803, + "learning_rate": 2.6249116350815917e-06, + "loss": 0.1409, + "step": 11032 + }, + { + "epoch": 2.3134829104634096, + "grad_norm": 4.330413912235461, + "learning_rate": 2.623382937962704e-06, + "loss": 0.1419, + "step": 11033 + }, + { + "epoch": 2.313692598028937, + "grad_norm": 4.086239909442471, + "learning_rate": 2.621854618908233e-06, + "loss": 0.1528, + "step": 11034 + }, + { + "epoch": 2.313902285594464, + "grad_norm": 4.246507720981344, + "learning_rate": 2.6203266779965074e-06, + "loss": 0.1434, + "step": 11035 + }, + { + "epoch": 2.3141119731599917, + "grad_norm": 4.498810340479034, + "learning_rate": 2.6187991153058434e-06, + "loss": 0.1534, + "step": 11036 + }, + { + "epoch": 2.314321660725519, + "grad_norm": 4.189964786486981, + "learning_rate": 2.6172719309145266e-06, + "loss": 0.1422, + "step": 11037 + }, + { + "epoch": 2.3145313482910463, + "grad_norm": 4.082382256871633, + "learning_rate": 2.615745124900826e-06, + "loss": 0.1569, + "step": 11038 + }, + { + "epoch": 2.314741035856574, + "grad_norm": 4.252897105437072, + "learning_rate": 2.614218697342994e-06, + "loss": 0.154, + "step": 11039 + }, + { + "epoch": 2.314950723422101, + "grad_norm": 5.2430805769654265, + "learning_rate": 2.612692648319267e-06, + "loss": 0.1504, + "step": 11040 + }, + { + "epoch": 2.3151604109876285, + "grad_norm": 4.808348435525007, + "learning_rate": 2.611166977907854e-06, + "loss": 0.176, + "step": 11041 + }, + { + "epoch": 2.3153700985531556, + "grad_norm": 4.689170116188598, + "learning_rate": 2.6096416861869467e-06, + "loss": 0.1421, + "step": 11042 + }, + { + "epoch": 2.315579786118683, + "grad_norm": 3.8015418691963143, + "learning_rate": 2.6081167732347224e-06, + "loss": 0.1157, + "step": 11043 + }, + { + "epoch": 2.3157894736842106, + "grad_norm": 5.230349188215154, + "learning_rate": 2.606592239129332e-06, + "loss": 0.1955, + "step": 11044 + }, + { + "epoch": 2.3159991612497377, + "grad_norm": 4.7429372302443635, + "learning_rate": 2.605068083948915e-06, + "loss": 0.1572, + "step": 11045 + }, + { + "epoch": 2.3162088488152652, + "grad_norm": 3.8242447164713815, + "learning_rate": 2.603544307771585e-06, + "loss": 0.1138, + "step": 11046 + }, + { + "epoch": 2.3164185363807928, + "grad_norm": 4.589727617072916, + "learning_rate": 2.6020209106754356e-06, + "loss": 0.1691, + "step": 11047 + }, + { + "epoch": 2.31662822394632, + "grad_norm": 5.5666875804948575, + "learning_rate": 2.600497892738545e-06, + "loss": 0.1719, + "step": 11048 + }, + { + "epoch": 2.3168379115118474, + "grad_norm": 4.79842266198786, + "learning_rate": 2.598975254038977e-06, + "loss": 0.1616, + "step": 11049 + }, + { + "epoch": 2.317047599077375, + "grad_norm": 3.721189596633461, + "learning_rate": 2.5974529946547565e-06, + "loss": 0.1276, + "step": 11050 + }, + { + "epoch": 2.317257286642902, + "grad_norm": 4.310452703709921, + "learning_rate": 2.595931114663911e-06, + "loss": 0.1516, + "step": 11051 + }, + { + "epoch": 2.3174669742084295, + "grad_norm": 4.800998941984439, + "learning_rate": 2.59440961414444e-06, + "loss": 0.1432, + "step": 11052 + }, + { + "epoch": 2.317676661773957, + "grad_norm": 4.176185711657633, + "learning_rate": 2.5928884931743205e-06, + "loss": 0.1413, + "step": 11053 + }, + { + "epoch": 2.317886349339484, + "grad_norm": 3.9365789723719775, + "learning_rate": 2.5913677518315093e-06, + "loss": 0.0952, + "step": 11054 + }, + { + "epoch": 2.3180960369050116, + "grad_norm": 4.347808106915857, + "learning_rate": 2.5898473901939535e-06, + "loss": 0.1899, + "step": 11055 + }, + { + "epoch": 2.3183057244705387, + "grad_norm": 4.366273051737066, + "learning_rate": 2.588327408339568e-06, + "loss": 0.1405, + "step": 11056 + }, + { + "epoch": 2.3185154120360663, + "grad_norm": 4.032375106953269, + "learning_rate": 2.586807806346262e-06, + "loss": 0.1594, + "step": 11057 + }, + { + "epoch": 2.318725099601594, + "grad_norm": 3.540466741180801, + "learning_rate": 2.585288584291912e-06, + "loss": 0.1186, + "step": 11058 + }, + { + "epoch": 2.318934787167121, + "grad_norm": 4.268923900692967, + "learning_rate": 2.583769742254381e-06, + "loss": 0.1781, + "step": 11059 + }, + { + "epoch": 2.3191444747326484, + "grad_norm": 5.8015503744344885, + "learning_rate": 2.5822512803115117e-06, + "loss": 0.1415, + "step": 11060 + }, + { + "epoch": 2.3193541622981755, + "grad_norm": 4.612179974168851, + "learning_rate": 2.5807331985411334e-06, + "loss": 0.1745, + "step": 11061 + }, + { + "epoch": 2.319563849863703, + "grad_norm": 5.5740521151437425, + "learning_rate": 2.579215497021046e-06, + "loss": 0.2312, + "step": 11062 + }, + { + "epoch": 2.3197735374292305, + "grad_norm": 5.004602740010987, + "learning_rate": 2.5776981758290332e-06, + "loss": 0.1742, + "step": 11063 + }, + { + "epoch": 2.3199832249947576, + "grad_norm": 4.339493179625758, + "learning_rate": 2.576181235042865e-06, + "loss": 0.153, + "step": 11064 + }, + { + "epoch": 2.320192912560285, + "grad_norm": 5.4524200210560485, + "learning_rate": 2.5746646747402815e-06, + "loss": 0.1695, + "step": 11065 + }, + { + "epoch": 2.3204026001258127, + "grad_norm": 5.389359063275125, + "learning_rate": 2.5731484949990147e-06, + "loss": 0.1899, + "step": 11066 + }, + { + "epoch": 2.3206122876913398, + "grad_norm": 4.216134597554432, + "learning_rate": 2.5716326958967684e-06, + "loss": 0.171, + "step": 11067 + }, + { + "epoch": 2.3208219752568673, + "grad_norm": 4.95813376127907, + "learning_rate": 2.570117277511228e-06, + "loss": 0.1367, + "step": 11068 + }, + { + "epoch": 2.321031662822395, + "grad_norm": 3.785872982925218, + "learning_rate": 2.568602239920063e-06, + "loss": 0.0974, + "step": 11069 + }, + { + "epoch": 2.321241350387922, + "grad_norm": 4.387770741695435, + "learning_rate": 2.567087583200929e-06, + "loss": 0.1548, + "step": 11070 + }, + { + "epoch": 2.3214510379534494, + "grad_norm": 4.529446370646171, + "learning_rate": 2.5655733074314416e-06, + "loss": 0.1587, + "step": 11071 + }, + { + "epoch": 2.321660725518977, + "grad_norm": 5.179315681511126, + "learning_rate": 2.564059412689217e-06, + "loss": 0.152, + "step": 11072 + }, + { + "epoch": 2.321870413084504, + "grad_norm": 4.050453092733685, + "learning_rate": 2.562545899051847e-06, + "loss": 0.1407, + "step": 11073 + }, + { + "epoch": 2.3220801006500316, + "grad_norm": 5.113077770356172, + "learning_rate": 2.5610327665968992e-06, + "loss": 0.1685, + "step": 11074 + }, + { + "epoch": 2.3222897882155586, + "grad_norm": 4.20621855664333, + "learning_rate": 2.559520015401922e-06, + "loss": 0.1179, + "step": 11075 + }, + { + "epoch": 2.322499475781086, + "grad_norm": 5.291327580309708, + "learning_rate": 2.558007645544451e-06, + "loss": 0.1675, + "step": 11076 + }, + { + "epoch": 2.3227091633466137, + "grad_norm": 4.2376471343947575, + "learning_rate": 2.556495657101994e-06, + "loss": 0.1519, + "step": 11077 + }, + { + "epoch": 2.322918850912141, + "grad_norm": 4.901123009345599, + "learning_rate": 2.5549840501520463e-06, + "loss": 0.1523, + "step": 11078 + }, + { + "epoch": 2.3231285384776683, + "grad_norm": 3.6203270072629885, + "learning_rate": 2.5534728247720785e-06, + "loss": 0.1021, + "step": 11079 + }, + { + "epoch": 2.3233382260431954, + "grad_norm": 4.173326123180353, + "learning_rate": 2.5519619810395424e-06, + "loss": 0.1583, + "step": 11080 + }, + { + "epoch": 2.323547913608723, + "grad_norm": 4.573039223975011, + "learning_rate": 2.5504515190318713e-06, + "loss": 0.1425, + "step": 11081 + }, + { + "epoch": 2.3237576011742505, + "grad_norm": 3.5597321829338986, + "learning_rate": 2.5489414388264844e-06, + "loss": 0.112, + "step": 11082 + }, + { + "epoch": 2.3239672887397775, + "grad_norm": 3.8548556232107187, + "learning_rate": 2.547431740500771e-06, + "loss": 0.1264, + "step": 11083 + }, + { + "epoch": 2.324176976305305, + "grad_norm": 4.258845044629492, + "learning_rate": 2.5459224241321046e-06, + "loss": 0.1179, + "step": 11084 + }, + { + "epoch": 2.3243866638708326, + "grad_norm": 3.9706791643882413, + "learning_rate": 2.5444134897978456e-06, + "loss": 0.1248, + "step": 11085 + }, + { + "epoch": 2.3245963514363597, + "grad_norm": 4.442730687173412, + "learning_rate": 2.5429049375753224e-06, + "loss": 0.1654, + "step": 11086 + }, + { + "epoch": 2.324806039001887, + "grad_norm": 4.493148525002239, + "learning_rate": 2.541396767541858e-06, + "loss": 0.1472, + "step": 11087 + }, + { + "epoch": 2.3250157265674147, + "grad_norm": 3.459810062691372, + "learning_rate": 2.5398889797747463e-06, + "loss": 0.126, + "step": 11088 + }, + { + "epoch": 2.325225414132942, + "grad_norm": 4.158404352130959, + "learning_rate": 2.5383815743512595e-06, + "loss": 0.1586, + "step": 11089 + }, + { + "epoch": 2.3254351016984693, + "grad_norm": 5.193345078114732, + "learning_rate": 2.5368745513486615e-06, + "loss": 0.1493, + "step": 11090 + }, + { + "epoch": 2.325644789263997, + "grad_norm": 4.2413299177364125, + "learning_rate": 2.5353679108441875e-06, + "loss": 0.0999, + "step": 11091 + }, + { + "epoch": 2.325854476829524, + "grad_norm": 4.201990052105541, + "learning_rate": 2.5338616529150516e-06, + "loss": 0.1558, + "step": 11092 + }, + { + "epoch": 2.3260641643950515, + "grad_norm": 4.343438918223055, + "learning_rate": 2.5323557776384554e-06, + "loss": 0.1697, + "step": 11093 + }, + { + "epoch": 2.3262738519605786, + "grad_norm": 4.844150750056045, + "learning_rate": 2.5308502850915805e-06, + "loss": 0.1671, + "step": 11094 + }, + { + "epoch": 2.326483539526106, + "grad_norm": 4.1837221109464675, + "learning_rate": 2.5293451753515828e-06, + "loss": 0.1594, + "step": 11095 + }, + { + "epoch": 2.3266932270916336, + "grad_norm": 4.346740812430216, + "learning_rate": 2.527840448495599e-06, + "loss": 0.1703, + "step": 11096 + }, + { + "epoch": 2.3269029146571607, + "grad_norm": 6.610197328359388, + "learning_rate": 2.526336104600755e-06, + "loss": 0.1604, + "step": 11097 + }, + { + "epoch": 2.3271126022226882, + "grad_norm": 4.509159284862679, + "learning_rate": 2.5248321437441457e-06, + "loss": 0.185, + "step": 11098 + }, + { + "epoch": 2.3273222897882153, + "grad_norm": 4.822726319811454, + "learning_rate": 2.523328566002857e-06, + "loss": 0.1701, + "step": 11099 + }, + { + "epoch": 2.327531977353743, + "grad_norm": 3.2019779670230304, + "learning_rate": 2.521825371453943e-06, + "loss": 0.108, + "step": 11100 + }, + { + "epoch": 2.3277416649192704, + "grad_norm": 4.492662695131915, + "learning_rate": 2.520322560174453e-06, + "loss": 0.1752, + "step": 11101 + }, + { + "epoch": 2.3279513524847975, + "grad_norm": 5.370252472622765, + "learning_rate": 2.5188201322414007e-06, + "loss": 0.1927, + "step": 11102 + }, + { + "epoch": 2.328161040050325, + "grad_norm": 5.014775960985252, + "learning_rate": 2.517318087731796e-06, + "loss": 0.1244, + "step": 11103 + }, + { + "epoch": 2.3283707276158525, + "grad_norm": 5.986217836138817, + "learning_rate": 2.515816426722617e-06, + "loss": 0.1733, + "step": 11104 + }, + { + "epoch": 2.3285804151813796, + "grad_norm": 3.3451895055116108, + "learning_rate": 2.5143151492908237e-06, + "loss": 0.1026, + "step": 11105 + }, + { + "epoch": 2.328790102746907, + "grad_norm": 3.5468633559378255, + "learning_rate": 2.5128142555133617e-06, + "loss": 0.1366, + "step": 11106 + }, + { + "epoch": 2.3289997903124346, + "grad_norm": 4.89089769554914, + "learning_rate": 2.511313745467162e-06, + "loss": 0.1815, + "step": 11107 + }, + { + "epoch": 2.3292094778779617, + "grad_norm": 3.619281794223496, + "learning_rate": 2.5098136192291144e-06, + "loss": 0.1118, + "step": 11108 + }, + { + "epoch": 2.3294191654434893, + "grad_norm": 4.641922129663621, + "learning_rate": 2.5083138768761104e-06, + "loss": 0.1807, + "step": 11109 + }, + { + "epoch": 2.329628853009017, + "grad_norm": 4.639482209149072, + "learning_rate": 2.506814518485017e-06, + "loss": 0.1421, + "step": 11110 + }, + { + "epoch": 2.329838540574544, + "grad_norm": 3.360695971494131, + "learning_rate": 2.505315544132676e-06, + "loss": 0.1165, + "step": 11111 + }, + { + "epoch": 2.3300482281400714, + "grad_norm": 3.451296025969975, + "learning_rate": 2.503816953895909e-06, + "loss": 0.1051, + "step": 11112 + }, + { + "epoch": 2.3302579157055985, + "grad_norm": 4.65163746412364, + "learning_rate": 2.5023187478515265e-06, + "loss": 0.1229, + "step": 11113 + }, + { + "epoch": 2.330467603271126, + "grad_norm": 4.441611796238513, + "learning_rate": 2.5008209260763106e-06, + "loss": 0.1266, + "step": 11114 + }, + { + "epoch": 2.3306772908366535, + "grad_norm": 3.674120525448985, + "learning_rate": 2.499323488647032e-06, + "loss": 0.1177, + "step": 11115 + }, + { + "epoch": 2.3308869784021806, + "grad_norm": 4.144654872796137, + "learning_rate": 2.4978264356404326e-06, + "loss": 0.1288, + "step": 11116 + }, + { + "epoch": 2.331096665967708, + "grad_norm": 6.127040321585709, + "learning_rate": 2.4963297671332385e-06, + "loss": 0.135, + "step": 11117 + }, + { + "epoch": 2.3313063535332352, + "grad_norm": 5.07939243125553, + "learning_rate": 2.4948334832021582e-06, + "loss": 0.1486, + "step": 11118 + }, + { + "epoch": 2.3315160410987628, + "grad_norm": 3.7141400132187083, + "learning_rate": 2.4933375839238804e-06, + "loss": 0.1527, + "step": 11119 + }, + { + "epoch": 2.3317257286642903, + "grad_norm": 4.070240532875839, + "learning_rate": 2.491842069375072e-06, + "loss": 0.1506, + "step": 11120 + }, + { + "epoch": 2.3319354162298174, + "grad_norm": 4.021320115602753, + "learning_rate": 2.4903469396323777e-06, + "loss": 0.1366, + "step": 11121 + }, + { + "epoch": 2.332145103795345, + "grad_norm": 4.60248002982288, + "learning_rate": 2.4888521947724297e-06, + "loss": 0.1398, + "step": 11122 + }, + { + "epoch": 2.3323547913608724, + "grad_norm": 4.2073382575084635, + "learning_rate": 2.487357834871831e-06, + "loss": 0.1157, + "step": 11123 + }, + { + "epoch": 2.3325644789263995, + "grad_norm": 3.923175511126156, + "learning_rate": 2.4858638600071773e-06, + "loss": 0.1194, + "step": 11124 + }, + { + "epoch": 2.332774166491927, + "grad_norm": 3.806675179879064, + "learning_rate": 2.484370270255032e-06, + "loss": 0.147, + "step": 11125 + }, + { + "epoch": 2.3329838540574546, + "grad_norm": 4.879890854695098, + "learning_rate": 2.482877065691943e-06, + "loss": 0.1658, + "step": 11126 + }, + { + "epoch": 2.3331935416229816, + "grad_norm": 4.558795994748064, + "learning_rate": 2.481384246394443e-06, + "loss": 0.1755, + "step": 11127 + }, + { + "epoch": 2.333403229188509, + "grad_norm": 5.464966829010082, + "learning_rate": 2.4798918124390457e-06, + "loss": 0.1264, + "step": 11128 + }, + { + "epoch": 2.3336129167540367, + "grad_norm": 4.785186881629059, + "learning_rate": 2.4783997639022306e-06, + "loss": 0.173, + "step": 11129 + }, + { + "epoch": 2.333822604319564, + "grad_norm": 5.786103214558417, + "learning_rate": 2.4769081008604734e-06, + "loss": 0.1663, + "step": 11130 + }, + { + "epoch": 2.3340322918850913, + "grad_norm": 5.021294785529363, + "learning_rate": 2.4754168233902266e-06, + "loss": 0.1492, + "step": 11131 + }, + { + "epoch": 2.3342419794506184, + "grad_norm": 4.458508579126794, + "learning_rate": 2.4739259315679177e-06, + "loss": 0.1361, + "step": 11132 + }, + { + "epoch": 2.334451667016146, + "grad_norm": 4.353386412620221, + "learning_rate": 2.4724354254699566e-06, + "loss": 0.1238, + "step": 11133 + }, + { + "epoch": 2.3346613545816735, + "grad_norm": 3.670337839950564, + "learning_rate": 2.4709453051727374e-06, + "loss": 0.1155, + "step": 11134 + }, + { + "epoch": 2.3348710421472005, + "grad_norm": 3.798832582505462, + "learning_rate": 2.469455570752628e-06, + "loss": 0.115, + "step": 11135 + }, + { + "epoch": 2.335080729712728, + "grad_norm": 5.335256265572671, + "learning_rate": 2.4679662222859833e-06, + "loss": 0.1785, + "step": 11136 + }, + { + "epoch": 2.3352904172782556, + "grad_norm": 3.3346472757433463, + "learning_rate": 2.466477259849135e-06, + "loss": 0.1265, + "step": 11137 + }, + { + "epoch": 2.3355001048437827, + "grad_norm": 3.2503663314227342, + "learning_rate": 2.46498868351839e-06, + "loss": 0.127, + "step": 11138 + }, + { + "epoch": 2.33570979240931, + "grad_norm": 3.9354092279268693, + "learning_rate": 2.4635004933700435e-06, + "loss": 0.1445, + "step": 11139 + }, + { + "epoch": 2.3359194799748373, + "grad_norm": 4.466760372413868, + "learning_rate": 2.4620126894803707e-06, + "loss": 0.1934, + "step": 11140 + }, + { + "epoch": 2.336129167540365, + "grad_norm": 4.837972487448911, + "learning_rate": 2.4605252719256214e-06, + "loss": 0.138, + "step": 11141 + }, + { + "epoch": 2.3363388551058923, + "grad_norm": 3.6373610076646883, + "learning_rate": 2.4590382407820256e-06, + "loss": 0.1216, + "step": 11142 + }, + { + "epoch": 2.3365485426714194, + "grad_norm": 3.642740655601929, + "learning_rate": 2.4575515961258023e-06, + "loss": 0.1203, + "step": 11143 + }, + { + "epoch": 2.336758230236947, + "grad_norm": 5.397866213767218, + "learning_rate": 2.456065338033139e-06, + "loss": 0.165, + "step": 11144 + }, + { + "epoch": 2.3369679178024745, + "grad_norm": 5.027257362815043, + "learning_rate": 2.4545794665802138e-06, + "loss": 0.1719, + "step": 11145 + }, + { + "epoch": 2.3371776053680016, + "grad_norm": 4.639955384818674, + "learning_rate": 2.4530939818431776e-06, + "loss": 0.143, + "step": 11146 + }, + { + "epoch": 2.337387292933529, + "grad_norm": 4.718114897204182, + "learning_rate": 2.451608883898162e-06, + "loss": 0.1484, + "step": 11147 + }, + { + "epoch": 2.3375969804990566, + "grad_norm": 3.320819271206125, + "learning_rate": 2.4501241728212832e-06, + "loss": 0.1325, + "step": 11148 + }, + { + "epoch": 2.3378066680645837, + "grad_norm": 4.944410894075514, + "learning_rate": 2.4486398486886398e-06, + "loss": 0.1665, + "step": 11149 + }, + { + "epoch": 2.3380163556301112, + "grad_norm": 4.064283298338463, + "learning_rate": 2.4471559115762965e-06, + "loss": 0.1423, + "step": 11150 + }, + { + "epoch": 2.3382260431956383, + "grad_norm": 3.8582038757927535, + "learning_rate": 2.445672361560313e-06, + "loss": 0.1369, + "step": 11151 + }, + { + "epoch": 2.338435730761166, + "grad_norm": 2.7660987789652647, + "learning_rate": 2.4441891987167244e-06, + "loss": 0.1054, + "step": 11152 + }, + { + "epoch": 2.3386454183266934, + "grad_norm": 3.5308357318584993, + "learning_rate": 2.442706423121546e-06, + "loss": 0.1393, + "step": 11153 + }, + { + "epoch": 2.3388551058922205, + "grad_norm": 4.888909342710058, + "learning_rate": 2.4412240348507664e-06, + "loss": 0.1786, + "step": 11154 + }, + { + "epoch": 2.339064793457748, + "grad_norm": 4.792551774103241, + "learning_rate": 2.439742033980369e-06, + "loss": 0.1774, + "step": 11155 + }, + { + "epoch": 2.3392744810232755, + "grad_norm": 4.151774811322938, + "learning_rate": 2.4382604205863027e-06, + "loss": 0.1133, + "step": 11156 + }, + { + "epoch": 2.3394841685888026, + "grad_norm": 4.069224381708613, + "learning_rate": 2.436779194744506e-06, + "loss": 0.1703, + "step": 11157 + }, + { + "epoch": 2.33969385615433, + "grad_norm": 4.240548417416079, + "learning_rate": 2.435298356530895e-06, + "loss": 0.1252, + "step": 11158 + }, + { + "epoch": 2.339903543719857, + "grad_norm": 4.653745844938985, + "learning_rate": 2.4338179060213595e-06, + "loss": 0.1577, + "step": 11159 + }, + { + "epoch": 2.3401132312853847, + "grad_norm": 4.531056040998752, + "learning_rate": 2.4323378432917797e-06, + "loss": 0.1476, + "step": 11160 + }, + { + "epoch": 2.3403229188509123, + "grad_norm": 5.011904639270193, + "learning_rate": 2.4308581684180143e-06, + "loss": 0.2101, + "step": 11161 + }, + { + "epoch": 2.3405326064164393, + "grad_norm": 4.190323218889102, + "learning_rate": 2.4293788814758944e-06, + "loss": 0.1607, + "step": 11162 + }, + { + "epoch": 2.340742293981967, + "grad_norm": 4.008122063183181, + "learning_rate": 2.4278999825412365e-06, + "loss": 0.1253, + "step": 11163 + }, + { + "epoch": 2.3409519815474944, + "grad_norm": 5.170535160916469, + "learning_rate": 2.426421471689836e-06, + "loss": 0.156, + "step": 11164 + }, + { + "epoch": 2.3411616691130215, + "grad_norm": 4.437653875537352, + "learning_rate": 2.424943348997474e-06, + "loss": 0.1445, + "step": 11165 + }, + { + "epoch": 2.341371356678549, + "grad_norm": 3.975341883673828, + "learning_rate": 2.4234656145399036e-06, + "loss": 0.1078, + "step": 11166 + }, + { + "epoch": 2.3415810442440765, + "grad_norm": 3.552419209079955, + "learning_rate": 2.4219882683928586e-06, + "loss": 0.122, + "step": 11167 + }, + { + "epoch": 2.3417907318096036, + "grad_norm": 4.642469040849319, + "learning_rate": 2.420511310632061e-06, + "loss": 0.1708, + "step": 11168 + }, + { + "epoch": 2.342000419375131, + "grad_norm": 4.069633032479241, + "learning_rate": 2.4190347413332047e-06, + "loss": 0.1548, + "step": 11169 + }, + { + "epoch": 2.3422101069406582, + "grad_norm": 4.734112694723988, + "learning_rate": 2.417558560571964e-06, + "loss": 0.1283, + "step": 11170 + }, + { + "epoch": 2.3424197945061858, + "grad_norm": 4.07517625565213, + "learning_rate": 2.4160827684240008e-06, + "loss": 0.1006, + "step": 11171 + }, + { + "epoch": 2.3426294820717133, + "grad_norm": 3.8345536193451273, + "learning_rate": 2.414607364964946e-06, + "loss": 0.1615, + "step": 11172 + }, + { + "epoch": 2.3428391696372404, + "grad_norm": 4.467830810595402, + "learning_rate": 2.4131323502704242e-06, + "loss": 0.1684, + "step": 11173 + }, + { + "epoch": 2.343048857202768, + "grad_norm": 3.7720735765431277, + "learning_rate": 2.4116577244160267e-06, + "loss": 0.1208, + "step": 11174 + }, + { + "epoch": 2.3432585447682954, + "grad_norm": 4.246121967852722, + "learning_rate": 2.41018348747733e-06, + "loss": 0.0985, + "step": 11175 + }, + { + "epoch": 2.3434682323338225, + "grad_norm": 3.800634027535383, + "learning_rate": 2.4087096395298926e-06, + "loss": 0.1613, + "step": 11176 + }, + { + "epoch": 2.34367791989935, + "grad_norm": 4.42094483434415, + "learning_rate": 2.4072361806492562e-06, + "loss": 0.1668, + "step": 11177 + }, + { + "epoch": 2.343887607464877, + "grad_norm": 5.520072812578669, + "learning_rate": 2.405763110910935e-06, + "loss": 0.1913, + "step": 11178 + }, + { + "epoch": 2.3440972950304046, + "grad_norm": 4.224236033367063, + "learning_rate": 2.4042904303904223e-06, + "loss": 0.1521, + "step": 11179 + }, + { + "epoch": 2.344306982595932, + "grad_norm": 3.3060666155429517, + "learning_rate": 2.402818139163202e-06, + "loss": 0.123, + "step": 11180 + }, + { + "epoch": 2.3445166701614593, + "grad_norm": 3.7952607261739235, + "learning_rate": 2.4013462373047257e-06, + "loss": 0.1432, + "step": 11181 + }, + { + "epoch": 2.344726357726987, + "grad_norm": 4.221770287683633, + "learning_rate": 2.3998747248904373e-06, + "loss": 0.1426, + "step": 11182 + }, + { + "epoch": 2.3449360452925143, + "grad_norm": 4.120804713608238, + "learning_rate": 2.3984036019957504e-06, + "loss": 0.137, + "step": 11183 + }, + { + "epoch": 2.3451457328580414, + "grad_norm": 4.447472762049035, + "learning_rate": 2.3969328686960604e-06, + "loss": 0.1479, + "step": 11184 + }, + { + "epoch": 2.345355420423569, + "grad_norm": 3.8457602646333062, + "learning_rate": 2.3954625250667484e-06, + "loss": 0.137, + "step": 11185 + }, + { + "epoch": 2.3455651079890965, + "grad_norm": 4.428127549534315, + "learning_rate": 2.393992571183177e-06, + "loss": 0.1601, + "step": 11186 + }, + { + "epoch": 2.3457747955546235, + "grad_norm": 3.8481284171334456, + "learning_rate": 2.392523007120672e-06, + "loss": 0.156, + "step": 11187 + }, + { + "epoch": 2.345984483120151, + "grad_norm": 3.9405610896941536, + "learning_rate": 2.3910538329545575e-06, + "loss": 0.1397, + "step": 11188 + }, + { + "epoch": 2.346194170685678, + "grad_norm": 4.135945331579041, + "learning_rate": 2.3895850487601345e-06, + "loss": 0.1469, + "step": 11189 + }, + { + "epoch": 2.3464038582512057, + "grad_norm": 5.304821708474476, + "learning_rate": 2.3881166546126764e-06, + "loss": 0.1515, + "step": 11190 + }, + { + "epoch": 2.346613545816733, + "grad_norm": 5.043851230035632, + "learning_rate": 2.3866486505874396e-06, + "loss": 0.1347, + "step": 11191 + }, + { + "epoch": 2.3468232333822603, + "grad_norm": 4.604645548923234, + "learning_rate": 2.3851810367596673e-06, + "loss": 0.1493, + "step": 11192 + }, + { + "epoch": 2.347032920947788, + "grad_norm": 4.567374445157383, + "learning_rate": 2.3837138132045713e-06, + "loss": 0.1587, + "step": 11193 + }, + { + "epoch": 2.3472426085133153, + "grad_norm": 3.9930927157004685, + "learning_rate": 2.3822469799973547e-06, + "loss": 0.1638, + "step": 11194 + }, + { + "epoch": 2.3474522960788424, + "grad_norm": 4.196364133885519, + "learning_rate": 2.380780537213192e-06, + "loss": 0.1076, + "step": 11195 + }, + { + "epoch": 2.34766198364437, + "grad_norm": 5.220985055518679, + "learning_rate": 2.3793144849272396e-06, + "loss": 0.1455, + "step": 11196 + }, + { + "epoch": 2.347871671209897, + "grad_norm": 4.994038567378567, + "learning_rate": 2.377848823214638e-06, + "loss": 0.1777, + "step": 11197 + }, + { + "epoch": 2.3480813587754246, + "grad_norm": 4.550143944029153, + "learning_rate": 2.3763835521505053e-06, + "loss": 0.132, + "step": 11198 + }, + { + "epoch": 2.348291046340952, + "grad_norm": 3.0942931355476104, + "learning_rate": 2.3749186718099392e-06, + "loss": 0.1384, + "step": 11199 + }, + { + "epoch": 2.348500733906479, + "grad_norm": 4.216851733223235, + "learning_rate": 2.3734541822680133e-06, + "loss": 0.1654, + "step": 11200 + }, + { + "epoch": 2.3487104214720067, + "grad_norm": 4.4839489504743035, + "learning_rate": 2.37199008359979e-06, + "loss": 0.1558, + "step": 11201 + }, + { + "epoch": 2.3489201090375342, + "grad_norm": 3.5548296669417856, + "learning_rate": 2.370526375880303e-06, + "loss": 0.1307, + "step": 11202 + }, + { + "epoch": 2.3491297966030613, + "grad_norm": 3.8556444621193, + "learning_rate": 2.3690630591845755e-06, + "loss": 0.1192, + "step": 11203 + }, + { + "epoch": 2.349339484168589, + "grad_norm": 4.683060179590733, + "learning_rate": 2.3676001335876e-06, + "loss": 0.1684, + "step": 11204 + }, + { + "epoch": 2.3495491717341164, + "grad_norm": 4.433197954025936, + "learning_rate": 2.366137599164353e-06, + "loss": 0.1745, + "step": 11205 + }, + { + "epoch": 2.3497588592996435, + "grad_norm": 4.3634272325132395, + "learning_rate": 2.3646754559897944e-06, + "loss": 0.163, + "step": 11206 + }, + { + "epoch": 2.349968546865171, + "grad_norm": 4.302417164974167, + "learning_rate": 2.363213704138867e-06, + "loss": 0.1817, + "step": 11207 + }, + { + "epoch": 2.350178234430698, + "grad_norm": 4.821051137584733, + "learning_rate": 2.3617523436864774e-06, + "loss": 0.1674, + "step": 11208 + }, + { + "epoch": 2.3503879219962256, + "grad_norm": 4.574793610296586, + "learning_rate": 2.3602913747075274e-06, + "loss": 0.1422, + "step": 11209 + }, + { + "epoch": 2.350597609561753, + "grad_norm": 4.706867981686905, + "learning_rate": 2.3588307972768976e-06, + "loss": 0.1545, + "step": 11210 + }, + { + "epoch": 2.35080729712728, + "grad_norm": 3.9910241863390024, + "learning_rate": 2.357370611469443e-06, + "loss": 0.1336, + "step": 11211 + }, + { + "epoch": 2.3510169846928077, + "grad_norm": 5.7962172419756826, + "learning_rate": 2.3559108173599974e-06, + "loss": 0.1698, + "step": 11212 + }, + { + "epoch": 2.3512266722583353, + "grad_norm": 4.603743608511823, + "learning_rate": 2.354451415023382e-06, + "loss": 0.1416, + "step": 11213 + }, + { + "epoch": 2.3514363598238623, + "grad_norm": 6.474164427921448, + "learning_rate": 2.352992404534391e-06, + "loss": 0.1804, + "step": 11214 + }, + { + "epoch": 2.35164604738939, + "grad_norm": 5.325426079370344, + "learning_rate": 2.3515337859678032e-06, + "loss": 0.1954, + "step": 11215 + }, + { + "epoch": 2.351855734954917, + "grad_norm": 4.0750923977894065, + "learning_rate": 2.350075559398376e-06, + "loss": 0.1405, + "step": 11216 + }, + { + "epoch": 2.3520654225204445, + "grad_norm": 3.957539865346227, + "learning_rate": 2.3486177249008413e-06, + "loss": 0.1328, + "step": 11217 + }, + { + "epoch": 2.352275110085972, + "grad_norm": 5.002165358277466, + "learning_rate": 2.3471602825499186e-06, + "loss": 0.156, + "step": 11218 + }, + { + "epoch": 2.352484797651499, + "grad_norm": 4.52823497995379, + "learning_rate": 2.3457032324203076e-06, + "loss": 0.1311, + "step": 11219 + }, + { + "epoch": 2.3526944852170266, + "grad_norm": 4.697879679746434, + "learning_rate": 2.3442465745866814e-06, + "loss": 0.1604, + "step": 11220 + }, + { + "epoch": 2.352904172782554, + "grad_norm": 4.166875455636941, + "learning_rate": 2.3427903091236936e-06, + "loss": 0.1494, + "step": 11221 + }, + { + "epoch": 2.3531138603480812, + "grad_norm": 3.0467560653214933, + "learning_rate": 2.341334436105984e-06, + "loss": 0.11, + "step": 11222 + }, + { + "epoch": 2.3533235479136088, + "grad_norm": 4.069252138371243, + "learning_rate": 2.3398789556081703e-06, + "loss": 0.1403, + "step": 11223 + }, + { + "epoch": 2.3535332354791363, + "grad_norm": 4.864315720825959, + "learning_rate": 2.3384238677048455e-06, + "loss": 0.1382, + "step": 11224 + }, + { + "epoch": 2.3537429230446634, + "grad_norm": 3.7488576398269746, + "learning_rate": 2.3369691724705834e-06, + "loss": 0.1221, + "step": 11225 + }, + { + "epoch": 2.353952610610191, + "grad_norm": 5.645603803680687, + "learning_rate": 2.3355148699799447e-06, + "loss": 0.1716, + "step": 11226 + }, + { + "epoch": 2.3541622981757184, + "grad_norm": 4.515834626835535, + "learning_rate": 2.3340609603074595e-06, + "loss": 0.1656, + "step": 11227 + }, + { + "epoch": 2.3543719857412455, + "grad_norm": 3.9994883561642482, + "learning_rate": 2.332607443527649e-06, + "loss": 0.144, + "step": 11228 + }, + { + "epoch": 2.354581673306773, + "grad_norm": 4.501584322690949, + "learning_rate": 2.3311543197150044e-06, + "loss": 0.1381, + "step": 11229 + }, + { + "epoch": 2.3547913608723, + "grad_norm": 4.768313099543709, + "learning_rate": 2.329701588944e-06, + "loss": 0.145, + "step": 11230 + }, + { + "epoch": 2.3550010484378276, + "grad_norm": 4.840034553262808, + "learning_rate": 2.3282492512890954e-06, + "loss": 0.1607, + "step": 11231 + }, + { + "epoch": 2.355210736003355, + "grad_norm": 4.142409711055538, + "learning_rate": 2.326797306824723e-06, + "loss": 0.1531, + "step": 11232 + }, + { + "epoch": 2.3554204235688823, + "grad_norm": 3.6331544244758955, + "learning_rate": 2.3253457556252933e-06, + "loss": 0.095, + "step": 11233 + }, + { + "epoch": 2.35563011113441, + "grad_norm": 3.8257343232881804, + "learning_rate": 2.3238945977652063e-06, + "loss": 0.1574, + "step": 11234 + }, + { + "epoch": 2.355839798699937, + "grad_norm": 3.918103285416472, + "learning_rate": 2.322443833318836e-06, + "loss": 0.1478, + "step": 11235 + }, + { + "epoch": 2.3560494862654644, + "grad_norm": 4.350963230021017, + "learning_rate": 2.320993462360536e-06, + "loss": 0.17, + "step": 11236 + }, + { + "epoch": 2.356259173830992, + "grad_norm": 5.2636941171505285, + "learning_rate": 2.319543484964636e-06, + "loss": 0.1694, + "step": 11237 + }, + { + "epoch": 2.356468861396519, + "grad_norm": 4.685358765623728, + "learning_rate": 2.3180939012054573e-06, + "loss": 0.1665, + "step": 11238 + }, + { + "epoch": 2.3566785489620465, + "grad_norm": 5.163522010701617, + "learning_rate": 2.316644711157285e-06, + "loss": 0.1799, + "step": 11239 + }, + { + "epoch": 2.356888236527574, + "grad_norm": 4.394498321197905, + "learning_rate": 2.3151959148944025e-06, + "loss": 0.1558, + "step": 11240 + }, + { + "epoch": 2.357097924093101, + "grad_norm": 4.408834672970853, + "learning_rate": 2.3137475124910567e-06, + "loss": 0.1504, + "step": 11241 + }, + { + "epoch": 2.3573076116586287, + "grad_norm": 3.5871876794783613, + "learning_rate": 2.31229950402148e-06, + "loss": 0.1275, + "step": 11242 + }, + { + "epoch": 2.357517299224156, + "grad_norm": 4.579123733696252, + "learning_rate": 2.3108518895598863e-06, + "loss": 0.1319, + "step": 11243 + }, + { + "epoch": 2.3577269867896833, + "grad_norm": 4.871814528670959, + "learning_rate": 2.309404669180473e-06, + "loss": 0.1557, + "step": 11244 + }, + { + "epoch": 2.357936674355211, + "grad_norm": 5.278381770408729, + "learning_rate": 2.307957842957409e-06, + "loss": 0.1701, + "step": 11245 + }, + { + "epoch": 2.3581463619207383, + "grad_norm": 4.1075139751892396, + "learning_rate": 2.3065114109648446e-06, + "loss": 0.14, + "step": 11246 + }, + { + "epoch": 2.3583560494862654, + "grad_norm": 4.896966515416556, + "learning_rate": 2.3050653732769156e-06, + "loss": 0.1332, + "step": 11247 + }, + { + "epoch": 2.358565737051793, + "grad_norm": 4.665470239518378, + "learning_rate": 2.303619729967733e-06, + "loss": 0.1679, + "step": 11248 + }, + { + "epoch": 2.35877542461732, + "grad_norm": 4.364236556757041, + "learning_rate": 2.3021744811113842e-06, + "loss": 0.1875, + "step": 11249 + }, + { + "epoch": 2.3589851121828476, + "grad_norm": 4.6961183711298, + "learning_rate": 2.300729626781949e-06, + "loss": 0.1535, + "step": 11250 + }, + { + "epoch": 2.359194799748375, + "grad_norm": 4.312771944096224, + "learning_rate": 2.29928516705347e-06, + "loss": 0.121, + "step": 11251 + }, + { + "epoch": 2.359404487313902, + "grad_norm": 4.488213940548144, + "learning_rate": 2.2978411019999857e-06, + "loss": 0.1395, + "step": 11252 + }, + { + "epoch": 2.3596141748794297, + "grad_norm": 4.599365551675834, + "learning_rate": 2.296397431695504e-06, + "loss": 0.1538, + "step": 11253 + }, + { + "epoch": 2.359823862444957, + "grad_norm": 4.048907471489787, + "learning_rate": 2.294954156214012e-06, + "loss": 0.1509, + "step": 11254 + }, + { + "epoch": 2.3600335500104843, + "grad_norm": 5.430306725560685, + "learning_rate": 2.293511275629483e-06, + "loss": 0.1932, + "step": 11255 + }, + { + "epoch": 2.360243237576012, + "grad_norm": 3.900783582179952, + "learning_rate": 2.29206879001587e-06, + "loss": 0.1101, + "step": 11256 + }, + { + "epoch": 2.360452925141539, + "grad_norm": 5.1910383344913225, + "learning_rate": 2.2906266994471005e-06, + "loss": 0.1816, + "step": 11257 + }, + { + "epoch": 2.3606626127070665, + "grad_norm": 5.132216885670787, + "learning_rate": 2.289185003997081e-06, + "loss": 0.1974, + "step": 11258 + }, + { + "epoch": 2.360872300272594, + "grad_norm": 4.284468269150263, + "learning_rate": 2.2877437037397066e-06, + "loss": 0.1422, + "step": 11259 + }, + { + "epoch": 2.361081987838121, + "grad_norm": 4.750823194590404, + "learning_rate": 2.2863027987488405e-06, + "loss": 0.1851, + "step": 11260 + }, + { + "epoch": 2.3612916754036486, + "grad_norm": 4.335870240201018, + "learning_rate": 2.2848622890983387e-06, + "loss": 0.1262, + "step": 11261 + }, + { + "epoch": 2.361501362969176, + "grad_norm": 4.441702641374946, + "learning_rate": 2.2834221748620244e-06, + "loss": 0.1463, + "step": 11262 + }, + { + "epoch": 2.361711050534703, + "grad_norm": 3.829497424626141, + "learning_rate": 2.281982456113705e-06, + "loss": 0.1641, + "step": 11263 + }, + { + "epoch": 2.3619207381002307, + "grad_norm": 5.0208891224968575, + "learning_rate": 2.280543132927172e-06, + "loss": 0.202, + "step": 11264 + }, + { + "epoch": 2.3621304256657583, + "grad_norm": 4.47880916261504, + "learning_rate": 2.279104205376197e-06, + "loss": 0.149, + "step": 11265 + }, + { + "epoch": 2.3623401132312853, + "grad_norm": 4.964393521400657, + "learning_rate": 2.277665673534516e-06, + "loss": 0.1506, + "step": 11266 + }, + { + "epoch": 2.362549800796813, + "grad_norm": 5.27619897170765, + "learning_rate": 2.2762275374758645e-06, + "loss": 0.2112, + "step": 11267 + }, + { + "epoch": 2.36275948836234, + "grad_norm": 4.231354239254277, + "learning_rate": 2.2747897972739495e-06, + "loss": 0.1351, + "step": 11268 + }, + { + "epoch": 2.3629691759278675, + "grad_norm": 3.6769278303354516, + "learning_rate": 2.2733524530024576e-06, + "loss": 0.1308, + "step": 11269 + }, + { + "epoch": 2.363178863493395, + "grad_norm": 3.721202586159344, + "learning_rate": 2.2719155047350494e-06, + "loss": 0.1404, + "step": 11270 + }, + { + "epoch": 2.363388551058922, + "grad_norm": 4.362660945352736, + "learning_rate": 2.2704789525453787e-06, + "loss": 0.1441, + "step": 11271 + }, + { + "epoch": 2.3635982386244496, + "grad_norm": 4.048020554720356, + "learning_rate": 2.269042796507065e-06, + "loss": 0.1562, + "step": 11272 + }, + { + "epoch": 2.3638079261899767, + "grad_norm": 3.9024374849643366, + "learning_rate": 2.2676070366937198e-06, + "loss": 0.1307, + "step": 11273 + }, + { + "epoch": 2.3640176137555042, + "grad_norm": 4.294012118742287, + "learning_rate": 2.266171673178924e-06, + "loss": 0.1134, + "step": 11274 + }, + { + "epoch": 2.3642273013210318, + "grad_norm": 3.539284346271426, + "learning_rate": 2.2647367060362425e-06, + "loss": 0.1316, + "step": 11275 + }, + { + "epoch": 2.364436988886559, + "grad_norm": 4.67855629688292, + "learning_rate": 2.263302135339219e-06, + "loss": 0.1788, + "step": 11276 + }, + { + "epoch": 2.3646466764520864, + "grad_norm": 3.8126406372524966, + "learning_rate": 2.2618679611613835e-06, + "loss": 0.1435, + "step": 11277 + }, + { + "epoch": 2.364856364017614, + "grad_norm": 4.489008585505364, + "learning_rate": 2.2604341835762365e-06, + "loss": 0.14, + "step": 11278 + }, + { + "epoch": 2.365066051583141, + "grad_norm": 4.168373813091158, + "learning_rate": 2.2590008026572574e-06, + "loss": 0.1752, + "step": 11279 + }, + { + "epoch": 2.3652757391486685, + "grad_norm": 3.59226251558483, + "learning_rate": 2.257567818477917e-06, + "loss": 0.1255, + "step": 11280 + }, + { + "epoch": 2.365485426714196, + "grad_norm": 3.7222066569754095, + "learning_rate": 2.2561352311116515e-06, + "loss": 0.1146, + "step": 11281 + }, + { + "epoch": 2.365695114279723, + "grad_norm": 3.4824075514027606, + "learning_rate": 2.2547030406318893e-06, + "loss": 0.119, + "step": 11282 + }, + { + "epoch": 2.3659048018452506, + "grad_norm": 5.130048207605561, + "learning_rate": 2.2532712471120265e-06, + "loss": 0.154, + "step": 11283 + }, + { + "epoch": 2.366114489410778, + "grad_norm": 3.908958754012853, + "learning_rate": 2.2518398506254514e-06, + "loss": 0.148, + "step": 11284 + }, + { + "epoch": 2.3663241769763053, + "grad_norm": 3.5358698543550147, + "learning_rate": 2.250408851245519e-06, + "loss": 0.1138, + "step": 11285 + }, + { + "epoch": 2.366533864541833, + "grad_norm": 4.2586166159511585, + "learning_rate": 2.248978249045578e-06, + "loss": 0.1298, + "step": 11286 + }, + { + "epoch": 2.36674355210736, + "grad_norm": 3.5956355255098633, + "learning_rate": 2.247548044098944e-06, + "loss": 0.1169, + "step": 11287 + }, + { + "epoch": 2.3669532396728874, + "grad_norm": 5.123813424369603, + "learning_rate": 2.2461182364789167e-06, + "loss": 0.1749, + "step": 11288 + }, + { + "epoch": 2.367162927238415, + "grad_norm": 3.8338985477232863, + "learning_rate": 2.2446888262587785e-06, + "loss": 0.1346, + "step": 11289 + }, + { + "epoch": 2.367372614803942, + "grad_norm": 6.150829654304728, + "learning_rate": 2.2432598135117934e-06, + "loss": 0.1625, + "step": 11290 + }, + { + "epoch": 2.3675823023694695, + "grad_norm": 4.407254728450832, + "learning_rate": 2.2418311983111916e-06, + "loss": 0.1243, + "step": 11291 + }, + { + "epoch": 2.3677919899349966, + "grad_norm": 3.8861953760367918, + "learning_rate": 2.240402980730196e-06, + "loss": 0.1438, + "step": 11292 + }, + { + "epoch": 2.368001677500524, + "grad_norm": 4.873657041011035, + "learning_rate": 2.23897516084201e-06, + "loss": 0.1774, + "step": 11293 + }, + { + "epoch": 2.3682113650660517, + "grad_norm": 5.103229242582608, + "learning_rate": 2.2375477387198064e-06, + "loss": 0.134, + "step": 11294 + }, + { + "epoch": 2.3684210526315788, + "grad_norm": 4.760257857212763, + "learning_rate": 2.2361207144367412e-06, + "loss": 0.1749, + "step": 11295 + }, + { + "epoch": 2.3686307401971063, + "grad_norm": 3.2985139816363027, + "learning_rate": 2.2346940880659594e-06, + "loss": 0.1215, + "step": 11296 + }, + { + "epoch": 2.368840427762634, + "grad_norm": 4.871932870780834, + "learning_rate": 2.2332678596805713e-06, + "loss": 0.1469, + "step": 11297 + }, + { + "epoch": 2.369050115328161, + "grad_norm": 4.815596081549712, + "learning_rate": 2.231842029353678e-06, + "loss": 0.1399, + "step": 11298 + }, + { + "epoch": 2.3692598028936884, + "grad_norm": 4.761103157052505, + "learning_rate": 2.2304165971583535e-06, + "loss": 0.1538, + "step": 11299 + }, + { + "epoch": 2.369469490459216, + "grad_norm": 4.404011998087831, + "learning_rate": 2.2289915631676517e-06, + "loss": 0.1286, + "step": 11300 + }, + { + "epoch": 2.369679178024743, + "grad_norm": 4.4000847308442745, + "learning_rate": 2.2275669274546107e-06, + "loss": 0.1421, + "step": 11301 + }, + { + "epoch": 2.3698888655902706, + "grad_norm": 4.448380977190623, + "learning_rate": 2.2261426900922476e-06, + "loss": 0.1549, + "step": 11302 + }, + { + "epoch": 2.370098553155798, + "grad_norm": 5.318164023679161, + "learning_rate": 2.2247188511535556e-06, + "loss": 0.1947, + "step": 11303 + }, + { + "epoch": 2.370308240721325, + "grad_norm": 4.039420609352076, + "learning_rate": 2.2232954107115047e-06, + "loss": 0.1657, + "step": 11304 + }, + { + "epoch": 2.3705179282868527, + "grad_norm": 3.8977114846128815, + "learning_rate": 2.2218723688390543e-06, + "loss": 0.1346, + "step": 11305 + }, + { + "epoch": 2.37072761585238, + "grad_norm": 4.013103170084518, + "learning_rate": 2.220449725609133e-06, + "loss": 0.1219, + "step": 11306 + }, + { + "epoch": 2.3709373034179073, + "grad_norm": 4.46814215352868, + "learning_rate": 2.21902748109466e-06, + "loss": 0.1515, + "step": 11307 + }, + { + "epoch": 2.371146990983435, + "grad_norm": 4.465558093998286, + "learning_rate": 2.217605635368523e-06, + "loss": 0.1346, + "step": 11308 + }, + { + "epoch": 2.371356678548962, + "grad_norm": 3.177579802145799, + "learning_rate": 2.216184188503592e-06, + "loss": 0.1074, + "step": 11309 + }, + { + "epoch": 2.3715663661144895, + "grad_norm": 5.203711256515378, + "learning_rate": 2.2147631405727244e-06, + "loss": 0.1604, + "step": 11310 + }, + { + "epoch": 2.3717760536800165, + "grad_norm": 5.374258829077727, + "learning_rate": 2.21334249164875e-06, + "loss": 0.1637, + "step": 11311 + }, + { + "epoch": 2.371985741245544, + "grad_norm": 4.0002657928439564, + "learning_rate": 2.211922241804474e-06, + "loss": 0.1753, + "step": 11312 + }, + { + "epoch": 2.3721954288110716, + "grad_norm": 3.7921261806721183, + "learning_rate": 2.210502391112691e-06, + "loss": 0.1189, + "step": 11313 + }, + { + "epoch": 2.3724051163765987, + "grad_norm": 3.4986333616972134, + "learning_rate": 2.2090829396461745e-06, + "loss": 0.1361, + "step": 11314 + }, + { + "epoch": 2.372614803942126, + "grad_norm": 4.660136249176361, + "learning_rate": 2.207663887477669e-06, + "loss": 0.1445, + "step": 11315 + }, + { + "epoch": 2.3728244915076537, + "grad_norm": 3.896538656855729, + "learning_rate": 2.2062452346799025e-06, + "loss": 0.1369, + "step": 11316 + }, + { + "epoch": 2.373034179073181, + "grad_norm": 4.475671952457565, + "learning_rate": 2.204826981325587e-06, + "loss": 0.175, + "step": 11317 + }, + { + "epoch": 2.3732438666387083, + "grad_norm": 3.70276803597874, + "learning_rate": 2.2034091274874058e-06, + "loss": 0.0964, + "step": 11318 + }, + { + "epoch": 2.373453554204236, + "grad_norm": 4.256423600081551, + "learning_rate": 2.2019916732380332e-06, + "loss": 0.1221, + "step": 11319 + }, + { + "epoch": 2.373663241769763, + "grad_norm": 3.885684905381357, + "learning_rate": 2.200574618650112e-06, + "loss": 0.124, + "step": 11320 + }, + { + "epoch": 2.3738729293352905, + "grad_norm": 4.16887216957322, + "learning_rate": 2.1991579637962657e-06, + "loss": 0.1557, + "step": 11321 + }, + { + "epoch": 2.374082616900818, + "grad_norm": 3.5837698356434484, + "learning_rate": 2.1977417087491048e-06, + "loss": 0.1222, + "step": 11322 + }, + { + "epoch": 2.374292304466345, + "grad_norm": 4.848672981230273, + "learning_rate": 2.1963258535812147e-06, + "loss": 0.1543, + "step": 11323 + }, + { + "epoch": 2.3745019920318726, + "grad_norm": 4.235418771268135, + "learning_rate": 2.194910398365161e-06, + "loss": 0.1363, + "step": 11324 + }, + { + "epoch": 2.3747116795973997, + "grad_norm": 4.435760444507633, + "learning_rate": 2.193495343173483e-06, + "loss": 0.1165, + "step": 11325 + }, + { + "epoch": 2.3749213671629272, + "grad_norm": 4.025518248985951, + "learning_rate": 2.1920806880787114e-06, + "loss": 0.1458, + "step": 11326 + }, + { + "epoch": 2.3751310547284548, + "grad_norm": 4.485366469424289, + "learning_rate": 2.1906664331533466e-06, + "loss": 0.1266, + "step": 11327 + }, + { + "epoch": 2.375340742293982, + "grad_norm": 3.8943925683870435, + "learning_rate": 2.189252578469869e-06, + "loss": 0.1374, + "step": 11328 + }, + { + "epoch": 2.3755504298595094, + "grad_norm": 3.248312120250793, + "learning_rate": 2.187839124100747e-06, + "loss": 0.1048, + "step": 11329 + }, + { + "epoch": 2.3757601174250365, + "grad_norm": 7.013154711627669, + "learning_rate": 2.186426070118416e-06, + "loss": 0.1974, + "step": 11330 + }, + { + "epoch": 2.375969804990564, + "grad_norm": 4.5600487610600355, + "learning_rate": 2.185013416595303e-06, + "loss": 0.1457, + "step": 11331 + }, + { + "epoch": 2.3761794925560915, + "grad_norm": 4.118171270966639, + "learning_rate": 2.183601163603808e-06, + "loss": 0.1589, + "step": 11332 + }, + { + "epoch": 2.3763891801216186, + "grad_norm": 5.5203675605358615, + "learning_rate": 2.182189311216307e-06, + "loss": 0.2032, + "step": 11333 + }, + { + "epoch": 2.376598867687146, + "grad_norm": 4.584080484818716, + "learning_rate": 2.180777859505163e-06, + "loss": 0.1354, + "step": 11334 + }, + { + "epoch": 2.3768085552526736, + "grad_norm": 4.484336487965244, + "learning_rate": 2.1793668085427187e-06, + "loss": 0.1821, + "step": 11335 + }, + { + "epoch": 2.3770182428182007, + "grad_norm": 3.824433790449907, + "learning_rate": 2.1779561584012888e-06, + "loss": 0.143, + "step": 11336 + }, + { + "epoch": 2.3772279303837283, + "grad_norm": 4.082295600417207, + "learning_rate": 2.1765459091531705e-06, + "loss": 0.1549, + "step": 11337 + }, + { + "epoch": 2.377437617949256, + "grad_norm": 5.1659436411971384, + "learning_rate": 2.175136060870646e-06, + "loss": 0.1764, + "step": 11338 + }, + { + "epoch": 2.377647305514783, + "grad_norm": 5.8472588511752726, + "learning_rate": 2.173726613625967e-06, + "loss": 0.1607, + "step": 11339 + }, + { + "epoch": 2.3778569930803104, + "grad_norm": 5.395234774935025, + "learning_rate": 2.1723175674913755e-06, + "loss": 0.1309, + "step": 11340 + }, + { + "epoch": 2.378066680645838, + "grad_norm": 3.802500942698872, + "learning_rate": 2.1709089225390854e-06, + "loss": 0.1156, + "step": 11341 + }, + { + "epoch": 2.378276368211365, + "grad_norm": 3.1416349046092207, + "learning_rate": 2.16950067884129e-06, + "loss": 0.0874, + "step": 11342 + }, + { + "epoch": 2.3784860557768925, + "grad_norm": 3.7829500043774127, + "learning_rate": 2.1680928364701646e-06, + "loss": 0.1447, + "step": 11343 + }, + { + "epoch": 2.3786957433424196, + "grad_norm": 4.429311859701148, + "learning_rate": 2.1666853954978695e-06, + "loss": 0.153, + "step": 11344 + }, + { + "epoch": 2.378905430907947, + "grad_norm": 3.449635741097212, + "learning_rate": 2.1652783559965328e-06, + "loss": 0.1229, + "step": 11345 + }, + { + "epoch": 2.3791151184734747, + "grad_norm": 4.167732533730095, + "learning_rate": 2.1638717180382662e-06, + "loss": 0.1148, + "step": 11346 + }, + { + "epoch": 2.3793248060390018, + "grad_norm": 4.230588248415238, + "learning_rate": 2.162465481695165e-06, + "loss": 0.1476, + "step": 11347 + }, + { + "epoch": 2.3795344936045293, + "grad_norm": 4.543630005415105, + "learning_rate": 2.1610596470393076e-06, + "loss": 0.1189, + "step": 11348 + }, + { + "epoch": 2.3797441811700564, + "grad_norm": 4.115427992809845, + "learning_rate": 2.1596542141427335e-06, + "loss": 0.1374, + "step": 11349 + }, + { + "epoch": 2.379953868735584, + "grad_norm": 3.18391371695857, + "learning_rate": 2.15824918307748e-06, + "loss": 0.1039, + "step": 11350 + }, + { + "epoch": 2.3801635563011114, + "grad_norm": 5.090937041776184, + "learning_rate": 2.156844553915558e-06, + "loss": 0.1573, + "step": 11351 + }, + { + "epoch": 2.3803732438666385, + "grad_norm": 6.802472181111065, + "learning_rate": 2.155440326728957e-06, + "loss": 0.2057, + "step": 11352 + }, + { + "epoch": 2.380582931432166, + "grad_norm": 5.572264663303104, + "learning_rate": 2.1540365015896427e-06, + "loss": 0.1783, + "step": 11353 + }, + { + "epoch": 2.3807926189976936, + "grad_norm": 4.984515869009512, + "learning_rate": 2.152633078569567e-06, + "loss": 0.1828, + "step": 11354 + }, + { + "epoch": 2.3810023065632206, + "grad_norm": 4.415013843638168, + "learning_rate": 2.151230057740654e-06, + "loss": 0.1307, + "step": 11355 + }, + { + "epoch": 2.381211994128748, + "grad_norm": 3.950344556764105, + "learning_rate": 2.1498274391748177e-06, + "loss": 0.1437, + "step": 11356 + }, + { + "epoch": 2.3814216816942757, + "grad_norm": 4.97114052206461, + "learning_rate": 2.1484252229439397e-06, + "loss": 0.1458, + "step": 11357 + }, + { + "epoch": 2.381631369259803, + "grad_norm": 4.13293003913876, + "learning_rate": 2.147023409119884e-06, + "loss": 0.1211, + "step": 11358 + }, + { + "epoch": 2.3818410568253303, + "grad_norm": 4.21688639233053, + "learning_rate": 2.1456219977744985e-06, + "loss": 0.1561, + "step": 11359 + }, + { + "epoch": 2.382050744390858, + "grad_norm": 5.493665500049063, + "learning_rate": 2.144220988979612e-06, + "loss": 0.1525, + "step": 11360 + }, + { + "epoch": 2.382260431956385, + "grad_norm": 3.3391852782402243, + "learning_rate": 2.1428203828070247e-06, + "loss": 0.107, + "step": 11361 + }, + { + "epoch": 2.3824701195219125, + "grad_norm": 4.338641993242093, + "learning_rate": 2.141420179328517e-06, + "loss": 0.1473, + "step": 11362 + }, + { + "epoch": 2.3826798070874395, + "grad_norm": 3.9598423837772954, + "learning_rate": 2.1400203786158593e-06, + "loss": 0.149, + "step": 11363 + }, + { + "epoch": 2.382889494652967, + "grad_norm": 4.159956637258988, + "learning_rate": 2.1386209807407854e-06, + "loss": 0.1229, + "step": 11364 + }, + { + "epoch": 2.3830991822184946, + "grad_norm": 3.997759874769324, + "learning_rate": 2.137221985775024e-06, + "loss": 0.1179, + "step": 11365 + }, + { + "epoch": 2.3833088697840217, + "grad_norm": 4.401835478252226, + "learning_rate": 2.135823393790274e-06, + "loss": 0.1627, + "step": 11366 + }, + { + "epoch": 2.383518557349549, + "grad_norm": 4.260673265776229, + "learning_rate": 2.1344252048582115e-06, + "loss": 0.1655, + "step": 11367 + }, + { + "epoch": 2.3837282449150763, + "grad_norm": 3.782840508042572, + "learning_rate": 2.1330274190504995e-06, + "loss": 0.1469, + "step": 11368 + }, + { + "epoch": 2.383937932480604, + "grad_norm": 4.261506147440046, + "learning_rate": 2.131630036438783e-06, + "loss": 0.1178, + "step": 11369 + }, + { + "epoch": 2.3841476200461313, + "grad_norm": 4.673290365816312, + "learning_rate": 2.130233057094668e-06, + "loss": 0.1854, + "step": 11370 + }, + { + "epoch": 2.3843573076116584, + "grad_norm": 5.916636588351488, + "learning_rate": 2.128836481089758e-06, + "loss": 0.1211, + "step": 11371 + }, + { + "epoch": 2.384566995177186, + "grad_norm": 4.73529721946947, + "learning_rate": 2.127440308495633e-06, + "loss": 0.1535, + "step": 11372 + }, + { + "epoch": 2.3847766827427135, + "grad_norm": 3.674921725905945, + "learning_rate": 2.126044539383847e-06, + "loss": 0.1276, + "step": 11373 + }, + { + "epoch": 2.3849863703082406, + "grad_norm": 4.876113723118431, + "learning_rate": 2.1246491738259324e-06, + "loss": 0.1861, + "step": 11374 + }, + { + "epoch": 2.385196057873768, + "grad_norm": 4.106452377394905, + "learning_rate": 2.1232542118934085e-06, + "loss": 0.153, + "step": 11375 + }, + { + "epoch": 2.3854057454392956, + "grad_norm": 5.321737130549812, + "learning_rate": 2.121859653657765e-06, + "loss": 0.1792, + "step": 11376 + }, + { + "epoch": 2.3856154330048227, + "grad_norm": 4.447496872863494, + "learning_rate": 2.1204654991904818e-06, + "loss": 0.1551, + "step": 11377 + }, + { + "epoch": 2.3858251205703502, + "grad_norm": 4.624052651765212, + "learning_rate": 2.1190717485630073e-06, + "loss": 0.1714, + "step": 11378 + }, + { + "epoch": 2.3860348081358778, + "grad_norm": 5.343328107944146, + "learning_rate": 2.1176784018467733e-06, + "loss": 0.17, + "step": 11379 + }, + { + "epoch": 2.386244495701405, + "grad_norm": 4.307459477668443, + "learning_rate": 2.116285459113191e-06, + "loss": 0.1401, + "step": 11380 + }, + { + "epoch": 2.3864541832669324, + "grad_norm": 4.082646392258286, + "learning_rate": 2.1148929204336555e-06, + "loss": 0.1741, + "step": 11381 + }, + { + "epoch": 2.3866638708324595, + "grad_norm": 4.264783186472907, + "learning_rate": 2.1135007858795344e-06, + "loss": 0.1255, + "step": 11382 + }, + { + "epoch": 2.386873558397987, + "grad_norm": 4.501729745039637, + "learning_rate": 2.1121090555221725e-06, + "loss": 0.1439, + "step": 11383 + }, + { + "epoch": 2.3870832459635145, + "grad_norm": 3.5376355657527356, + "learning_rate": 2.110717729432906e-06, + "loss": 0.1314, + "step": 11384 + }, + { + "epoch": 2.3872929335290416, + "grad_norm": 4.4515916047713056, + "learning_rate": 2.1093268076830366e-06, + "loss": 0.1441, + "step": 11385 + }, + { + "epoch": 2.387502621094569, + "grad_norm": 5.06269533628943, + "learning_rate": 2.107936290343856e-06, + "loss": 0.1307, + "step": 11386 + }, + { + "epoch": 2.3877123086600966, + "grad_norm": 4.25874185087666, + "learning_rate": 2.106546177486629e-06, + "loss": 0.1351, + "step": 11387 + }, + { + "epoch": 2.3879219962256237, + "grad_norm": 5.065248224682678, + "learning_rate": 2.105156469182599e-06, + "loss": 0.1561, + "step": 11388 + }, + { + "epoch": 2.3881316837911513, + "grad_norm": 5.028934161245213, + "learning_rate": 2.1037671655029947e-06, + "loss": 0.1482, + "step": 11389 + }, + { + "epoch": 2.3883413713566783, + "grad_norm": 3.787209473207291, + "learning_rate": 2.1023782665190174e-06, + "loss": 0.1209, + "step": 11390 + }, + { + "epoch": 2.388551058922206, + "grad_norm": 3.940795359785619, + "learning_rate": 2.1009897723018504e-06, + "loss": 0.1384, + "step": 11391 + }, + { + "epoch": 2.3887607464877334, + "grad_norm": 4.1603141896842875, + "learning_rate": 2.0996016829226575e-06, + "loss": 0.1617, + "step": 11392 + }, + { + "epoch": 2.3889704340532605, + "grad_norm": 4.1160318716928135, + "learning_rate": 2.098213998452584e-06, + "loss": 0.1543, + "step": 11393 + }, + { + "epoch": 2.389180121618788, + "grad_norm": 4.73458972853516, + "learning_rate": 2.0968267189627467e-06, + "loss": 0.1601, + "step": 11394 + }, + { + "epoch": 2.3893898091843155, + "grad_norm": 4.622585266131426, + "learning_rate": 2.0954398445242465e-06, + "loss": 0.1662, + "step": 11395 + }, + { + "epoch": 2.3895994967498426, + "grad_norm": 4.5333241776694475, + "learning_rate": 2.094053375208165e-06, + "loss": 0.1284, + "step": 11396 + }, + { + "epoch": 2.38980918431537, + "grad_norm": 4.900293764428171, + "learning_rate": 2.092667311085559e-06, + "loss": 0.1406, + "step": 11397 + }, + { + "epoch": 2.3900188718808977, + "grad_norm": 3.5778838554273467, + "learning_rate": 2.091281652227469e-06, + "loss": 0.1105, + "step": 11398 + }, + { + "epoch": 2.3902285594464248, + "grad_norm": 4.248576431372044, + "learning_rate": 2.0898963987049126e-06, + "loss": 0.13, + "step": 11399 + }, + { + "epoch": 2.3904382470119523, + "grad_norm": 4.186385611051059, + "learning_rate": 2.088511550588882e-06, + "loss": 0.1509, + "step": 11400 + }, + { + "epoch": 2.3906479345774794, + "grad_norm": 4.162849241691574, + "learning_rate": 2.0871271079503562e-06, + "loss": 0.1305, + "step": 11401 + }, + { + "epoch": 2.390857622143007, + "grad_norm": 3.673584043444878, + "learning_rate": 2.085743070860293e-06, + "loss": 0.1118, + "step": 11402 + }, + { + "epoch": 2.3910673097085344, + "grad_norm": 6.927093421897617, + "learning_rate": 2.084359439389624e-06, + "loss": 0.136, + "step": 11403 + }, + { + "epoch": 2.3912769972740615, + "grad_norm": 3.5282438111794847, + "learning_rate": 2.08297621360926e-06, + "loss": 0.1164, + "step": 11404 + }, + { + "epoch": 2.391486684839589, + "grad_norm": 4.342567716477776, + "learning_rate": 2.0815933935900967e-06, + "loss": 0.1507, + "step": 11405 + }, + { + "epoch": 2.3916963724051166, + "grad_norm": 4.346386007382226, + "learning_rate": 2.08021097940301e-06, + "loss": 0.111, + "step": 11406 + }, + { + "epoch": 2.3919060599706436, + "grad_norm": 4.4997643715996976, + "learning_rate": 2.0788289711188426e-06, + "loss": 0.1312, + "step": 11407 + }, + { + "epoch": 2.392115747536171, + "grad_norm": 4.546432927933047, + "learning_rate": 2.0774473688084282e-06, + "loss": 0.1278, + "step": 11408 + }, + { + "epoch": 2.3923254351016983, + "grad_norm": 3.8990030364645683, + "learning_rate": 2.0760661725425814e-06, + "loss": 0.1277, + "step": 11409 + }, + { + "epoch": 2.392535122667226, + "grad_norm": 4.100017993630564, + "learning_rate": 2.0746853823920856e-06, + "loss": 0.1711, + "step": 11410 + }, + { + "epoch": 2.3927448102327533, + "grad_norm": 4.328232125327681, + "learning_rate": 2.0733049984277066e-06, + "loss": 0.1321, + "step": 11411 + }, + { + "epoch": 2.3929544977982804, + "grad_norm": 4.840124516390636, + "learning_rate": 2.071925020720198e-06, + "loss": 0.1613, + "step": 11412 + }, + { + "epoch": 2.393164185363808, + "grad_norm": 4.519081450246461, + "learning_rate": 2.070545449340279e-06, + "loss": 0.1172, + "step": 11413 + }, + { + "epoch": 2.3933738729293355, + "grad_norm": 3.3685284372121838, + "learning_rate": 2.0691662843586622e-06, + "loss": 0.1151, + "step": 11414 + }, + { + "epoch": 2.3935835604948625, + "grad_norm": 5.652389583849986, + "learning_rate": 2.067787525846028e-06, + "loss": 0.197, + "step": 11415 + }, + { + "epoch": 2.39379324806039, + "grad_norm": 4.134359512670427, + "learning_rate": 2.0664091738730385e-06, + "loss": 0.1316, + "step": 11416 + }, + { + "epoch": 2.3940029356259176, + "grad_norm": 3.242631990343553, + "learning_rate": 2.065031228510339e-06, + "loss": 0.1207, + "step": 11417 + }, + { + "epoch": 2.3942126231914447, + "grad_norm": 3.7407478336153672, + "learning_rate": 2.0636536898285544e-06, + "loss": 0.1295, + "step": 11418 + }, + { + "epoch": 2.394422310756972, + "grad_norm": 4.278124123850817, + "learning_rate": 2.062276557898283e-06, + "loss": 0.184, + "step": 11419 + }, + { + "epoch": 2.3946319983224993, + "grad_norm": 6.019886705361297, + "learning_rate": 2.0608998327901033e-06, + "loss": 0.1577, + "step": 11420 + }, + { + "epoch": 2.394841685888027, + "grad_norm": 4.269722429717349, + "learning_rate": 2.0595235145745783e-06, + "loss": 0.1244, + "step": 11421 + }, + { + "epoch": 2.3950513734535543, + "grad_norm": 4.618525136019961, + "learning_rate": 2.0581476033222425e-06, + "loss": 0.1656, + "step": 11422 + }, + { + "epoch": 2.3952610610190814, + "grad_norm": 4.267773007934097, + "learning_rate": 2.05677209910362e-06, + "loss": 0.1314, + "step": 11423 + }, + { + "epoch": 2.395470748584609, + "grad_norm": 4.877740221286599, + "learning_rate": 2.0553970019892035e-06, + "loss": 0.1708, + "step": 11424 + }, + { + "epoch": 2.3956804361501365, + "grad_norm": 4.445038521792703, + "learning_rate": 2.0540223120494673e-06, + "loss": 0.1465, + "step": 11425 + }, + { + "epoch": 2.3958901237156636, + "grad_norm": 4.462170813606795, + "learning_rate": 2.05264802935487e-06, + "loss": 0.1584, + "step": 11426 + }, + { + "epoch": 2.396099811281191, + "grad_norm": 4.469517611365437, + "learning_rate": 2.0512741539758496e-06, + "loss": 0.1416, + "step": 11427 + }, + { + "epoch": 2.396309498846718, + "grad_norm": 4.3615891500851385, + "learning_rate": 2.049900685982811e-06, + "loss": 0.1524, + "step": 11428 + }, + { + "epoch": 2.3965191864122457, + "grad_norm": 4.3709207912636945, + "learning_rate": 2.0485276254461493e-06, + "loss": 0.1384, + "step": 11429 + }, + { + "epoch": 2.3967288739777732, + "grad_norm": 3.9168267165179333, + "learning_rate": 2.047154972436243e-06, + "loss": 0.1537, + "step": 11430 + }, + { + "epoch": 2.3969385615433003, + "grad_norm": 4.838711024804494, + "learning_rate": 2.0457827270234354e-06, + "loss": 0.1908, + "step": 11431 + }, + { + "epoch": 2.397148249108828, + "grad_norm": 4.460202403931419, + "learning_rate": 2.044410889278058e-06, + "loss": 0.1106, + "step": 11432 + }, + { + "epoch": 2.3973579366743554, + "grad_norm": 3.4130607815526264, + "learning_rate": 2.043039459270422e-06, + "loss": 0.1339, + "step": 11433 + }, + { + "epoch": 2.3975676242398825, + "grad_norm": 4.209958485952366, + "learning_rate": 2.041668437070813e-06, + "loss": 0.1111, + "step": 11434 + }, + { + "epoch": 2.39777731180541, + "grad_norm": 4.41147714653408, + "learning_rate": 2.040297822749502e-06, + "loss": 0.1685, + "step": 11435 + }, + { + "epoch": 2.3979869993709375, + "grad_norm": 3.733852203362261, + "learning_rate": 2.0389276163767326e-06, + "loss": 0.1117, + "step": 11436 + }, + { + "epoch": 2.3981966869364646, + "grad_norm": 3.2002577989476846, + "learning_rate": 2.0375578180227274e-06, + "loss": 0.1255, + "step": 11437 + }, + { + "epoch": 2.398406374501992, + "grad_norm": 4.4309621013440585, + "learning_rate": 2.036188427757695e-06, + "loss": 0.1391, + "step": 11438 + }, + { + "epoch": 2.398616062067519, + "grad_norm": 6.349846989759825, + "learning_rate": 2.034819445651821e-06, + "loss": 0.1683, + "step": 11439 + }, + { + "epoch": 2.3988257496330467, + "grad_norm": 5.70336901652201, + "learning_rate": 2.0334508717752646e-06, + "loss": 0.1562, + "step": 11440 + }, + { + "epoch": 2.3990354371985743, + "grad_norm": 4.0296362329724085, + "learning_rate": 2.0320827061981665e-06, + "loss": 0.1582, + "step": 11441 + }, + { + "epoch": 2.3992451247641013, + "grad_norm": 3.3567670685125783, + "learning_rate": 2.030714948990652e-06, + "loss": 0.1137, + "step": 11442 + }, + { + "epoch": 2.399454812329629, + "grad_norm": 4.282495544580757, + "learning_rate": 2.029347600222815e-06, + "loss": 0.1662, + "step": 11443 + }, + { + "epoch": 2.3996644998951564, + "grad_norm": 3.68115551077214, + "learning_rate": 2.027980659964741e-06, + "loss": 0.1312, + "step": 11444 + }, + { + "epoch": 2.3998741874606835, + "grad_norm": 3.699127362803209, + "learning_rate": 2.026614128286485e-06, + "loss": 0.1221, + "step": 11445 + }, + { + "epoch": 2.400083875026211, + "grad_norm": 4.500273948371631, + "learning_rate": 2.0252480052580804e-06, + "loss": 0.159, + "step": 11446 + }, + { + "epoch": 2.400293562591738, + "grad_norm": 4.995282686499235, + "learning_rate": 2.0238822909495482e-06, + "loss": 0.168, + "step": 11447 + }, + { + "epoch": 2.4005032501572656, + "grad_norm": 6.759062377194313, + "learning_rate": 2.022516985430888e-06, + "loss": 0.141, + "step": 11448 + }, + { + "epoch": 2.400712937722793, + "grad_norm": 5.918246469024353, + "learning_rate": 2.0211520887720626e-06, + "loss": 0.1346, + "step": 11449 + }, + { + "epoch": 2.4009226252883202, + "grad_norm": 6.923755534804394, + "learning_rate": 2.0197876010430305e-06, + "loss": 0.2228, + "step": 11450 + }, + { + "epoch": 2.4011323128538478, + "grad_norm": 4.340432402031409, + "learning_rate": 2.0184235223137283e-06, + "loss": 0.1493, + "step": 11451 + }, + { + "epoch": 2.4013420004193753, + "grad_norm": 3.4915353747652533, + "learning_rate": 2.017059852654064e-06, + "loss": 0.1006, + "step": 11452 + }, + { + "epoch": 2.4015516879849024, + "grad_norm": 4.673543733894659, + "learning_rate": 2.0156965921339256e-06, + "loss": 0.1653, + "step": 11453 + }, + { + "epoch": 2.40176137555043, + "grad_norm": 4.1003943044441185, + "learning_rate": 2.0143337408231868e-06, + "loss": 0.1461, + "step": 11454 + }, + { + "epoch": 2.4019710631159574, + "grad_norm": 4.376403938617877, + "learning_rate": 2.012971298791693e-06, + "loss": 0.1584, + "step": 11455 + }, + { + "epoch": 2.4021807506814845, + "grad_norm": 4.066676680163921, + "learning_rate": 2.0116092661092735e-06, + "loss": 0.1505, + "step": 11456 + }, + { + "epoch": 2.402390438247012, + "grad_norm": 4.0670722778198, + "learning_rate": 2.010247642845735e-06, + "loss": 0.1473, + "step": 11457 + }, + { + "epoch": 2.402600125812539, + "grad_norm": 5.110728670920862, + "learning_rate": 2.008886429070861e-06, + "loss": 0.1413, + "step": 11458 + }, + { + "epoch": 2.4028098133780667, + "grad_norm": 4.269787250327614, + "learning_rate": 2.007525624854416e-06, + "loss": 0.1374, + "step": 11459 + }, + { + "epoch": 2.403019500943594, + "grad_norm": 3.9932069343500736, + "learning_rate": 2.006165230266148e-06, + "loss": 0.1251, + "step": 11460 + }, + { + "epoch": 2.4032291885091213, + "grad_norm": 3.664657598078533, + "learning_rate": 2.004805245375777e-06, + "loss": 0.1283, + "step": 11461 + }, + { + "epoch": 2.403438876074649, + "grad_norm": 2.878912428334256, + "learning_rate": 2.0034456702530003e-06, + "loss": 0.0843, + "step": 11462 + }, + { + "epoch": 2.4036485636401763, + "grad_norm": 3.870955256401124, + "learning_rate": 2.0020865049675065e-06, + "loss": 0.1336, + "step": 11463 + }, + { + "epoch": 2.4038582512057034, + "grad_norm": 5.563738492685919, + "learning_rate": 2.000727749588948e-06, + "loss": 0.1756, + "step": 11464 + }, + { + "epoch": 2.404067938771231, + "grad_norm": 4.506380017819829, + "learning_rate": 1.9993694041869684e-06, + "loss": 0.13, + "step": 11465 + }, + { + "epoch": 2.404277626336758, + "grad_norm": 3.2383140499319802, + "learning_rate": 1.99801146883118e-06, + "loss": 0.1237, + "step": 11466 + }, + { + "epoch": 2.4044873139022855, + "grad_norm": 4.192634138357596, + "learning_rate": 1.996653943591186e-06, + "loss": 0.1475, + "step": 11467 + }, + { + "epoch": 2.404697001467813, + "grad_norm": 3.52117273596207, + "learning_rate": 1.995296828536559e-06, + "loss": 0.1349, + "step": 11468 + }, + { + "epoch": 2.40490668903334, + "grad_norm": 4.591496306345599, + "learning_rate": 1.9939401237368493e-06, + "loss": 0.1733, + "step": 11469 + }, + { + "epoch": 2.4051163765988677, + "grad_norm": 4.239550845333539, + "learning_rate": 1.9925838292615975e-06, + "loss": 0.1372, + "step": 11470 + }, + { + "epoch": 2.405326064164395, + "grad_norm": 3.7184402561968986, + "learning_rate": 1.9912279451803097e-06, + "loss": 0.1535, + "step": 11471 + }, + { + "epoch": 2.4055357517299223, + "grad_norm": 4.516264710314278, + "learning_rate": 1.989872471562483e-06, + "loss": 0.1189, + "step": 11472 + }, + { + "epoch": 2.40574543929545, + "grad_norm": 4.860444897327862, + "learning_rate": 1.988517408477585e-06, + "loss": 0.1164, + "step": 11473 + }, + { + "epoch": 2.4059551268609773, + "grad_norm": 4.028536062763015, + "learning_rate": 1.9871627559950624e-06, + "loss": 0.1411, + "step": 11474 + }, + { + "epoch": 2.4061648144265044, + "grad_norm": 3.578787247828352, + "learning_rate": 1.985808514184345e-06, + "loss": 0.1224, + "step": 11475 + }, + { + "epoch": 2.406374501992032, + "grad_norm": 3.4048350340908367, + "learning_rate": 1.984454683114846e-06, + "loss": 0.1004, + "step": 11476 + }, + { + "epoch": 2.4065841895575595, + "grad_norm": 3.4537123628553097, + "learning_rate": 1.9831012628559453e-06, + "loss": 0.1181, + "step": 11477 + }, + { + "epoch": 2.4067938771230866, + "grad_norm": 3.820163674785773, + "learning_rate": 1.981748253477007e-06, + "loss": 0.1375, + "step": 11478 + }, + { + "epoch": 2.407003564688614, + "grad_norm": 5.391529433941355, + "learning_rate": 1.98039565504738e-06, + "loss": 0.1498, + "step": 11479 + }, + { + "epoch": 2.407213252254141, + "grad_norm": 4.247035697643328, + "learning_rate": 1.9790434676363825e-06, + "loss": 0.1267, + "step": 11480 + }, + { + "epoch": 2.4074229398196687, + "grad_norm": 4.595513143211677, + "learning_rate": 1.9776916913133215e-06, + "loss": 0.1407, + "step": 11481 + }, + { + "epoch": 2.4076326273851962, + "grad_norm": 4.500333988958926, + "learning_rate": 1.9763403261474756e-06, + "loss": 0.1424, + "step": 11482 + }, + { + "epoch": 2.4078423149507233, + "grad_norm": 4.907634201009687, + "learning_rate": 1.9749893722081026e-06, + "loss": 0.122, + "step": 11483 + }, + { + "epoch": 2.408052002516251, + "grad_norm": 3.512054274838209, + "learning_rate": 1.9736388295644413e-06, + "loss": 0.1229, + "step": 11484 + }, + { + "epoch": 2.408261690081778, + "grad_norm": 6.235302119705029, + "learning_rate": 1.972288698285717e-06, + "loss": 0.1705, + "step": 11485 + }, + { + "epoch": 2.4084713776473055, + "grad_norm": 3.451189742314622, + "learning_rate": 1.9709389784411146e-06, + "loss": 0.1302, + "step": 11486 + }, + { + "epoch": 2.408681065212833, + "grad_norm": 3.981396903585127, + "learning_rate": 1.9695896700998162e-06, + "loss": 0.1389, + "step": 11487 + }, + { + "epoch": 2.40889075277836, + "grad_norm": 4.824694066435134, + "learning_rate": 1.9682407733309773e-06, + "loss": 0.1703, + "step": 11488 + }, + { + "epoch": 2.4091004403438876, + "grad_norm": 4.997114448024437, + "learning_rate": 1.966892288203729e-06, + "loss": 0.1439, + "step": 11489 + }, + { + "epoch": 2.409310127909415, + "grad_norm": 4.5509291088658435, + "learning_rate": 1.9655442147871818e-06, + "loss": 0.1408, + "step": 11490 + }, + { + "epoch": 2.409519815474942, + "grad_norm": 3.8545978549538567, + "learning_rate": 1.9641965531504304e-06, + "loss": 0.1119, + "step": 11491 + }, + { + "epoch": 2.4097295030404697, + "grad_norm": 4.802458653596581, + "learning_rate": 1.9628493033625406e-06, + "loss": 0.1442, + "step": 11492 + }, + { + "epoch": 2.4099391906059973, + "grad_norm": 3.7853202941746265, + "learning_rate": 1.9615024654925663e-06, + "loss": 0.1016, + "step": 11493 + }, + { + "epoch": 2.4101488781715243, + "grad_norm": 3.6928092437390667, + "learning_rate": 1.960156039609533e-06, + "loss": 0.118, + "step": 11494 + }, + { + "epoch": 2.410358565737052, + "grad_norm": 3.831769847557559, + "learning_rate": 1.9588100257824438e-06, + "loss": 0.1225, + "step": 11495 + }, + { + "epoch": 2.4105682533025794, + "grad_norm": 3.5504085280365416, + "learning_rate": 1.9574644240802886e-06, + "loss": 0.1177, + "step": 11496 + }, + { + "epoch": 2.4107779408681065, + "grad_norm": 4.409311481631019, + "learning_rate": 1.9561192345720326e-06, + "loss": 0.1392, + "step": 11497 + }, + { + "epoch": 2.410987628433634, + "grad_norm": 3.2102298837153476, + "learning_rate": 1.954774457326617e-06, + "loss": 0.0828, + "step": 11498 + }, + { + "epoch": 2.411197315999161, + "grad_norm": 5.153629986384603, + "learning_rate": 1.953430092412962e-06, + "loss": 0.164, + "step": 11499 + }, + { + "epoch": 2.4114070035646886, + "grad_norm": 4.403420521189587, + "learning_rate": 1.9520861398999735e-06, + "loss": 0.1596, + "step": 11500 + }, + { + "epoch": 2.411616691130216, + "grad_norm": 4.438476879761939, + "learning_rate": 1.9507425998565268e-06, + "loss": 0.1503, + "step": 11501 + }, + { + "epoch": 2.4118263786957432, + "grad_norm": 4.619566409163347, + "learning_rate": 1.949399472351484e-06, + "loss": 0.1464, + "step": 11502 + }, + { + "epoch": 2.4120360662612708, + "grad_norm": 4.942655211386968, + "learning_rate": 1.948056757453681e-06, + "loss": 0.1509, + "step": 11503 + }, + { + "epoch": 2.412245753826798, + "grad_norm": 5.237247911650881, + "learning_rate": 1.946714455231934e-06, + "loss": 0.1692, + "step": 11504 + }, + { + "epoch": 2.4124554413923254, + "grad_norm": 4.4252204177736285, + "learning_rate": 1.945372565755037e-06, + "loss": 0.1611, + "step": 11505 + }, + { + "epoch": 2.412665128957853, + "grad_norm": 3.358056187973224, + "learning_rate": 1.9440310890917726e-06, + "loss": 0.1274, + "step": 11506 + }, + { + "epoch": 2.41287481652338, + "grad_norm": 3.055634201741452, + "learning_rate": 1.9426900253108803e-06, + "loss": 0.0825, + "step": 11507 + }, + { + "epoch": 2.4130845040889075, + "grad_norm": 3.7323797757306134, + "learning_rate": 1.9413493744810996e-06, + "loss": 0.1231, + "step": 11508 + }, + { + "epoch": 2.413294191654435, + "grad_norm": 3.8998550964059198, + "learning_rate": 1.9400091366711438e-06, + "loss": 0.1375, + "step": 11509 + }, + { + "epoch": 2.413503879219962, + "grad_norm": 3.490424782197383, + "learning_rate": 1.938669311949699e-06, + "loss": 0.1205, + "step": 11510 + }, + { + "epoch": 2.4137135667854897, + "grad_norm": 4.894179441895004, + "learning_rate": 1.9373299003854306e-06, + "loss": 0.1181, + "step": 11511 + }, + { + "epoch": 2.413923254351017, + "grad_norm": 5.526502008675969, + "learning_rate": 1.9359909020469913e-06, + "loss": 0.1615, + "step": 11512 + }, + { + "epoch": 2.4141329419165443, + "grad_norm": 5.1778528258589835, + "learning_rate": 1.934652317003002e-06, + "loss": 0.1366, + "step": 11513 + }, + { + "epoch": 2.414342629482072, + "grad_norm": 5.391625647861582, + "learning_rate": 1.9333141453220726e-06, + "loss": 0.1791, + "step": 11514 + }, + { + "epoch": 2.4145523170475993, + "grad_norm": 3.8537333118109562, + "learning_rate": 1.9319763870727846e-06, + "loss": 0.1142, + "step": 11515 + }, + { + "epoch": 2.4147620046131264, + "grad_norm": 4.483120008508993, + "learning_rate": 1.9306390423236976e-06, + "loss": 0.1506, + "step": 11516 + }, + { + "epoch": 2.414971692178654, + "grad_norm": 5.387763111627581, + "learning_rate": 1.9293021111433553e-06, + "loss": 0.1606, + "step": 11517 + }, + { + "epoch": 2.415181379744181, + "grad_norm": 3.8602041090411285, + "learning_rate": 1.9279655936002796e-06, + "loss": 0.1416, + "step": 11518 + }, + { + "epoch": 2.4153910673097085, + "grad_norm": 4.518715966468808, + "learning_rate": 1.926629489762968e-06, + "loss": 0.1424, + "step": 11519 + }, + { + "epoch": 2.415600754875236, + "grad_norm": 4.074535804199905, + "learning_rate": 1.9252937996998954e-06, + "loss": 0.1508, + "step": 11520 + }, + { + "epoch": 2.415810442440763, + "grad_norm": 4.360592726929684, + "learning_rate": 1.923958523479522e-06, + "loss": 0.1045, + "step": 11521 + }, + { + "epoch": 2.4160201300062907, + "grad_norm": 5.057960642591632, + "learning_rate": 1.9226236611702797e-06, + "loss": 0.1813, + "step": 11522 + }, + { + "epoch": 2.4162298175718178, + "grad_norm": 3.8854775163669584, + "learning_rate": 1.9212892128405857e-06, + "loss": 0.1212, + "step": 11523 + }, + { + "epoch": 2.4164395051373453, + "grad_norm": 3.9785381479426363, + "learning_rate": 1.919955178558832e-06, + "loss": 0.1331, + "step": 11524 + }, + { + "epoch": 2.416649192702873, + "grad_norm": 5.301440180779231, + "learning_rate": 1.9186215583933877e-06, + "loss": 0.1786, + "step": 11525 + }, + { + "epoch": 2.4168588802684, + "grad_norm": 4.6769825934515055, + "learning_rate": 1.917288352412604e-06, + "loss": 0.1316, + "step": 11526 + }, + { + "epoch": 2.4170685678339274, + "grad_norm": 3.351400773552443, + "learning_rate": 1.9159555606848144e-06, + "loss": 0.1155, + "step": 11527 + }, + { + "epoch": 2.417278255399455, + "grad_norm": 4.0787514179447495, + "learning_rate": 1.9146231832783234e-06, + "loss": 0.1513, + "step": 11528 + }, + { + "epoch": 2.417487942964982, + "grad_norm": 3.8377240176352605, + "learning_rate": 1.913291220261414e-06, + "loss": 0.121, + "step": 11529 + }, + { + "epoch": 2.4176976305305096, + "grad_norm": 4.530193763261716, + "learning_rate": 1.9119596717023593e-06, + "loss": 0.158, + "step": 11530 + }, + { + "epoch": 2.417907318096037, + "grad_norm": 3.571589664579052, + "learning_rate": 1.910628537669398e-06, + "loss": 0.125, + "step": 11531 + }, + { + "epoch": 2.418117005661564, + "grad_norm": 4.046000040405084, + "learning_rate": 1.909297818230752e-06, + "loss": 0.1585, + "step": 11532 + }, + { + "epoch": 2.4183266932270917, + "grad_norm": 4.099647154459045, + "learning_rate": 1.907967513454626e-06, + "loss": 0.1574, + "step": 11533 + }, + { + "epoch": 2.4185363807926192, + "grad_norm": 4.140979480189836, + "learning_rate": 1.9066376234092021e-06, + "loss": 0.1425, + "step": 11534 + }, + { + "epoch": 2.4187460683581463, + "grad_norm": 3.9554759533095316, + "learning_rate": 1.9053081481626379e-06, + "loss": 0.1323, + "step": 11535 + }, + { + "epoch": 2.418955755923674, + "grad_norm": 3.589748420736371, + "learning_rate": 1.9039790877830677e-06, + "loss": 0.1319, + "step": 11536 + }, + { + "epoch": 2.419165443489201, + "grad_norm": 3.796030733400382, + "learning_rate": 1.9026504423386127e-06, + "loss": 0.1294, + "step": 11537 + }, + { + "epoch": 2.4193751310547285, + "grad_norm": 4.459965383244707, + "learning_rate": 1.901322211897365e-06, + "loss": 0.119, + "step": 11538 + }, + { + "epoch": 2.419584818620256, + "grad_norm": 4.805784570267907, + "learning_rate": 1.8999943965274036e-06, + "loss": 0.153, + "step": 11539 + }, + { + "epoch": 2.419794506185783, + "grad_norm": 4.8437393146058625, + "learning_rate": 1.8986669962967773e-06, + "loss": 0.1836, + "step": 11540 + }, + { + "epoch": 2.4200041937513106, + "grad_norm": 3.672803682336201, + "learning_rate": 1.8973400112735152e-06, + "loss": 0.1023, + "step": 11541 + }, + { + "epoch": 2.4202138813168377, + "grad_norm": 5.4751474122569705, + "learning_rate": 1.8960134415256325e-06, + "loss": 0.1621, + "step": 11542 + }, + { + "epoch": 2.420423568882365, + "grad_norm": 3.981012327132323, + "learning_rate": 1.8946872871211185e-06, + "loss": 0.1335, + "step": 11543 + }, + { + "epoch": 2.4206332564478927, + "grad_norm": 5.222317740462597, + "learning_rate": 1.8933615481279388e-06, + "loss": 0.1587, + "step": 11544 + }, + { + "epoch": 2.42084294401342, + "grad_norm": 4.334153410153875, + "learning_rate": 1.8920362246140366e-06, + "loss": 0.1298, + "step": 11545 + }, + { + "epoch": 2.4210526315789473, + "grad_norm": 3.966153720263226, + "learning_rate": 1.8907113166473444e-06, + "loss": 0.1425, + "step": 11546 + }, + { + "epoch": 2.421262319144475, + "grad_norm": 5.0173451169824, + "learning_rate": 1.8893868242957614e-06, + "loss": 0.1734, + "step": 11547 + }, + { + "epoch": 2.421472006710002, + "grad_norm": 4.104423916512362, + "learning_rate": 1.8880627476271684e-06, + "loss": 0.1625, + "step": 11548 + }, + { + "epoch": 2.4216816942755295, + "grad_norm": 4.677099452799848, + "learning_rate": 1.8867390867094316e-06, + "loss": 0.1556, + "step": 11549 + }, + { + "epoch": 2.421891381841057, + "grad_norm": 4.218474731361375, + "learning_rate": 1.8854158416103852e-06, + "loss": 0.1301, + "step": 11550 + }, + { + "epoch": 2.422101069406584, + "grad_norm": 3.8116331369338243, + "learning_rate": 1.884093012397854e-06, + "loss": 0.1491, + "step": 11551 + }, + { + "epoch": 2.4223107569721116, + "grad_norm": 5.279907844456461, + "learning_rate": 1.8827705991396327e-06, + "loss": 0.1791, + "step": 11552 + }, + { + "epoch": 2.422520444537639, + "grad_norm": 4.275882782585302, + "learning_rate": 1.881448601903494e-06, + "loss": 0.1426, + "step": 11553 + }, + { + "epoch": 2.4227301321031662, + "grad_norm": 3.7942859000921616, + "learning_rate": 1.8801270207571953e-06, + "loss": 0.0933, + "step": 11554 + }, + { + "epoch": 2.4229398196686938, + "grad_norm": 4.17888874628217, + "learning_rate": 1.8788058557684718e-06, + "loss": 0.1657, + "step": 11555 + }, + { + "epoch": 2.423149507234221, + "grad_norm": 4.8385185436556695, + "learning_rate": 1.8774851070050349e-06, + "loss": 0.1639, + "step": 11556 + }, + { + "epoch": 2.4233591947997484, + "grad_norm": 3.648020519856198, + "learning_rate": 1.8761647745345712e-06, + "loss": 0.1371, + "step": 11557 + }, + { + "epoch": 2.423568882365276, + "grad_norm": 4.438671663400621, + "learning_rate": 1.8748448584247547e-06, + "loss": 0.1784, + "step": 11558 + }, + { + "epoch": 2.423778569930803, + "grad_norm": 4.724968961977691, + "learning_rate": 1.8735253587432301e-06, + "loss": 0.1543, + "step": 11559 + }, + { + "epoch": 2.4239882574963305, + "grad_norm": 4.7704754760958155, + "learning_rate": 1.8722062755576277e-06, + "loss": 0.1645, + "step": 11560 + }, + { + "epoch": 2.4241979450618576, + "grad_norm": 3.884495031465616, + "learning_rate": 1.87088760893555e-06, + "loss": 0.1529, + "step": 11561 + }, + { + "epoch": 2.424407632627385, + "grad_norm": 4.170765093902806, + "learning_rate": 1.8695693589445797e-06, + "loss": 0.1341, + "step": 11562 + }, + { + "epoch": 2.4246173201929127, + "grad_norm": 3.6493213125385418, + "learning_rate": 1.8682515256522804e-06, + "loss": 0.1053, + "step": 11563 + }, + { + "epoch": 2.4248270077584397, + "grad_norm": 4.0516949637488135, + "learning_rate": 1.8669341091262005e-06, + "loss": 0.149, + "step": 11564 + }, + { + "epoch": 2.4250366953239673, + "grad_norm": 3.917069968248065, + "learning_rate": 1.8656171094338472e-06, + "loss": 0.119, + "step": 11565 + }, + { + "epoch": 2.425246382889495, + "grad_norm": 4.661929632093208, + "learning_rate": 1.8643005266427262e-06, + "loss": 0.1653, + "step": 11566 + }, + { + "epoch": 2.425456070455022, + "grad_norm": 4.587427803078533, + "learning_rate": 1.862984360820317e-06, + "loss": 0.1725, + "step": 11567 + }, + { + "epoch": 2.4256657580205494, + "grad_norm": 5.219391024111514, + "learning_rate": 1.861668612034071e-06, + "loss": 0.1495, + "step": 11568 + }, + { + "epoch": 2.425875445586077, + "grad_norm": 3.9288525951703788, + "learning_rate": 1.8603532803514225e-06, + "loss": 0.1328, + "step": 11569 + }, + { + "epoch": 2.426085133151604, + "grad_norm": 4.411072407997811, + "learning_rate": 1.859038365839787e-06, + "loss": 0.1409, + "step": 11570 + }, + { + "epoch": 2.4262948207171315, + "grad_norm": 4.802286478172426, + "learning_rate": 1.8577238685665532e-06, + "loss": 0.1659, + "step": 11571 + }, + { + "epoch": 2.426504508282659, + "grad_norm": 4.46855863915047, + "learning_rate": 1.856409788599096e-06, + "loss": 0.1217, + "step": 11572 + }, + { + "epoch": 2.426714195848186, + "grad_norm": 4.276144776655258, + "learning_rate": 1.8550961260047618e-06, + "loss": 0.1495, + "step": 11573 + }, + { + "epoch": 2.4269238834137137, + "grad_norm": 4.19834843643309, + "learning_rate": 1.853782880850874e-06, + "loss": 0.1741, + "step": 11574 + }, + { + "epoch": 2.4271335709792408, + "grad_norm": 5.535748703773797, + "learning_rate": 1.8524700532047435e-06, + "loss": 0.1669, + "step": 11575 + }, + { + "epoch": 2.4273432585447683, + "grad_norm": 4.9004793071666555, + "learning_rate": 1.8511576431336564e-06, + "loss": 0.1383, + "step": 11576 + }, + { + "epoch": 2.427552946110296, + "grad_norm": 3.8938871754300837, + "learning_rate": 1.8498456507048734e-06, + "loss": 0.1485, + "step": 11577 + }, + { + "epoch": 2.427762633675823, + "grad_norm": 3.8075118637340415, + "learning_rate": 1.848534075985634e-06, + "loss": 0.1247, + "step": 11578 + }, + { + "epoch": 2.4279723212413504, + "grad_norm": 4.546713399719087, + "learning_rate": 1.8472229190431634e-06, + "loss": 0.1447, + "step": 11579 + }, + { + "epoch": 2.4281820088068775, + "grad_norm": 3.8183160676963706, + "learning_rate": 1.8459121799446567e-06, + "loss": 0.1499, + "step": 11580 + }, + { + "epoch": 2.428391696372405, + "grad_norm": 3.6908101921378824, + "learning_rate": 1.8446018587572945e-06, + "loss": 0.1485, + "step": 11581 + }, + { + "epoch": 2.4286013839379326, + "grad_norm": 4.3182762037486455, + "learning_rate": 1.8432919555482331e-06, + "loss": 0.1142, + "step": 11582 + }, + { + "epoch": 2.4288110715034597, + "grad_norm": 5.244094145869, + "learning_rate": 1.8419824703846034e-06, + "loss": 0.1668, + "step": 11583 + }, + { + "epoch": 2.429020759068987, + "grad_norm": 4.4707026346222225, + "learning_rate": 1.8406734033335205e-06, + "loss": 0.1416, + "step": 11584 + }, + { + "epoch": 2.4292304466345147, + "grad_norm": 5.644774200398411, + "learning_rate": 1.839364754462084e-06, + "loss": 0.193, + "step": 11585 + }, + { + "epoch": 2.429440134200042, + "grad_norm": 3.421698419337396, + "learning_rate": 1.8380565238373515e-06, + "loss": 0.1233, + "step": 11586 + }, + { + "epoch": 2.4296498217655693, + "grad_norm": 4.270402520694216, + "learning_rate": 1.8367487115263782e-06, + "loss": 0.1313, + "step": 11587 + }, + { + "epoch": 2.429859509331097, + "grad_norm": 4.758688444196812, + "learning_rate": 1.8354413175961927e-06, + "loss": 0.1459, + "step": 11588 + }, + { + "epoch": 2.430069196896624, + "grad_norm": 3.9493660715397336, + "learning_rate": 1.8341343421138046e-06, + "loss": 0.1261, + "step": 11589 + }, + { + "epoch": 2.4302788844621515, + "grad_norm": 3.0969277406306164, + "learning_rate": 1.8328277851461896e-06, + "loss": 0.1015, + "step": 11590 + }, + { + "epoch": 2.430488572027679, + "grad_norm": 3.86641177599819, + "learning_rate": 1.831521646760316e-06, + "loss": 0.1376, + "step": 11591 + }, + { + "epoch": 2.430698259593206, + "grad_norm": 3.285235110605288, + "learning_rate": 1.830215927023128e-06, + "loss": 0.1121, + "step": 11592 + }, + { + "epoch": 2.4309079471587336, + "grad_norm": 5.338087363722873, + "learning_rate": 1.8289106260015443e-06, + "loss": 0.188, + "step": 11593 + }, + { + "epoch": 2.4311176347242607, + "grad_norm": 4.103614429871245, + "learning_rate": 1.8276057437624595e-06, + "loss": 0.1497, + "step": 11594 + }, + { + "epoch": 2.431327322289788, + "grad_norm": 5.219029463349436, + "learning_rate": 1.8263012803727576e-06, + "loss": 0.1236, + "step": 11595 + }, + { + "epoch": 2.4315370098553157, + "grad_norm": 3.509086954565987, + "learning_rate": 1.8249972358992884e-06, + "loss": 0.1201, + "step": 11596 + }, + { + "epoch": 2.431746697420843, + "grad_norm": 4.467464805532375, + "learning_rate": 1.8236936104088932e-06, + "loss": 0.1506, + "step": 11597 + }, + { + "epoch": 2.4319563849863703, + "grad_norm": 4.114036883713786, + "learning_rate": 1.8223904039683816e-06, + "loss": 0.1327, + "step": 11598 + }, + { + "epoch": 2.4321660725518974, + "grad_norm": 3.716981971534856, + "learning_rate": 1.8210876166445412e-06, + "loss": 0.1318, + "step": 11599 + }, + { + "epoch": 2.432375760117425, + "grad_norm": 5.047725538649979, + "learning_rate": 1.8197852485041479e-06, + "loss": 0.1637, + "step": 11600 + }, + { + "epoch": 2.4325854476829525, + "grad_norm": 3.9681602273539385, + "learning_rate": 1.8184832996139501e-06, + "loss": 0.1274, + "step": 11601 + }, + { + "epoch": 2.4327951352484796, + "grad_norm": 3.789515392775734, + "learning_rate": 1.8171817700406736e-06, + "loss": 0.159, + "step": 11602 + }, + { + "epoch": 2.433004822814007, + "grad_norm": 4.015874604666652, + "learning_rate": 1.8158806598510204e-06, + "loss": 0.1092, + "step": 11603 + }, + { + "epoch": 2.4332145103795346, + "grad_norm": 4.872027005285861, + "learning_rate": 1.8145799691116805e-06, + "loss": 0.1606, + "step": 11604 + }, + { + "epoch": 2.4334241979450617, + "grad_norm": 3.6391003911024296, + "learning_rate": 1.813279697889313e-06, + "loss": 0.1051, + "step": 11605 + }, + { + "epoch": 2.4336338855105892, + "grad_norm": 7.050559231916166, + "learning_rate": 1.8119798462505612e-06, + "loss": 0.1811, + "step": 11606 + }, + { + "epoch": 2.4338435730761168, + "grad_norm": 5.102285905243485, + "learning_rate": 1.8106804142620449e-06, + "loss": 0.1499, + "step": 11607 + }, + { + "epoch": 2.434053260641644, + "grad_norm": 4.318424877069384, + "learning_rate": 1.8093814019903577e-06, + "loss": 0.1405, + "step": 11608 + }, + { + "epoch": 2.4342629482071714, + "grad_norm": 4.874774410112969, + "learning_rate": 1.8080828095020819e-06, + "loss": 0.1543, + "step": 11609 + }, + { + "epoch": 2.434472635772699, + "grad_norm": 4.289148330915365, + "learning_rate": 1.806784636863771e-06, + "loss": 0.1558, + "step": 11610 + }, + { + "epoch": 2.434682323338226, + "grad_norm": 3.552246822868351, + "learning_rate": 1.8054868841419549e-06, + "loss": 0.1493, + "step": 11611 + }, + { + "epoch": 2.4348920109037535, + "grad_norm": 3.7513602592361384, + "learning_rate": 1.8041895514031493e-06, + "loss": 0.1298, + "step": 11612 + }, + { + "epoch": 2.4351016984692806, + "grad_norm": 4.905802152210184, + "learning_rate": 1.8028926387138456e-06, + "loss": 0.1328, + "step": 11613 + }, + { + "epoch": 2.435311386034808, + "grad_norm": 4.170007899332841, + "learning_rate": 1.8015961461405118e-06, + "loss": 0.1453, + "step": 11614 + }, + { + "epoch": 2.4355210736003357, + "grad_norm": 3.5796486863115904, + "learning_rate": 1.8003000737495923e-06, + "loss": 0.1101, + "step": 11615 + }, + { + "epoch": 2.4357307611658627, + "grad_norm": 3.947394739230773, + "learning_rate": 1.7990044216075186e-06, + "loss": 0.1382, + "step": 11616 + }, + { + "epoch": 2.4359404487313903, + "grad_norm": 3.783281274405878, + "learning_rate": 1.7977091897806886e-06, + "loss": 0.1009, + "step": 11617 + }, + { + "epoch": 2.4361501362969173, + "grad_norm": 5.792851576131324, + "learning_rate": 1.7964143783354925e-06, + "loss": 0.1719, + "step": 11618 + }, + { + "epoch": 2.436359823862445, + "grad_norm": 3.5958710621670336, + "learning_rate": 1.7951199873382873e-06, + "loss": 0.1394, + "step": 11619 + }, + { + "epoch": 2.4365695114279724, + "grad_norm": 5.096906359643023, + "learning_rate": 1.7938260168554112e-06, + "loss": 0.1876, + "step": 11620 + }, + { + "epoch": 2.4367791989934995, + "grad_norm": 4.62530718972109, + "learning_rate": 1.7925324669531853e-06, + "loss": 0.1422, + "step": 11621 + }, + { + "epoch": 2.436988886559027, + "grad_norm": 3.084964368189745, + "learning_rate": 1.7912393376979064e-06, + "loss": 0.1235, + "step": 11622 + }, + { + "epoch": 2.4371985741245545, + "grad_norm": 4.563698145888654, + "learning_rate": 1.7899466291558499e-06, + "loss": 0.1476, + "step": 11623 + }, + { + "epoch": 2.4374082616900816, + "grad_norm": 4.126461705233828, + "learning_rate": 1.7886543413932655e-06, + "loss": 0.142, + "step": 11624 + }, + { + "epoch": 2.437617949255609, + "grad_norm": 3.659989552813751, + "learning_rate": 1.7873624744763907e-06, + "loss": 0.1474, + "step": 11625 + }, + { + "epoch": 2.4378276368211367, + "grad_norm": 3.50288718532517, + "learning_rate": 1.7860710284714334e-06, + "loss": 0.0753, + "step": 11626 + }, + { + "epoch": 2.4380373243866638, + "grad_norm": 3.6278997167168217, + "learning_rate": 1.7847800034445784e-06, + "loss": 0.1217, + "step": 11627 + }, + { + "epoch": 2.4382470119521913, + "grad_norm": 4.662845521691005, + "learning_rate": 1.7834893994620007e-06, + "loss": 0.1611, + "step": 11628 + }, + { + "epoch": 2.438456699517719, + "grad_norm": 4.617434001085566, + "learning_rate": 1.7821992165898394e-06, + "loss": 0.1549, + "step": 11629 + }, + { + "epoch": 2.438666387083246, + "grad_norm": 6.2793805770872035, + "learning_rate": 1.780909454894223e-06, + "loss": 0.2151, + "step": 11630 + }, + { + "epoch": 2.4388760746487734, + "grad_norm": 4.174854093002755, + "learning_rate": 1.779620114441253e-06, + "loss": 0.1312, + "step": 11631 + }, + { + "epoch": 2.4390857622143005, + "grad_norm": 5.598812009577553, + "learning_rate": 1.7783311952970072e-06, + "loss": 0.1582, + "step": 11632 + }, + { + "epoch": 2.439295449779828, + "grad_norm": 4.533099904707504, + "learning_rate": 1.7770426975275468e-06, + "loss": 0.1523, + "step": 11633 + }, + { + "epoch": 2.4395051373453556, + "grad_norm": 3.5368892804206116, + "learning_rate": 1.775754621198914e-06, + "loss": 0.1339, + "step": 11634 + }, + { + "epoch": 2.4397148249108827, + "grad_norm": 4.094024981456998, + "learning_rate": 1.7744669663771198e-06, + "loss": 0.1623, + "step": 11635 + }, + { + "epoch": 2.43992451247641, + "grad_norm": 4.536694867982867, + "learning_rate": 1.7731797331281574e-06, + "loss": 0.1901, + "step": 11636 + }, + { + "epoch": 2.4401342000419377, + "grad_norm": 2.9432722757813945, + "learning_rate": 1.771892921518006e-06, + "loss": 0.1033, + "step": 11637 + }, + { + "epoch": 2.440343887607465, + "grad_norm": 5.398432769752616, + "learning_rate": 1.7706065316126097e-06, + "loss": 0.1388, + "step": 11638 + }, + { + "epoch": 2.4405535751729923, + "grad_norm": 3.6611296986210267, + "learning_rate": 1.7693205634779054e-06, + "loss": 0.1087, + "step": 11639 + }, + { + "epoch": 2.4407632627385194, + "grad_norm": 3.815553052200072, + "learning_rate": 1.7680350171797966e-06, + "loss": 0.1199, + "step": 11640 + }, + { + "epoch": 2.440972950304047, + "grad_norm": 3.746946118409033, + "learning_rate": 1.7667498927841686e-06, + "loss": 0.1459, + "step": 11641 + }, + { + "epoch": 2.4411826378695745, + "grad_norm": 3.50954775367525, + "learning_rate": 1.7654651903568887e-06, + "loss": 0.1268, + "step": 11642 + }, + { + "epoch": 2.4413923254351015, + "grad_norm": 4.338792428491563, + "learning_rate": 1.7641809099638052e-06, + "loss": 0.1437, + "step": 11643 + }, + { + "epoch": 2.441602013000629, + "grad_norm": 4.251083704147908, + "learning_rate": 1.7628970516707288e-06, + "loss": 0.151, + "step": 11644 + }, + { + "epoch": 2.4418117005661566, + "grad_norm": 3.5984200401758395, + "learning_rate": 1.761613615543465e-06, + "loss": 0.1294, + "step": 11645 + }, + { + "epoch": 2.4420213881316837, + "grad_norm": 4.069162601956504, + "learning_rate": 1.760330601647794e-06, + "loss": 0.15, + "step": 11646 + }, + { + "epoch": 2.442231075697211, + "grad_norm": 4.284091545195634, + "learning_rate": 1.7590480100494723e-06, + "loss": 0.1333, + "step": 11647 + }, + { + "epoch": 2.4424407632627387, + "grad_norm": 4.436216482089664, + "learning_rate": 1.7577658408142296e-06, + "loss": 0.1411, + "step": 11648 + }, + { + "epoch": 2.442650450828266, + "grad_norm": 4.114813369168413, + "learning_rate": 1.7564840940077831e-06, + "loss": 0.131, + "step": 11649 + }, + { + "epoch": 2.4428601383937933, + "grad_norm": 3.9510720990415544, + "learning_rate": 1.7552027696958273e-06, + "loss": 0.1407, + "step": 11650 + }, + { + "epoch": 2.4430698259593204, + "grad_norm": 4.254572443994705, + "learning_rate": 1.7539218679440295e-06, + "loss": 0.134, + "step": 11651 + }, + { + "epoch": 2.443279513524848, + "grad_norm": 5.052339382643115, + "learning_rate": 1.752641388818035e-06, + "loss": 0.1316, + "step": 11652 + }, + { + "epoch": 2.4434892010903755, + "grad_norm": 3.4801378040600297, + "learning_rate": 1.7513613323834766e-06, + "loss": 0.1496, + "step": 11653 + }, + { + "epoch": 2.4436988886559026, + "grad_norm": 4.474093318666903, + "learning_rate": 1.7500816987059533e-06, + "loss": 0.1641, + "step": 11654 + }, + { + "epoch": 2.44390857622143, + "grad_norm": 3.569491462606839, + "learning_rate": 1.7488024878510546e-06, + "loss": 0.1394, + "step": 11655 + }, + { + "epoch": 2.4441182637869576, + "grad_norm": 5.171505995008727, + "learning_rate": 1.7475236998843404e-06, + "loss": 0.151, + "step": 11656 + }, + { + "epoch": 2.4443279513524847, + "grad_norm": 3.8850534474147698, + "learning_rate": 1.746245334871346e-06, + "loss": 0.1212, + "step": 11657 + }, + { + "epoch": 2.4445376389180122, + "grad_norm": 5.3519759157699385, + "learning_rate": 1.7449673928775934e-06, + "loss": 0.1692, + "step": 11658 + }, + { + "epoch": 2.4447473264835393, + "grad_norm": 4.2457292960975925, + "learning_rate": 1.7436898739685836e-06, + "loss": 0.1013, + "step": 11659 + }, + { + "epoch": 2.444957014049067, + "grad_norm": 4.157043468905097, + "learning_rate": 1.7424127782097866e-06, + "loss": 0.1495, + "step": 11660 + }, + { + "epoch": 2.4451667016145944, + "grad_norm": 4.393449682972221, + "learning_rate": 1.7411361056666543e-06, + "loss": 0.1655, + "step": 11661 + }, + { + "epoch": 2.4453763891801215, + "grad_norm": 4.414633089717411, + "learning_rate": 1.7398598564046231e-06, + "loss": 0.1447, + "step": 11662 + }, + { + "epoch": 2.445586076745649, + "grad_norm": 5.586456908572827, + "learning_rate": 1.7385840304890978e-06, + "loss": 0.1731, + "step": 11663 + }, + { + "epoch": 2.4457957643111765, + "grad_norm": 2.948747927618289, + "learning_rate": 1.7373086279854734e-06, + "loss": 0.108, + "step": 11664 + }, + { + "epoch": 2.4460054518767036, + "grad_norm": 5.526856821536726, + "learning_rate": 1.7360336489591122e-06, + "loss": 0.1774, + "step": 11665 + }, + { + "epoch": 2.446215139442231, + "grad_norm": 3.6775186599390577, + "learning_rate": 1.7347590934753566e-06, + "loss": 0.1115, + "step": 11666 + }, + { + "epoch": 2.4464248270077587, + "grad_norm": 4.178274685953776, + "learning_rate": 1.733484961599533e-06, + "loss": 0.1372, + "step": 11667 + }, + { + "epoch": 2.4466345145732857, + "grad_norm": 4.910626158879536, + "learning_rate": 1.7322112533969482e-06, + "loss": 0.1535, + "step": 11668 + }, + { + "epoch": 2.4468442021388133, + "grad_norm": 4.471037286616529, + "learning_rate": 1.730937968932871e-06, + "loss": 0.1402, + "step": 11669 + }, + { + "epoch": 2.4470538897043403, + "grad_norm": 3.588733974631398, + "learning_rate": 1.7296651082725657e-06, + "loss": 0.1008, + "step": 11670 + }, + { + "epoch": 2.447263577269868, + "grad_norm": 4.311689490493596, + "learning_rate": 1.7283926714812693e-06, + "loss": 0.1601, + "step": 11671 + }, + { + "epoch": 2.4474732648353954, + "grad_norm": 4.002955842729539, + "learning_rate": 1.727120658624195e-06, + "loss": 0.1578, + "step": 11672 + }, + { + "epoch": 2.4476829524009225, + "grad_norm": 4.291165418814481, + "learning_rate": 1.7258490697665331e-06, + "loss": 0.1432, + "step": 11673 + }, + { + "epoch": 2.44789263996645, + "grad_norm": 5.26831003810725, + "learning_rate": 1.72457790497346e-06, + "loss": 0.184, + "step": 11674 + }, + { + "epoch": 2.4481023275319775, + "grad_norm": 3.570986338916178, + "learning_rate": 1.7233071643101195e-06, + "loss": 0.1128, + "step": 11675 + }, + { + "epoch": 2.4483120150975046, + "grad_norm": 3.9987271604371264, + "learning_rate": 1.722036847841645e-06, + "loss": 0.1418, + "step": 11676 + }, + { + "epoch": 2.448521702663032, + "grad_norm": 3.76620197488452, + "learning_rate": 1.7207669556331385e-06, + "loss": 0.134, + "step": 11677 + }, + { + "epoch": 2.4487313902285592, + "grad_norm": 3.698337422628714, + "learning_rate": 1.719497487749683e-06, + "loss": 0.1195, + "step": 11678 + }, + { + "epoch": 2.4489410777940868, + "grad_norm": 4.235277695964923, + "learning_rate": 1.7182284442563435e-06, + "loss": 0.1561, + "step": 11679 + }, + { + "epoch": 2.4491507653596143, + "grad_norm": 3.6130341874533607, + "learning_rate": 1.716959825218162e-06, + "loss": 0.1433, + "step": 11680 + }, + { + "epoch": 2.4493604529251414, + "grad_norm": 4.189201744914129, + "learning_rate": 1.7156916307001558e-06, + "loss": 0.1545, + "step": 11681 + }, + { + "epoch": 2.449570140490669, + "grad_norm": 3.5592516219431922, + "learning_rate": 1.7144238607673202e-06, + "loss": 0.0971, + "step": 11682 + }, + { + "epoch": 2.4497798280561964, + "grad_norm": 4.409566367736808, + "learning_rate": 1.7131565154846342e-06, + "loss": 0.1576, + "step": 11683 + }, + { + "epoch": 2.4499895156217235, + "grad_norm": 4.795557195698707, + "learning_rate": 1.7118895949170466e-06, + "loss": 0.1423, + "step": 11684 + }, + { + "epoch": 2.450199203187251, + "grad_norm": 3.838838929487421, + "learning_rate": 1.710623099129496e-06, + "loss": 0.1372, + "step": 11685 + }, + { + "epoch": 2.4504088907527786, + "grad_norm": 4.748323924801688, + "learning_rate": 1.7093570281868877e-06, + "loss": 0.1089, + "step": 11686 + }, + { + "epoch": 2.4506185783183057, + "grad_norm": 5.3271736737080975, + "learning_rate": 1.708091382154109e-06, + "loss": 0.1601, + "step": 11687 + }, + { + "epoch": 2.450828265883833, + "grad_norm": 4.455878469225267, + "learning_rate": 1.7068261610960313e-06, + "loss": 0.1468, + "step": 11688 + }, + { + "epoch": 2.4510379534493603, + "grad_norm": 3.5607846836403882, + "learning_rate": 1.705561365077496e-06, + "loss": 0.1262, + "step": 11689 + }, + { + "epoch": 2.451247641014888, + "grad_norm": 3.7169329686739547, + "learning_rate": 1.7042969941633247e-06, + "loss": 0.1213, + "step": 11690 + }, + { + "epoch": 2.4514573285804153, + "grad_norm": 4.43642304382663, + "learning_rate": 1.703033048418321e-06, + "loss": 0.1482, + "step": 11691 + }, + { + "epoch": 2.4516670161459424, + "grad_norm": 5.800553639759782, + "learning_rate": 1.7017695279072655e-06, + "loss": 0.1445, + "step": 11692 + }, + { + "epoch": 2.45187670371147, + "grad_norm": 3.820853437237589, + "learning_rate": 1.700506432694915e-06, + "loss": 0.1373, + "step": 11693 + }, + { + "epoch": 2.4520863912769975, + "grad_norm": 3.574471631119868, + "learning_rate": 1.6992437628460012e-06, + "loss": 0.0946, + "step": 11694 + }, + { + "epoch": 2.4522960788425245, + "grad_norm": 3.8550870460287596, + "learning_rate": 1.6979815184252446e-06, + "loss": 0.122, + "step": 11695 + }, + { + "epoch": 2.452505766408052, + "grad_norm": 4.39650800220711, + "learning_rate": 1.6967196994973322e-06, + "loss": 0.1391, + "step": 11696 + }, + { + "epoch": 2.452715453973579, + "grad_norm": 4.405001305073273, + "learning_rate": 1.6954583061269381e-06, + "loss": 0.1417, + "step": 11697 + }, + { + "epoch": 2.4529251415391067, + "grad_norm": 3.62682088591341, + "learning_rate": 1.6941973383787102e-06, + "loss": 0.1689, + "step": 11698 + }, + { + "epoch": 2.453134829104634, + "grad_norm": 4.326610062497214, + "learning_rate": 1.6929367963172716e-06, + "loss": 0.1487, + "step": 11699 + }, + { + "epoch": 2.4533445166701613, + "grad_norm": 4.446067010239069, + "learning_rate": 1.6916766800072294e-06, + "loss": 0.1643, + "step": 11700 + }, + { + "epoch": 2.453554204235689, + "grad_norm": 3.733085875432683, + "learning_rate": 1.6904169895131705e-06, + "loss": 0.1421, + "step": 11701 + }, + { + "epoch": 2.4537638918012163, + "grad_norm": 4.979338794226402, + "learning_rate": 1.6891577248996528e-06, + "loss": 0.1677, + "step": 11702 + }, + { + "epoch": 2.4539735793667434, + "grad_norm": 4.564529022072222, + "learning_rate": 1.6878988862312141e-06, + "loss": 0.1345, + "step": 11703 + }, + { + "epoch": 2.454183266932271, + "grad_norm": 3.7017955948351617, + "learning_rate": 1.6866404735723762e-06, + "loss": 0.124, + "step": 11704 + }, + { + "epoch": 2.4543929544977985, + "grad_norm": 4.31256264758994, + "learning_rate": 1.6853824869876334e-06, + "loss": 0.1445, + "step": 11705 + }, + { + "epoch": 2.4546026420633256, + "grad_norm": 4.726608581160705, + "learning_rate": 1.6841249265414562e-06, + "loss": 0.1892, + "step": 11706 + }, + { + "epoch": 2.454812329628853, + "grad_norm": 3.951149583094605, + "learning_rate": 1.6828677922983016e-06, + "loss": 0.1359, + "step": 11707 + }, + { + "epoch": 2.45502201719438, + "grad_norm": 4.723496675944823, + "learning_rate": 1.6816110843225964e-06, + "loss": 0.1588, + "step": 11708 + }, + { + "epoch": 2.4552317047599077, + "grad_norm": 3.5675918903297243, + "learning_rate": 1.6803548026787531e-06, + "loss": 0.1402, + "step": 11709 + }, + { + "epoch": 2.4554413923254352, + "grad_norm": 3.5477777399110297, + "learning_rate": 1.6790989474311525e-06, + "loss": 0.103, + "step": 11710 + }, + { + "epoch": 2.4556510798909623, + "grad_norm": 5.043903318896898, + "learning_rate": 1.6778435186441655e-06, + "loss": 0.1773, + "step": 11711 + }, + { + "epoch": 2.45586076745649, + "grad_norm": 3.9463900449576945, + "learning_rate": 1.6765885163821305e-06, + "loss": 0.1436, + "step": 11712 + }, + { + "epoch": 2.4560704550220174, + "grad_norm": 4.345611977317838, + "learning_rate": 1.6753339407093716e-06, + "loss": 0.1377, + "step": 11713 + }, + { + "epoch": 2.4562801425875445, + "grad_norm": 4.484473670769457, + "learning_rate": 1.6740797916901875e-06, + "loss": 0.1516, + "step": 11714 + }, + { + "epoch": 2.456489830153072, + "grad_norm": 3.988734787420637, + "learning_rate": 1.672826069388852e-06, + "loss": 0.1699, + "step": 11715 + }, + { + "epoch": 2.456699517718599, + "grad_norm": 3.379006212056817, + "learning_rate": 1.671572773869623e-06, + "loss": 0.1209, + "step": 11716 + }, + { + "epoch": 2.4569092052841266, + "grad_norm": 3.2239407754406737, + "learning_rate": 1.6703199051967368e-06, + "loss": 0.1114, + "step": 11717 + }, + { + "epoch": 2.457118892849654, + "grad_norm": 4.609798421444875, + "learning_rate": 1.6690674634344028e-06, + "loss": 0.1484, + "step": 11718 + }, + { + "epoch": 2.457328580415181, + "grad_norm": 3.825286887146204, + "learning_rate": 1.667815448646808e-06, + "loss": 0.1322, + "step": 11719 + }, + { + "epoch": 2.4575382679807087, + "grad_norm": 4.019395236386977, + "learning_rate": 1.666563860898125e-06, + "loss": 0.1508, + "step": 11720 + }, + { + "epoch": 2.4577479555462363, + "grad_norm": 3.8144259825413203, + "learning_rate": 1.665312700252496e-06, + "loss": 0.1356, + "step": 11721 + }, + { + "epoch": 2.4579576431117633, + "grad_norm": 3.903707468966791, + "learning_rate": 1.6640619667740487e-06, + "loss": 0.148, + "step": 11722 + }, + { + "epoch": 2.458167330677291, + "grad_norm": 5.474609011800797, + "learning_rate": 1.662811660526884e-06, + "loss": 0.1788, + "step": 11723 + }, + { + "epoch": 2.4583770182428184, + "grad_norm": 3.795217878221053, + "learning_rate": 1.6615617815750785e-06, + "loss": 0.1367, + "step": 11724 + }, + { + "epoch": 2.4585867058083455, + "grad_norm": 4.186073310279482, + "learning_rate": 1.6603123299826952e-06, + "loss": 0.1121, + "step": 11725 + }, + { + "epoch": 2.458796393373873, + "grad_norm": 5.051326217628293, + "learning_rate": 1.6590633058137739e-06, + "loss": 0.1771, + "step": 11726 + }, + { + "epoch": 2.4590060809394005, + "grad_norm": 2.859400860563314, + "learning_rate": 1.6578147091323194e-06, + "loss": 0.1112, + "step": 11727 + }, + { + "epoch": 2.4592157685049276, + "grad_norm": 4.076957558715104, + "learning_rate": 1.65656654000233e-06, + "loss": 0.1504, + "step": 11728 + }, + { + "epoch": 2.459425456070455, + "grad_norm": 4.974530500416484, + "learning_rate": 1.6553187984877794e-06, + "loss": 0.1432, + "step": 11729 + }, + { + "epoch": 2.4596351436359822, + "grad_norm": 6.832899218295099, + "learning_rate": 1.654071484652613e-06, + "loss": 0.2114, + "step": 11730 + }, + { + "epoch": 2.4598448312015098, + "grad_norm": 3.425279238123437, + "learning_rate": 1.652824598560756e-06, + "loss": 0.1393, + "step": 11731 + }, + { + "epoch": 2.4600545187670373, + "grad_norm": 4.798244783013604, + "learning_rate": 1.6515781402761178e-06, + "loss": 0.1795, + "step": 11732 + }, + { + "epoch": 2.4602642063325644, + "grad_norm": 5.160155677326122, + "learning_rate": 1.650332109862578e-06, + "loss": 0.1755, + "step": 11733 + }, + { + "epoch": 2.460473893898092, + "grad_norm": 4.55573336884829, + "learning_rate": 1.6490865073840012e-06, + "loss": 0.1371, + "step": 11734 + }, + { + "epoch": 2.460683581463619, + "grad_norm": 4.689444648769969, + "learning_rate": 1.6478413329042241e-06, + "loss": 0.1581, + "step": 11735 + }, + { + "epoch": 2.4608932690291465, + "grad_norm": 5.4201384224461435, + "learning_rate": 1.6465965864870636e-06, + "loss": 0.1497, + "step": 11736 + }, + { + "epoch": 2.461102956594674, + "grad_norm": 3.6287365004064465, + "learning_rate": 1.6453522681963164e-06, + "loss": 0.1536, + "step": 11737 + }, + { + "epoch": 2.461312644160201, + "grad_norm": 3.6616532466385783, + "learning_rate": 1.6441083780957568e-06, + "loss": 0.1278, + "step": 11738 + }, + { + "epoch": 2.4615223317257287, + "grad_norm": 3.6914578769747046, + "learning_rate": 1.642864916249136e-06, + "loss": 0.1445, + "step": 11739 + }, + { + "epoch": 2.461732019291256, + "grad_norm": 3.6663745713038933, + "learning_rate": 1.6416218827201813e-06, + "loss": 0.1338, + "step": 11740 + }, + { + "epoch": 2.4619417068567833, + "grad_norm": 3.833988929727368, + "learning_rate": 1.6403792775726035e-06, + "loss": 0.1182, + "step": 11741 + }, + { + "epoch": 2.462151394422311, + "grad_norm": 4.423237581815354, + "learning_rate": 1.6391371008700852e-06, + "loss": 0.1395, + "step": 11742 + }, + { + "epoch": 2.4623610819878383, + "grad_norm": 4.189313452066283, + "learning_rate": 1.637895352676293e-06, + "loss": 0.1387, + "step": 11743 + }, + { + "epoch": 2.4625707695533654, + "grad_norm": 4.3764190239440515, + "learning_rate": 1.6366540330548686e-06, + "loss": 0.1372, + "step": 11744 + }, + { + "epoch": 2.462780457118893, + "grad_norm": 5.014165012194932, + "learning_rate": 1.6354131420694276e-06, + "loss": 0.1585, + "step": 11745 + }, + { + "epoch": 2.4629901446844205, + "grad_norm": 5.23412055920183, + "learning_rate": 1.63417267978357e-06, + "loss": 0.1455, + "step": 11746 + }, + { + "epoch": 2.4631998322499475, + "grad_norm": 3.5166038732592604, + "learning_rate": 1.6329326462608774e-06, + "loss": 0.154, + "step": 11747 + }, + { + "epoch": 2.463409519815475, + "grad_norm": 3.724216827066388, + "learning_rate": 1.6316930415648946e-06, + "loss": 0.1353, + "step": 11748 + }, + { + "epoch": 2.463619207381002, + "grad_norm": 3.095427442032372, + "learning_rate": 1.6304538657591562e-06, + "loss": 0.096, + "step": 11749 + }, + { + "epoch": 2.4638288949465297, + "grad_norm": 4.933847781004414, + "learning_rate": 1.629215118907177e-06, + "loss": 0.1683, + "step": 11750 + }, + { + "epoch": 2.464038582512057, + "grad_norm": 4.997692151281342, + "learning_rate": 1.6279768010724405e-06, + "loss": 0.1609, + "step": 11751 + }, + { + "epoch": 2.4642482700775843, + "grad_norm": 3.474351926937195, + "learning_rate": 1.626738912318412e-06, + "loss": 0.1134, + "step": 11752 + }, + { + "epoch": 2.464457957643112, + "grad_norm": 3.8562070006652234, + "learning_rate": 1.625501452708539e-06, + "loss": 0.153, + "step": 11753 + }, + { + "epoch": 2.464667645208639, + "grad_norm": 3.0809851411868796, + "learning_rate": 1.6242644223062388e-06, + "loss": 0.0937, + "step": 11754 + }, + { + "epoch": 2.4648773327741664, + "grad_norm": 4.23543577723989, + "learning_rate": 1.623027821174916e-06, + "loss": 0.1313, + "step": 11755 + }, + { + "epoch": 2.465087020339694, + "grad_norm": 5.446591071000503, + "learning_rate": 1.6217916493779473e-06, + "loss": 0.1768, + "step": 11756 + }, + { + "epoch": 2.465296707905221, + "grad_norm": 3.1987959570885462, + "learning_rate": 1.6205559069786847e-06, + "loss": 0.1279, + "step": 11757 + }, + { + "epoch": 2.4655063954707486, + "grad_norm": 4.173271586019623, + "learning_rate": 1.6193205940404656e-06, + "loss": 0.1459, + "step": 11758 + }, + { + "epoch": 2.465716083036276, + "grad_norm": 3.8808073971218806, + "learning_rate": 1.6180857106266045e-06, + "loss": 0.1158, + "step": 11759 + }, + { + "epoch": 2.465925770601803, + "grad_norm": 3.890823200143949, + "learning_rate": 1.616851256800387e-06, + "loss": 0.0957, + "step": 11760 + }, + { + "epoch": 2.4661354581673307, + "grad_norm": 4.7218539451389105, + "learning_rate": 1.6156172326250818e-06, + "loss": 0.1753, + "step": 11761 + }, + { + "epoch": 2.4663451457328582, + "grad_norm": 6.244042513007644, + "learning_rate": 1.6143836381639366e-06, + "loss": 0.1826, + "step": 11762 + }, + { + "epoch": 2.4665548332983853, + "grad_norm": 3.9792557462414972, + "learning_rate": 1.6131504734801716e-06, + "loss": 0.1677, + "step": 11763 + }, + { + "epoch": 2.466764520863913, + "grad_norm": 3.9360357046038206, + "learning_rate": 1.6119177386369943e-06, + "loss": 0.155, + "step": 11764 + }, + { + "epoch": 2.4669742084294404, + "grad_norm": 4.0894509689794685, + "learning_rate": 1.6106854336975808e-06, + "loss": 0.1581, + "step": 11765 + }, + { + "epoch": 2.4671838959949675, + "grad_norm": 4.56045207919616, + "learning_rate": 1.609453558725087e-06, + "loss": 0.1204, + "step": 11766 + }, + { + "epoch": 2.467393583560495, + "grad_norm": 4.113545949616134, + "learning_rate": 1.6082221137826538e-06, + "loss": 0.1618, + "step": 11767 + }, + { + "epoch": 2.467603271126022, + "grad_norm": 3.3431955946339924, + "learning_rate": 1.6069910989333915e-06, + "loss": 0.1302, + "step": 11768 + }, + { + "epoch": 2.4678129586915496, + "grad_norm": 3.881445070057582, + "learning_rate": 1.6057605142403897e-06, + "loss": 0.1468, + "step": 11769 + }, + { + "epoch": 2.468022646257077, + "grad_norm": 5.8557843195805415, + "learning_rate": 1.6045303597667206e-06, + "loss": 0.187, + "step": 11770 + }, + { + "epoch": 2.468232333822604, + "grad_norm": 5.615903884614249, + "learning_rate": 1.6033006355754355e-06, + "loss": 0.1734, + "step": 11771 + }, + { + "epoch": 2.4684420213881317, + "grad_norm": 4.480712096008919, + "learning_rate": 1.6020713417295541e-06, + "loss": 0.1416, + "step": 11772 + }, + { + "epoch": 2.468651708953659, + "grad_norm": 4.551764679829709, + "learning_rate": 1.6008424782920806e-06, + "loss": 0.1397, + "step": 11773 + }, + { + "epoch": 2.4688613965191863, + "grad_norm": 4.021038244891588, + "learning_rate": 1.599614045325998e-06, + "loss": 0.1084, + "step": 11774 + }, + { + "epoch": 2.469071084084714, + "grad_norm": 3.7796055465638956, + "learning_rate": 1.5983860428942677e-06, + "loss": 0.1123, + "step": 11775 + }, + { + "epoch": 2.469280771650241, + "grad_norm": 3.734566099884102, + "learning_rate": 1.5971584710598242e-06, + "loss": 0.1061, + "step": 11776 + }, + { + "epoch": 2.4694904592157685, + "grad_norm": 4.178463492697262, + "learning_rate": 1.5959313298855804e-06, + "loss": 0.1362, + "step": 11777 + }, + { + "epoch": 2.469700146781296, + "grad_norm": 4.903862944623397, + "learning_rate": 1.594704619434435e-06, + "loss": 0.1884, + "step": 11778 + }, + { + "epoch": 2.469909834346823, + "grad_norm": 4.299595522877735, + "learning_rate": 1.5934783397692533e-06, + "loss": 0.1291, + "step": 11779 + }, + { + "epoch": 2.4701195219123506, + "grad_norm": 4.632466432960411, + "learning_rate": 1.5922524909528903e-06, + "loss": 0.1561, + "step": 11780 + }, + { + "epoch": 2.470329209477878, + "grad_norm": 3.5322922683881894, + "learning_rate": 1.5910270730481692e-06, + "loss": 0.1169, + "step": 11781 + }, + { + "epoch": 2.4705388970434052, + "grad_norm": 3.4997735443839293, + "learning_rate": 1.589802086117892e-06, + "loss": 0.1294, + "step": 11782 + }, + { + "epoch": 2.4707485846089328, + "grad_norm": 3.252062100082361, + "learning_rate": 1.5885775302248453e-06, + "loss": 0.1285, + "step": 11783 + }, + { + "epoch": 2.4709582721744603, + "grad_norm": 3.5285278911751115, + "learning_rate": 1.587353405431794e-06, + "loss": 0.1332, + "step": 11784 + }, + { + "epoch": 2.4711679597399874, + "grad_norm": 4.447736339511368, + "learning_rate": 1.5861297118014662e-06, + "loss": 0.1495, + "step": 11785 + }, + { + "epoch": 2.471377647305515, + "grad_norm": 4.266219649048172, + "learning_rate": 1.584906449396585e-06, + "loss": 0.1455, + "step": 11786 + }, + { + "epoch": 2.471587334871042, + "grad_norm": 4.684846693159901, + "learning_rate": 1.5836836182798443e-06, + "loss": 0.1557, + "step": 11787 + }, + { + "epoch": 2.4717970224365695, + "grad_norm": 4.63365780390233, + "learning_rate": 1.5824612185139165e-06, + "loss": 0.1536, + "step": 11788 + }, + { + "epoch": 2.472006710002097, + "grad_norm": 4.358589877591667, + "learning_rate": 1.5812392501614482e-06, + "loss": 0.1143, + "step": 11789 + }, + { + "epoch": 2.472216397567624, + "grad_norm": 5.777124349478835, + "learning_rate": 1.5800177132850724e-06, + "loss": 0.142, + "step": 11790 + }, + { + "epoch": 2.4724260851331517, + "grad_norm": 4.58126395578425, + "learning_rate": 1.5787966079473904e-06, + "loss": 0.1856, + "step": 11791 + }, + { + "epoch": 2.4726357726986787, + "grad_norm": 5.759565198721991, + "learning_rate": 1.5775759342109899e-06, + "loss": 0.1697, + "step": 11792 + }, + { + "epoch": 2.4728454602642063, + "grad_norm": 3.56304060604929, + "learning_rate": 1.5763556921384315e-06, + "loss": 0.1041, + "step": 11793 + }, + { + "epoch": 2.473055147829734, + "grad_norm": 4.688031824460209, + "learning_rate": 1.5751358817922502e-06, + "loss": 0.1671, + "step": 11794 + }, + { + "epoch": 2.473264835395261, + "grad_norm": 3.817424067047199, + "learning_rate": 1.573916503234968e-06, + "loss": 0.1183, + "step": 11795 + }, + { + "epoch": 2.4734745229607884, + "grad_norm": 3.8570270234838753, + "learning_rate": 1.5726975565290824e-06, + "loss": 0.1163, + "step": 11796 + }, + { + "epoch": 2.473684210526316, + "grad_norm": 5.62404102091517, + "learning_rate": 1.5714790417370618e-06, + "loss": 0.2112, + "step": 11797 + }, + { + "epoch": 2.473893898091843, + "grad_norm": 4.6524004494361275, + "learning_rate": 1.570260958921358e-06, + "loss": 0.1735, + "step": 11798 + }, + { + "epoch": 2.4741035856573705, + "grad_norm": 2.9688582184165977, + "learning_rate": 1.5690433081444023e-06, + "loss": 0.1113, + "step": 11799 + }, + { + "epoch": 2.474313273222898, + "grad_norm": 4.994284119966936, + "learning_rate": 1.567826089468597e-06, + "loss": 0.1499, + "step": 11800 + }, + { + "epoch": 2.474522960788425, + "grad_norm": 3.590827167692221, + "learning_rate": 1.5666093029563335e-06, + "loss": 0.1172, + "step": 11801 + }, + { + "epoch": 2.4747326483539527, + "grad_norm": 4.764100956917982, + "learning_rate": 1.5653929486699682e-06, + "loss": 0.1549, + "step": 11802 + }, + { + "epoch": 2.47494233591948, + "grad_norm": 4.000926603922949, + "learning_rate": 1.5641770266718425e-06, + "loss": 0.1548, + "step": 11803 + }, + { + "epoch": 2.4751520234850073, + "grad_norm": 4.9638337343009145, + "learning_rate": 1.5629615370242745e-06, + "loss": 0.1026, + "step": 11804 + }, + { + "epoch": 2.475361711050535, + "grad_norm": 3.7887276998061523, + "learning_rate": 1.5617464797895666e-06, + "loss": 0.1393, + "step": 11805 + }, + { + "epoch": 2.475571398616062, + "grad_norm": 3.179956073337639, + "learning_rate": 1.560531855029981e-06, + "loss": 0.1147, + "step": 11806 + }, + { + "epoch": 2.4757810861815894, + "grad_norm": 3.484309836679592, + "learning_rate": 1.5593176628077767e-06, + "loss": 0.1108, + "step": 11807 + }, + { + "epoch": 2.475990773747117, + "grad_norm": 4.287122212979466, + "learning_rate": 1.558103903185183e-06, + "loss": 0.1515, + "step": 11808 + }, + { + "epoch": 2.476200461312644, + "grad_norm": 3.9042309204573415, + "learning_rate": 1.556890576224407e-06, + "loss": 0.1579, + "step": 11809 + }, + { + "epoch": 2.4764101488781716, + "grad_norm": 4.274468797760751, + "learning_rate": 1.5556776819876296e-06, + "loss": 0.1374, + "step": 11810 + }, + { + "epoch": 2.4766198364436987, + "grad_norm": 5.566457632731147, + "learning_rate": 1.5544652205370203e-06, + "loss": 0.1575, + "step": 11811 + }, + { + "epoch": 2.476829524009226, + "grad_norm": 3.213799707234014, + "learning_rate": 1.5532531919347127e-06, + "loss": 0.118, + "step": 11812 + }, + { + "epoch": 2.4770392115747537, + "grad_norm": 4.3541484114463485, + "learning_rate": 1.5520415962428325e-06, + "loss": 0.1322, + "step": 11813 + }, + { + "epoch": 2.477248899140281, + "grad_norm": 4.855383855939923, + "learning_rate": 1.5508304335234725e-06, + "loss": 0.0966, + "step": 11814 + }, + { + "epoch": 2.4774585867058083, + "grad_norm": 5.367344856384804, + "learning_rate": 1.549619703838705e-06, + "loss": 0.2098, + "step": 11815 + }, + { + "epoch": 2.477668274271336, + "grad_norm": 5.015746142285111, + "learning_rate": 1.548409407250584e-06, + "loss": 0.1828, + "step": 11816 + }, + { + "epoch": 2.477877961836863, + "grad_norm": 3.39963965597521, + "learning_rate": 1.5471995438211418e-06, + "loss": 0.1025, + "step": 11817 + }, + { + "epoch": 2.4780876494023905, + "grad_norm": 7.232819013434011, + "learning_rate": 1.5459901136123834e-06, + "loss": 0.2417, + "step": 11818 + }, + { + "epoch": 2.478297336967918, + "grad_norm": 3.0867581015834715, + "learning_rate": 1.5447811166862925e-06, + "loss": 0.1142, + "step": 11819 + }, + { + "epoch": 2.478507024533445, + "grad_norm": 4.298037871665368, + "learning_rate": 1.5435725531048362e-06, + "loss": 0.1267, + "step": 11820 + }, + { + "epoch": 2.4787167120989726, + "grad_norm": 4.373892603946846, + "learning_rate": 1.5423644229299506e-06, + "loss": 0.143, + "step": 11821 + }, + { + "epoch": 2.4789263996645, + "grad_norm": 3.265878466045089, + "learning_rate": 1.5411567262235605e-06, + "loss": 0.1321, + "step": 11822 + }, + { + "epoch": 2.479136087230027, + "grad_norm": 3.799821116839136, + "learning_rate": 1.5399494630475586e-06, + "loss": 0.1276, + "step": 11823 + }, + { + "epoch": 2.4793457747955547, + "grad_norm": 4.910500616301495, + "learning_rate": 1.5387426334638189e-06, + "loss": 0.191, + "step": 11824 + }, + { + "epoch": 2.479555462361082, + "grad_norm": 4.966184556499949, + "learning_rate": 1.537536237534193e-06, + "loss": 0.1638, + "step": 11825 + }, + { + "epoch": 2.4797651499266093, + "grad_norm": 4.947293974233818, + "learning_rate": 1.5363302753205177e-06, + "loss": 0.1394, + "step": 11826 + }, + { + "epoch": 2.479974837492137, + "grad_norm": 4.420435070328749, + "learning_rate": 1.535124746884591e-06, + "loss": 0.1535, + "step": 11827 + }, + { + "epoch": 2.480184525057664, + "grad_norm": 4.69978911875586, + "learning_rate": 1.5339196522882017e-06, + "loss": 0.1538, + "step": 11828 + }, + { + "epoch": 2.4803942126231915, + "grad_norm": 5.175169276279862, + "learning_rate": 1.5327149915931173e-06, + "loss": 0.1257, + "step": 11829 + }, + { + "epoch": 2.4806039001887186, + "grad_norm": 4.999418976485478, + "learning_rate": 1.531510764861075e-06, + "loss": 0.1281, + "step": 11830 + }, + { + "epoch": 2.480813587754246, + "grad_norm": 3.854125899961667, + "learning_rate": 1.5303069721537912e-06, + "loss": 0.121, + "step": 11831 + }, + { + "epoch": 2.4810232753197736, + "grad_norm": 4.366764438261112, + "learning_rate": 1.5291036135329652e-06, + "loss": 0.1414, + "step": 11832 + }, + { + "epoch": 2.4812329628853007, + "grad_norm": 4.18102729171108, + "learning_rate": 1.5279006890602733e-06, + "loss": 0.1194, + "step": 11833 + }, + { + "epoch": 2.4814426504508282, + "grad_norm": 4.979768063492309, + "learning_rate": 1.526698198797365e-06, + "loss": 0.1517, + "step": 11834 + }, + { + "epoch": 2.4816523380163558, + "grad_norm": 3.9488939343662173, + "learning_rate": 1.5254961428058678e-06, + "loss": 0.1446, + "step": 11835 + }, + { + "epoch": 2.481862025581883, + "grad_norm": 4.59018665311151, + "learning_rate": 1.5242945211473936e-06, + "loss": 0.1635, + "step": 11836 + }, + { + "epoch": 2.4820717131474104, + "grad_norm": 4.224787443030674, + "learning_rate": 1.5230933338835241e-06, + "loss": 0.1556, + "step": 11837 + }, + { + "epoch": 2.482281400712938, + "grad_norm": 4.279839485963888, + "learning_rate": 1.5218925810758256e-06, + "loss": 0.1331, + "step": 11838 + }, + { + "epoch": 2.482491088278465, + "grad_norm": 5.407492332100062, + "learning_rate": 1.5206922627858367e-06, + "loss": 0.1292, + "step": 11839 + }, + { + "epoch": 2.4827007758439925, + "grad_norm": 4.429061773970704, + "learning_rate": 1.5194923790750738e-06, + "loss": 0.1212, + "step": 11840 + }, + { + "epoch": 2.48291046340952, + "grad_norm": 4.020767133826447, + "learning_rate": 1.5182929300050353e-06, + "loss": 0.1429, + "step": 11841 + }, + { + "epoch": 2.483120150975047, + "grad_norm": 5.1474317333802055, + "learning_rate": 1.5170939156371978e-06, + "loss": 0.1577, + "step": 11842 + }, + { + "epoch": 2.4833298385405747, + "grad_norm": 5.22423217552654, + "learning_rate": 1.5158953360330086e-06, + "loss": 0.1802, + "step": 11843 + }, + { + "epoch": 2.4835395261061017, + "grad_norm": 4.353578637429943, + "learning_rate": 1.5146971912538977e-06, + "loss": 0.1526, + "step": 11844 + }, + { + "epoch": 2.4837492136716293, + "grad_norm": 4.8014925208867005, + "learning_rate": 1.5134994813612737e-06, + "loss": 0.1463, + "step": 11845 + }, + { + "epoch": 2.483958901237157, + "grad_norm": 3.2016024822144376, + "learning_rate": 1.5123022064165205e-06, + "loss": 0.1063, + "step": 11846 + }, + { + "epoch": 2.484168588802684, + "grad_norm": 3.061946591652013, + "learning_rate": 1.5111053664809983e-06, + "loss": 0.0954, + "step": 11847 + }, + { + "epoch": 2.4843782763682114, + "grad_norm": 3.8862065827236787, + "learning_rate": 1.5099089616160523e-06, + "loss": 0.122, + "step": 11848 + }, + { + "epoch": 2.4845879639337385, + "grad_norm": 3.1609369013371373, + "learning_rate": 1.5087129918829946e-06, + "loss": 0.1002, + "step": 11849 + }, + { + "epoch": 2.484797651499266, + "grad_norm": 5.022728445571421, + "learning_rate": 1.507517457343125e-06, + "loss": 0.141, + "step": 11850 + }, + { + "epoch": 2.4850073390647935, + "grad_norm": 3.16948114759684, + "learning_rate": 1.5063223580577159e-06, + "loss": 0.1174, + "step": 11851 + }, + { + "epoch": 2.4852170266303206, + "grad_norm": 4.12757541917744, + "learning_rate": 1.5051276940880155e-06, + "loss": 0.1424, + "step": 11852 + }, + { + "epoch": 2.485426714195848, + "grad_norm": 3.2653015269687353, + "learning_rate": 1.503933465495253e-06, + "loss": 0.1106, + "step": 11853 + }, + { + "epoch": 2.4856364017613757, + "grad_norm": 4.643528641388025, + "learning_rate": 1.5027396723406395e-06, + "loss": 0.1647, + "step": 11854 + }, + { + "epoch": 2.4858460893269028, + "grad_norm": 5.062840533062073, + "learning_rate": 1.501546314685356e-06, + "loss": 0.1413, + "step": 11855 + }, + { + "epoch": 2.4860557768924303, + "grad_norm": 4.374360883172957, + "learning_rate": 1.500353392590561e-06, + "loss": 0.1626, + "step": 11856 + }, + { + "epoch": 2.486265464457958, + "grad_norm": 3.851082609914037, + "learning_rate": 1.4991609061173984e-06, + "loss": 0.1338, + "step": 11857 + }, + { + "epoch": 2.486475152023485, + "grad_norm": 5.105289973432092, + "learning_rate": 1.497968855326981e-06, + "loss": 0.1429, + "step": 11858 + }, + { + "epoch": 2.4866848395890124, + "grad_norm": 4.803652980933307, + "learning_rate": 1.4967772402804082e-06, + "loss": 0.1351, + "step": 11859 + }, + { + "epoch": 2.48689452715454, + "grad_norm": 4.417975671406087, + "learning_rate": 1.4955860610387495e-06, + "loss": 0.128, + "step": 11860 + }, + { + "epoch": 2.487104214720067, + "grad_norm": 4.452248138964775, + "learning_rate": 1.4943953176630543e-06, + "loss": 0.1191, + "step": 11861 + }, + { + "epoch": 2.4873139022855946, + "grad_norm": 4.5293179279593145, + "learning_rate": 1.4932050102143503e-06, + "loss": 0.1287, + "step": 11862 + }, + { + "epoch": 2.4875235898511217, + "grad_norm": 3.9126790070487965, + "learning_rate": 1.4920151387536486e-06, + "loss": 0.137, + "step": 11863 + }, + { + "epoch": 2.487733277416649, + "grad_norm": 4.332492138687773, + "learning_rate": 1.4908257033419226e-06, + "loss": 0.1715, + "step": 11864 + }, + { + "epoch": 2.4879429649821767, + "grad_norm": 4.8286363169286215, + "learning_rate": 1.4896367040401382e-06, + "loss": 0.1195, + "step": 11865 + }, + { + "epoch": 2.488152652547704, + "grad_norm": 4.52475730780462, + "learning_rate": 1.4884481409092356e-06, + "loss": 0.1755, + "step": 11866 + }, + { + "epoch": 2.4883623401132313, + "grad_norm": 5.062774352010064, + "learning_rate": 1.487260014010128e-06, + "loss": 0.1607, + "step": 11867 + }, + { + "epoch": 2.4885720276787584, + "grad_norm": 3.65963675108377, + "learning_rate": 1.4860723234037066e-06, + "loss": 0.1371, + "step": 11868 + }, + { + "epoch": 2.488781715244286, + "grad_norm": 3.7262309680921453, + "learning_rate": 1.4848850691508476e-06, + "loss": 0.1408, + "step": 11869 + }, + { + "epoch": 2.4889914028098135, + "grad_norm": 7.619352746497935, + "learning_rate": 1.4836982513123965e-06, + "loss": 0.1534, + "step": 11870 + }, + { + "epoch": 2.4892010903753405, + "grad_norm": 5.388045217848504, + "learning_rate": 1.4825118699491815e-06, + "loss": 0.1636, + "step": 11871 + }, + { + "epoch": 2.489410777940868, + "grad_norm": 3.792800012433815, + "learning_rate": 1.4813259251220069e-06, + "loss": 0.1424, + "step": 11872 + }, + { + "epoch": 2.4896204655063956, + "grad_norm": 3.612821643827426, + "learning_rate": 1.4801404168916511e-06, + "loss": 0.1419, + "step": 11873 + }, + { + "epoch": 2.4898301530719227, + "grad_norm": 5.651389763679717, + "learning_rate": 1.4789553453188754e-06, + "loss": 0.1651, + "step": 11874 + }, + { + "epoch": 2.49003984063745, + "grad_norm": 4.0074406403922955, + "learning_rate": 1.47777071046442e-06, + "loss": 0.1398, + "step": 11875 + }, + { + "epoch": 2.4902495282029777, + "grad_norm": 5.202403629707815, + "learning_rate": 1.4765865123889967e-06, + "loss": 0.1329, + "step": 11876 + }, + { + "epoch": 2.490459215768505, + "grad_norm": 4.365099978031514, + "learning_rate": 1.4754027511532954e-06, + "loss": 0.1515, + "step": 11877 + }, + { + "epoch": 2.4906689033340323, + "grad_norm": 4.6235639230024805, + "learning_rate": 1.4742194268179911e-06, + "loss": 0.1631, + "step": 11878 + }, + { + "epoch": 2.49087859089956, + "grad_norm": 4.549393359040409, + "learning_rate": 1.4730365394437252e-06, + "loss": 0.1711, + "step": 11879 + }, + { + "epoch": 2.491088278465087, + "grad_norm": 3.9769602005948723, + "learning_rate": 1.471854089091128e-06, + "loss": 0.1573, + "step": 11880 + }, + { + "epoch": 2.4912979660306145, + "grad_norm": 4.178709977139378, + "learning_rate": 1.4706720758208016e-06, + "loss": 0.1628, + "step": 11881 + }, + { + "epoch": 2.4915076535961416, + "grad_norm": 5.693388954694277, + "learning_rate": 1.4694904996933213e-06, + "loss": 0.1667, + "step": 11882 + }, + { + "epoch": 2.491717341161669, + "grad_norm": 4.541180690923713, + "learning_rate": 1.4683093607692479e-06, + "loss": 0.156, + "step": 11883 + }, + { + "epoch": 2.4919270287271966, + "grad_norm": 3.6722427203059875, + "learning_rate": 1.4671286591091228e-06, + "loss": 0.1314, + "step": 11884 + }, + { + "epoch": 2.4921367162927237, + "grad_norm": 4.889179305131772, + "learning_rate": 1.4659483947734487e-06, + "loss": 0.1641, + "step": 11885 + }, + { + "epoch": 2.4923464038582512, + "grad_norm": 3.4429981792259916, + "learning_rate": 1.4647685678227197e-06, + "loss": 0.1204, + "step": 11886 + }, + { + "epoch": 2.4925560914237788, + "grad_norm": 3.944329213724358, + "learning_rate": 1.4635891783174095e-06, + "loss": 0.1109, + "step": 11887 + }, + { + "epoch": 2.492765778989306, + "grad_norm": 4.061479007705403, + "learning_rate": 1.4624102263179584e-06, + "loss": 0.1287, + "step": 11888 + }, + { + "epoch": 2.4929754665548334, + "grad_norm": 4.143768006344261, + "learning_rate": 1.4612317118847886e-06, + "loss": 0.1354, + "step": 11889 + }, + { + "epoch": 2.4931851541203605, + "grad_norm": 3.965578465906405, + "learning_rate": 1.4600536350783057e-06, + "loss": 0.1164, + "step": 11890 + }, + { + "epoch": 2.493394841685888, + "grad_norm": 4.235821489849507, + "learning_rate": 1.4588759959588828e-06, + "loss": 0.1588, + "step": 11891 + }, + { + "epoch": 2.4936045292514155, + "grad_norm": 7.824522161100078, + "learning_rate": 1.4576987945868826e-06, + "loss": 0.1308, + "step": 11892 + }, + { + "epoch": 2.4938142168169426, + "grad_norm": 4.8478841524399225, + "learning_rate": 1.4565220310226314e-06, + "loss": 0.1325, + "step": 11893 + }, + { + "epoch": 2.49402390438247, + "grad_norm": 6.190552596379382, + "learning_rate": 1.4553457053264475e-06, + "loss": 0.25, + "step": 11894 + }, + { + "epoch": 2.4942335919479977, + "grad_norm": 5.6059324514679005, + "learning_rate": 1.4541698175586128e-06, + "loss": 0.2003, + "step": 11895 + }, + { + "epoch": 2.4944432795135247, + "grad_norm": 3.6088258137245575, + "learning_rate": 1.4529943677794e-06, + "loss": 0.0791, + "step": 11896 + }, + { + "epoch": 2.4946529670790523, + "grad_norm": 4.841589191709809, + "learning_rate": 1.4518193560490501e-06, + "loss": 0.1397, + "step": 11897 + }, + { + "epoch": 2.49486265464458, + "grad_norm": 4.014471890546808, + "learning_rate": 1.4506447824277815e-06, + "loss": 0.1332, + "step": 11898 + }, + { + "epoch": 2.495072342210107, + "grad_norm": 3.254902344881698, + "learning_rate": 1.4494706469757958e-06, + "loss": 0.1072, + "step": 11899 + }, + { + "epoch": 2.4952820297756344, + "grad_norm": 5.16455998807153, + "learning_rate": 1.4482969497532717e-06, + "loss": 0.1644, + "step": 11900 + }, + { + "epoch": 2.4954917173411615, + "grad_norm": 4.645343865534257, + "learning_rate": 1.4471236908203623e-06, + "loss": 0.1583, + "step": 11901 + }, + { + "epoch": 2.495701404906689, + "grad_norm": 3.996074222818098, + "learning_rate": 1.4459508702371939e-06, + "loss": 0.1125, + "step": 11902 + }, + { + "epoch": 2.4959110924722165, + "grad_norm": 3.4354103713489246, + "learning_rate": 1.4447784880638837e-06, + "loss": 0.0866, + "step": 11903 + }, + { + "epoch": 2.4961207800377436, + "grad_norm": 4.6672239621200085, + "learning_rate": 1.4436065443605108e-06, + "loss": 0.1495, + "step": 11904 + }, + { + "epoch": 2.496330467603271, + "grad_norm": 3.5709244690887085, + "learning_rate": 1.4424350391871455e-06, + "loss": 0.1466, + "step": 11905 + }, + { + "epoch": 2.4965401551687987, + "grad_norm": 4.081820902926082, + "learning_rate": 1.441263972603827e-06, + "loss": 0.1474, + "step": 11906 + }, + { + "epoch": 2.4967498427343258, + "grad_norm": 3.2126615183171428, + "learning_rate": 1.4400933446705712e-06, + "loss": 0.1117, + "step": 11907 + }, + { + "epoch": 2.4969595302998533, + "grad_norm": 3.203532200101062, + "learning_rate": 1.4389231554473803e-06, + "loss": 0.1104, + "step": 11908 + }, + { + "epoch": 2.4971692178653804, + "grad_norm": 3.929159555677046, + "learning_rate": 1.4377534049942265e-06, + "loss": 0.1301, + "step": 11909 + }, + { + "epoch": 2.497378905430908, + "grad_norm": 4.429654200956873, + "learning_rate": 1.4365840933710594e-06, + "loss": 0.147, + "step": 11910 + }, + { + "epoch": 2.4975885929964354, + "grad_norm": 3.69432365517285, + "learning_rate": 1.4354152206378092e-06, + "loss": 0.1379, + "step": 11911 + }, + { + "epoch": 2.4977982805619625, + "grad_norm": 4.5691739791666315, + "learning_rate": 1.4342467868543863e-06, + "loss": 0.1538, + "step": 11912 + }, + { + "epoch": 2.49800796812749, + "grad_norm": 5.3423270004317045, + "learning_rate": 1.433078792080672e-06, + "loss": 0.1998, + "step": 11913 + }, + { + "epoch": 2.4982176556930176, + "grad_norm": 4.046194460253145, + "learning_rate": 1.4319112363765253e-06, + "loss": 0.1309, + "step": 11914 + }, + { + "epoch": 2.4984273432585447, + "grad_norm": 4.337037682355158, + "learning_rate": 1.4307441198017912e-06, + "loss": 0.1344, + "step": 11915 + }, + { + "epoch": 2.498637030824072, + "grad_norm": 5.701306146940049, + "learning_rate": 1.429577442416281e-06, + "loss": 0.1413, + "step": 11916 + }, + { + "epoch": 2.4988467183895997, + "grad_norm": 4.030430621545223, + "learning_rate": 1.4284112042797938e-06, + "loss": 0.1435, + "step": 11917 + }, + { + "epoch": 2.499056405955127, + "grad_norm": 4.4463605065432015, + "learning_rate": 1.4272454054520989e-06, + "loss": 0.1432, + "step": 11918 + }, + { + "epoch": 2.4992660935206543, + "grad_norm": 5.363547812191036, + "learning_rate": 1.4260800459929447e-06, + "loss": 0.1782, + "step": 11919 + }, + { + "epoch": 2.4994757810861814, + "grad_norm": 6.644405939881724, + "learning_rate": 1.4249151259620575e-06, + "loss": 0.1386, + "step": 11920 + }, + { + "epoch": 2.499685468651709, + "grad_norm": 3.6327690025771116, + "learning_rate": 1.4237506454191452e-06, + "loss": 0.1134, + "step": 11921 + }, + { + "epoch": 2.4998951562172365, + "grad_norm": 4.015831231529636, + "learning_rate": 1.422586604423888e-06, + "loss": 0.1499, + "step": 11922 + }, + { + "epoch": 2.5001048437827635, + "grad_norm": 4.13129517418543, + "learning_rate": 1.421423003035941e-06, + "loss": 0.1274, + "step": 11923 + }, + { + "epoch": 2.500314531348291, + "grad_norm": 3.805114010358794, + "learning_rate": 1.420259841314946e-06, + "loss": 0.1407, + "step": 11924 + }, + { + "epoch": 2.500524218913818, + "grad_norm": 5.7804665427700614, + "learning_rate": 1.419097119320515e-06, + "loss": 0.1982, + "step": 11925 + }, + { + "epoch": 2.5007339064793457, + "grad_norm": 5.501558371091093, + "learning_rate": 1.417934837112237e-06, + "loss": 0.1747, + "step": 11926 + }, + { + "epoch": 2.500943594044873, + "grad_norm": 3.276467546611582, + "learning_rate": 1.4167729947496866e-06, + "loss": 0.1309, + "step": 11927 + }, + { + "epoch": 2.5011532816104003, + "grad_norm": 4.0470134797024055, + "learning_rate": 1.415611592292404e-06, + "loss": 0.1292, + "step": 11928 + }, + { + "epoch": 2.501362969175928, + "grad_norm": 4.432152411732758, + "learning_rate": 1.4144506297999194e-06, + "loss": 0.1484, + "step": 11929 + }, + { + "epoch": 2.5015726567414553, + "grad_norm": 3.6012859174280756, + "learning_rate": 1.4132901073317295e-06, + "loss": 0.1143, + "step": 11930 + }, + { + "epoch": 2.5017823443069824, + "grad_norm": 4.273226823443075, + "learning_rate": 1.4121300249473135e-06, + "loss": 0.1488, + "step": 11931 + }, + { + "epoch": 2.50199203187251, + "grad_norm": 4.432798716182812, + "learning_rate": 1.4109703827061282e-06, + "loss": 0.1273, + "step": 11932 + }, + { + "epoch": 2.5022017194380375, + "grad_norm": 4.79172093440161, + "learning_rate": 1.4098111806676095e-06, + "loss": 0.1554, + "step": 11933 + }, + { + "epoch": 2.5024114070035646, + "grad_norm": 4.341485738331202, + "learning_rate": 1.4086524188911677e-06, + "loss": 0.1349, + "step": 11934 + }, + { + "epoch": 2.502621094569092, + "grad_norm": 3.915140412456316, + "learning_rate": 1.4074940974361872e-06, + "loss": 0.1279, + "step": 11935 + }, + { + "epoch": 2.5028307821346196, + "grad_norm": 4.968978162466264, + "learning_rate": 1.4063362163620397e-06, + "loss": 0.1358, + "step": 11936 + }, + { + "epoch": 2.5030404697001467, + "grad_norm": 4.410629895339175, + "learning_rate": 1.4051787757280632e-06, + "loss": 0.1137, + "step": 11937 + }, + { + "epoch": 2.5032501572656742, + "grad_norm": 4.811616931515607, + "learning_rate": 1.4040217755935847e-06, + "loss": 0.1526, + "step": 11938 + }, + { + "epoch": 2.5034598448312018, + "grad_norm": 5.096220643223814, + "learning_rate": 1.402865216017899e-06, + "loss": 0.1355, + "step": 11939 + }, + { + "epoch": 2.503669532396729, + "grad_norm": 3.1627960000059105, + "learning_rate": 1.4017090970602786e-06, + "loss": 0.0897, + "step": 11940 + }, + { + "epoch": 2.5038792199622564, + "grad_norm": 4.976500763802138, + "learning_rate": 1.4005534187799807e-06, + "loss": 0.1876, + "step": 11941 + }, + { + "epoch": 2.5040889075277835, + "grad_norm": 5.218924275003913, + "learning_rate": 1.39939818123624e-06, + "loss": 0.1678, + "step": 11942 + }, + { + "epoch": 2.504298595093311, + "grad_norm": 3.584119962900178, + "learning_rate": 1.3982433844882548e-06, + "loss": 0.1082, + "step": 11943 + }, + { + "epoch": 2.504508282658838, + "grad_norm": 4.01421327292469, + "learning_rate": 1.3970890285952155e-06, + "loss": 0.1482, + "step": 11944 + }, + { + "epoch": 2.5047179702243656, + "grad_norm": 3.562057101592667, + "learning_rate": 1.3959351136162858e-06, + "loss": 0.1407, + "step": 11945 + }, + { + "epoch": 2.504927657789893, + "grad_norm": 3.533009160033275, + "learning_rate": 1.394781639610605e-06, + "loss": 0.133, + "step": 11946 + }, + { + "epoch": 2.50513734535542, + "grad_norm": 3.8199708111494934, + "learning_rate": 1.3936286066372874e-06, + "loss": 0.101, + "step": 11947 + }, + { + "epoch": 2.5053470329209477, + "grad_norm": 3.9010696471935677, + "learning_rate": 1.392476014755434e-06, + "loss": 0.1213, + "step": 11948 + }, + { + "epoch": 2.5055567204864753, + "grad_norm": 4.568045863687571, + "learning_rate": 1.391323864024111e-06, + "loss": 0.1584, + "step": 11949 + }, + { + "epoch": 2.5057664080520023, + "grad_norm": 4.273464364393656, + "learning_rate": 1.3901721545023727e-06, + "loss": 0.135, + "step": 11950 + }, + { + "epoch": 2.50597609561753, + "grad_norm": 4.957609641711073, + "learning_rate": 1.3890208862492449e-06, + "loss": 0.1421, + "step": 11951 + }, + { + "epoch": 2.5061857831830574, + "grad_norm": 3.625858105636269, + "learning_rate": 1.3878700593237305e-06, + "loss": 0.1555, + "step": 11952 + }, + { + "epoch": 2.5063954707485845, + "grad_norm": 3.7182816054675873, + "learning_rate": 1.386719673784811e-06, + "loss": 0.1388, + "step": 11953 + }, + { + "epoch": 2.506605158314112, + "grad_norm": 3.916620171108524, + "learning_rate": 1.3855697296914494e-06, + "loss": 0.1454, + "step": 11954 + }, + { + "epoch": 2.5068148458796395, + "grad_norm": 4.053578213200887, + "learning_rate": 1.3844202271025808e-06, + "loss": 0.1579, + "step": 11955 + }, + { + "epoch": 2.5070245334451666, + "grad_norm": 3.7737396159521586, + "learning_rate": 1.383271166077117e-06, + "loss": 0.1041, + "step": 11956 + }, + { + "epoch": 2.507234221010694, + "grad_norm": 5.14656676329622, + "learning_rate": 1.3821225466739497e-06, + "loss": 0.1447, + "step": 11957 + }, + { + "epoch": 2.5074439085762217, + "grad_norm": 5.132166351965843, + "learning_rate": 1.3809743689519517e-06, + "loss": 0.1484, + "step": 11958 + }, + { + "epoch": 2.5076535961417488, + "grad_norm": 4.81106712073615, + "learning_rate": 1.3798266329699662e-06, + "loss": 0.14, + "step": 11959 + }, + { + "epoch": 2.5078632837072763, + "grad_norm": 5.235037893516624, + "learning_rate": 1.3786793387868136e-06, + "loss": 0.1846, + "step": 11960 + }, + { + "epoch": 2.5080729712728034, + "grad_norm": 4.257054935877029, + "learning_rate": 1.3775324864613005e-06, + "loss": 0.1157, + "step": 11961 + }, + { + "epoch": 2.508282658838331, + "grad_norm": 4.103525328318305, + "learning_rate": 1.3763860760522007e-06, + "loss": 0.1356, + "step": 11962 + }, + { + "epoch": 2.508492346403858, + "grad_norm": 3.149713995941829, + "learning_rate": 1.3752401076182732e-06, + "loss": 0.0879, + "step": 11963 + }, + { + "epoch": 2.5087020339693855, + "grad_norm": 4.429226229382079, + "learning_rate": 1.3740945812182493e-06, + "loss": 0.1267, + "step": 11964 + }, + { + "epoch": 2.508911721534913, + "grad_norm": 6.547600176378466, + "learning_rate": 1.372949496910837e-06, + "loss": 0.1851, + "step": 11965 + }, + { + "epoch": 2.50912140910044, + "grad_norm": 5.478268930419371, + "learning_rate": 1.3718048547547247e-06, + "loss": 0.1578, + "step": 11966 + }, + { + "epoch": 2.5093310966659677, + "grad_norm": 3.4089511869476508, + "learning_rate": 1.3706606548085843e-06, + "loss": 0.1025, + "step": 11967 + }, + { + "epoch": 2.509540784231495, + "grad_norm": 4.272993198373761, + "learning_rate": 1.3695168971310468e-06, + "loss": 0.1182, + "step": 11968 + }, + { + "epoch": 2.5097504717970223, + "grad_norm": 3.8687937158312473, + "learning_rate": 1.3683735817807375e-06, + "loss": 0.123, + "step": 11969 + }, + { + "epoch": 2.50996015936255, + "grad_norm": 5.116617846085351, + "learning_rate": 1.3672307088162552e-06, + "loss": 0.1805, + "step": 11970 + }, + { + "epoch": 2.5101698469280773, + "grad_norm": 3.416887576970437, + "learning_rate": 1.3660882782961716e-06, + "loss": 0.1152, + "step": 11971 + }, + { + "epoch": 2.5103795344936044, + "grad_norm": 4.35034445318283, + "learning_rate": 1.3649462902790366e-06, + "loss": 0.1452, + "step": 11972 + }, + { + "epoch": 2.510589222059132, + "grad_norm": 5.142296630161542, + "learning_rate": 1.3638047448233828e-06, + "loss": 0.1513, + "step": 11973 + }, + { + "epoch": 2.5107989096246595, + "grad_norm": 4.3035813605968425, + "learning_rate": 1.3626636419877127e-06, + "loss": 0.15, + "step": 11974 + }, + { + "epoch": 2.5110085971901865, + "grad_norm": 4.168059628049141, + "learning_rate": 1.361522981830513e-06, + "loss": 0.1372, + "step": 11975 + }, + { + "epoch": 2.511218284755714, + "grad_norm": 4.294870549754423, + "learning_rate": 1.3603827644102442e-06, + "loss": 0.1195, + "step": 11976 + }, + { + "epoch": 2.5114279723212416, + "grad_norm": 4.442401668544624, + "learning_rate": 1.359242989785341e-06, + "loss": 0.1594, + "step": 11977 + }, + { + "epoch": 2.5116376598867687, + "grad_norm": 4.309957658452955, + "learning_rate": 1.3581036580142203e-06, + "loss": 0.1546, + "step": 11978 + }, + { + "epoch": 2.511847347452296, + "grad_norm": 3.985671009795197, + "learning_rate": 1.356964769155279e-06, + "loss": 0.1014, + "step": 11979 + }, + { + "epoch": 2.5120570350178233, + "grad_norm": 4.412188089368526, + "learning_rate": 1.3558263232668834e-06, + "loss": 0.1645, + "step": 11980 + }, + { + "epoch": 2.512266722583351, + "grad_norm": 5.79969767940507, + "learning_rate": 1.3546883204073791e-06, + "loss": 0.1546, + "step": 11981 + }, + { + "epoch": 2.512476410148878, + "grad_norm": 4.120770821190215, + "learning_rate": 1.353550760635095e-06, + "loss": 0.1303, + "step": 11982 + }, + { + "epoch": 2.5126860977144054, + "grad_norm": 4.666419260275762, + "learning_rate": 1.352413644008329e-06, + "loss": 0.1216, + "step": 11983 + }, + { + "epoch": 2.512895785279933, + "grad_norm": 4.828422451771381, + "learning_rate": 1.3512769705853645e-06, + "loss": 0.1379, + "step": 11984 + }, + { + "epoch": 2.51310547284546, + "grad_norm": 5.226512129823166, + "learning_rate": 1.350140740424456e-06, + "loss": 0.1472, + "step": 11985 + }, + { + "epoch": 2.5133151604109876, + "grad_norm": 3.553453144006635, + "learning_rate": 1.3490049535838346e-06, + "loss": 0.1275, + "step": 11986 + }, + { + "epoch": 2.513524847976515, + "grad_norm": 3.909624118065923, + "learning_rate": 1.3478696101217147e-06, + "loss": 0.1281, + "step": 11987 + }, + { + "epoch": 2.513734535542042, + "grad_norm": 3.401753074454119, + "learning_rate": 1.3467347100962846e-06, + "loss": 0.139, + "step": 11988 + }, + { + "epoch": 2.5139442231075697, + "grad_norm": 4.843046943435912, + "learning_rate": 1.3456002535657075e-06, + "loss": 0.1664, + "step": 11989 + }, + { + "epoch": 2.5141539106730972, + "grad_norm": 4.106305895195221, + "learning_rate": 1.344466240588127e-06, + "loss": 0.1149, + "step": 11990 + }, + { + "epoch": 2.5143635982386243, + "grad_norm": 3.5149617416525523, + "learning_rate": 1.3433326712216655e-06, + "loss": 0.1361, + "step": 11991 + }, + { + "epoch": 2.514573285804152, + "grad_norm": 4.009176114200098, + "learning_rate": 1.3421995455244197e-06, + "loss": 0.1114, + "step": 11992 + }, + { + "epoch": 2.5147829733696794, + "grad_norm": 6.195707950592568, + "learning_rate": 1.3410668635544611e-06, + "loss": 0.1424, + "step": 11993 + }, + { + "epoch": 2.5149926609352065, + "grad_norm": 4.342843562877368, + "learning_rate": 1.3399346253698453e-06, + "loss": 0.1674, + "step": 11994 + }, + { + "epoch": 2.515202348500734, + "grad_norm": 4.2814391036323824, + "learning_rate": 1.3388028310285984e-06, + "loss": 0.1472, + "step": 11995 + }, + { + "epoch": 2.5154120360662615, + "grad_norm": 3.8543805964456577, + "learning_rate": 1.3376714805887313e-06, + "loss": 0.1184, + "step": 11996 + }, + { + "epoch": 2.5156217236317886, + "grad_norm": 5.104708674824505, + "learning_rate": 1.3365405741082238e-06, + "loss": 0.1436, + "step": 11997 + }, + { + "epoch": 2.515831411197316, + "grad_norm": 4.311931555366447, + "learning_rate": 1.3354101116450358e-06, + "loss": 0.1526, + "step": 11998 + }, + { + "epoch": 2.516041098762843, + "grad_norm": 4.714281472228413, + "learning_rate": 1.3342800932571088e-06, + "loss": 0.1513, + "step": 11999 + }, + { + "epoch": 2.5162507863283707, + "grad_norm": 4.154490004019689, + "learning_rate": 1.3331505190023575e-06, + "loss": 0.0979, + "step": 12000 + }, + { + "epoch": 2.516460473893898, + "grad_norm": 3.556873884857982, + "learning_rate": 1.3320213889386758e-06, + "loss": 0.0952, + "step": 12001 + }, + { + "epoch": 2.5166701614594253, + "grad_norm": 6.4603490518250855, + "learning_rate": 1.3308927031239283e-06, + "loss": 0.1621, + "step": 12002 + }, + { + "epoch": 2.516879849024953, + "grad_norm": 3.5325152788296665, + "learning_rate": 1.3297644616159678e-06, + "loss": 0.1358, + "step": 12003 + }, + { + "epoch": 2.51708953659048, + "grad_norm": 4.93648754299256, + "learning_rate": 1.3286366644726167e-06, + "loss": 0.1642, + "step": 12004 + }, + { + "epoch": 2.5172992241560075, + "grad_norm": 4.869240160726903, + "learning_rate": 1.3275093117516736e-06, + "loss": 0.1486, + "step": 12005 + }, + { + "epoch": 2.517508911721535, + "grad_norm": 4.149418153152551, + "learning_rate": 1.326382403510923e-06, + "loss": 0.1435, + "step": 12006 + }, + { + "epoch": 2.517718599287062, + "grad_norm": 3.2486215991616367, + "learning_rate": 1.3252559398081143e-06, + "loss": 0.1182, + "step": 12007 + }, + { + "epoch": 2.5179282868525896, + "grad_norm": 4.42344339710403, + "learning_rate": 1.3241299207009884e-06, + "loss": 0.1441, + "step": 12008 + }, + { + "epoch": 2.518137974418117, + "grad_norm": 3.5949422929091317, + "learning_rate": 1.3230043462472497e-06, + "loss": 0.1115, + "step": 12009 + }, + { + "epoch": 2.5183476619836442, + "grad_norm": 3.7304959039049366, + "learning_rate": 1.3218792165045868e-06, + "loss": 0.1156, + "step": 12010 + }, + { + "epoch": 2.5185573495491718, + "grad_norm": 4.615123594995057, + "learning_rate": 1.3207545315306647e-06, + "loss": 0.1259, + "step": 12011 + }, + { + "epoch": 2.5187670371146993, + "grad_norm": 4.589079645450028, + "learning_rate": 1.3196302913831294e-06, + "loss": 0.1156, + "step": 12012 + }, + { + "epoch": 2.5189767246802264, + "grad_norm": 4.313588161138957, + "learning_rate": 1.318506496119596e-06, + "loss": 0.1217, + "step": 12013 + }, + { + "epoch": 2.519186412245754, + "grad_norm": 3.4927513074928407, + "learning_rate": 1.3173831457976605e-06, + "loss": 0.1244, + "step": 12014 + }, + { + "epoch": 2.5193960998112814, + "grad_norm": 4.340219023502884, + "learning_rate": 1.316260240474897e-06, + "loss": 0.1357, + "step": 12015 + }, + { + "epoch": 2.5196057873768085, + "grad_norm": 4.089764962489509, + "learning_rate": 1.3151377802088605e-06, + "loss": 0.1335, + "step": 12016 + }, + { + "epoch": 2.519815474942336, + "grad_norm": 4.450427337379712, + "learning_rate": 1.3140157650570762e-06, + "loss": 0.1692, + "step": 12017 + }, + { + "epoch": 2.520025162507863, + "grad_norm": 4.7051551218925844, + "learning_rate": 1.3128941950770457e-06, + "loss": 0.145, + "step": 12018 + }, + { + "epoch": 2.5202348500733907, + "grad_norm": 5.663195163320106, + "learning_rate": 1.3117730703262576e-06, + "loss": 0.1676, + "step": 12019 + }, + { + "epoch": 2.5204445376389177, + "grad_norm": 4.918368963329637, + "learning_rate": 1.310652390862166e-06, + "loss": 0.1763, + "step": 12020 + }, + { + "epoch": 2.5206542252044453, + "grad_norm": 4.095045628502151, + "learning_rate": 1.3095321567422125e-06, + "loss": 0.1267, + "step": 12021 + }, + { + "epoch": 2.520863912769973, + "grad_norm": 4.391280609356593, + "learning_rate": 1.3084123680238093e-06, + "loss": 0.1359, + "step": 12022 + }, + { + "epoch": 2.5210736003355, + "grad_norm": 3.9288519600029908, + "learning_rate": 1.3072930247643446e-06, + "loss": 0.1255, + "step": 12023 + }, + { + "epoch": 2.5212832879010274, + "grad_norm": 4.777140562168834, + "learning_rate": 1.306174127021189e-06, + "loss": 0.1261, + "step": 12024 + }, + { + "epoch": 2.521492975466555, + "grad_norm": 4.355173517921103, + "learning_rate": 1.3050556748516928e-06, + "loss": 0.1759, + "step": 12025 + }, + { + "epoch": 2.521702663032082, + "grad_norm": 4.45975007443483, + "learning_rate": 1.3039376683131699e-06, + "loss": 0.1403, + "step": 12026 + }, + { + "epoch": 2.5219123505976095, + "grad_norm": 4.598514329648075, + "learning_rate": 1.3028201074629232e-06, + "loss": 0.162, + "step": 12027 + }, + { + "epoch": 2.522122038163137, + "grad_norm": 4.305533104953591, + "learning_rate": 1.3017029923582325e-06, + "loss": 0.1732, + "step": 12028 + }, + { + "epoch": 2.522331725728664, + "grad_norm": 3.3156434984605694, + "learning_rate": 1.300586323056351e-06, + "loss": 0.1259, + "step": 12029 + }, + { + "epoch": 2.5225414132941917, + "grad_norm": 5.20202719323522, + "learning_rate": 1.299470099614506e-06, + "loss": 0.1702, + "step": 12030 + }, + { + "epoch": 2.522751100859719, + "grad_norm": 5.03580238929334, + "learning_rate": 1.2983543220899109e-06, + "loss": 0.1698, + "step": 12031 + }, + { + "epoch": 2.5229607884252463, + "grad_norm": 3.654331092047329, + "learning_rate": 1.2972389905397464e-06, + "loss": 0.1164, + "step": 12032 + }, + { + "epoch": 2.523170475990774, + "grad_norm": 3.9086765493936197, + "learning_rate": 1.2961241050211803e-06, + "loss": 0.1512, + "step": 12033 + }, + { + "epoch": 2.5233801635563013, + "grad_norm": 5.749241166765074, + "learning_rate": 1.2950096655913502e-06, + "loss": 0.19, + "step": 12034 + }, + { + "epoch": 2.5235898511218284, + "grad_norm": 4.504616013744337, + "learning_rate": 1.2938956723073702e-06, + "loss": 0.1533, + "step": 12035 + }, + { + "epoch": 2.523799538687356, + "grad_norm": 4.314661691316422, + "learning_rate": 1.2927821252263374e-06, + "loss": 0.1602, + "step": 12036 + }, + { + "epoch": 2.524009226252883, + "grad_norm": 4.136352361865108, + "learning_rate": 1.2916690244053242e-06, + "loss": 0.1492, + "step": 12037 + }, + { + "epoch": 2.5242189138184106, + "grad_norm": 4.595024843800442, + "learning_rate": 1.290556369901378e-06, + "loss": 0.1283, + "step": 12038 + }, + { + "epoch": 2.524428601383938, + "grad_norm": 3.8727865206151737, + "learning_rate": 1.2894441617715203e-06, + "loss": 0.1525, + "step": 12039 + }, + { + "epoch": 2.524638288949465, + "grad_norm": 3.9519770389262066, + "learning_rate": 1.2883324000727604e-06, + "loss": 0.1117, + "step": 12040 + }, + { + "epoch": 2.5248479765149927, + "grad_norm": 4.368750688028554, + "learning_rate": 1.2872210848620714e-06, + "loss": 0.1499, + "step": 12041 + }, + { + "epoch": 2.52505766408052, + "grad_norm": 3.555174894899924, + "learning_rate": 1.2861102161964168e-06, + "loss": 0.123, + "step": 12042 + }, + { + "epoch": 2.5252673516460473, + "grad_norm": 3.9178413015490037, + "learning_rate": 1.2849997941327263e-06, + "loss": 0.1205, + "step": 12043 + }, + { + "epoch": 2.525477039211575, + "grad_norm": 4.067045514018608, + "learning_rate": 1.2838898187279092e-06, + "loss": 0.1321, + "step": 12044 + }, + { + "epoch": 2.525686726777102, + "grad_norm": 6.014739319544709, + "learning_rate": 1.2827802900388565e-06, + "loss": 0.1853, + "step": 12045 + }, + { + "epoch": 2.5258964143426295, + "grad_norm": 4.17066000362069, + "learning_rate": 1.281671208122438e-06, + "loss": 0.1544, + "step": 12046 + }, + { + "epoch": 2.526106101908157, + "grad_norm": 4.341134418176675, + "learning_rate": 1.2805625730354864e-06, + "loss": 0.1242, + "step": 12047 + }, + { + "epoch": 2.526315789473684, + "grad_norm": 3.922347868807532, + "learning_rate": 1.2794543848348263e-06, + "loss": 0.1365, + "step": 12048 + }, + { + "epoch": 2.5265254770392116, + "grad_norm": 4.054015972457748, + "learning_rate": 1.2783466435772561e-06, + "loss": 0.1409, + "step": 12049 + }, + { + "epoch": 2.526735164604739, + "grad_norm": 4.0070372787037165, + "learning_rate": 1.2772393493195478e-06, + "loss": 0.1191, + "step": 12050 + }, + { + "epoch": 2.526944852170266, + "grad_norm": 5.828187932896821, + "learning_rate": 1.276132502118449e-06, + "loss": 0.1759, + "step": 12051 + }, + { + "epoch": 2.5271545397357937, + "grad_norm": 4.536732875710181, + "learning_rate": 1.2750261020306932e-06, + "loss": 0.1559, + "step": 12052 + }, + { + "epoch": 2.5273642273013213, + "grad_norm": 4.102793797323526, + "learning_rate": 1.2739201491129793e-06, + "loss": 0.1697, + "step": 12053 + }, + { + "epoch": 2.5275739148668483, + "grad_norm": 3.5724096703256745, + "learning_rate": 1.2728146434219945e-06, + "loss": 0.1179, + "step": 12054 + }, + { + "epoch": 2.527783602432376, + "grad_norm": 4.669263104624574, + "learning_rate": 1.2717095850143957e-06, + "loss": 0.163, + "step": 12055 + }, + { + "epoch": 2.527993289997903, + "grad_norm": 4.6980404046320885, + "learning_rate": 1.2706049739468173e-06, + "loss": 0.1585, + "step": 12056 + }, + { + "epoch": 2.5282029775634305, + "grad_norm": 3.466171322114327, + "learning_rate": 1.2695008102758745e-06, + "loss": 0.1099, + "step": 12057 + }, + { + "epoch": 2.528412665128958, + "grad_norm": 3.8391463490694577, + "learning_rate": 1.2683970940581592e-06, + "loss": 0.1324, + "step": 12058 + }, + { + "epoch": 2.528622352694485, + "grad_norm": 3.974302202206801, + "learning_rate": 1.2672938253502376e-06, + "loss": 0.1326, + "step": 12059 + }, + { + "epoch": 2.5288320402600126, + "grad_norm": 2.9930757205397143, + "learning_rate": 1.2661910042086511e-06, + "loss": 0.09, + "step": 12060 + }, + { + "epoch": 2.5290417278255397, + "grad_norm": 4.293084581468155, + "learning_rate": 1.2650886306899257e-06, + "loss": 0.1623, + "step": 12061 + }, + { + "epoch": 2.5292514153910672, + "grad_norm": 4.316655061744194, + "learning_rate": 1.2639867048505561e-06, + "loss": 0.1454, + "step": 12062 + }, + { + "epoch": 2.5294611029565948, + "grad_norm": 4.052800192011256, + "learning_rate": 1.2628852267470214e-06, + "loss": 0.1489, + "step": 12063 + }, + { + "epoch": 2.529670790522122, + "grad_norm": 4.266121938206998, + "learning_rate": 1.2617841964357735e-06, + "loss": 0.1304, + "step": 12064 + }, + { + "epoch": 2.5298804780876494, + "grad_norm": 4.010083714910853, + "learning_rate": 1.2606836139732381e-06, + "loss": 0.1256, + "step": 12065 + }, + { + "epoch": 2.530090165653177, + "grad_norm": 5.1010578544502465, + "learning_rate": 1.2595834794158278e-06, + "loss": 0.1878, + "step": 12066 + }, + { + "epoch": 2.530299853218704, + "grad_norm": 4.626352693511925, + "learning_rate": 1.2584837928199234e-06, + "loss": 0.1639, + "step": 12067 + }, + { + "epoch": 2.5305095407842315, + "grad_norm": 4.109030503348539, + "learning_rate": 1.257384554241884e-06, + "loss": 0.1417, + "step": 12068 + }, + { + "epoch": 2.530719228349759, + "grad_norm": 4.345874135370887, + "learning_rate": 1.2562857637380487e-06, + "loss": 0.1285, + "step": 12069 + }, + { + "epoch": 2.530928915915286, + "grad_norm": 3.897247403588684, + "learning_rate": 1.2551874213647363e-06, + "loss": 0.127, + "step": 12070 + }, + { + "epoch": 2.5311386034808137, + "grad_norm": 5.046250749443973, + "learning_rate": 1.254089527178235e-06, + "loss": 0.178, + "step": 12071 + }, + { + "epoch": 2.531348291046341, + "grad_norm": 3.923858141060383, + "learning_rate": 1.2529920812348129e-06, + "loss": 0.1153, + "step": 12072 + }, + { + "epoch": 2.5315579786118683, + "grad_norm": 4.857445168353048, + "learning_rate": 1.2518950835907196e-06, + "loss": 0.1494, + "step": 12073 + }, + { + "epoch": 2.531767666177396, + "grad_norm": 3.876565023185596, + "learning_rate": 1.2507985343021734e-06, + "loss": 0.1229, + "step": 12074 + }, + { + "epoch": 2.5319773537429233, + "grad_norm": 4.286467944324081, + "learning_rate": 1.2497024334253794e-06, + "loss": 0.1652, + "step": 12075 + }, + { + "epoch": 2.5321870413084504, + "grad_norm": 5.175026986386072, + "learning_rate": 1.2486067810165103e-06, + "loss": 0.1579, + "step": 12076 + }, + { + "epoch": 2.532396728873978, + "grad_norm": 5.199711775349933, + "learning_rate": 1.2475115771317247e-06, + "loss": 0.1594, + "step": 12077 + }, + { + "epoch": 2.532606416439505, + "grad_norm": 3.231076942436497, + "learning_rate": 1.2464168218271488e-06, + "loss": 0.1059, + "step": 12078 + }, + { + "epoch": 2.5328161040050325, + "grad_norm": 4.880412867082489, + "learning_rate": 1.2453225151588955e-06, + "loss": 0.1153, + "step": 12079 + }, + { + "epoch": 2.5330257915705596, + "grad_norm": 4.1682662607423895, + "learning_rate": 1.2442286571830463e-06, + "loss": 0.1456, + "step": 12080 + }, + { + "epoch": 2.533235479136087, + "grad_norm": 4.932167973036506, + "learning_rate": 1.2431352479556635e-06, + "loss": 0.1463, + "step": 12081 + }, + { + "epoch": 2.5334451667016147, + "grad_norm": 3.106961178091361, + "learning_rate": 1.2420422875327876e-06, + "loss": 0.137, + "step": 12082 + }, + { + "epoch": 2.5336548542671418, + "grad_norm": 4.5231630360642185, + "learning_rate": 1.2409497759704381e-06, + "loss": 0.1101, + "step": 12083 + }, + { + "epoch": 2.5338645418326693, + "grad_norm": 3.9527462531187547, + "learning_rate": 1.2398577133245993e-06, + "loss": 0.1304, + "step": 12084 + }, + { + "epoch": 2.534074229398197, + "grad_norm": 3.710421360035279, + "learning_rate": 1.2387660996512475e-06, + "loss": 0.1222, + "step": 12085 + }, + { + "epoch": 2.534283916963724, + "grad_norm": 4.980933200086237, + "learning_rate": 1.23767493500633e-06, + "loss": 0.1441, + "step": 12086 + }, + { + "epoch": 2.5344936045292514, + "grad_norm": 4.464371317468486, + "learning_rate": 1.23658421944577e-06, + "loss": 0.1289, + "step": 12087 + }, + { + "epoch": 2.534703292094779, + "grad_norm": 4.246290248094093, + "learning_rate": 1.235493953025466e-06, + "loss": 0.0971, + "step": 12088 + }, + { + "epoch": 2.534912979660306, + "grad_norm": 3.75869695858341, + "learning_rate": 1.2344041358013004e-06, + "loss": 0.1259, + "step": 12089 + }, + { + "epoch": 2.5351226672258336, + "grad_norm": 3.9317376316210533, + "learning_rate": 1.2333147678291246e-06, + "loss": 0.1158, + "step": 12090 + }, + { + "epoch": 2.535332354791361, + "grad_norm": 5.026491155850308, + "learning_rate": 1.2322258491647732e-06, + "loss": 0.1732, + "step": 12091 + }, + { + "epoch": 2.535542042356888, + "grad_norm": 4.072500172156615, + "learning_rate": 1.2311373798640536e-06, + "loss": 0.1289, + "step": 12092 + }, + { + "epoch": 2.5357517299224157, + "grad_norm": 3.6114886183997132, + "learning_rate": 1.2300493599827512e-06, + "loss": 0.1215, + "step": 12093 + }, + { + "epoch": 2.5359614174879432, + "grad_norm": 4.924318047779965, + "learning_rate": 1.228961789576628e-06, + "loss": 0.1841, + "step": 12094 + }, + { + "epoch": 2.5361711050534703, + "grad_norm": 3.660489197343683, + "learning_rate": 1.2278746687014298e-06, + "loss": 0.1366, + "step": 12095 + }, + { + "epoch": 2.536380792618998, + "grad_norm": 3.075210888560592, + "learning_rate": 1.2267879974128682e-06, + "loss": 0.1183, + "step": 12096 + }, + { + "epoch": 2.536590480184525, + "grad_norm": 4.842141918057536, + "learning_rate": 1.2257017757666357e-06, + "loss": 0.1768, + "step": 12097 + }, + { + "epoch": 2.5368001677500525, + "grad_norm": 4.5117673317039975, + "learning_rate": 1.2246160038184086e-06, + "loss": 0.172, + "step": 12098 + }, + { + "epoch": 2.5370098553155795, + "grad_norm": 3.2133546925982905, + "learning_rate": 1.223530681623829e-06, + "loss": 0.1076, + "step": 12099 + }, + { + "epoch": 2.537219542881107, + "grad_norm": 4.121729677409907, + "learning_rate": 1.2224458092385249e-06, + "loss": 0.1068, + "step": 12100 + }, + { + "epoch": 2.5374292304466346, + "grad_norm": 3.1666248258739556, + "learning_rate": 1.2213613867180984e-06, + "loss": 0.1027, + "step": 12101 + }, + { + "epoch": 2.5376389180121617, + "grad_norm": 3.9782939538031443, + "learning_rate": 1.220277414118124e-06, + "loss": 0.1402, + "step": 12102 + }, + { + "epoch": 2.537848605577689, + "grad_norm": 4.721290246767434, + "learning_rate": 1.2191938914941604e-06, + "loss": 0.1643, + "step": 12103 + }, + { + "epoch": 2.5380582931432167, + "grad_norm": 3.7179590897315387, + "learning_rate": 1.218110818901742e-06, + "loss": 0.1446, + "step": 12104 + }, + { + "epoch": 2.538267980708744, + "grad_norm": 3.887924404743961, + "learning_rate": 1.2170281963963726e-06, + "loss": 0.133, + "step": 12105 + }, + { + "epoch": 2.5384776682742713, + "grad_norm": 3.842412955451362, + "learning_rate": 1.2159460240335418e-06, + "loss": 0.1034, + "step": 12106 + }, + { + "epoch": 2.538687355839799, + "grad_norm": 3.248488634442362, + "learning_rate": 1.2148643018687134e-06, + "loss": 0.0961, + "step": 12107 + }, + { + "epoch": 2.538897043405326, + "grad_norm": 2.800003844753062, + "learning_rate": 1.2137830299573273e-06, + "loss": 0.1049, + "step": 12108 + }, + { + "epoch": 2.5391067309708535, + "grad_norm": 3.9780201446554218, + "learning_rate": 1.212702208354798e-06, + "loss": 0.1589, + "step": 12109 + }, + { + "epoch": 2.539316418536381, + "grad_norm": 4.836320944876949, + "learning_rate": 1.2116218371165222e-06, + "loss": 0.1555, + "step": 12110 + }, + { + "epoch": 2.539526106101908, + "grad_norm": 3.4471527807202156, + "learning_rate": 1.2105419162978681e-06, + "loss": 0.1064, + "step": 12111 + }, + { + "epoch": 2.5397357936674356, + "grad_norm": 3.2225002560285274, + "learning_rate": 1.2094624459541882e-06, + "loss": 0.1312, + "step": 12112 + }, + { + "epoch": 2.539945481232963, + "grad_norm": 4.685250277471805, + "learning_rate": 1.2083834261408046e-06, + "loss": 0.1749, + "step": 12113 + }, + { + "epoch": 2.5401551687984902, + "grad_norm": 3.822468710446993, + "learning_rate": 1.2073048569130152e-06, + "loss": 0.1239, + "step": 12114 + }, + { + "epoch": 2.5403648563640178, + "grad_norm": 4.116738197839818, + "learning_rate": 1.2062267383261039e-06, + "loss": 0.1265, + "step": 12115 + }, + { + "epoch": 2.540574543929545, + "grad_norm": 5.412632927717294, + "learning_rate": 1.2051490704353252e-06, + "loss": 0.1638, + "step": 12116 + }, + { + "epoch": 2.5407842314950724, + "grad_norm": 6.344457023928824, + "learning_rate": 1.2040718532959106e-06, + "loss": 0.1685, + "step": 12117 + }, + { + "epoch": 2.5409939190605995, + "grad_norm": 3.1100604668164613, + "learning_rate": 1.2029950869630669e-06, + "loss": 0.1045, + "step": 12118 + }, + { + "epoch": 2.541203606626127, + "grad_norm": 4.019892926529375, + "learning_rate": 1.2019187714919856e-06, + "loss": 0.1236, + "step": 12119 + }, + { + "epoch": 2.5414132941916545, + "grad_norm": 3.6330324999787353, + "learning_rate": 1.2008429069378237e-06, + "loss": 0.12, + "step": 12120 + }, + { + "epoch": 2.5416229817571816, + "grad_norm": 3.963197008644183, + "learning_rate": 1.1997674933557268e-06, + "loss": 0.1182, + "step": 12121 + }, + { + "epoch": 2.541832669322709, + "grad_norm": 4.492945230785115, + "learning_rate": 1.1986925308008092e-06, + "loss": 0.1607, + "step": 12122 + }, + { + "epoch": 2.5420423568882367, + "grad_norm": 4.552721754242347, + "learning_rate": 1.1976180193281618e-06, + "loss": 0.1298, + "step": 12123 + }, + { + "epoch": 2.5422520444537637, + "grad_norm": 4.630313105402612, + "learning_rate": 1.1965439589928585e-06, + "loss": 0.1133, + "step": 12124 + }, + { + "epoch": 2.5424617320192913, + "grad_norm": 3.703271664132475, + "learning_rate": 1.1954703498499499e-06, + "loss": 0.101, + "step": 12125 + }, + { + "epoch": 2.542671419584819, + "grad_norm": 3.426792288780992, + "learning_rate": 1.1943971919544528e-06, + "loss": 0.1219, + "step": 12126 + }, + { + "epoch": 2.542881107150346, + "grad_norm": 4.877799172989464, + "learning_rate": 1.1933244853613723e-06, + "loss": 0.1536, + "step": 12127 + }, + { + "epoch": 2.5430907947158734, + "grad_norm": 3.6176260210292006, + "learning_rate": 1.1922522301256878e-06, + "loss": 0.1422, + "step": 12128 + }, + { + "epoch": 2.543300482281401, + "grad_norm": 4.322031373852679, + "learning_rate": 1.191180426302353e-06, + "loss": 0.121, + "step": 12129 + }, + { + "epoch": 2.543510169846928, + "grad_norm": 3.6169406834211695, + "learning_rate": 1.1901090739462972e-06, + "loss": 0.1128, + "step": 12130 + }, + { + "epoch": 2.5437198574124555, + "grad_norm": 3.9027743174794405, + "learning_rate": 1.1890381731124335e-06, + "loss": 0.1409, + "step": 12131 + }, + { + "epoch": 2.543929544977983, + "grad_norm": 4.355749160780657, + "learning_rate": 1.1879677238556441e-06, + "loss": 0.1397, + "step": 12132 + }, + { + "epoch": 2.54413923254351, + "grad_norm": 4.913415519397946, + "learning_rate": 1.1868977262307934e-06, + "loss": 0.1789, + "step": 12133 + }, + { + "epoch": 2.5443489201090377, + "grad_norm": 5.227240550887243, + "learning_rate": 1.1858281802927207e-06, + "loss": 0.1592, + "step": 12134 + }, + { + "epoch": 2.5445586076745648, + "grad_norm": 3.613132505834895, + "learning_rate": 1.1847590860962387e-06, + "loss": 0.1316, + "step": 12135 + }, + { + "epoch": 2.5447682952400923, + "grad_norm": 3.5862282596674095, + "learning_rate": 1.1836904436961427e-06, + "loss": 0.1332, + "step": 12136 + }, + { + "epoch": 2.5449779828056194, + "grad_norm": 4.509062968467781, + "learning_rate": 1.1826222531472042e-06, + "loss": 0.1508, + "step": 12137 + }, + { + "epoch": 2.545187670371147, + "grad_norm": 4.084265435634725, + "learning_rate": 1.1815545145041684e-06, + "loss": 0.0946, + "step": 12138 + }, + { + "epoch": 2.5453973579366744, + "grad_norm": 3.3336203354876597, + "learning_rate": 1.1804872278217572e-06, + "loss": 0.09, + "step": 12139 + }, + { + "epoch": 2.5456070455022015, + "grad_norm": 4.291465095902692, + "learning_rate": 1.179420393154672e-06, + "loss": 0.1372, + "step": 12140 + }, + { + "epoch": 2.545816733067729, + "grad_norm": 3.8620663893691805, + "learning_rate": 1.1783540105575919e-06, + "loss": 0.1385, + "step": 12141 + }, + { + "epoch": 2.5460264206332566, + "grad_norm": 3.8750848466434524, + "learning_rate": 1.1772880800851693e-06, + "loss": 0.1278, + "step": 12142 + }, + { + "epoch": 2.5462361081987837, + "grad_norm": 4.490364689655418, + "learning_rate": 1.176222601792032e-06, + "loss": 0.108, + "step": 12143 + }, + { + "epoch": 2.546445795764311, + "grad_norm": 5.810093459382098, + "learning_rate": 1.1751575757327938e-06, + "loss": 0.1535, + "step": 12144 + }, + { + "epoch": 2.5466554833298387, + "grad_norm": 4.040677427863925, + "learning_rate": 1.1740930019620344e-06, + "loss": 0.1357, + "step": 12145 + }, + { + "epoch": 2.546865170895366, + "grad_norm": 4.856363106439736, + "learning_rate": 1.173028880534316e-06, + "loss": 0.1551, + "step": 12146 + }, + { + "epoch": 2.5470748584608933, + "grad_norm": 4.619290062541604, + "learning_rate": 1.171965211504178e-06, + "loss": 0.1591, + "step": 12147 + }, + { + "epoch": 2.547284546026421, + "grad_norm": 3.1866159584941838, + "learning_rate": 1.1709019949261324e-06, + "loss": 0.105, + "step": 12148 + }, + { + "epoch": 2.547494233591948, + "grad_norm": 3.8019816413326537, + "learning_rate": 1.1698392308546747e-06, + "loss": 0.1136, + "step": 12149 + }, + { + "epoch": 2.5477039211574755, + "grad_norm": 4.153000596744503, + "learning_rate": 1.1687769193442723e-06, + "loss": 0.116, + "step": 12150 + }, + { + "epoch": 2.547913608723003, + "grad_norm": 4.438509064105294, + "learning_rate": 1.167715060449367e-06, + "loss": 0.1697, + "step": 12151 + }, + { + "epoch": 2.54812329628853, + "grad_norm": 5.281248480874893, + "learning_rate": 1.1666536542243845e-06, + "loss": 0.1282, + "step": 12152 + }, + { + "epoch": 2.5483329838540576, + "grad_norm": 4.527590737916612, + "learning_rate": 1.1655927007237245e-06, + "loss": 0.1192, + "step": 12153 + }, + { + "epoch": 2.5485426714195847, + "grad_norm": 4.2638813752373474, + "learning_rate": 1.1645322000017611e-06, + "loss": 0.1312, + "step": 12154 + }, + { + "epoch": 2.548752358985112, + "grad_norm": 4.775688618058503, + "learning_rate": 1.1634721521128444e-06, + "loss": 0.1444, + "step": 12155 + }, + { + "epoch": 2.5489620465506393, + "grad_norm": 6.646884400612622, + "learning_rate": 1.1624125571113075e-06, + "loss": 0.1615, + "step": 12156 + }, + { + "epoch": 2.549171734116167, + "grad_norm": 3.667316037949729, + "learning_rate": 1.161353415051455e-06, + "loss": 0.1395, + "step": 12157 + }, + { + "epoch": 2.5493814216816943, + "grad_norm": 4.5606721567798285, + "learning_rate": 1.1602947259875707e-06, + "loss": 0.1377, + "step": 12158 + }, + { + "epoch": 2.5495911092472214, + "grad_norm": 4.8617299094134125, + "learning_rate": 1.1592364899739127e-06, + "loss": 0.1427, + "step": 12159 + }, + { + "epoch": 2.549800796812749, + "grad_norm": 4.074640328585298, + "learning_rate": 1.1581787070647177e-06, + "loss": 0.1475, + "step": 12160 + }, + { + "epoch": 2.5500104843782765, + "grad_norm": 4.661976486560679, + "learning_rate": 1.1571213773141976e-06, + "loss": 0.1389, + "step": 12161 + }, + { + "epoch": 2.5502201719438036, + "grad_norm": 4.080649597937662, + "learning_rate": 1.1560645007765492e-06, + "loss": 0.1506, + "step": 12162 + }, + { + "epoch": 2.550429859509331, + "grad_norm": 4.26948789915272, + "learning_rate": 1.1550080775059302e-06, + "loss": 0.154, + "step": 12163 + }, + { + "epoch": 2.5506395470748586, + "grad_norm": 4.09343523487981, + "learning_rate": 1.1539521075564874e-06, + "loss": 0.1558, + "step": 12164 + }, + { + "epoch": 2.5508492346403857, + "grad_norm": 5.117742504350601, + "learning_rate": 1.1528965909823443e-06, + "loss": 0.13, + "step": 12165 + }, + { + "epoch": 2.5510589222059132, + "grad_norm": 3.3540629526650645, + "learning_rate": 1.151841527837596e-06, + "loss": 0.1015, + "step": 12166 + }, + { + "epoch": 2.5512686097714408, + "grad_norm": 4.435219058861561, + "learning_rate": 1.1507869181763131e-06, + "loss": 0.1565, + "step": 12167 + }, + { + "epoch": 2.551478297336968, + "grad_norm": 5.601116834718981, + "learning_rate": 1.14973276205255e-06, + "loss": 0.1488, + "step": 12168 + }, + { + "epoch": 2.5516879849024954, + "grad_norm": 3.9945961107232746, + "learning_rate": 1.1486790595203324e-06, + "loss": 0.1186, + "step": 12169 + }, + { + "epoch": 2.551897672468023, + "grad_norm": 5.422935968288645, + "learning_rate": 1.1476258106336658e-06, + "loss": 0.1851, + "step": 12170 + }, + { + "epoch": 2.55210736003355, + "grad_norm": 3.768837945187381, + "learning_rate": 1.1465730154465304e-06, + "loss": 0.1171, + "step": 12171 + }, + { + "epoch": 2.5523170475990775, + "grad_norm": 4.774831157902279, + "learning_rate": 1.1455206740128822e-06, + "loss": 0.1569, + "step": 12172 + }, + { + "epoch": 2.5525267351646046, + "grad_norm": 5.162741883923306, + "learning_rate": 1.1444687863866556e-06, + "loss": 0.1473, + "step": 12173 + }, + { + "epoch": 2.552736422730132, + "grad_norm": 4.466673355371275, + "learning_rate": 1.1434173526217652e-06, + "loss": 0.1723, + "step": 12174 + }, + { + "epoch": 2.552946110295659, + "grad_norm": 4.600500063498581, + "learning_rate": 1.1423663727720968e-06, + "loss": 0.1406, + "step": 12175 + }, + { + "epoch": 2.5531557978611867, + "grad_norm": 3.1679131878414806, + "learning_rate": 1.141315846891512e-06, + "loss": 0.1024, + "step": 12176 + }, + { + "epoch": 2.5533654854267143, + "grad_norm": 4.55054469783078, + "learning_rate": 1.1402657750338563e-06, + "loss": 0.1567, + "step": 12177 + }, + { + "epoch": 2.5535751729922413, + "grad_norm": 3.6596099490095693, + "learning_rate": 1.1392161572529447e-06, + "loss": 0.1109, + "step": 12178 + }, + { + "epoch": 2.553784860557769, + "grad_norm": 4.326735151682969, + "learning_rate": 1.1381669936025752e-06, + "loss": 0.1238, + "step": 12179 + }, + { + "epoch": 2.5539945481232964, + "grad_norm": 4.698726728378975, + "learning_rate": 1.1371182841365181e-06, + "loss": 0.1501, + "step": 12180 + }, + { + "epoch": 2.5542042356888235, + "grad_norm": 4.62920152682113, + "learning_rate": 1.136070028908518e-06, + "loss": 0.142, + "step": 12181 + }, + { + "epoch": 2.554413923254351, + "grad_norm": 4.708000937078805, + "learning_rate": 1.1350222279723034e-06, + "loss": 0.1473, + "step": 12182 + }, + { + "epoch": 2.5546236108198785, + "grad_norm": 3.394441095714444, + "learning_rate": 1.1339748813815777e-06, + "loss": 0.1157, + "step": 12183 + }, + { + "epoch": 2.5548332983854056, + "grad_norm": 6.067154093817903, + "learning_rate": 1.1329279891900136e-06, + "loss": 0.1245, + "step": 12184 + }, + { + "epoch": 2.555042985950933, + "grad_norm": 4.518632434887595, + "learning_rate": 1.1318815514512693e-06, + "loss": 0.152, + "step": 12185 + }, + { + "epoch": 2.5552526735164607, + "grad_norm": 4.916971049978362, + "learning_rate": 1.1308355682189787e-06, + "loss": 0.1526, + "step": 12186 + }, + { + "epoch": 2.5554623610819878, + "grad_norm": 4.108214294436182, + "learning_rate": 1.1297900395467476e-06, + "loss": 0.1436, + "step": 12187 + }, + { + "epoch": 2.5556720486475153, + "grad_norm": 3.965589090580005, + "learning_rate": 1.128744965488159e-06, + "loss": 0.1286, + "step": 12188 + }, + { + "epoch": 2.555881736213043, + "grad_norm": 4.645654937237945, + "learning_rate": 1.1277003460967794e-06, + "loss": 0.1335, + "step": 12189 + }, + { + "epoch": 2.55609142377857, + "grad_norm": 4.019856458206179, + "learning_rate": 1.1266561814261445e-06, + "loss": 0.1522, + "step": 12190 + }, + { + "epoch": 2.5563011113440974, + "grad_norm": 4.660677481605267, + "learning_rate": 1.1256124715297711e-06, + "loss": 0.1362, + "step": 12191 + }, + { + "epoch": 2.5565107989096245, + "grad_norm": 4.876085646494219, + "learning_rate": 1.124569216461151e-06, + "loss": 0.169, + "step": 12192 + }, + { + "epoch": 2.556720486475152, + "grad_norm": 4.122476044013269, + "learning_rate": 1.1235264162737503e-06, + "loss": 0.1372, + "step": 12193 + }, + { + "epoch": 2.556930174040679, + "grad_norm": 4.0528665743947965, + "learning_rate": 1.1224840710210161e-06, + "loss": 0.1081, + "step": 12194 + }, + { + "epoch": 2.5571398616062067, + "grad_norm": 4.019532730094852, + "learning_rate": 1.1214421807563726e-06, + "loss": 0.1217, + "step": 12195 + }, + { + "epoch": 2.557349549171734, + "grad_norm": 3.5413099709455973, + "learning_rate": 1.1204007455332154e-06, + "loss": 0.1075, + "step": 12196 + }, + { + "epoch": 2.5575592367372613, + "grad_norm": 5.388968882960073, + "learning_rate": 1.1193597654049204e-06, + "loss": 0.1746, + "step": 12197 + }, + { + "epoch": 2.557768924302789, + "grad_norm": 3.946131583797869, + "learning_rate": 1.1183192404248388e-06, + "loss": 0.1282, + "step": 12198 + }, + { + "epoch": 2.5579786118683163, + "grad_norm": 5.113424632449606, + "learning_rate": 1.1172791706463038e-06, + "loss": 0.1532, + "step": 12199 + }, + { + "epoch": 2.5581882994338434, + "grad_norm": 4.214236422415214, + "learning_rate": 1.1162395561226168e-06, + "loss": 0.1298, + "step": 12200 + }, + { + "epoch": 2.558397986999371, + "grad_norm": 4.206197625949368, + "learning_rate": 1.115200396907059e-06, + "loss": 0.1476, + "step": 12201 + }, + { + "epoch": 2.5586076745648985, + "grad_norm": 4.713390017189801, + "learning_rate": 1.114161693052893e-06, + "loss": 0.1489, + "step": 12202 + }, + { + "epoch": 2.5588173621304255, + "grad_norm": 4.192222933383105, + "learning_rate": 1.1131234446133498e-06, + "loss": 0.1418, + "step": 12203 + }, + { + "epoch": 2.559027049695953, + "grad_norm": 5.337746146445536, + "learning_rate": 1.1120856516416455e-06, + "loss": 0.1692, + "step": 12204 + }, + { + "epoch": 2.5592367372614806, + "grad_norm": 4.7023551210195285, + "learning_rate": 1.1110483141909667e-06, + "loss": 0.1312, + "step": 12205 + }, + { + "epoch": 2.5594464248270077, + "grad_norm": 3.778900162503842, + "learning_rate": 1.1100114323144772e-06, + "loss": 0.113, + "step": 12206 + }, + { + "epoch": 2.559656112392535, + "grad_norm": 5.119940671582861, + "learning_rate": 1.1089750060653216e-06, + "loss": 0.1881, + "step": 12207 + }, + { + "epoch": 2.5598657999580627, + "grad_norm": 3.8437742240024204, + "learning_rate": 1.107939035496619e-06, + "loss": 0.1068, + "step": 12208 + }, + { + "epoch": 2.56007548752359, + "grad_norm": 4.278520482989656, + "learning_rate": 1.106903520661461e-06, + "loss": 0.1406, + "step": 12209 + }, + { + "epoch": 2.5602851750891173, + "grad_norm": 3.412520794337883, + "learning_rate": 1.1058684616129212e-06, + "loss": 0.1085, + "step": 12210 + }, + { + "epoch": 2.5604948626546444, + "grad_norm": 3.6774329992886505, + "learning_rate": 1.1048338584040519e-06, + "loss": 0.1357, + "step": 12211 + }, + { + "epoch": 2.560704550220172, + "grad_norm": 3.9448686457466104, + "learning_rate": 1.103799711087874e-06, + "loss": 0.1328, + "step": 12212 + }, + { + "epoch": 2.560914237785699, + "grad_norm": 5.905435880051503, + "learning_rate": 1.1027660197173894e-06, + "loss": 0.1606, + "step": 12213 + }, + { + "epoch": 2.5611239253512266, + "grad_norm": 3.0940281823139086, + "learning_rate": 1.1017327843455794e-06, + "loss": 0.095, + "step": 12214 + }, + { + "epoch": 2.561333612916754, + "grad_norm": 4.072362399838126, + "learning_rate": 1.1007000050253958e-06, + "loss": 0.0924, + "step": 12215 + }, + { + "epoch": 2.561543300482281, + "grad_norm": 4.1515451672202595, + "learning_rate": 1.099667681809774e-06, + "loss": 0.1333, + "step": 12216 + }, + { + "epoch": 2.5617529880478087, + "grad_norm": 4.208427931873919, + "learning_rate": 1.0986358147516207e-06, + "loss": 0.1257, + "step": 12217 + }, + { + "epoch": 2.5619626756133362, + "grad_norm": 5.05182675256152, + "learning_rate": 1.0976044039038181e-06, + "loss": 0.18, + "step": 12218 + }, + { + "epoch": 2.5621723631788633, + "grad_norm": 4.015620119489866, + "learning_rate": 1.0965734493192293e-06, + "loss": 0.119, + "step": 12219 + }, + { + "epoch": 2.562382050744391, + "grad_norm": 3.9258383541556316, + "learning_rate": 1.0955429510506966e-06, + "loss": 0.1142, + "step": 12220 + }, + { + "epoch": 2.5625917383099184, + "grad_norm": 4.6704836551966915, + "learning_rate": 1.0945129091510308e-06, + "loss": 0.1177, + "step": 12221 + }, + { + "epoch": 2.5628014258754455, + "grad_norm": 3.693553465065929, + "learning_rate": 1.0934833236730235e-06, + "loss": 0.1128, + "step": 12222 + }, + { + "epoch": 2.563011113440973, + "grad_norm": 4.291241429776918, + "learning_rate": 1.0924541946694444e-06, + "loss": 0.1356, + "step": 12223 + }, + { + "epoch": 2.5632208010065005, + "grad_norm": 3.559320233172159, + "learning_rate": 1.091425522193037e-06, + "loss": 0.1146, + "step": 12224 + }, + { + "epoch": 2.5634304885720276, + "grad_norm": 4.704572678966715, + "learning_rate": 1.0903973062965212e-06, + "loss": 0.1563, + "step": 12225 + }, + { + "epoch": 2.563640176137555, + "grad_norm": 4.144404446116386, + "learning_rate": 1.0893695470325994e-06, + "loss": 0.1301, + "step": 12226 + }, + { + "epoch": 2.5638498637030827, + "grad_norm": 3.935310163786143, + "learning_rate": 1.0883422444539393e-06, + "loss": 0.1316, + "step": 12227 + }, + { + "epoch": 2.5640595512686097, + "grad_norm": 4.017918351176443, + "learning_rate": 1.087315398613198e-06, + "loss": 0.1503, + "step": 12228 + }, + { + "epoch": 2.5642692388341373, + "grad_norm": 4.518043726409943, + "learning_rate": 1.086289009563002e-06, + "loss": 0.134, + "step": 12229 + }, + { + "epoch": 2.5644789263996643, + "grad_norm": 3.79619339797227, + "learning_rate": 1.0852630773559514e-06, + "loss": 0.1245, + "step": 12230 + }, + { + "epoch": 2.564688613965192, + "grad_norm": 3.7546005057328546, + "learning_rate": 1.08423760204463e-06, + "loss": 0.1236, + "step": 12231 + }, + { + "epoch": 2.564898301530719, + "grad_norm": 4.071772273456543, + "learning_rate": 1.0832125836815965e-06, + "loss": 0.1223, + "step": 12232 + }, + { + "epoch": 2.5651079890962465, + "grad_norm": 3.872014189381969, + "learning_rate": 1.0821880223193837e-06, + "loss": 0.134, + "step": 12233 + }, + { + "epoch": 2.565317676661774, + "grad_norm": 3.194526472651351, + "learning_rate": 1.0811639180105006e-06, + "loss": 0.1125, + "step": 12234 + }, + { + "epoch": 2.565527364227301, + "grad_norm": 3.702729304993627, + "learning_rate": 1.0801402708074371e-06, + "loss": 0.1153, + "step": 12235 + }, + { + "epoch": 2.5657370517928286, + "grad_norm": 5.725080956473324, + "learning_rate": 1.079117080762654e-06, + "loss": 0.197, + "step": 12236 + }, + { + "epoch": 2.565946739358356, + "grad_norm": 3.8026105029586894, + "learning_rate": 1.0780943479285944e-06, + "loss": 0.0891, + "step": 12237 + }, + { + "epoch": 2.5661564269238832, + "grad_norm": 3.6374062786869104, + "learning_rate": 1.0770720723576734e-06, + "loss": 0.1262, + "step": 12238 + }, + { + "epoch": 2.5663661144894108, + "grad_norm": 4.014396149426894, + "learning_rate": 1.0760502541022832e-06, + "loss": 0.1004, + "step": 12239 + }, + { + "epoch": 2.5665758020549383, + "grad_norm": 5.320708273599552, + "learning_rate": 1.0750288932147957e-06, + "loss": 0.1724, + "step": 12240 + }, + { + "epoch": 2.5667854896204654, + "grad_norm": 4.322144374254234, + "learning_rate": 1.0740079897475609e-06, + "loss": 0.1601, + "step": 12241 + }, + { + "epoch": 2.566995177185993, + "grad_norm": 4.1033927148071685, + "learning_rate": 1.0729875437528937e-06, + "loss": 0.1251, + "step": 12242 + }, + { + "epoch": 2.5672048647515204, + "grad_norm": 3.6541957358274852, + "learning_rate": 1.0719675552830977e-06, + "loss": 0.138, + "step": 12243 + }, + { + "epoch": 2.5674145523170475, + "grad_norm": 4.2540995389288065, + "learning_rate": 1.0709480243904514e-06, + "loss": 0.1332, + "step": 12244 + }, + { + "epoch": 2.567624239882575, + "grad_norm": 5.301371938737536, + "learning_rate": 1.069928951127206e-06, + "loss": 0.1603, + "step": 12245 + }, + { + "epoch": 2.5678339274481026, + "grad_norm": 4.56346530533326, + "learning_rate": 1.0689103355455887e-06, + "loss": 0.1339, + "step": 12246 + }, + { + "epoch": 2.5680436150136297, + "grad_norm": 4.690190320853709, + "learning_rate": 1.067892177697808e-06, + "loss": 0.1705, + "step": 12247 + }, + { + "epoch": 2.568253302579157, + "grad_norm": 4.791196674976329, + "learning_rate": 1.066874477636043e-06, + "loss": 0.1866, + "step": 12248 + }, + { + "epoch": 2.5684629901446843, + "grad_norm": 3.529413524804673, + "learning_rate": 1.0658572354124574e-06, + "loss": 0.1407, + "step": 12249 + }, + { + "epoch": 2.568672677710212, + "grad_norm": 4.4444206608379195, + "learning_rate": 1.0648404510791832e-06, + "loss": 0.1312, + "step": 12250 + }, + { + "epoch": 2.568882365275739, + "grad_norm": 4.10217111465742, + "learning_rate": 1.063824124688332e-06, + "loss": 0.1349, + "step": 12251 + }, + { + "epoch": 2.5690920528412664, + "grad_norm": 5.416873948388693, + "learning_rate": 1.062808256291993e-06, + "loss": 0.1667, + "step": 12252 + }, + { + "epoch": 2.569301740406794, + "grad_norm": 4.193222655873493, + "learning_rate": 1.0617928459422345e-06, + "loss": 0.1358, + "step": 12253 + }, + { + "epoch": 2.569511427972321, + "grad_norm": 3.342166632981273, + "learning_rate": 1.060777893691094e-06, + "loss": 0.1009, + "step": 12254 + }, + { + "epoch": 2.5697211155378485, + "grad_norm": 3.553759240487575, + "learning_rate": 1.0597633995905887e-06, + "loss": 0.1199, + "step": 12255 + }, + { + "epoch": 2.569930803103376, + "grad_norm": 3.375077053990747, + "learning_rate": 1.058749363692718e-06, + "loss": 0.1234, + "step": 12256 + }, + { + "epoch": 2.570140490668903, + "grad_norm": 3.6939890540088736, + "learning_rate": 1.0577357860494475e-06, + "loss": 0.1194, + "step": 12257 + }, + { + "epoch": 2.5703501782344307, + "grad_norm": 3.920686872268463, + "learning_rate": 1.0567226667127306e-06, + "loss": 0.0996, + "step": 12258 + }, + { + "epoch": 2.570559865799958, + "grad_norm": 4.526284188267804, + "learning_rate": 1.0557100057344871e-06, + "loss": 0.1646, + "step": 12259 + }, + { + "epoch": 2.5707695533654853, + "grad_norm": 4.077610365346558, + "learning_rate": 1.0546978031666177e-06, + "loss": 0.1617, + "step": 12260 + }, + { + "epoch": 2.570979240931013, + "grad_norm": 4.556265039249083, + "learning_rate": 1.0536860590610009e-06, + "loss": 0.1324, + "step": 12261 + }, + { + "epoch": 2.5711889284965403, + "grad_norm": 3.7457350186463954, + "learning_rate": 1.0526747734694908e-06, + "loss": 0.1504, + "step": 12262 + }, + { + "epoch": 2.5713986160620674, + "grad_norm": 3.8296243671396617, + "learning_rate": 1.0516639464439182e-06, + "loss": 0.1423, + "step": 12263 + }, + { + "epoch": 2.571608303627595, + "grad_norm": 4.285112907804235, + "learning_rate": 1.050653578036086e-06, + "loss": 0.1323, + "step": 12264 + }, + { + "epoch": 2.5718179911931225, + "grad_norm": 6.78428986667062, + "learning_rate": 1.0496436682977807e-06, + "loss": 0.1788, + "step": 12265 + }, + { + "epoch": 2.5720276787586496, + "grad_norm": 4.808252569237714, + "learning_rate": 1.048634217280764e-06, + "loss": 0.1827, + "step": 12266 + }, + { + "epoch": 2.572237366324177, + "grad_norm": 3.1081839596677767, + "learning_rate": 1.0476252250367647e-06, + "loss": 0.0751, + "step": 12267 + }, + { + "epoch": 2.572447053889704, + "grad_norm": 5.2551302669926745, + "learning_rate": 1.0466166916175003e-06, + "loss": 0.1568, + "step": 12268 + }, + { + "epoch": 2.5726567414552317, + "grad_norm": 4.567603992500484, + "learning_rate": 1.0456086170746615e-06, + "loss": 0.1528, + "step": 12269 + }, + { + "epoch": 2.572866429020759, + "grad_norm": 4.920386645666056, + "learning_rate": 1.0446010014599116e-06, + "loss": 0.168, + "step": 12270 + }, + { + "epoch": 2.5730761165862863, + "grad_norm": 4.14092376919901, + "learning_rate": 1.0435938448248906e-06, + "loss": 0.1486, + "step": 12271 + }, + { + "epoch": 2.573285804151814, + "grad_norm": 3.389975600827999, + "learning_rate": 1.0425871472212211e-06, + "loss": 0.1034, + "step": 12272 + }, + { + "epoch": 2.573495491717341, + "grad_norm": 3.810779613429181, + "learning_rate": 1.0415809087004936e-06, + "loss": 0.103, + "step": 12273 + }, + { + "epoch": 2.5737051792828685, + "grad_norm": 4.147369823663081, + "learning_rate": 1.0405751293142851e-06, + "loss": 0.1617, + "step": 12274 + }, + { + "epoch": 2.573914866848396, + "grad_norm": 3.5918291470590544, + "learning_rate": 1.0395698091141405e-06, + "loss": 0.1301, + "step": 12275 + }, + { + "epoch": 2.574124554413923, + "grad_norm": 5.702992779361755, + "learning_rate": 1.0385649481515814e-06, + "loss": 0.2085, + "step": 12276 + }, + { + "epoch": 2.5743342419794506, + "grad_norm": 3.4956848584711424, + "learning_rate": 1.0375605464781125e-06, + "loss": 0.1432, + "step": 12277 + }, + { + "epoch": 2.574543929544978, + "grad_norm": 4.6571068116420475, + "learning_rate": 1.0365566041452114e-06, + "loss": 0.1495, + "step": 12278 + }, + { + "epoch": 2.574753617110505, + "grad_norm": 4.936938095977498, + "learning_rate": 1.0355531212043312e-06, + "loss": 0.1333, + "step": 12279 + }, + { + "epoch": 2.5749633046760327, + "grad_norm": 4.785161220976136, + "learning_rate": 1.034550097706899e-06, + "loss": 0.1607, + "step": 12280 + }, + { + "epoch": 2.5751729922415603, + "grad_norm": 4.802038718101305, + "learning_rate": 1.0335475337043266e-06, + "loss": 0.1736, + "step": 12281 + }, + { + "epoch": 2.5753826798070873, + "grad_norm": 4.615253150500586, + "learning_rate": 1.032545429247992e-06, + "loss": 0.1694, + "step": 12282 + }, + { + "epoch": 2.575592367372615, + "grad_norm": 5.385899732711564, + "learning_rate": 1.0315437843892595e-06, + "loss": 0.2194, + "step": 12283 + }, + { + "epoch": 2.5758020549381424, + "grad_norm": 4.670532837635864, + "learning_rate": 1.0305425991794638e-06, + "loss": 0.1502, + "step": 12284 + }, + { + "epoch": 2.5760117425036695, + "grad_norm": 5.528930059006826, + "learning_rate": 1.0295418736699137e-06, + "loss": 0.1756, + "step": 12285 + }, + { + "epoch": 2.576221430069197, + "grad_norm": 3.6247727438738417, + "learning_rate": 1.0285416079119036e-06, + "loss": 0.1202, + "step": 12286 + }, + { + "epoch": 2.576431117634724, + "grad_norm": 4.327164974280052, + "learning_rate": 1.0275418019566962e-06, + "loss": 0.1117, + "step": 12287 + }, + { + "epoch": 2.5766408052002516, + "grad_norm": 5.33326951502281, + "learning_rate": 1.0265424558555314e-06, + "loss": 0.1522, + "step": 12288 + }, + { + "epoch": 2.576850492765779, + "grad_norm": 4.883478154967185, + "learning_rate": 1.0255435696596294e-06, + "loss": 0.2121, + "step": 12289 + }, + { + "epoch": 2.5770601803313062, + "grad_norm": 4.216356297517448, + "learning_rate": 1.024545143420187e-06, + "loss": 0.1219, + "step": 12290 + }, + { + "epoch": 2.5772698678968338, + "grad_norm": 4.424947280455243, + "learning_rate": 1.023547177188373e-06, + "loss": 0.149, + "step": 12291 + }, + { + "epoch": 2.577479555462361, + "grad_norm": 4.420572162774648, + "learning_rate": 1.0225496710153337e-06, + "loss": 0.1287, + "step": 12292 + }, + { + "epoch": 2.5776892430278884, + "grad_norm": 4.611145224532912, + "learning_rate": 1.0215526249521957e-06, + "loss": 0.1359, + "step": 12293 + }, + { + "epoch": 2.577898930593416, + "grad_norm": 3.6589471589406437, + "learning_rate": 1.0205560390500557e-06, + "loss": 0.1239, + "step": 12294 + }, + { + "epoch": 2.578108618158943, + "grad_norm": 3.4525076688009344, + "learning_rate": 1.0195599133599955e-06, + "loss": 0.1111, + "step": 12295 + }, + { + "epoch": 2.5783183057244705, + "grad_norm": 6.611862239505533, + "learning_rate": 1.018564247933065e-06, + "loss": 0.1724, + "step": 12296 + }, + { + "epoch": 2.578527993289998, + "grad_norm": 4.123234539355565, + "learning_rate": 1.0175690428202923e-06, + "loss": 0.152, + "step": 12297 + }, + { + "epoch": 2.578737680855525, + "grad_norm": 4.1703687536714416, + "learning_rate": 1.0165742980726857e-06, + "loss": 0.1642, + "step": 12298 + }, + { + "epoch": 2.5789473684210527, + "grad_norm": 3.9378480112927687, + "learning_rate": 1.0155800137412287e-06, + "loss": 0.1306, + "step": 12299 + }, + { + "epoch": 2.57915705598658, + "grad_norm": 4.524950133211574, + "learning_rate": 1.0145861898768794e-06, + "loss": 0.1629, + "step": 12300 + }, + { + "epoch": 2.5793667435521073, + "grad_norm": 4.2750388273436934, + "learning_rate": 1.0135928265305695e-06, + "loss": 0.1238, + "step": 12301 + }, + { + "epoch": 2.579576431117635, + "grad_norm": 5.15976927076516, + "learning_rate": 1.0125999237532159e-06, + "loss": 0.1354, + "step": 12302 + }, + { + "epoch": 2.5797861186831623, + "grad_norm": 3.6394099688880814, + "learning_rate": 1.0116074815957044e-06, + "loss": 0.1509, + "step": 12303 + }, + { + "epoch": 2.5799958062486894, + "grad_norm": 4.993492256314699, + "learning_rate": 1.010615500108897e-06, + "loss": 0.1544, + "step": 12304 + }, + { + "epoch": 2.580205493814217, + "grad_norm": 3.7232198597526147, + "learning_rate": 1.009623979343638e-06, + "loss": 0.1122, + "step": 12305 + }, + { + "epoch": 2.580415181379744, + "grad_norm": 4.037771377627534, + "learning_rate": 1.0086329193507405e-06, + "loss": 0.1027, + "step": 12306 + }, + { + "epoch": 2.5806248689452715, + "grad_norm": 6.990721620577069, + "learning_rate": 1.0076423201810036e-06, + "loss": 0.1271, + "step": 12307 + }, + { + "epoch": 2.580834556510799, + "grad_norm": 4.1410746106497935, + "learning_rate": 1.0066521818851938e-06, + "loss": 0.1276, + "step": 12308 + }, + { + "epoch": 2.581044244076326, + "grad_norm": 4.774771339537855, + "learning_rate": 1.0056625045140566e-06, + "loss": 0.1472, + "step": 12309 + }, + { + "epoch": 2.5812539316418537, + "grad_norm": 3.6652852329458665, + "learning_rate": 1.0046732881183152e-06, + "loss": 0.1173, + "step": 12310 + }, + { + "epoch": 2.5814636192073808, + "grad_norm": 3.266901874208994, + "learning_rate": 1.003684532748671e-06, + "loss": 0.139, + "step": 12311 + }, + { + "epoch": 2.5816733067729083, + "grad_norm": 3.171937295050627, + "learning_rate": 1.0026962384557993e-06, + "loss": 0.0748, + "step": 12312 + }, + { + "epoch": 2.581882994338436, + "grad_norm": 3.891274431280533, + "learning_rate": 1.001708405290347e-06, + "loss": 0.1602, + "step": 12313 + }, + { + "epoch": 2.582092681903963, + "grad_norm": 4.587337410199784, + "learning_rate": 1.0007210333029483e-06, + "loss": 0.1533, + "step": 12314 + }, + { + "epoch": 2.5823023694694904, + "grad_norm": 3.2107477587422926, + "learning_rate": 9.997341225442026e-07, + "loss": 0.0839, + "step": 12315 + }, + { + "epoch": 2.582512057035018, + "grad_norm": 3.635218908219605, + "learning_rate": 9.98747673064696e-07, + "loss": 0.1458, + "step": 12316 + }, + { + "epoch": 2.582721744600545, + "grad_norm": 3.3544491051596212, + "learning_rate": 9.97761684914983e-07, + "loss": 0.0909, + "step": 12317 + }, + { + "epoch": 2.5829314321660726, + "grad_norm": 5.018317903454361, + "learning_rate": 9.967761581455948e-07, + "loss": 0.1633, + "step": 12318 + }, + { + "epoch": 2.5831411197316, + "grad_norm": 5.542974511527306, + "learning_rate": 9.957910928070436e-07, + "loss": 0.1783, + "step": 12319 + }, + { + "epoch": 2.583350807297127, + "grad_norm": 5.166646153995886, + "learning_rate": 9.9480648894982e-07, + "loss": 0.1569, + "step": 12320 + }, + { + "epoch": 2.5835604948626547, + "grad_norm": 3.0579825928907676, + "learning_rate": 9.93822346624379e-07, + "loss": 0.1201, + "step": 12321 + }, + { + "epoch": 2.5837701824281822, + "grad_norm": 5.0069304113773585, + "learning_rate": 9.928386658811629e-07, + "loss": 0.0922, + "step": 12322 + }, + { + "epoch": 2.5839798699937093, + "grad_norm": 4.820975851444204, + "learning_rate": 9.918554467705877e-07, + "loss": 0.1669, + "step": 12323 + }, + { + "epoch": 2.584189557559237, + "grad_norm": 3.336565209904173, + "learning_rate": 9.908726893430475e-07, + "loss": 0.111, + "step": 12324 + }, + { + "epoch": 2.5843992451247644, + "grad_norm": 4.531203786433443, + "learning_rate": 9.898903936489046e-07, + "loss": 0.1486, + "step": 12325 + }, + { + "epoch": 2.5846089326902915, + "grad_norm": 4.778731013707345, + "learning_rate": 9.889085597385062e-07, + "loss": 0.1316, + "step": 12326 + }, + { + "epoch": 2.584818620255819, + "grad_norm": 3.461830449944108, + "learning_rate": 9.879271876621743e-07, + "loss": 0.1098, + "step": 12327 + }, + { + "epoch": 2.585028307821346, + "grad_norm": 4.587505915874927, + "learning_rate": 9.869462774702054e-07, + "loss": 0.1163, + "step": 12328 + }, + { + "epoch": 2.5852379953868736, + "grad_norm": 5.625875259226333, + "learning_rate": 9.859658292128704e-07, + "loss": 0.1676, + "step": 12329 + }, + { + "epoch": 2.5854476829524007, + "grad_norm": 8.017736701930426, + "learning_rate": 9.849858429404223e-07, + "loss": 0.1161, + "step": 12330 + }, + { + "epoch": 2.585657370517928, + "grad_norm": 3.573495855668602, + "learning_rate": 9.840063187030836e-07, + "loss": 0.1138, + "step": 12331 + }, + { + "epoch": 2.5858670580834557, + "grad_norm": 3.4503796062193133, + "learning_rate": 9.8302725655106e-07, + "loss": 0.0979, + "step": 12332 + }, + { + "epoch": 2.586076745648983, + "grad_norm": 4.621692222897732, + "learning_rate": 9.820486565345288e-07, + "loss": 0.1685, + "step": 12333 + }, + { + "epoch": 2.5862864332145103, + "grad_norm": 4.726241924314743, + "learning_rate": 9.81070518703643e-07, + "loss": 0.14, + "step": 12334 + }, + { + "epoch": 2.586496120780038, + "grad_norm": 4.494331973336337, + "learning_rate": 9.800928431085366e-07, + "loss": 0.142, + "step": 12335 + }, + { + "epoch": 2.586705808345565, + "grad_norm": 4.267128428153658, + "learning_rate": 9.791156297993165e-07, + "loss": 0.1631, + "step": 12336 + }, + { + "epoch": 2.5869154959110925, + "grad_norm": 3.83347913844968, + "learning_rate": 9.78138878826067e-07, + "loss": 0.1095, + "step": 12337 + }, + { + "epoch": 2.58712518347662, + "grad_norm": 4.911332265194813, + "learning_rate": 9.77162590238846e-07, + "loss": 0.182, + "step": 12338 + }, + { + "epoch": 2.587334871042147, + "grad_norm": 4.350936330948426, + "learning_rate": 9.761867640876932e-07, + "loss": 0.1395, + "step": 12339 + }, + { + "epoch": 2.5875445586076746, + "grad_norm": 3.988771498604097, + "learning_rate": 9.752114004226176e-07, + "loss": 0.1337, + "step": 12340 + }, + { + "epoch": 2.587754246173202, + "grad_norm": 5.569465467641354, + "learning_rate": 9.742364992936126e-07, + "loss": 0.1452, + "step": 12341 + }, + { + "epoch": 2.5879639337387292, + "grad_norm": 3.6471334386970233, + "learning_rate": 9.732620607506415e-07, + "loss": 0.1285, + "step": 12342 + }, + { + "epoch": 2.5881736213042568, + "grad_norm": 5.384274756106724, + "learning_rate": 9.722880848436444e-07, + "loss": 0.1293, + "step": 12343 + }, + { + "epoch": 2.5883833088697843, + "grad_norm": 4.488561480088919, + "learning_rate": 9.713145716225404e-07, + "loss": 0.1369, + "step": 12344 + }, + { + "epoch": 2.5885929964353114, + "grad_norm": 4.471419853164163, + "learning_rate": 9.70341521137227e-07, + "loss": 0.1469, + "step": 12345 + }, + { + "epoch": 2.588802684000839, + "grad_norm": 4.847982399486932, + "learning_rate": 9.693689334375688e-07, + "loss": 0.1433, + "step": 12346 + }, + { + "epoch": 2.589012371566366, + "grad_norm": 3.7620133454695317, + "learning_rate": 9.683968085734152e-07, + "loss": 0.1224, + "step": 12347 + }, + { + "epoch": 2.5892220591318935, + "grad_norm": 3.2875661580763658, + "learning_rate": 9.674251465945917e-07, + "loss": 0.1143, + "step": 12348 + }, + { + "epoch": 2.5894317466974206, + "grad_norm": 3.9115075483117043, + "learning_rate": 9.664539475508961e-07, + "loss": 0.1311, + "step": 12349 + }, + { + "epoch": 2.589641434262948, + "grad_norm": 4.343426686907019, + "learning_rate": 9.65483211492102e-07, + "loss": 0.1538, + "step": 12350 + }, + { + "epoch": 2.5898511218284757, + "grad_norm": 4.62105227687183, + "learning_rate": 9.64512938467964e-07, + "loss": 0.1285, + "step": 12351 + }, + { + "epoch": 2.5900608093940027, + "grad_norm": 3.7838127011193947, + "learning_rate": 9.635431285282071e-07, + "loss": 0.1186, + "step": 12352 + }, + { + "epoch": 2.5902704969595303, + "grad_norm": 4.176735639346055, + "learning_rate": 9.6257378172254e-07, + "loss": 0.1485, + "step": 12353 + }, + { + "epoch": 2.590480184525058, + "grad_norm": 4.227462236275158, + "learning_rate": 9.616048981006409e-07, + "loss": 0.1398, + "step": 12354 + }, + { + "epoch": 2.590689872090585, + "grad_norm": 4.2016849265512715, + "learning_rate": 9.606364777121656e-07, + "loss": 0.1512, + "step": 12355 + }, + { + "epoch": 2.5908995596561124, + "grad_norm": 3.050513212011917, + "learning_rate": 9.59668520606749e-07, + "loss": 0.1074, + "step": 12356 + }, + { + "epoch": 2.59110924722164, + "grad_norm": 5.191146736102522, + "learning_rate": 9.587010268340013e-07, + "loss": 0.1552, + "step": 12357 + }, + { + "epoch": 2.591318934787167, + "grad_norm": 4.291988724326341, + "learning_rate": 9.577339964435084e-07, + "loss": 0.132, + "step": 12358 + }, + { + "epoch": 2.5915286223526945, + "grad_norm": 3.433029391661557, + "learning_rate": 9.567674294848294e-07, + "loss": 0.11, + "step": 12359 + }, + { + "epoch": 2.591738309918222, + "grad_norm": 2.8292189033213, + "learning_rate": 9.558013260075049e-07, + "loss": 0.1062, + "step": 12360 + }, + { + "epoch": 2.591947997483749, + "grad_norm": 4.0476764291349765, + "learning_rate": 9.548356860610486e-07, + "loss": 0.1146, + "step": 12361 + }, + { + "epoch": 2.5921576850492767, + "grad_norm": 3.401370098423267, + "learning_rate": 9.53870509694952e-07, + "loss": 0.1213, + "step": 12362 + }, + { + "epoch": 2.592367372614804, + "grad_norm": 4.292439880066373, + "learning_rate": 9.529057969586819e-07, + "loss": 0.1383, + "step": 12363 + }, + { + "epoch": 2.5925770601803313, + "grad_norm": 3.969880981773326, + "learning_rate": 9.519415479016792e-07, + "loss": 0.1363, + "step": 12364 + }, + { + "epoch": 2.592786747745859, + "grad_norm": 4.634108141293236, + "learning_rate": 9.509777625733674e-07, + "loss": 0.1, + "step": 12365 + }, + { + "epoch": 2.592996435311386, + "grad_norm": 3.8278376763399, + "learning_rate": 9.500144410231393e-07, + "loss": 0.099, + "step": 12366 + }, + { + "epoch": 2.5932061228769134, + "grad_norm": 2.8536345560223273, + "learning_rate": 9.490515833003666e-07, + "loss": 0.0887, + "step": 12367 + }, + { + "epoch": 2.5934158104424405, + "grad_norm": 3.8841582914818025, + "learning_rate": 9.480891894543975e-07, + "loss": 0.1327, + "step": 12368 + }, + { + "epoch": 2.593625498007968, + "grad_norm": 4.258583417054837, + "learning_rate": 9.471272595345593e-07, + "loss": 0.1438, + "step": 12369 + }, + { + "epoch": 2.5938351855734956, + "grad_norm": 3.931479907221189, + "learning_rate": 9.461657935901502e-07, + "loss": 0.1666, + "step": 12370 + }, + { + "epoch": 2.5940448731390227, + "grad_norm": 3.829509438421232, + "learning_rate": 9.452047916704466e-07, + "loss": 0.1088, + "step": 12371 + }, + { + "epoch": 2.59425456070455, + "grad_norm": 4.69056035375459, + "learning_rate": 9.442442538247032e-07, + "loss": 0.1538, + "step": 12372 + }, + { + "epoch": 2.5944642482700777, + "grad_norm": 5.926750805684561, + "learning_rate": 9.432841801021453e-07, + "loss": 0.1447, + "step": 12373 + }, + { + "epoch": 2.594673935835605, + "grad_norm": 5.4662174324588335, + "learning_rate": 9.423245705519845e-07, + "loss": 0.1932, + "step": 12374 + }, + { + "epoch": 2.5948836234011323, + "grad_norm": 4.349322831822157, + "learning_rate": 9.413654252233995e-07, + "loss": 0.1582, + "step": 12375 + }, + { + "epoch": 2.59509331096666, + "grad_norm": 4.201727018760815, + "learning_rate": 9.404067441655451e-07, + "loss": 0.1341, + "step": 12376 + }, + { + "epoch": 2.595302998532187, + "grad_norm": 2.768546211643465, + "learning_rate": 9.394485274275578e-07, + "loss": 0.0868, + "step": 12377 + }, + { + "epoch": 2.5955126860977145, + "grad_norm": 4.541204083028868, + "learning_rate": 9.384907750585504e-07, + "loss": 0.1391, + "step": 12378 + }, + { + "epoch": 2.595722373663242, + "grad_norm": 3.9930183684231673, + "learning_rate": 9.375334871076081e-07, + "loss": 0.1312, + "step": 12379 + }, + { + "epoch": 2.595932061228769, + "grad_norm": 4.659188083864193, + "learning_rate": 9.365766636237894e-07, + "loss": 0.0996, + "step": 12380 + }, + { + "epoch": 2.5961417487942966, + "grad_norm": 4.915618001560934, + "learning_rate": 9.356203046561396e-07, + "loss": 0.1794, + "step": 12381 + }, + { + "epoch": 2.596351436359824, + "grad_norm": 3.7286374243684497, + "learning_rate": 9.346644102536695e-07, + "loss": 0.1158, + "step": 12382 + }, + { + "epoch": 2.596561123925351, + "grad_norm": 4.860765299376346, + "learning_rate": 9.337089804653699e-07, + "loss": 0.1788, + "step": 12383 + }, + { + "epoch": 2.5967708114908787, + "grad_norm": 3.875729838813462, + "learning_rate": 9.327540153402092e-07, + "loss": 0.1374, + "step": 12384 + }, + { + "epoch": 2.596980499056406, + "grad_norm": 4.869701193710267, + "learning_rate": 9.317995149271342e-07, + "loss": 0.1449, + "step": 12385 + }, + { + "epoch": 2.5971901866219333, + "grad_norm": 4.197722150672071, + "learning_rate": 9.30845479275062e-07, + "loss": 0.1553, + "step": 12386 + }, + { + "epoch": 2.5973998741874604, + "grad_norm": 5.244335472138452, + "learning_rate": 9.298919084328861e-07, + "loss": 0.1667, + "step": 12387 + }, + { + "epoch": 2.597609561752988, + "grad_norm": 5.631474119833839, + "learning_rate": 9.289388024494839e-07, + "loss": 0.172, + "step": 12388 + }, + { + "epoch": 2.5978192493185155, + "grad_norm": 4.833733407258562, + "learning_rate": 9.279861613736985e-07, + "loss": 0.1447, + "step": 12389 + }, + { + "epoch": 2.5980289368840426, + "grad_norm": 4.373315789365453, + "learning_rate": 9.270339852543598e-07, + "loss": 0.1529, + "step": 12390 + }, + { + "epoch": 2.59823862444957, + "grad_norm": 3.4214431413838233, + "learning_rate": 9.260822741402652e-07, + "loss": 0.1207, + "step": 12391 + }, + { + "epoch": 2.5984483120150976, + "grad_norm": 4.946652803292614, + "learning_rate": 9.251310280801895e-07, + "loss": 0.1349, + "step": 12392 + }, + { + "epoch": 2.5986579995806247, + "grad_norm": 4.043580326065742, + "learning_rate": 9.241802471228889e-07, + "loss": 0.1337, + "step": 12393 + }, + { + "epoch": 2.5988676871461522, + "grad_norm": 4.528488112469475, + "learning_rate": 9.232299313170922e-07, + "loss": 0.1345, + "step": 12394 + }, + { + "epoch": 2.5990773747116798, + "grad_norm": 4.327430887308288, + "learning_rate": 9.222800807115062e-07, + "loss": 0.1019, + "step": 12395 + }, + { + "epoch": 2.599287062277207, + "grad_norm": 5.17068866271806, + "learning_rate": 9.213306953548085e-07, + "loss": 0.1661, + "step": 12396 + }, + { + "epoch": 2.5994967498427344, + "grad_norm": 3.5983958628270947, + "learning_rate": 9.203817752956601e-07, + "loss": 0.1242, + "step": 12397 + }, + { + "epoch": 2.599706437408262, + "grad_norm": 5.728291676075231, + "learning_rate": 9.194333205826911e-07, + "loss": 0.1451, + "step": 12398 + }, + { + "epoch": 2.599916124973789, + "grad_norm": 3.245235570748017, + "learning_rate": 9.18485331264517e-07, + "loss": 0.1171, + "step": 12399 + }, + { + "epoch": 2.6001258125393165, + "grad_norm": 4.471708132896414, + "learning_rate": 9.175378073897201e-07, + "loss": 0.1254, + "step": 12400 + }, + { + "epoch": 2.600335500104844, + "grad_norm": 3.108410809324801, + "learning_rate": 9.165907490068615e-07, + "loss": 0.1036, + "step": 12401 + }, + { + "epoch": 2.600545187670371, + "grad_norm": 3.8026375521905784, + "learning_rate": 9.156441561644814e-07, + "loss": 0.1171, + "step": 12402 + }, + { + "epoch": 2.6007548752358987, + "grad_norm": 5.045848602372193, + "learning_rate": 9.146980289110974e-07, + "loss": 0.1558, + "step": 12403 + }, + { + "epoch": 2.6009645628014257, + "grad_norm": 3.949997308953114, + "learning_rate": 9.137523672951942e-07, + "loss": 0.1273, + "step": 12404 + }, + { + "epoch": 2.6011742503669533, + "grad_norm": 3.7647474361271924, + "learning_rate": 9.128071713652409e-07, + "loss": 0.1194, + "step": 12405 + }, + { + "epoch": 2.6013839379324804, + "grad_norm": 3.7728221583604156, + "learning_rate": 9.118624411696831e-07, + "loss": 0.0999, + "step": 12406 + }, + { + "epoch": 2.601593625498008, + "grad_norm": 4.050426188940153, + "learning_rate": 9.109181767569374e-07, + "loss": 0.1195, + "step": 12407 + }, + { + "epoch": 2.6018033130635354, + "grad_norm": 3.560871521697155, + "learning_rate": 9.099743781753989e-07, + "loss": 0.1239, + "step": 12408 + }, + { + "epoch": 2.6020130006290625, + "grad_norm": 4.5823091385496095, + "learning_rate": 9.090310454734397e-07, + "loss": 0.1534, + "step": 12409 + }, + { + "epoch": 2.60222268819459, + "grad_norm": 4.783878513355387, + "learning_rate": 9.080881786994056e-07, + "loss": 0.146, + "step": 12410 + }, + { + "epoch": 2.6024323757601175, + "grad_norm": 3.619155984823816, + "learning_rate": 9.071457779016235e-07, + "loss": 0.1148, + "step": 12411 + }, + { + "epoch": 2.6026420633256446, + "grad_norm": 4.638483909010977, + "learning_rate": 9.062038431283904e-07, + "loss": 0.1649, + "step": 12412 + }, + { + "epoch": 2.602851750891172, + "grad_norm": 3.8700880135710323, + "learning_rate": 9.052623744279821e-07, + "loss": 0.1296, + "step": 12413 + }, + { + "epoch": 2.6030614384566997, + "grad_norm": 5.579768659617796, + "learning_rate": 9.043213718486499e-07, + "loss": 0.1646, + "step": 12414 + }, + { + "epoch": 2.6032711260222268, + "grad_norm": 4.695633423052056, + "learning_rate": 9.033808354386254e-07, + "loss": 0.1278, + "step": 12415 + }, + { + "epoch": 2.6034808135877543, + "grad_norm": 3.5694107387026333, + "learning_rate": 9.024407652461109e-07, + "loss": 0.1067, + "step": 12416 + }, + { + "epoch": 2.603690501153282, + "grad_norm": 4.241945709666966, + "learning_rate": 9.015011613192837e-07, + "loss": 0.1389, + "step": 12417 + }, + { + "epoch": 2.603900188718809, + "grad_norm": 5.160062312585765, + "learning_rate": 9.005620237063039e-07, + "loss": 0.1396, + "step": 12418 + }, + { + "epoch": 2.6041098762843364, + "grad_norm": 4.413744077337538, + "learning_rate": 8.996233524553011e-07, + "loss": 0.1828, + "step": 12419 + }, + { + "epoch": 2.604319563849864, + "grad_norm": 5.503388522295018, + "learning_rate": 8.986851476143876e-07, + "loss": 0.1681, + "step": 12420 + }, + { + "epoch": 2.604529251415391, + "grad_norm": 4.384169036915855, + "learning_rate": 8.977474092316451e-07, + "loss": 0.1366, + "step": 12421 + }, + { + "epoch": 2.6047389389809186, + "grad_norm": 4.654877196466101, + "learning_rate": 8.96810137355133e-07, + "loss": 0.1479, + "step": 12422 + }, + { + "epoch": 2.6049486265464457, + "grad_norm": 4.512400556318743, + "learning_rate": 8.958733320328905e-07, + "loss": 0.1757, + "step": 12423 + }, + { + "epoch": 2.605158314111973, + "grad_norm": 3.2754901763946016, + "learning_rate": 8.949369933129337e-07, + "loss": 0.103, + "step": 12424 + }, + { + "epoch": 2.6053680016775003, + "grad_norm": 4.38492029252869, + "learning_rate": 8.940011212432442e-07, + "loss": 0.1485, + "step": 12425 + }, + { + "epoch": 2.605577689243028, + "grad_norm": 4.649245406715256, + "learning_rate": 8.930657158717904e-07, + "loss": 0.1422, + "step": 12426 + }, + { + "epoch": 2.6057873768085553, + "grad_norm": 4.18344797086107, + "learning_rate": 8.92130777246516e-07, + "loss": 0.1659, + "step": 12427 + }, + { + "epoch": 2.6059970643740824, + "grad_norm": 3.346866309431157, + "learning_rate": 8.911963054153361e-07, + "loss": 0.1166, + "step": 12428 + }, + { + "epoch": 2.60620675193961, + "grad_norm": 4.2906166017233796, + "learning_rate": 8.902623004261424e-07, + "loss": 0.1462, + "step": 12429 + }, + { + "epoch": 2.6064164395051375, + "grad_norm": 5.793009929066904, + "learning_rate": 8.893287623268077e-07, + "loss": 0.2097, + "step": 12430 + }, + { + "epoch": 2.6066261270706645, + "grad_norm": 3.5577772408429493, + "learning_rate": 8.883956911651736e-07, + "loss": 0.1532, + "step": 12431 + }, + { + "epoch": 2.606835814636192, + "grad_norm": 4.462176542416087, + "learning_rate": 8.874630869890644e-07, + "loss": 0.1426, + "step": 12432 + }, + { + "epoch": 2.6070455022017196, + "grad_norm": 6.613579030776804, + "learning_rate": 8.865309498462782e-07, + "loss": 0.2295, + "step": 12433 + }, + { + "epoch": 2.6072551897672467, + "grad_norm": 4.006961337434616, + "learning_rate": 8.855992797845847e-07, + "loss": 0.1384, + "step": 12434 + }, + { + "epoch": 2.607464877332774, + "grad_norm": 4.765849714961155, + "learning_rate": 8.846680768517368e-07, + "loss": 0.1457, + "step": 12435 + }, + { + "epoch": 2.6076745648983017, + "grad_norm": 4.105845866749517, + "learning_rate": 8.837373410954619e-07, + "loss": 0.1446, + "step": 12436 + }, + { + "epoch": 2.607884252463829, + "grad_norm": 3.455321983963246, + "learning_rate": 8.828070725634586e-07, + "loss": 0.117, + "step": 12437 + }, + { + "epoch": 2.6080939400293564, + "grad_norm": 5.790581351749775, + "learning_rate": 8.81877271303404e-07, + "loss": 0.1644, + "step": 12438 + }, + { + "epoch": 2.608303627594884, + "grad_norm": 3.9772472301459705, + "learning_rate": 8.80947937362957e-07, + "loss": 0.1214, + "step": 12439 + }, + { + "epoch": 2.608513315160411, + "grad_norm": 4.219530312022978, + "learning_rate": 8.800190707897416e-07, + "loss": 0.1482, + "step": 12440 + }, + { + "epoch": 2.6087230027259385, + "grad_norm": 5.5839233088214755, + "learning_rate": 8.790906716313685e-07, + "loss": 0.1486, + "step": 12441 + }, + { + "epoch": 2.6089326902914656, + "grad_norm": 3.9812559393177094, + "learning_rate": 8.781627399354176e-07, + "loss": 0.137, + "step": 12442 + }, + { + "epoch": 2.609142377856993, + "grad_norm": 4.326362108177064, + "learning_rate": 8.772352757494462e-07, + "loss": 0.1336, + "step": 12443 + }, + { + "epoch": 2.60935206542252, + "grad_norm": 4.791230319862026, + "learning_rate": 8.76308279120991e-07, + "loss": 0.1724, + "step": 12444 + }, + { + "epoch": 2.6095617529880477, + "grad_norm": 3.0948779613655124, + "learning_rate": 8.753817500975581e-07, + "loss": 0.1006, + "step": 12445 + }, + { + "epoch": 2.6097714405535752, + "grad_norm": 3.581805385345101, + "learning_rate": 8.744556887266387e-07, + "loss": 0.1366, + "step": 12446 + }, + { + "epoch": 2.6099811281191023, + "grad_norm": 4.6527008048543825, + "learning_rate": 8.735300950556913e-07, + "loss": 0.1407, + "step": 12447 + }, + { + "epoch": 2.61019081568463, + "grad_norm": 3.829342947287962, + "learning_rate": 8.72604969132157e-07, + "loss": 0.1378, + "step": 12448 + }, + { + "epoch": 2.6104005032501574, + "grad_norm": 4.096005873798838, + "learning_rate": 8.7168031100345e-07, + "loss": 0.1303, + "step": 12449 + }, + { + "epoch": 2.6106101908156845, + "grad_norm": 4.516398020101538, + "learning_rate": 8.707561207169557e-07, + "loss": 0.1563, + "step": 12450 + }, + { + "epoch": 2.610819878381212, + "grad_norm": 4.939284191470693, + "learning_rate": 8.698323983200452e-07, + "loss": 0.1592, + "step": 12451 + }, + { + "epoch": 2.6110295659467395, + "grad_norm": 5.964570621914853, + "learning_rate": 8.689091438600605e-07, + "loss": 0.198, + "step": 12452 + }, + { + "epoch": 2.6112392535122666, + "grad_norm": 4.124011358812057, + "learning_rate": 8.679863573843206e-07, + "loss": 0.1313, + "step": 12453 + }, + { + "epoch": 2.611448941077794, + "grad_norm": 4.582397874529155, + "learning_rate": 8.670640389401164e-07, + "loss": 0.1341, + "step": 12454 + }, + { + "epoch": 2.6116586286433217, + "grad_norm": 4.775223860432403, + "learning_rate": 8.661421885747212e-07, + "loss": 0.1781, + "step": 12455 + }, + { + "epoch": 2.6118683162088487, + "grad_norm": 5.23509432187901, + "learning_rate": 8.652208063353795e-07, + "loss": 0.1237, + "step": 12456 + }, + { + "epoch": 2.6120780037743763, + "grad_norm": 5.143009841000892, + "learning_rate": 8.64299892269318e-07, + "loss": 0.1729, + "step": 12457 + }, + { + "epoch": 2.612287691339904, + "grad_norm": 4.09327804630886, + "learning_rate": 8.6337944642373e-07, + "loss": 0.1272, + "step": 12458 + }, + { + "epoch": 2.612497378905431, + "grad_norm": 3.9598890836619622, + "learning_rate": 8.624594688457921e-07, + "loss": 0.1353, + "step": 12459 + }, + { + "epoch": 2.6127070664709584, + "grad_norm": 4.51636056535437, + "learning_rate": 8.615399595826535e-07, + "loss": 0.1312, + "step": 12460 + }, + { + "epoch": 2.6129167540364855, + "grad_norm": 3.5740134690305867, + "learning_rate": 8.606209186814463e-07, + "loss": 0.0903, + "step": 12461 + }, + { + "epoch": 2.613126441602013, + "grad_norm": 3.084324205276457, + "learning_rate": 8.597023461892639e-07, + "loss": 0.0935, + "step": 12462 + }, + { + "epoch": 2.61333612916754, + "grad_norm": 4.359064753941047, + "learning_rate": 8.5878424215319e-07, + "loss": 0.1419, + "step": 12463 + }, + { + "epoch": 2.6135458167330676, + "grad_norm": 4.774579910488419, + "learning_rate": 8.57866606620279e-07, + "loss": 0.1353, + "step": 12464 + }, + { + "epoch": 2.613755504298595, + "grad_norm": 3.750609031161766, + "learning_rate": 8.569494396375622e-07, + "loss": 0.1125, + "step": 12465 + }, + { + "epoch": 2.6139651918641222, + "grad_norm": 3.2346064639082424, + "learning_rate": 8.56032741252042e-07, + "loss": 0.0885, + "step": 12466 + }, + { + "epoch": 2.6141748794296498, + "grad_norm": 4.276055853770689, + "learning_rate": 8.551165115107052e-07, + "loss": 0.1365, + "step": 12467 + }, + { + "epoch": 2.6143845669951773, + "grad_norm": 3.8059509768508955, + "learning_rate": 8.542007504605054e-07, + "loss": 0.1393, + "step": 12468 + }, + { + "epoch": 2.6145942545607044, + "grad_norm": 3.6922418189855772, + "learning_rate": 8.532854581483829e-07, + "loss": 0.1143, + "step": 12469 + }, + { + "epoch": 2.614803942126232, + "grad_norm": 4.2704346072824615, + "learning_rate": 8.523706346212435e-07, + "loss": 0.1346, + "step": 12470 + }, + { + "epoch": 2.6150136296917594, + "grad_norm": 3.8199798360161026, + "learning_rate": 8.51456279925974e-07, + "loss": 0.1222, + "step": 12471 + }, + { + "epoch": 2.6152233172572865, + "grad_norm": 8.225860711129755, + "learning_rate": 8.505423941094371e-07, + "loss": 0.1786, + "step": 12472 + }, + { + "epoch": 2.615433004822814, + "grad_norm": 3.803775858172037, + "learning_rate": 8.49628977218474e-07, + "loss": 0.1052, + "step": 12473 + }, + { + "epoch": 2.6156426923883416, + "grad_norm": 4.596830418451265, + "learning_rate": 8.487160292998952e-07, + "loss": 0.1323, + "step": 12474 + }, + { + "epoch": 2.6158523799538687, + "grad_norm": 5.650414174904719, + "learning_rate": 8.478035504004911e-07, + "loss": 0.2002, + "step": 12475 + }, + { + "epoch": 2.616062067519396, + "grad_norm": 4.73651913267565, + "learning_rate": 8.468915405670297e-07, + "loss": 0.1895, + "step": 12476 + }, + { + "epoch": 2.6162717550849237, + "grad_norm": 5.242246273403459, + "learning_rate": 8.459799998462503e-07, + "loss": 0.1698, + "step": 12477 + }, + { + "epoch": 2.616481442650451, + "grad_norm": 4.070876270443945, + "learning_rate": 8.450689282848756e-07, + "loss": 0.1311, + "step": 12478 + }, + { + "epoch": 2.6166911302159783, + "grad_norm": 4.769807613842107, + "learning_rate": 8.44158325929596e-07, + "loss": 0.1574, + "step": 12479 + }, + { + "epoch": 2.6169008177815054, + "grad_norm": 3.9581053882038018, + "learning_rate": 8.432481928270808e-07, + "loss": 0.1196, + "step": 12480 + }, + { + "epoch": 2.617110505347033, + "grad_norm": 4.475019044180424, + "learning_rate": 8.423385290239771e-07, + "loss": 0.1407, + "step": 12481 + }, + { + "epoch": 2.61732019291256, + "grad_norm": 5.1372302667320575, + "learning_rate": 8.4142933456691e-07, + "loss": 0.1281, + "step": 12482 + }, + { + "epoch": 2.6175298804780875, + "grad_norm": 2.9060303425745007, + "learning_rate": 8.405206095024709e-07, + "loss": 0.0918, + "step": 12483 + }, + { + "epoch": 2.617739568043615, + "grad_norm": 4.251609892652858, + "learning_rate": 8.396123538772372e-07, + "loss": 0.1608, + "step": 12484 + }, + { + "epoch": 2.617949255609142, + "grad_norm": 4.84092217200265, + "learning_rate": 8.387045677377604e-07, + "loss": 0.1572, + "step": 12485 + }, + { + "epoch": 2.6181589431746697, + "grad_norm": 4.940211935234428, + "learning_rate": 8.377972511305632e-07, + "loss": 0.1551, + "step": 12486 + }, + { + "epoch": 2.618368630740197, + "grad_norm": 4.760245197568732, + "learning_rate": 8.368904041021464e-07, + "loss": 0.1664, + "step": 12487 + }, + { + "epoch": 2.6185783183057243, + "grad_norm": 4.781642444834001, + "learning_rate": 8.359840266989905e-07, + "loss": 0.1219, + "step": 12488 + }, + { + "epoch": 2.618788005871252, + "grad_norm": 3.9956764061989447, + "learning_rate": 8.350781189675461e-07, + "loss": 0.1309, + "step": 12489 + }, + { + "epoch": 2.6189976934367794, + "grad_norm": 4.060513848445209, + "learning_rate": 8.34172680954245e-07, + "loss": 0.1454, + "step": 12490 + }, + { + "epoch": 2.6192073810023064, + "grad_norm": 4.118826553694415, + "learning_rate": 8.332677127054923e-07, + "loss": 0.1438, + "step": 12491 + }, + { + "epoch": 2.619417068567834, + "grad_norm": 3.677533333769006, + "learning_rate": 8.323632142676663e-07, + "loss": 0.1232, + "step": 12492 + }, + { + "epoch": 2.6196267561333615, + "grad_norm": 3.850450374530168, + "learning_rate": 8.314591856871257e-07, + "loss": 0.1265, + "step": 12493 + }, + { + "epoch": 2.6198364436988886, + "grad_norm": 4.431314143124674, + "learning_rate": 8.305556270102055e-07, + "loss": 0.1307, + "step": 12494 + }, + { + "epoch": 2.620046131264416, + "grad_norm": 4.108188412282608, + "learning_rate": 8.296525382832121e-07, + "loss": 0.1716, + "step": 12495 + }, + { + "epoch": 2.6202558188299436, + "grad_norm": 5.462928380927774, + "learning_rate": 8.287499195524307e-07, + "loss": 0.1609, + "step": 12496 + }, + { + "epoch": 2.6204655063954707, + "grad_norm": 3.5094391061153773, + "learning_rate": 8.278477708641242e-07, + "loss": 0.1134, + "step": 12497 + }, + { + "epoch": 2.6206751939609982, + "grad_norm": 5.346461282040452, + "learning_rate": 8.269460922645267e-07, + "loss": 0.1701, + "step": 12498 + }, + { + "epoch": 2.6208848815265253, + "grad_norm": 4.146829504603081, + "learning_rate": 8.260448837998525e-07, + "loss": 0.1485, + "step": 12499 + }, + { + "epoch": 2.621094569092053, + "grad_norm": 4.262607551820511, + "learning_rate": 8.251441455162901e-07, + "loss": 0.1106, + "step": 12500 + }, + { + "epoch": 2.62130425665758, + "grad_norm": 3.2909705199283388, + "learning_rate": 8.242438774600003e-07, + "loss": 0.131, + "step": 12501 + }, + { + "epoch": 2.6215139442231075, + "grad_norm": 4.875331850173637, + "learning_rate": 8.233440796771275e-07, + "loss": 0.1744, + "step": 12502 + }, + { + "epoch": 2.621723631788635, + "grad_norm": 3.632698854596682, + "learning_rate": 8.224447522137901e-07, + "loss": 0.1221, + "step": 12503 + }, + { + "epoch": 2.621933319354162, + "grad_norm": 3.7479758677085226, + "learning_rate": 8.215458951160727e-07, + "loss": 0.1246, + "step": 12504 + }, + { + "epoch": 2.6221430069196896, + "grad_norm": 4.204097612004744, + "learning_rate": 8.20647508430048e-07, + "loss": 0.1624, + "step": 12505 + }, + { + "epoch": 2.622352694485217, + "grad_norm": 3.1665800207515074, + "learning_rate": 8.197495922017595e-07, + "loss": 0.1014, + "step": 12506 + }, + { + "epoch": 2.622562382050744, + "grad_norm": 4.042634420346649, + "learning_rate": 8.188521464772292e-07, + "loss": 0.1383, + "step": 12507 + }, + { + "epoch": 2.6227720696162717, + "grad_norm": 3.500808733611115, + "learning_rate": 8.179551713024481e-07, + "loss": 0.1146, + "step": 12508 + }, + { + "epoch": 2.6229817571817993, + "grad_norm": 4.929193474972663, + "learning_rate": 8.170586667233893e-07, + "loss": 0.1407, + "step": 12509 + }, + { + "epoch": 2.6231914447473264, + "grad_norm": 4.319142756884631, + "learning_rate": 8.161626327860017e-07, + "loss": 0.1345, + "step": 12510 + }, + { + "epoch": 2.623401132312854, + "grad_norm": 4.650440397641079, + "learning_rate": 8.152670695362086e-07, + "loss": 0.1279, + "step": 12511 + }, + { + "epoch": 2.6236108198783814, + "grad_norm": 4.6412856363463195, + "learning_rate": 8.143719770199065e-07, + "loss": 0.2197, + "step": 12512 + }, + { + "epoch": 2.6238205074439085, + "grad_norm": 4.545146468358653, + "learning_rate": 8.134773552829745e-07, + "loss": 0.1401, + "step": 12513 + }, + { + "epoch": 2.624030195009436, + "grad_norm": 5.120914083900178, + "learning_rate": 8.125832043712589e-07, + "loss": 0.1554, + "step": 12514 + }, + { + "epoch": 2.6242398825749635, + "grad_norm": 4.378461140513062, + "learning_rate": 8.116895243305911e-07, + "loss": 0.1472, + "step": 12515 + }, + { + "epoch": 2.6244495701404906, + "grad_norm": 5.6040548756185995, + "learning_rate": 8.10796315206771e-07, + "loss": 0.1384, + "step": 12516 + }, + { + "epoch": 2.624659257706018, + "grad_norm": 5.939318288629242, + "learning_rate": 8.099035770455755e-07, + "loss": 0.1552, + "step": 12517 + }, + { + "epoch": 2.6248689452715452, + "grad_norm": 4.564307909350276, + "learning_rate": 8.090113098927621e-07, + "loss": 0.1674, + "step": 12518 + }, + { + "epoch": 2.6250786328370728, + "grad_norm": 4.137203442560623, + "learning_rate": 8.081195137940612e-07, + "loss": 0.1049, + "step": 12519 + }, + { + "epoch": 2.6252883204026, + "grad_norm": 6.779648545852885, + "learning_rate": 8.072281887951772e-07, + "loss": 0.175, + "step": 12520 + }, + { + "epoch": 2.6254980079681274, + "grad_norm": 4.414722299635095, + "learning_rate": 8.063373349417913e-07, + "loss": 0.1508, + "step": 12521 + }, + { + "epoch": 2.625707695533655, + "grad_norm": 3.3749954709407874, + "learning_rate": 8.054469522795649e-07, + "loss": 0.1081, + "step": 12522 + }, + { + "epoch": 2.625917383099182, + "grad_norm": 3.617552737021408, + "learning_rate": 8.045570408541281e-07, + "loss": 0.1029, + "step": 12523 + }, + { + "epoch": 2.6261270706647095, + "grad_norm": 4.251296366565555, + "learning_rate": 8.036676007110911e-07, + "loss": 0.147, + "step": 12524 + }, + { + "epoch": 2.626336758230237, + "grad_norm": 4.465259091723742, + "learning_rate": 8.027786318960406e-07, + "loss": 0.1379, + "step": 12525 + }, + { + "epoch": 2.626546445795764, + "grad_norm": 4.769412162924134, + "learning_rate": 8.01890134454536e-07, + "loss": 0.1581, + "step": 12526 + }, + { + "epoch": 2.6267561333612917, + "grad_norm": 2.9418264988990708, + "learning_rate": 8.010021084321173e-07, + "loss": 0.111, + "step": 12527 + }, + { + "epoch": 2.626965820926819, + "grad_norm": 6.149070947200199, + "learning_rate": 8.00114553874295e-07, + "loss": 0.203, + "step": 12528 + }, + { + "epoch": 2.6271755084923463, + "grad_norm": 4.342435726620971, + "learning_rate": 7.992274708265557e-07, + "loss": 0.1373, + "step": 12529 + }, + { + "epoch": 2.627385196057874, + "grad_norm": 4.134785012878603, + "learning_rate": 7.983408593343667e-07, + "loss": 0.1419, + "step": 12530 + }, + { + "epoch": 2.6275948836234013, + "grad_norm": 3.8979868974839325, + "learning_rate": 7.974547194431703e-07, + "loss": 0.1328, + "step": 12531 + }, + { + "epoch": 2.6278045711889284, + "grad_norm": 4.020821633691476, + "learning_rate": 7.965690511983804e-07, + "loss": 0.1142, + "step": 12532 + }, + { + "epoch": 2.628014258754456, + "grad_norm": 4.261088972036191, + "learning_rate": 7.956838546453882e-07, + "loss": 0.1318, + "step": 12533 + }, + { + "epoch": 2.6282239463199835, + "grad_norm": 3.8096513757879467, + "learning_rate": 7.947991298295632e-07, + "loss": 0.1003, + "step": 12534 + }, + { + "epoch": 2.6284336338855105, + "grad_norm": 3.594314400652386, + "learning_rate": 7.939148767962468e-07, + "loss": 0.0999, + "step": 12535 + }, + { + "epoch": 2.628643321451038, + "grad_norm": 2.895019326971635, + "learning_rate": 7.930310955907627e-07, + "loss": 0.1011, + "step": 12536 + }, + { + "epoch": 2.628853009016565, + "grad_norm": 4.74189094495242, + "learning_rate": 7.921477862584026e-07, + "loss": 0.1578, + "step": 12537 + }, + { + "epoch": 2.6290626965820927, + "grad_norm": 3.602380171658708, + "learning_rate": 7.912649488444379e-07, + "loss": 0.1544, + "step": 12538 + }, + { + "epoch": 2.62927238414762, + "grad_norm": 3.099961957096034, + "learning_rate": 7.903825833941159e-07, + "loss": 0.1107, + "step": 12539 + }, + { + "epoch": 2.6294820717131473, + "grad_norm": 4.46631195366002, + "learning_rate": 7.895006899526625e-07, + "loss": 0.1621, + "step": 12540 + }, + { + "epoch": 2.629691759278675, + "grad_norm": 3.9501169651971084, + "learning_rate": 7.886192685652705e-07, + "loss": 0.1244, + "step": 12541 + }, + { + "epoch": 2.629901446844202, + "grad_norm": 4.409815621166538, + "learning_rate": 7.877383192771182e-07, + "loss": 0.1206, + "step": 12542 + }, + { + "epoch": 2.6301111344097294, + "grad_norm": 4.317304772394556, + "learning_rate": 7.868578421333551e-07, + "loss": 0.1413, + "step": 12543 + }, + { + "epoch": 2.630320821975257, + "grad_norm": 5.035071277937222, + "learning_rate": 7.859778371791082e-07, + "loss": 0.165, + "step": 12544 + }, + { + "epoch": 2.630530509540784, + "grad_norm": 6.519378639038373, + "learning_rate": 7.850983044594762e-07, + "loss": 0.1659, + "step": 12545 + }, + { + "epoch": 2.6307401971063116, + "grad_norm": 3.97971816819629, + "learning_rate": 7.842192440195395e-07, + "loss": 0.1222, + "step": 12546 + }, + { + "epoch": 2.630949884671839, + "grad_norm": 5.123931401558346, + "learning_rate": 7.833406559043499e-07, + "loss": 0.1702, + "step": 12547 + }, + { + "epoch": 2.631159572237366, + "grad_norm": 3.6534674317095397, + "learning_rate": 7.824625401589381e-07, + "loss": 0.12, + "step": 12548 + }, + { + "epoch": 2.6313692598028937, + "grad_norm": 3.481611349926908, + "learning_rate": 7.815848968283079e-07, + "loss": 0.124, + "step": 12549 + }, + { + "epoch": 2.6315789473684212, + "grad_norm": 3.885767296328627, + "learning_rate": 7.807077259574392e-07, + "loss": 0.1168, + "step": 12550 + }, + { + "epoch": 2.6317886349339483, + "grad_norm": 4.879011229400413, + "learning_rate": 7.798310275912891e-07, + "loss": 0.1953, + "step": 12551 + }, + { + "epoch": 2.631998322499476, + "grad_norm": 4.134315971715776, + "learning_rate": 7.789548017747928e-07, + "loss": 0.1244, + "step": 12552 + }, + { + "epoch": 2.6322080100650034, + "grad_norm": 3.8777831491062313, + "learning_rate": 7.780790485528556e-07, + "loss": 0.1286, + "step": 12553 + }, + { + "epoch": 2.6324176976305305, + "grad_norm": 4.235887096372328, + "learning_rate": 7.772037679703592e-07, + "loss": 0.1521, + "step": 12554 + }, + { + "epoch": 2.632627385196058, + "grad_norm": 4.817236419914835, + "learning_rate": 7.763289600721691e-07, + "loss": 0.182, + "step": 12555 + }, + { + "epoch": 2.632837072761585, + "grad_norm": 6.711570814223188, + "learning_rate": 7.754546249031148e-07, + "loss": 0.2651, + "step": 12556 + }, + { + "epoch": 2.6330467603271126, + "grad_norm": 3.966812115564374, + "learning_rate": 7.745807625080116e-07, + "loss": 0.1414, + "step": 12557 + }, + { + "epoch": 2.63325644789264, + "grad_norm": 3.8359922899948224, + "learning_rate": 7.737073729316458e-07, + "loss": 0.1635, + "step": 12558 + }, + { + "epoch": 2.633466135458167, + "grad_norm": 3.5525722346077595, + "learning_rate": 7.728344562187773e-07, + "loss": 0.1022, + "step": 12559 + }, + { + "epoch": 2.6336758230236947, + "grad_norm": 4.019947955577282, + "learning_rate": 7.719620124141469e-07, + "loss": 0.1483, + "step": 12560 + }, + { + "epoch": 2.633885510589222, + "grad_norm": 4.687015854350539, + "learning_rate": 7.710900415624712e-07, + "loss": 0.147, + "step": 12561 + }, + { + "epoch": 2.6340951981547494, + "grad_norm": 4.141118190628882, + "learning_rate": 7.702185437084342e-07, + "loss": 0.1352, + "step": 12562 + }, + { + "epoch": 2.634304885720277, + "grad_norm": 4.014687286259511, + "learning_rate": 7.693475188967059e-07, + "loss": 0.152, + "step": 12563 + }, + { + "epoch": 2.634514573285804, + "grad_norm": 4.808634429731679, + "learning_rate": 7.684769671719283e-07, + "loss": 0.1679, + "step": 12564 + }, + { + "epoch": 2.6347242608513315, + "grad_norm": 4.707119438128783, + "learning_rate": 7.67606888578718e-07, + "loss": 0.1444, + "step": 12565 + }, + { + "epoch": 2.634933948416859, + "grad_norm": 3.6564576575132626, + "learning_rate": 7.667372831616649e-07, + "loss": 0.1595, + "step": 12566 + }, + { + "epoch": 2.635143635982386, + "grad_norm": 4.722708372273352, + "learning_rate": 7.658681509653409e-07, + "loss": 0.144, + "step": 12567 + }, + { + "epoch": 2.6353533235479136, + "grad_norm": 4.268815605778567, + "learning_rate": 7.649994920342907e-07, + "loss": 0.139, + "step": 12568 + }, + { + "epoch": 2.635563011113441, + "grad_norm": 3.337035122797393, + "learning_rate": 7.641313064130352e-07, + "loss": 0.1076, + "step": 12569 + }, + { + "epoch": 2.6357726986789682, + "grad_norm": 4.691405160540121, + "learning_rate": 7.632635941460664e-07, + "loss": 0.1419, + "step": 12570 + }, + { + "epoch": 2.6359823862444958, + "grad_norm": 4.217483711603581, + "learning_rate": 7.623963552778602e-07, + "loss": 0.1073, + "step": 12571 + }, + { + "epoch": 2.6361920738100233, + "grad_norm": 4.614364095806026, + "learning_rate": 7.615295898528618e-07, + "loss": 0.1682, + "step": 12572 + }, + { + "epoch": 2.6364017613755504, + "grad_norm": 4.913581886994403, + "learning_rate": 7.60663297915496e-07, + "loss": 0.1317, + "step": 12573 + }, + { + "epoch": 2.636611448941078, + "grad_norm": 5.3735997710444225, + "learning_rate": 7.597974795101615e-07, + "loss": 0.1002, + "step": 12574 + }, + { + "epoch": 2.6368211365066054, + "grad_norm": 4.628727167423716, + "learning_rate": 7.589321346812306e-07, + "loss": 0.1617, + "step": 12575 + }, + { + "epoch": 2.6370308240721325, + "grad_norm": 5.420235085587457, + "learning_rate": 7.580672634730557e-07, + "loss": 0.1453, + "step": 12576 + }, + { + "epoch": 2.63724051163766, + "grad_norm": 5.408780639193367, + "learning_rate": 7.572028659299646e-07, + "loss": 0.1667, + "step": 12577 + }, + { + "epoch": 2.637450199203187, + "grad_norm": 3.68061344580538, + "learning_rate": 7.563389420962586e-07, + "loss": 0.1101, + "step": 12578 + }, + { + "epoch": 2.6376598867687147, + "grad_norm": 5.467383371983453, + "learning_rate": 7.554754920162111e-07, + "loss": 0.1564, + "step": 12579 + }, + { + "epoch": 2.6378695743342417, + "grad_norm": 4.673193172097822, + "learning_rate": 7.5461251573408e-07, + "loss": 0.1345, + "step": 12580 + }, + { + "epoch": 2.6380792618997693, + "grad_norm": 4.000929465795318, + "learning_rate": 7.537500132940911e-07, + "loss": 0.1336, + "step": 12581 + }, + { + "epoch": 2.638288949465297, + "grad_norm": 5.229897156964368, + "learning_rate": 7.528879847404535e-07, + "loss": 0.1812, + "step": 12582 + }, + { + "epoch": 2.638498637030824, + "grad_norm": 4.2210304307883595, + "learning_rate": 7.520264301173441e-07, + "loss": 0.1259, + "step": 12583 + }, + { + "epoch": 2.6387083245963514, + "grad_norm": 4.428285619233624, + "learning_rate": 7.511653494689186e-07, + "loss": 0.1627, + "step": 12584 + }, + { + "epoch": 2.638918012161879, + "grad_norm": 2.9515018033350633, + "learning_rate": 7.503047428393106e-07, + "loss": 0.095, + "step": 12585 + }, + { + "epoch": 2.639127699727406, + "grad_norm": 4.310322306287954, + "learning_rate": 7.494446102726294e-07, + "loss": 0.1432, + "step": 12586 + }, + { + "epoch": 2.6393373872929335, + "grad_norm": 3.7895239461777614, + "learning_rate": 7.48584951812954e-07, + "loss": 0.1395, + "step": 12587 + }, + { + "epoch": 2.639547074858461, + "grad_norm": 4.978326091435996, + "learning_rate": 7.477257675043447e-07, + "loss": 0.1922, + "step": 12588 + }, + { + "epoch": 2.639756762423988, + "grad_norm": 4.797021525588315, + "learning_rate": 7.468670573908399e-07, + "loss": 0.1517, + "step": 12589 + }, + { + "epoch": 2.6399664499895157, + "grad_norm": 4.035058321323052, + "learning_rate": 7.460088215164463e-07, + "loss": 0.1509, + "step": 12590 + }, + { + "epoch": 2.640176137555043, + "grad_norm": 5.504647747994489, + "learning_rate": 7.451510599251499e-07, + "loss": 0.1732, + "step": 12591 + }, + { + "epoch": 2.6403858251205703, + "grad_norm": 3.2438899010942945, + "learning_rate": 7.442937726609145e-07, + "loss": 0.1079, + "step": 12592 + }, + { + "epoch": 2.640595512686098, + "grad_norm": 3.209178690851853, + "learning_rate": 7.434369597676749e-07, + "loss": 0.1206, + "step": 12593 + }, + { + "epoch": 2.6408052002516254, + "grad_norm": 4.292244295781463, + "learning_rate": 7.425806212893472e-07, + "loss": 0.1551, + "step": 12594 + }, + { + "epoch": 2.6410148878171524, + "grad_norm": 4.8402474699134945, + "learning_rate": 7.417247572698194e-07, + "loss": 0.1883, + "step": 12595 + }, + { + "epoch": 2.64122457538268, + "grad_norm": 4.698461006950638, + "learning_rate": 7.408693677529532e-07, + "loss": 0.1461, + "step": 12596 + }, + { + "epoch": 2.641434262948207, + "grad_norm": 5.04896906133982, + "learning_rate": 7.400144527825914e-07, + "loss": 0.176, + "step": 12597 + }, + { + "epoch": 2.6416439505137346, + "grad_norm": 3.650307667148311, + "learning_rate": 7.391600124025511e-07, + "loss": 0.0971, + "step": 12598 + }, + { + "epoch": 2.6418536380792617, + "grad_norm": 3.2041939416460403, + "learning_rate": 7.383060466566228e-07, + "loss": 0.1025, + "step": 12599 + }, + { + "epoch": 2.642063325644789, + "grad_norm": 5.763439474652581, + "learning_rate": 7.374525555885704e-07, + "loss": 0.1508, + "step": 12600 + }, + { + "epoch": 2.6422730132103167, + "grad_norm": 5.21435687581906, + "learning_rate": 7.365995392421421e-07, + "loss": 0.1503, + "step": 12601 + }, + { + "epoch": 2.642482700775844, + "grad_norm": 3.6231930976983664, + "learning_rate": 7.357469976610521e-07, + "loss": 0.1258, + "step": 12602 + }, + { + "epoch": 2.6426923883413713, + "grad_norm": 3.877325125174339, + "learning_rate": 7.348949308889985e-07, + "loss": 0.1147, + "step": 12603 + }, + { + "epoch": 2.642902075906899, + "grad_norm": 4.122724696455569, + "learning_rate": 7.340433389696489e-07, + "loss": 0.1542, + "step": 12604 + }, + { + "epoch": 2.643111763472426, + "grad_norm": 3.972453923474521, + "learning_rate": 7.33192221946647e-07, + "loss": 0.1429, + "step": 12605 + }, + { + "epoch": 2.6433214510379535, + "grad_norm": 4.015269816390465, + "learning_rate": 7.32341579863618e-07, + "loss": 0.1034, + "step": 12606 + }, + { + "epoch": 2.643531138603481, + "grad_norm": 5.513555813843799, + "learning_rate": 7.314914127641582e-07, + "loss": 0.1506, + "step": 12607 + }, + { + "epoch": 2.643740826169008, + "grad_norm": 3.521344413504259, + "learning_rate": 7.306417206918359e-07, + "loss": 0.1096, + "step": 12608 + }, + { + "epoch": 2.6439505137345356, + "grad_norm": 5.540244187609464, + "learning_rate": 7.297925036902032e-07, + "loss": 0.1595, + "step": 12609 + }, + { + "epoch": 2.644160201300063, + "grad_norm": 4.040151491461108, + "learning_rate": 7.289437618027839e-07, + "loss": 0.1299, + "step": 12610 + }, + { + "epoch": 2.64436988886559, + "grad_norm": 4.109575021407578, + "learning_rate": 7.280954950730767e-07, + "loss": 0.1356, + "step": 12611 + }, + { + "epoch": 2.6445795764311177, + "grad_norm": 4.171848492814981, + "learning_rate": 7.272477035445557e-07, + "loss": 0.1168, + "step": 12612 + }, + { + "epoch": 2.6447892639966453, + "grad_norm": 4.27842625088676, + "learning_rate": 7.264003872606739e-07, + "loss": 0.1406, + "step": 12613 + }, + { + "epoch": 2.6449989515621724, + "grad_norm": 4.035926692442725, + "learning_rate": 7.255535462648544e-07, + "loss": 0.1465, + "step": 12614 + }, + { + "epoch": 2.6452086391277, + "grad_norm": 4.152846517543206, + "learning_rate": 7.247071806005035e-07, + "loss": 0.1259, + "step": 12615 + }, + { + "epoch": 2.645418326693227, + "grad_norm": 5.093508811225395, + "learning_rate": 7.238612903109976e-07, + "loss": 0.1167, + "step": 12616 + }, + { + "epoch": 2.6456280142587545, + "grad_norm": 4.769817633079835, + "learning_rate": 7.230158754396865e-07, + "loss": 0.1552, + "step": 12617 + }, + { + "epoch": 2.6458377018242816, + "grad_norm": 4.710167524175901, + "learning_rate": 7.221709360299022e-07, + "loss": 0.1795, + "step": 12618 + }, + { + "epoch": 2.646047389389809, + "grad_norm": 3.9346658771277654, + "learning_rate": 7.213264721249525e-07, + "loss": 0.155, + "step": 12619 + }, + { + "epoch": 2.6462570769553366, + "grad_norm": 3.7665936598770284, + "learning_rate": 7.204824837681101e-07, + "loss": 0.136, + "step": 12620 + }, + { + "epoch": 2.6464667645208637, + "grad_norm": 5.413114293709111, + "learning_rate": 7.196389710026352e-07, + "loss": 0.13, + "step": 12621 + }, + { + "epoch": 2.6466764520863912, + "grad_norm": 5.430464437355707, + "learning_rate": 7.18795933871761e-07, + "loss": 0.1967, + "step": 12622 + }, + { + "epoch": 2.6468861396519188, + "grad_norm": 4.838773723627348, + "learning_rate": 7.179533724186927e-07, + "loss": 0.1518, + "step": 12623 + }, + { + "epoch": 2.647095827217446, + "grad_norm": 3.260822521718474, + "learning_rate": 7.171112866866104e-07, + "loss": 0.105, + "step": 12624 + }, + { + "epoch": 2.6473055147829734, + "grad_norm": 4.065326315896353, + "learning_rate": 7.162696767186772e-07, + "loss": 0.133, + "step": 12625 + }, + { + "epoch": 2.647515202348501, + "grad_norm": 3.7634625146516454, + "learning_rate": 7.154285425580243e-07, + "loss": 0.1294, + "step": 12626 + }, + { + "epoch": 2.647724889914028, + "grad_norm": 3.5424257777676527, + "learning_rate": 7.145878842477627e-07, + "loss": 0.0962, + "step": 12627 + }, + { + "epoch": 2.6479345774795555, + "grad_norm": 5.2674700860643755, + "learning_rate": 7.137477018309747e-07, + "loss": 0.1415, + "step": 12628 + }, + { + "epoch": 2.648144265045083, + "grad_norm": 5.2204793287333136, + "learning_rate": 7.129079953507246e-07, + "loss": 0.1164, + "step": 12629 + }, + { + "epoch": 2.64835395261061, + "grad_norm": 4.315467933145188, + "learning_rate": 7.120687648500468e-07, + "loss": 0.1293, + "step": 12630 + }, + { + "epoch": 2.6485636401761377, + "grad_norm": 5.265944376939534, + "learning_rate": 7.112300103719549e-07, + "loss": 0.1709, + "step": 12631 + }, + { + "epoch": 2.648773327741665, + "grad_norm": 5.447535489681987, + "learning_rate": 7.103917319594356e-07, + "loss": 0.1571, + "step": 12632 + }, + { + "epoch": 2.6489830153071923, + "grad_norm": 3.4988078747793954, + "learning_rate": 7.095539296554499e-07, + "loss": 0.1077, + "step": 12633 + }, + { + "epoch": 2.64919270287272, + "grad_norm": 2.588474898374482, + "learning_rate": 7.087166035029403e-07, + "loss": 0.1007, + "step": 12634 + }, + { + "epoch": 2.649402390438247, + "grad_norm": 4.364946780875802, + "learning_rate": 7.078797535448201e-07, + "loss": 0.1645, + "step": 12635 + }, + { + "epoch": 2.6496120780037744, + "grad_norm": 4.099121305160108, + "learning_rate": 7.070433798239784e-07, + "loss": 0.1449, + "step": 12636 + }, + { + "epoch": 2.6498217655693015, + "grad_norm": 3.974024835337795, + "learning_rate": 7.06207482383281e-07, + "loss": 0.1328, + "step": 12637 + }, + { + "epoch": 2.650031453134829, + "grad_norm": 5.845957287062704, + "learning_rate": 7.053720612655702e-07, + "loss": 0.1725, + "step": 12638 + }, + { + "epoch": 2.6502411407003565, + "grad_norm": 3.9062197769903197, + "learning_rate": 7.045371165136605e-07, + "loss": 0.1205, + "step": 12639 + }, + { + "epoch": 2.6504508282658836, + "grad_norm": 4.53164629909146, + "learning_rate": 7.03702648170348e-07, + "loss": 0.1354, + "step": 12640 + }, + { + "epoch": 2.650660515831411, + "grad_norm": 3.609244672478305, + "learning_rate": 7.028686562783971e-07, + "loss": 0.1, + "step": 12641 + }, + { + "epoch": 2.6508702033969387, + "grad_norm": 4.362939927326942, + "learning_rate": 7.020351408805514e-07, + "loss": 0.1348, + "step": 12642 + }, + { + "epoch": 2.6510798909624658, + "grad_norm": 4.21682711838217, + "learning_rate": 7.012021020195325e-07, + "loss": 0.1095, + "step": 12643 + }, + { + "epoch": 2.6512895785279933, + "grad_norm": 4.907432698337553, + "learning_rate": 7.003695397380362e-07, + "loss": 0.1695, + "step": 12644 + }, + { + "epoch": 2.651499266093521, + "grad_norm": 5.5229083108521, + "learning_rate": 6.99537454078727e-07, + "loss": 0.1726, + "step": 12645 + }, + { + "epoch": 2.651708953659048, + "grad_norm": 4.352157879708616, + "learning_rate": 6.987058450842532e-07, + "loss": 0.1528, + "step": 12646 + }, + { + "epoch": 2.6519186412245754, + "grad_norm": 4.2400275325578844, + "learning_rate": 6.978747127972407e-07, + "loss": 0.1368, + "step": 12647 + }, + { + "epoch": 2.652128328790103, + "grad_norm": 4.484034089802241, + "learning_rate": 6.97044057260281e-07, + "loss": 0.1357, + "step": 12648 + }, + { + "epoch": 2.65233801635563, + "grad_norm": 4.878218402757219, + "learning_rate": 6.962138785159478e-07, + "loss": 0.1388, + "step": 12649 + }, + { + "epoch": 2.6525477039211576, + "grad_norm": 4.26261262452982, + "learning_rate": 6.953841766067915e-07, + "loss": 0.1444, + "step": 12650 + }, + { + "epoch": 2.652757391486685, + "grad_norm": 4.387552773337927, + "learning_rate": 6.945549515753325e-07, + "loss": 0.122, + "step": 12651 + }, + { + "epoch": 2.652967079052212, + "grad_norm": 4.691733644259645, + "learning_rate": 6.937262034640735e-07, + "loss": 0.1498, + "step": 12652 + }, + { + "epoch": 2.6531767666177397, + "grad_norm": 4.026961630370885, + "learning_rate": 6.928979323154883e-07, + "loss": 0.1117, + "step": 12653 + }, + { + "epoch": 2.653386454183267, + "grad_norm": 3.535103968634144, + "learning_rate": 6.92070138172023e-07, + "loss": 0.1106, + "step": 12654 + }, + { + "epoch": 2.6535961417487943, + "grad_norm": 4.340422957226782, + "learning_rate": 6.91242821076108e-07, + "loss": 0.1239, + "step": 12655 + }, + { + "epoch": 2.6538058293143214, + "grad_norm": 3.8862852530530785, + "learning_rate": 6.90415981070145e-07, + "loss": 0.1342, + "step": 12656 + }, + { + "epoch": 2.654015516879849, + "grad_norm": 3.8735343998502287, + "learning_rate": 6.8958961819651e-07, + "loss": 0.1172, + "step": 12657 + }, + { + "epoch": 2.6542252044453765, + "grad_norm": 4.196653747846812, + "learning_rate": 6.887637324975538e-07, + "loss": 0.1305, + "step": 12658 + }, + { + "epoch": 2.6544348920109035, + "grad_norm": 4.675004291138012, + "learning_rate": 6.879383240156068e-07, + "loss": 0.1309, + "step": 12659 + }, + { + "epoch": 2.654644579576431, + "grad_norm": 4.3536023105224615, + "learning_rate": 6.871133927929696e-07, + "loss": 0.1332, + "step": 12660 + }, + { + "epoch": 2.6548542671419586, + "grad_norm": 3.6891008354533508, + "learning_rate": 6.862889388719251e-07, + "loss": 0.1258, + "step": 12661 + }, + { + "epoch": 2.6550639547074857, + "grad_norm": 3.9009692119192416, + "learning_rate": 6.854649622947273e-07, + "loss": 0.1534, + "step": 12662 + }, + { + "epoch": 2.655273642273013, + "grad_norm": 5.0894974924162595, + "learning_rate": 6.846414631036036e-07, + "loss": 0.1627, + "step": 12663 + }, + { + "epoch": 2.6554833298385407, + "grad_norm": 5.441259647127946, + "learning_rate": 6.838184413407611e-07, + "loss": 0.1455, + "step": 12664 + }, + { + "epoch": 2.655693017404068, + "grad_norm": 6.744767318163683, + "learning_rate": 6.829958970483852e-07, + "loss": 0.1948, + "step": 12665 + }, + { + "epoch": 2.6559027049695954, + "grad_norm": 5.481736694650218, + "learning_rate": 6.821738302686254e-07, + "loss": 0.1704, + "step": 12666 + }, + { + "epoch": 2.656112392535123, + "grad_norm": 3.883459572991206, + "learning_rate": 6.81352241043618e-07, + "loss": 0.1121, + "step": 12667 + }, + { + "epoch": 2.65632208010065, + "grad_norm": 4.376530323647416, + "learning_rate": 6.805311294154715e-07, + "loss": 0.1206, + "step": 12668 + }, + { + "epoch": 2.6565317676661775, + "grad_norm": 5.019842802019677, + "learning_rate": 6.79710495426269e-07, + "loss": 0.1577, + "step": 12669 + }, + { + "epoch": 2.656741455231705, + "grad_norm": 3.8781944141075715, + "learning_rate": 6.788903391180679e-07, + "loss": 0.1237, + "step": 12670 + }, + { + "epoch": 2.656951142797232, + "grad_norm": 3.23722075228387, + "learning_rate": 6.780706605329046e-07, + "loss": 0.1042, + "step": 12671 + }, + { + "epoch": 2.6571608303627596, + "grad_norm": 3.920396995546099, + "learning_rate": 6.772514597127866e-07, + "loss": 0.1144, + "step": 12672 + }, + { + "epoch": 2.6573705179282867, + "grad_norm": 6.093583052591885, + "learning_rate": 6.764327366997015e-07, + "loss": 0.1982, + "step": 12673 + }, + { + "epoch": 2.6575802054938142, + "grad_norm": 3.384891574803331, + "learning_rate": 6.756144915356111e-07, + "loss": 0.1259, + "step": 12674 + }, + { + "epoch": 2.6577898930593413, + "grad_norm": 3.8154105391508852, + "learning_rate": 6.747967242624476e-07, + "loss": 0.1371, + "step": 12675 + }, + { + "epoch": 2.657999580624869, + "grad_norm": 3.393146078772084, + "learning_rate": 6.739794349221263e-07, + "loss": 0.1171, + "step": 12676 + }, + { + "epoch": 2.6582092681903964, + "grad_norm": 4.530025571915646, + "learning_rate": 6.731626235565348e-07, + "loss": 0.1183, + "step": 12677 + }, + { + "epoch": 2.6584189557559235, + "grad_norm": 3.3788694885492245, + "learning_rate": 6.723462902075362e-07, + "loss": 0.13, + "step": 12678 + }, + { + "epoch": 2.658628643321451, + "grad_norm": 6.18797116219666, + "learning_rate": 6.71530434916966e-07, + "loss": 0.182, + "step": 12679 + }, + { + "epoch": 2.6588383308869785, + "grad_norm": 3.8852023143682013, + "learning_rate": 6.707150577266431e-07, + "loss": 0.138, + "step": 12680 + }, + { + "epoch": 2.6590480184525056, + "grad_norm": 3.3361076186950527, + "learning_rate": 6.699001586783516e-07, + "loss": 0.1148, + "step": 12681 + }, + { + "epoch": 2.659257706018033, + "grad_norm": 5.3163228970901795, + "learning_rate": 6.690857378138616e-07, + "loss": 0.1367, + "step": 12682 + }, + { + "epoch": 2.6594673935835607, + "grad_norm": 4.691092042699991, + "learning_rate": 6.682717951749107e-07, + "loss": 0.1526, + "step": 12683 + }, + { + "epoch": 2.6596770811490877, + "grad_norm": 3.061695848672184, + "learning_rate": 6.674583308032146e-07, + "loss": 0.1128, + "step": 12684 + }, + { + "epoch": 2.6598867687146153, + "grad_norm": 4.352388606275134, + "learning_rate": 6.666453447404664e-07, + "loss": 0.1455, + "step": 12685 + }, + { + "epoch": 2.660096456280143, + "grad_norm": 3.7763577046927597, + "learning_rate": 6.658328370283318e-07, + "loss": 0.1566, + "step": 12686 + }, + { + "epoch": 2.66030614384567, + "grad_norm": 5.383267788256102, + "learning_rate": 6.65020807708453e-07, + "loss": 0.165, + "step": 12687 + }, + { + "epoch": 2.6605158314111974, + "grad_norm": 4.0731842092449835, + "learning_rate": 6.642092568224479e-07, + "loss": 0.1476, + "step": 12688 + }, + { + "epoch": 2.660725518976725, + "grad_norm": 5.148919127682927, + "learning_rate": 6.63398184411912e-07, + "loss": 0.1216, + "step": 12689 + }, + { + "epoch": 2.660935206542252, + "grad_norm": 5.3853521539558225, + "learning_rate": 6.62587590518412e-07, + "loss": 0.195, + "step": 12690 + }, + { + "epoch": 2.6611448941077795, + "grad_norm": 3.7112029207200585, + "learning_rate": 6.617774751834926e-07, + "loss": 0.1273, + "step": 12691 + }, + { + "epoch": 2.6613545816733066, + "grad_norm": 5.6438188086742365, + "learning_rate": 6.609678384486728e-07, + "loss": 0.1895, + "step": 12692 + }, + { + "epoch": 2.661564269238834, + "grad_norm": 3.585877251274759, + "learning_rate": 6.601586803554516e-07, + "loss": 0.079, + "step": 12693 + }, + { + "epoch": 2.6617739568043612, + "grad_norm": 4.7552206384993845, + "learning_rate": 6.593500009452969e-07, + "loss": 0.1839, + "step": 12694 + }, + { + "epoch": 2.6619836443698888, + "grad_norm": 3.4343884122717214, + "learning_rate": 6.585418002596533e-07, + "loss": 0.0836, + "step": 12695 + }, + { + "epoch": 2.6621933319354163, + "grad_norm": 4.960552437173351, + "learning_rate": 6.577340783399466e-07, + "loss": 0.2143, + "step": 12696 + }, + { + "epoch": 2.6624030195009434, + "grad_norm": 5.3369331504361766, + "learning_rate": 6.569268352275693e-07, + "loss": 0.1636, + "step": 12697 + }, + { + "epoch": 2.662612707066471, + "grad_norm": 5.773745303408101, + "learning_rate": 6.561200709638982e-07, + "loss": 0.2039, + "step": 12698 + }, + { + "epoch": 2.6628223946319984, + "grad_norm": 4.01504535482958, + "learning_rate": 6.553137855902802e-07, + "loss": 0.1053, + "step": 12699 + }, + { + "epoch": 2.6630320821975255, + "grad_norm": 3.3015730143836386, + "learning_rate": 6.545079791480369e-07, + "loss": 0.1182, + "step": 12700 + }, + { + "epoch": 2.663241769763053, + "grad_norm": 3.8974302878975218, + "learning_rate": 6.537026516784672e-07, + "loss": 0.1272, + "step": 12701 + }, + { + "epoch": 2.6634514573285806, + "grad_norm": 3.4911772367544787, + "learning_rate": 6.528978032228506e-07, + "loss": 0.0909, + "step": 12702 + }, + { + "epoch": 2.6636611448941077, + "grad_norm": 4.311480612882354, + "learning_rate": 6.520934338224305e-07, + "loss": 0.1431, + "step": 12703 + }, + { + "epoch": 2.663870832459635, + "grad_norm": 4.51776396315153, + "learning_rate": 6.512895435184352e-07, + "loss": 0.1716, + "step": 12704 + }, + { + "epoch": 2.6640805200251627, + "grad_norm": 4.578104621472425, + "learning_rate": 6.504861323520673e-07, + "loss": 0.1566, + "step": 12705 + }, + { + "epoch": 2.66429020759069, + "grad_norm": 5.210668449304779, + "learning_rate": 6.496832003644993e-07, + "loss": 0.1581, + "step": 12706 + }, + { + "epoch": 2.6644998951562173, + "grad_norm": 5.803265303991464, + "learning_rate": 6.48880747596885e-07, + "loss": 0.1366, + "step": 12707 + }, + { + "epoch": 2.664709582721745, + "grad_norm": 4.1144914060496065, + "learning_rate": 6.480787740903516e-07, + "loss": 0.141, + "step": 12708 + }, + { + "epoch": 2.664919270287272, + "grad_norm": 4.952466455763738, + "learning_rate": 6.472772798859994e-07, + "loss": 0.1361, + "step": 12709 + }, + { + "epoch": 2.6651289578527995, + "grad_norm": 4.669109916458371, + "learning_rate": 6.464762650249101e-07, + "loss": 0.1349, + "step": 12710 + }, + { + "epoch": 2.6653386454183265, + "grad_norm": 3.683386887504382, + "learning_rate": 6.456757295481341e-07, + "loss": 0.1353, + "step": 12711 + }, + { + "epoch": 2.665548332983854, + "grad_norm": 4.820214735219044, + "learning_rate": 6.44875673496701e-07, + "loss": 0.1694, + "step": 12712 + }, + { + "epoch": 2.665758020549381, + "grad_norm": 3.55495293406985, + "learning_rate": 6.440760969116145e-07, + "loss": 0.1116, + "step": 12713 + }, + { + "epoch": 2.6659677081149087, + "grad_norm": 3.7451704581002034, + "learning_rate": 6.432769998338573e-07, + "loss": 0.1501, + "step": 12714 + }, + { + "epoch": 2.666177395680436, + "grad_norm": 5.135749258720896, + "learning_rate": 6.424783823043812e-07, + "loss": 0.1345, + "step": 12715 + }, + { + "epoch": 2.6663870832459633, + "grad_norm": 4.696369053061642, + "learning_rate": 6.416802443641168e-07, + "loss": 0.1583, + "step": 12716 + }, + { + "epoch": 2.666596770811491, + "grad_norm": 3.5069320664738832, + "learning_rate": 6.408825860539736e-07, + "loss": 0.1461, + "step": 12717 + }, + { + "epoch": 2.6668064583770184, + "grad_norm": 6.527851261359122, + "learning_rate": 6.400854074148277e-07, + "loss": 0.1388, + "step": 12718 + }, + { + "epoch": 2.6670161459425454, + "grad_norm": 3.5905993534070486, + "learning_rate": 6.392887084875399e-07, + "loss": 0.1031, + "step": 12719 + }, + { + "epoch": 2.667225833508073, + "grad_norm": 4.572313198679673, + "learning_rate": 6.384924893129418e-07, + "loss": 0.1564, + "step": 12720 + }, + { + "epoch": 2.6674355210736005, + "grad_norm": 5.669899754180489, + "learning_rate": 6.376967499318387e-07, + "loss": 0.1828, + "step": 12721 + }, + { + "epoch": 2.6676452086391276, + "grad_norm": 3.63017708414544, + "learning_rate": 6.369014903850146e-07, + "loss": 0.1203, + "step": 12722 + }, + { + "epoch": 2.667854896204655, + "grad_norm": 3.5330825384954263, + "learning_rate": 6.361067107132324e-07, + "loss": 0.101, + "step": 12723 + }, + { + "epoch": 2.6680645837701826, + "grad_norm": 3.2888326149734115, + "learning_rate": 6.353124109572195e-07, + "loss": 0.1054, + "step": 12724 + }, + { + "epoch": 2.6682742713357097, + "grad_norm": 4.189764617241636, + "learning_rate": 6.345185911576867e-07, + "loss": 0.1526, + "step": 12725 + }, + { + "epoch": 2.6684839589012372, + "grad_norm": 4.35691479083874, + "learning_rate": 6.337252513553216e-07, + "loss": 0.1482, + "step": 12726 + }, + { + "epoch": 2.6686936464667648, + "grad_norm": 6.134140266964438, + "learning_rate": 6.329323915907826e-07, + "loss": 0.1883, + "step": 12727 + }, + { + "epoch": 2.668903334032292, + "grad_norm": 9.93398649479604, + "learning_rate": 6.321400119047039e-07, + "loss": 0.1593, + "step": 12728 + }, + { + "epoch": 2.6691130215978194, + "grad_norm": 3.371213532349761, + "learning_rate": 6.313481123376975e-07, + "loss": 0.116, + "step": 12729 + }, + { + "epoch": 2.6693227091633465, + "grad_norm": 4.218904502896088, + "learning_rate": 6.305566929303486e-07, + "loss": 0.1059, + "step": 12730 + }, + { + "epoch": 2.669532396728874, + "grad_norm": 4.809198942611835, + "learning_rate": 6.297657537232216e-07, + "loss": 0.144, + "step": 12731 + }, + { + "epoch": 2.669742084294401, + "grad_norm": 5.84224414795106, + "learning_rate": 6.289752947568506e-07, + "loss": 0.1245, + "step": 12732 + }, + { + "epoch": 2.6699517718599286, + "grad_norm": 4.716220099170551, + "learning_rate": 6.281853160717477e-07, + "loss": 0.12, + "step": 12733 + }, + { + "epoch": 2.670161459425456, + "grad_norm": 4.383744220988682, + "learning_rate": 6.273958177084028e-07, + "loss": 0.1442, + "step": 12734 + }, + { + "epoch": 2.670371146990983, + "grad_norm": 4.991446289480173, + "learning_rate": 6.266067997072789e-07, + "loss": 0.1545, + "step": 12735 + }, + { + "epoch": 2.6705808345565107, + "grad_norm": 3.658300588274846, + "learning_rate": 6.258182621088138e-07, + "loss": 0.1404, + "step": 12736 + }, + { + "epoch": 2.6707905221220383, + "grad_norm": 3.8698224827336096, + "learning_rate": 6.250302049534207e-07, + "loss": 0.1376, + "step": 12737 + }, + { + "epoch": 2.6710002096875654, + "grad_norm": 5.400434770282956, + "learning_rate": 6.242426282814906e-07, + "loss": 0.1515, + "step": 12738 + }, + { + "epoch": 2.671209897253093, + "grad_norm": 4.198740784257216, + "learning_rate": 6.234555321333858e-07, + "loss": 0.1677, + "step": 12739 + }, + { + "epoch": 2.6714195848186204, + "grad_norm": 3.306817489466332, + "learning_rate": 6.226689165494493e-07, + "loss": 0.1101, + "step": 12740 + }, + { + "epoch": 2.6716292723841475, + "grad_norm": 4.004903447252203, + "learning_rate": 6.218827815699946e-07, + "loss": 0.1272, + "step": 12741 + }, + { + "epoch": 2.671838959949675, + "grad_norm": 4.6228323884893765, + "learning_rate": 6.210971272353117e-07, + "loss": 0.1407, + "step": 12742 + }, + { + "epoch": 2.6720486475152025, + "grad_norm": 3.642884107118108, + "learning_rate": 6.203119535856683e-07, + "loss": 0.1039, + "step": 12743 + }, + { + "epoch": 2.6722583350807296, + "grad_norm": 4.644345393985968, + "learning_rate": 6.195272606613079e-07, + "loss": 0.1555, + "step": 12744 + }, + { + "epoch": 2.672468022646257, + "grad_norm": 4.834913009397528, + "learning_rate": 6.187430485024426e-07, + "loss": 0.1531, + "step": 12745 + }, + { + "epoch": 2.6726777102117847, + "grad_norm": 3.776297970259029, + "learning_rate": 6.179593171492671e-07, + "loss": 0.1471, + "step": 12746 + }, + { + "epoch": 2.6728873977773118, + "grad_norm": 4.369053425172367, + "learning_rate": 6.171760666419491e-07, + "loss": 0.1271, + "step": 12747 + }, + { + "epoch": 2.6730970853428393, + "grad_norm": 5.215239506788188, + "learning_rate": 6.163932970206333e-07, + "loss": 0.1631, + "step": 12748 + }, + { + "epoch": 2.6733067729083664, + "grad_norm": 4.809793509421329, + "learning_rate": 6.15611008325433e-07, + "loss": 0.1637, + "step": 12749 + }, + { + "epoch": 2.673516460473894, + "grad_norm": 4.483612475357762, + "learning_rate": 6.148292005964451e-07, + "loss": 0.1445, + "step": 12750 + }, + { + "epoch": 2.673726148039421, + "grad_norm": 5.147480928074263, + "learning_rate": 6.140478738737409e-07, + "loss": 0.1354, + "step": 12751 + }, + { + "epoch": 2.6739358356049485, + "grad_norm": 4.156572352274291, + "learning_rate": 6.132670281973618e-07, + "loss": 0.144, + "step": 12752 + }, + { + "epoch": 2.674145523170476, + "grad_norm": 4.906318160000465, + "learning_rate": 6.124866636073257e-07, + "loss": 0.1418, + "step": 12753 + }, + { + "epoch": 2.674355210736003, + "grad_norm": 4.469233705111311, + "learning_rate": 6.117067801436327e-07, + "loss": 0.1651, + "step": 12754 + }, + { + "epoch": 2.6745648983015307, + "grad_norm": 3.8880825248755406, + "learning_rate": 6.109273778462477e-07, + "loss": 0.1033, + "step": 12755 + }, + { + "epoch": 2.674774585867058, + "grad_norm": 3.4653958656996733, + "learning_rate": 6.101484567551208e-07, + "loss": 0.1273, + "step": 12756 + }, + { + "epoch": 2.6749842734325853, + "grad_norm": 3.8904701515917064, + "learning_rate": 6.093700169101724e-07, + "loss": 0.1174, + "step": 12757 + }, + { + "epoch": 2.675193960998113, + "grad_norm": 4.025990802211541, + "learning_rate": 6.08592058351295e-07, + "loss": 0.1202, + "step": 12758 + }, + { + "epoch": 2.6754036485636403, + "grad_norm": 3.784841784852634, + "learning_rate": 6.078145811183645e-07, + "loss": 0.1089, + "step": 12759 + }, + { + "epoch": 2.6756133361291674, + "grad_norm": 6.200025015982894, + "learning_rate": 6.070375852512267e-07, + "loss": 0.1646, + "step": 12760 + }, + { + "epoch": 2.675823023694695, + "grad_norm": 4.557343211336646, + "learning_rate": 6.062610707897054e-07, + "loss": 0.1013, + "step": 12761 + }, + { + "epoch": 2.6760327112602225, + "grad_norm": 4.492603514523045, + "learning_rate": 6.054850377735943e-07, + "loss": 0.1446, + "step": 12762 + }, + { + "epoch": 2.6762423988257495, + "grad_norm": 5.19013901288648, + "learning_rate": 6.047094862426706e-07, + "loss": 0.137, + "step": 12763 + }, + { + "epoch": 2.676452086391277, + "grad_norm": 3.906319130881846, + "learning_rate": 6.039344162366812e-07, + "loss": 0.13, + "step": 12764 + }, + { + "epoch": 2.6766617739568046, + "grad_norm": 5.2150404717990755, + "learning_rate": 6.031598277953488e-07, + "loss": 0.1175, + "step": 12765 + }, + { + "epoch": 2.6768714615223317, + "grad_norm": 4.797787609990199, + "learning_rate": 6.023857209583739e-07, + "loss": 0.1495, + "step": 12766 + }, + { + "epoch": 2.677081149087859, + "grad_norm": 4.897544976879718, + "learning_rate": 6.016120957654281e-07, + "loss": 0.1622, + "step": 12767 + }, + { + "epoch": 2.6772908366533863, + "grad_norm": 3.597809091248149, + "learning_rate": 6.008389522561653e-07, + "loss": 0.1368, + "step": 12768 + }, + { + "epoch": 2.677500524218914, + "grad_norm": 5.400485795253416, + "learning_rate": 6.000662904702081e-07, + "loss": 0.1931, + "step": 12769 + }, + { + "epoch": 2.677710211784441, + "grad_norm": 4.042867120557339, + "learning_rate": 5.992941104471562e-07, + "loss": 0.1386, + "step": 12770 + }, + { + "epoch": 2.6779198993499684, + "grad_norm": 3.124741202754535, + "learning_rate": 5.985224122265854e-07, + "loss": 0.1037, + "step": 12771 + }, + { + "epoch": 2.678129586915496, + "grad_norm": 4.703141696993759, + "learning_rate": 5.977511958480475e-07, + "loss": 0.1326, + "step": 12772 + }, + { + "epoch": 2.678339274481023, + "grad_norm": 3.5240236315409783, + "learning_rate": 5.969804613510688e-07, + "loss": 0.1021, + "step": 12773 + }, + { + "epoch": 2.6785489620465506, + "grad_norm": 6.268208339195901, + "learning_rate": 5.962102087751487e-07, + "loss": 0.1636, + "step": 12774 + }, + { + "epoch": 2.678758649612078, + "grad_norm": 4.3717317607729935, + "learning_rate": 5.954404381597678e-07, + "loss": 0.1277, + "step": 12775 + }, + { + "epoch": 2.678968337177605, + "grad_norm": 5.739817259074334, + "learning_rate": 5.946711495443735e-07, + "loss": 0.1535, + "step": 12776 + }, + { + "epoch": 2.6791780247431327, + "grad_norm": 4.049907933444462, + "learning_rate": 5.939023429683965e-07, + "loss": 0.1112, + "step": 12777 + }, + { + "epoch": 2.6793877123086602, + "grad_norm": 3.5271443630655255, + "learning_rate": 5.931340184712387e-07, + "loss": 0.1149, + "step": 12778 + }, + { + "epoch": 2.6795973998741873, + "grad_norm": 3.9077488641426994, + "learning_rate": 5.923661760922772e-07, + "loss": 0.1387, + "step": 12779 + }, + { + "epoch": 2.679807087439715, + "grad_norm": 3.5995636807479054, + "learning_rate": 5.915988158708652e-07, + "loss": 0.1163, + "step": 12780 + }, + { + "epoch": 2.6800167750052424, + "grad_norm": 4.515959490940363, + "learning_rate": 5.908319378463345e-07, + "loss": 0.1526, + "step": 12781 + }, + { + "epoch": 2.6802264625707695, + "grad_norm": 3.6460911096116213, + "learning_rate": 5.900655420579848e-07, + "loss": 0.1096, + "step": 12782 + }, + { + "epoch": 2.680436150136297, + "grad_norm": 4.64507960612798, + "learning_rate": 5.892996285450958e-07, + "loss": 0.1556, + "step": 12783 + }, + { + "epoch": 2.6806458377018245, + "grad_norm": 5.354660174517429, + "learning_rate": 5.885341973469239e-07, + "loss": 0.1343, + "step": 12784 + }, + { + "epoch": 2.6808555252673516, + "grad_norm": 4.0555097606634165, + "learning_rate": 5.877692485026986e-07, + "loss": 0.1192, + "step": 12785 + }, + { + "epoch": 2.681065212832879, + "grad_norm": 4.209022904040481, + "learning_rate": 5.870047820516222e-07, + "loss": 0.0976, + "step": 12786 + }, + { + "epoch": 2.681274900398406, + "grad_norm": 4.025159569595992, + "learning_rate": 5.862407980328788e-07, + "loss": 0.1392, + "step": 12787 + }, + { + "epoch": 2.6814845879639337, + "grad_norm": 4.091986784021727, + "learning_rate": 5.854772964856192e-07, + "loss": 0.1421, + "step": 12788 + }, + { + "epoch": 2.6816942755294613, + "grad_norm": 3.5470673576811294, + "learning_rate": 5.847142774489789e-07, + "loss": 0.1271, + "step": 12789 + }, + { + "epoch": 2.6819039630949884, + "grad_norm": 3.55769407996581, + "learning_rate": 5.839517409620621e-07, + "loss": 0.1004, + "step": 12790 + }, + { + "epoch": 2.682113650660516, + "grad_norm": 3.7104387138851522, + "learning_rate": 5.831896870639475e-07, + "loss": 0.1103, + "step": 12791 + }, + { + "epoch": 2.682323338226043, + "grad_norm": 3.469020609432843, + "learning_rate": 5.82428115793695e-07, + "loss": 0.1193, + "step": 12792 + }, + { + "epoch": 2.6825330257915705, + "grad_norm": 4.22258198449591, + "learning_rate": 5.816670271903369e-07, + "loss": 0.1549, + "step": 12793 + }, + { + "epoch": 2.682742713357098, + "grad_norm": 4.228763201405874, + "learning_rate": 5.809064212928783e-07, + "loss": 0.1227, + "step": 12794 + }, + { + "epoch": 2.682952400922625, + "grad_norm": 3.5002008437960215, + "learning_rate": 5.801462981403017e-07, + "loss": 0.1132, + "step": 12795 + }, + { + "epoch": 2.6831620884881526, + "grad_norm": 6.756367338151461, + "learning_rate": 5.793866577715657e-07, + "loss": 0.172, + "step": 12796 + }, + { + "epoch": 2.68337177605368, + "grad_norm": 4.344034258328609, + "learning_rate": 5.786275002256025e-07, + "loss": 0.1567, + "step": 12797 + }, + { + "epoch": 2.6835814636192072, + "grad_norm": 4.408965548937051, + "learning_rate": 5.778688255413212e-07, + "loss": 0.1343, + "step": 12798 + }, + { + "epoch": 2.6837911511847348, + "grad_norm": 4.07839422563627, + "learning_rate": 5.771106337576038e-07, + "loss": 0.1454, + "step": 12799 + }, + { + "epoch": 2.6840008387502623, + "grad_norm": 5.8816202036413205, + "learning_rate": 5.763529249133093e-07, + "loss": 0.1507, + "step": 12800 + }, + { + "epoch": 2.6842105263157894, + "grad_norm": 4.86602304472139, + "learning_rate": 5.755956990472711e-07, + "loss": 0.1468, + "step": 12801 + }, + { + "epoch": 2.684420213881317, + "grad_norm": 3.500719007491239, + "learning_rate": 5.748389561983015e-07, + "loss": 0.1136, + "step": 12802 + }, + { + "epoch": 2.6846299014468444, + "grad_norm": 4.989813071866303, + "learning_rate": 5.740826964051805e-07, + "loss": 0.1539, + "step": 12803 + }, + { + "epoch": 2.6848395890123715, + "grad_norm": 4.654407712606749, + "learning_rate": 5.733269197066693e-07, + "loss": 0.1429, + "step": 12804 + }, + { + "epoch": 2.685049276577899, + "grad_norm": 4.5511256857098035, + "learning_rate": 5.72571626141506e-07, + "loss": 0.1456, + "step": 12805 + }, + { + "epoch": 2.685258964143426, + "grad_norm": 3.4385160112379682, + "learning_rate": 5.71816815748396e-07, + "loss": 0.1246, + "step": 12806 + }, + { + "epoch": 2.6854686517089537, + "grad_norm": 3.699521419193321, + "learning_rate": 5.710624885660265e-07, + "loss": 0.098, + "step": 12807 + }, + { + "epoch": 2.685678339274481, + "grad_norm": 5.1769483042255375, + "learning_rate": 5.703086446330585e-07, + "loss": 0.1509, + "step": 12808 + }, + { + "epoch": 2.6858880268400083, + "grad_norm": 3.9930895513634233, + "learning_rate": 5.695552839881268e-07, + "loss": 0.1271, + "step": 12809 + }, + { + "epoch": 2.686097714405536, + "grad_norm": 4.952225650666098, + "learning_rate": 5.688024066698439e-07, + "loss": 0.1403, + "step": 12810 + }, + { + "epoch": 2.686307401971063, + "grad_norm": 5.121689585926485, + "learning_rate": 5.680500127167943e-07, + "loss": 0.1411, + "step": 12811 + }, + { + "epoch": 2.6865170895365904, + "grad_norm": 5.195136479841332, + "learning_rate": 5.672981021675417e-07, + "loss": 0.1624, + "step": 12812 + }, + { + "epoch": 2.686726777102118, + "grad_norm": 5.398556005844397, + "learning_rate": 5.665466750606196e-07, + "loss": 0.1826, + "step": 12813 + }, + { + "epoch": 2.686936464667645, + "grad_norm": 4.575079031151572, + "learning_rate": 5.657957314345441e-07, + "loss": 0.136, + "step": 12814 + }, + { + "epoch": 2.6871461522331725, + "grad_norm": 4.432476278777013, + "learning_rate": 5.650452713277998e-07, + "loss": 0.133, + "step": 12815 + }, + { + "epoch": 2.6873558397987, + "grad_norm": 3.3534408261339337, + "learning_rate": 5.642952947788483e-07, + "loss": 0.0958, + "step": 12816 + }, + { + "epoch": 2.687565527364227, + "grad_norm": 6.880326841928252, + "learning_rate": 5.635458018261286e-07, + "loss": 0.0986, + "step": 12817 + }, + { + "epoch": 2.6877752149297547, + "grad_norm": 4.499830380610673, + "learning_rate": 5.627967925080535e-07, + "loss": 0.1437, + "step": 12818 + }, + { + "epoch": 2.687984902495282, + "grad_norm": 4.553806998262058, + "learning_rate": 5.620482668630123e-07, + "loss": 0.1292, + "step": 12819 + }, + { + "epoch": 2.6881945900608093, + "grad_norm": 4.025919344498264, + "learning_rate": 5.613002249293642e-07, + "loss": 0.1542, + "step": 12820 + }, + { + "epoch": 2.688404277626337, + "grad_norm": 5.257580708811834, + "learning_rate": 5.605526667454519e-07, + "loss": 0.1421, + "step": 12821 + }, + { + "epoch": 2.6886139651918644, + "grad_norm": 3.9473085604016718, + "learning_rate": 5.598055923495859e-07, + "loss": 0.1185, + "step": 12822 + }, + { + "epoch": 2.6888236527573914, + "grad_norm": 4.085600488961794, + "learning_rate": 5.590590017800578e-07, + "loss": 0.1379, + "step": 12823 + }, + { + "epoch": 2.689033340322919, + "grad_norm": 5.6836744464965925, + "learning_rate": 5.583128950751304e-07, + "loss": 0.1305, + "step": 12824 + }, + { + "epoch": 2.6892430278884465, + "grad_norm": 3.81804350719673, + "learning_rate": 5.575672722730419e-07, + "loss": 0.1348, + "step": 12825 + }, + { + "epoch": 2.6894527154539736, + "grad_norm": 3.356325668359058, + "learning_rate": 5.568221334120083e-07, + "loss": 0.1257, + "step": 12826 + }, + { + "epoch": 2.689662403019501, + "grad_norm": 4.394239455814096, + "learning_rate": 5.560774785302192e-07, + "loss": 0.1522, + "step": 12827 + }, + { + "epoch": 2.689872090585028, + "grad_norm": 4.575388451390649, + "learning_rate": 5.553333076658374e-07, + "loss": 0.1185, + "step": 12828 + }, + { + "epoch": 2.6900817781505557, + "grad_norm": 3.303928367948205, + "learning_rate": 5.545896208570045e-07, + "loss": 0.1082, + "step": 12829 + }, + { + "epoch": 2.690291465716083, + "grad_norm": 3.1300046728189206, + "learning_rate": 5.538464181418379e-07, + "loss": 0.0983, + "step": 12830 + }, + { + "epoch": 2.6905011532816103, + "grad_norm": 4.059315491534546, + "learning_rate": 5.531036995584249e-07, + "loss": 0.138, + "step": 12831 + }, + { + "epoch": 2.690710840847138, + "grad_norm": 4.377515590100833, + "learning_rate": 5.523614651448317e-07, + "loss": 0.1374, + "step": 12832 + }, + { + "epoch": 2.690920528412665, + "grad_norm": 4.4728844546941, + "learning_rate": 5.516197149391e-07, + "loss": 0.1153, + "step": 12833 + }, + { + "epoch": 2.6911302159781925, + "grad_norm": 4.05538839842716, + "learning_rate": 5.508784489792451e-07, + "loss": 0.1164, + "step": 12834 + }, + { + "epoch": 2.69133990354372, + "grad_norm": 4.944304712714176, + "learning_rate": 5.501376673032587e-07, + "loss": 0.1545, + "step": 12835 + }, + { + "epoch": 2.691549591109247, + "grad_norm": 3.770859336744159, + "learning_rate": 5.493973699491073e-07, + "loss": 0.1195, + "step": 12836 + }, + { + "epoch": 2.6917592786747746, + "grad_norm": 3.223311776695362, + "learning_rate": 5.486575569547303e-07, + "loss": 0.0854, + "step": 12837 + }, + { + "epoch": 2.691968966240302, + "grad_norm": 4.158575029817456, + "learning_rate": 5.479182283580453e-07, + "loss": 0.1228, + "step": 12838 + }, + { + "epoch": 2.692178653805829, + "grad_norm": 4.733046280896581, + "learning_rate": 5.471793841969475e-07, + "loss": 0.1618, + "step": 12839 + }, + { + "epoch": 2.6923883413713567, + "grad_norm": 3.862882112461269, + "learning_rate": 5.46441024509301e-07, + "loss": 0.1417, + "step": 12840 + }, + { + "epoch": 2.6925980289368843, + "grad_norm": 5.5670852325662326, + "learning_rate": 5.457031493329467e-07, + "loss": 0.169, + "step": 12841 + }, + { + "epoch": 2.6928077165024114, + "grad_norm": 4.034730287070835, + "learning_rate": 5.449657587057045e-07, + "loss": 0.1325, + "step": 12842 + }, + { + "epoch": 2.693017404067939, + "grad_norm": 5.410612560818423, + "learning_rate": 5.442288526653672e-07, + "loss": 0.1788, + "step": 12843 + }, + { + "epoch": 2.6932270916334664, + "grad_norm": 3.6000222066663747, + "learning_rate": 5.434924312496992e-07, + "loss": 0.1164, + "step": 12844 + }, + { + "epoch": 2.6934367791989935, + "grad_norm": 4.131473401384019, + "learning_rate": 5.427564944964481e-07, + "loss": 0.144, + "step": 12845 + }, + { + "epoch": 2.693646466764521, + "grad_norm": 3.5502653029057347, + "learning_rate": 5.420210424433269e-07, + "loss": 0.1273, + "step": 12846 + }, + { + "epoch": 2.693856154330048, + "grad_norm": 4.313079841865814, + "learning_rate": 5.412860751280324e-07, + "loss": 0.1164, + "step": 12847 + }, + { + "epoch": 2.6940658418955756, + "grad_norm": 4.19684441087359, + "learning_rate": 5.405515925882332e-07, + "loss": 0.126, + "step": 12848 + }, + { + "epoch": 2.6942755294611027, + "grad_norm": 4.2494456077744704, + "learning_rate": 5.398175948615692e-07, + "loss": 0.1489, + "step": 12849 + }, + { + "epoch": 2.6944852170266302, + "grad_norm": 4.390633000272372, + "learning_rate": 5.390840819856613e-07, + "loss": 0.1167, + "step": 12850 + }, + { + "epoch": 2.6946949045921578, + "grad_norm": 3.465133406252074, + "learning_rate": 5.383510539981052e-07, + "loss": 0.137, + "step": 12851 + }, + { + "epoch": 2.694904592157685, + "grad_norm": 4.228916769183762, + "learning_rate": 5.376185109364673e-07, + "loss": 0.1255, + "step": 12852 + }, + { + "epoch": 2.6951142797232124, + "grad_norm": 4.289390887741756, + "learning_rate": 5.368864528382922e-07, + "loss": 0.1704, + "step": 12853 + }, + { + "epoch": 2.69532396728874, + "grad_norm": 4.242634837374145, + "learning_rate": 5.361548797410998e-07, + "loss": 0.1114, + "step": 12854 + }, + { + "epoch": 2.695533654854267, + "grad_norm": 4.071015545282936, + "learning_rate": 5.354237916823823e-07, + "loss": 0.1512, + "step": 12855 + }, + { + "epoch": 2.6957433424197945, + "grad_norm": 3.7171100702745927, + "learning_rate": 5.346931886996132e-07, + "loss": 0.1163, + "step": 12856 + }, + { + "epoch": 2.695953029985322, + "grad_norm": 4.252449940361792, + "learning_rate": 5.339630708302346e-07, + "loss": 0.1277, + "step": 12857 + }, + { + "epoch": 2.696162717550849, + "grad_norm": 4.250896799262427, + "learning_rate": 5.332334381116643e-07, + "loss": 0.1438, + "step": 12858 + }, + { + "epoch": 2.6963724051163767, + "grad_norm": 5.318366338604121, + "learning_rate": 5.325042905813005e-07, + "loss": 0.1616, + "step": 12859 + }, + { + "epoch": 2.696582092681904, + "grad_norm": 4.427708656007492, + "learning_rate": 5.317756282765152e-07, + "loss": 0.1507, + "step": 12860 + }, + { + "epoch": 2.6967917802474313, + "grad_norm": 3.9280451614176455, + "learning_rate": 5.310474512346476e-07, + "loss": 0.1354, + "step": 12861 + }, + { + "epoch": 2.697001467812959, + "grad_norm": 4.451344699274738, + "learning_rate": 5.30319759493022e-07, + "loss": 0.1324, + "step": 12862 + }, + { + "epoch": 2.6972111553784863, + "grad_norm": 3.806971984511964, + "learning_rate": 5.295925530889334e-07, + "loss": 0.1178, + "step": 12863 + }, + { + "epoch": 2.6974208429440134, + "grad_norm": 3.6706076985361378, + "learning_rate": 5.288658320596529e-07, + "loss": 0.1107, + "step": 12864 + }, + { + "epoch": 2.697630530509541, + "grad_norm": 3.4757671497873095, + "learning_rate": 5.281395964424229e-07, + "loss": 0.1085, + "step": 12865 + }, + { + "epoch": 2.697840218075068, + "grad_norm": 4.528489561745475, + "learning_rate": 5.274138462744693e-07, + "loss": 0.1654, + "step": 12866 + }, + { + "epoch": 2.6980499056405955, + "grad_norm": 4.336963742398315, + "learning_rate": 5.266885815929835e-07, + "loss": 0.121, + "step": 12867 + }, + { + "epoch": 2.6982595932061226, + "grad_norm": 3.4332250539813054, + "learning_rate": 5.259638024351399e-07, + "loss": 0.0873, + "step": 12868 + }, + { + "epoch": 2.69846928077165, + "grad_norm": 5.916331043663303, + "learning_rate": 5.252395088380835e-07, + "loss": 0.1738, + "step": 12869 + }, + { + "epoch": 2.6986789683371777, + "grad_norm": 4.143655914420538, + "learning_rate": 5.245157008389334e-07, + "loss": 0.1601, + "step": 12870 + }, + { + "epoch": 2.6988886559027048, + "grad_norm": 4.426412851082678, + "learning_rate": 5.237923784747889e-07, + "loss": 0.1307, + "step": 12871 + }, + { + "epoch": 2.6990983434682323, + "grad_norm": 4.576500674665621, + "learning_rate": 5.230695417827214e-07, + "loss": 0.1608, + "step": 12872 + }, + { + "epoch": 2.69930803103376, + "grad_norm": 3.3190453237441653, + "learning_rate": 5.223471907997768e-07, + "loss": 0.1002, + "step": 12873 + }, + { + "epoch": 2.699517718599287, + "grad_norm": 3.603233066777662, + "learning_rate": 5.216253255629756e-07, + "loss": 0.1408, + "step": 12874 + }, + { + "epoch": 2.6997274061648144, + "grad_norm": 4.110136446441054, + "learning_rate": 5.209039461093157e-07, + "loss": 0.1849, + "step": 12875 + }, + { + "epoch": 2.699937093730342, + "grad_norm": 4.168173204893433, + "learning_rate": 5.201830524757712e-07, + "loss": 0.1303, + "step": 12876 + }, + { + "epoch": 2.700146781295869, + "grad_norm": 3.9801289094666448, + "learning_rate": 5.194626446992868e-07, + "loss": 0.1358, + "step": 12877 + }, + { + "epoch": 2.7003564688613966, + "grad_norm": 4.7962538078344625, + "learning_rate": 5.187427228167841e-07, + "loss": 0.1613, + "step": 12878 + }, + { + "epoch": 2.700566156426924, + "grad_norm": 4.021730804720595, + "learning_rate": 5.180232868651614e-07, + "loss": 0.1301, + "step": 12879 + }, + { + "epoch": 2.700775843992451, + "grad_norm": 4.2870120910781315, + "learning_rate": 5.173043368812902e-07, + "loss": 0.1079, + "step": 12880 + }, + { + "epoch": 2.7009855315579787, + "grad_norm": 4.022749962440249, + "learning_rate": 5.165858729020202e-07, + "loss": 0.118, + "step": 12881 + }, + { + "epoch": 2.7011952191235062, + "grad_norm": 3.845429217342205, + "learning_rate": 5.158678949641726e-07, + "loss": 0.1293, + "step": 12882 + }, + { + "epoch": 2.7014049066890333, + "grad_norm": 4.0569265027943, + "learning_rate": 5.151504031045429e-07, + "loss": 0.1618, + "step": 12883 + }, + { + "epoch": 2.701614594254561, + "grad_norm": 4.721182374593468, + "learning_rate": 5.144333973599058e-07, + "loss": 0.1602, + "step": 12884 + }, + { + "epoch": 2.701824281820088, + "grad_norm": 4.386778487990207, + "learning_rate": 5.13716877767012e-07, + "loss": 0.1631, + "step": 12885 + }, + { + "epoch": 2.7020339693856155, + "grad_norm": 5.736770148841213, + "learning_rate": 5.130008443625778e-07, + "loss": 0.1844, + "step": 12886 + }, + { + "epoch": 2.7022436569511425, + "grad_norm": 3.796202617425379, + "learning_rate": 5.122852971833058e-07, + "loss": 0.121, + "step": 12887 + }, + { + "epoch": 2.70245334451667, + "grad_norm": 3.4456789062199387, + "learning_rate": 5.115702362658681e-07, + "loss": 0.1126, + "step": 12888 + }, + { + "epoch": 2.7026630320821976, + "grad_norm": 4.908146967473637, + "learning_rate": 5.10855661646914e-07, + "loss": 0.1407, + "step": 12889 + }, + { + "epoch": 2.7028727196477247, + "grad_norm": 4.280959328408613, + "learning_rate": 5.101415733630632e-07, + "loss": 0.1458, + "step": 12890 + }, + { + "epoch": 2.703082407213252, + "grad_norm": 3.5326011427196864, + "learning_rate": 5.094279714509176e-07, + "loss": 0.139, + "step": 12891 + }, + { + "epoch": 2.7032920947787797, + "grad_norm": 4.588449235259221, + "learning_rate": 5.087148559470468e-07, + "loss": 0.1419, + "step": 12892 + }, + { + "epoch": 2.703501782344307, + "grad_norm": 5.423659056032981, + "learning_rate": 5.080022268880036e-07, + "loss": 0.1306, + "step": 12893 + }, + { + "epoch": 2.7037114699098344, + "grad_norm": 3.712798224548057, + "learning_rate": 5.072900843103091e-07, + "loss": 0.1293, + "step": 12894 + }, + { + "epoch": 2.703921157475362, + "grad_norm": 4.73864505983658, + "learning_rate": 5.065784282504615e-07, + "loss": 0.1371, + "step": 12895 + }, + { + "epoch": 2.704130845040889, + "grad_norm": 4.123363263628095, + "learning_rate": 5.058672587449343e-07, + "loss": 0.0808, + "step": 12896 + }, + { + "epoch": 2.7043405326064165, + "grad_norm": 4.354155265861245, + "learning_rate": 5.051565758301779e-07, + "loss": 0.1286, + "step": 12897 + }, + { + "epoch": 2.704550220171944, + "grad_norm": 3.7863848836974277, + "learning_rate": 5.044463795426147e-07, + "loss": 0.1254, + "step": 12898 + }, + { + "epoch": 2.704759907737471, + "grad_norm": 3.787978147289437, + "learning_rate": 5.03736669918643e-07, + "loss": 0.1086, + "step": 12899 + }, + { + "epoch": 2.7049695953029986, + "grad_norm": 4.110695690613969, + "learning_rate": 5.030274469946395e-07, + "loss": 0.1532, + "step": 12900 + }, + { + "epoch": 2.705179282868526, + "grad_norm": 4.465080941296165, + "learning_rate": 5.023187108069483e-07, + "loss": 0.1264, + "step": 12901 + }, + { + "epoch": 2.7053889704340532, + "grad_norm": 4.80143661057014, + "learning_rate": 5.016104613918981e-07, + "loss": 0.1335, + "step": 12902 + }, + { + "epoch": 2.7055986579995808, + "grad_norm": 5.0381109938023405, + "learning_rate": 5.009026987857857e-07, + "loss": 0.1586, + "step": 12903 + }, + { + "epoch": 2.705808345565108, + "grad_norm": 4.802065550838619, + "learning_rate": 5.00195423024884e-07, + "loss": 0.1476, + "step": 12904 + }, + { + "epoch": 2.7060180331306354, + "grad_norm": 4.853637498825989, + "learning_rate": 4.994886341454452e-07, + "loss": 0.1077, + "step": 12905 + }, + { + "epoch": 2.7062277206961625, + "grad_norm": 3.9431367171723077, + "learning_rate": 4.987823321836904e-07, + "loss": 0.1307, + "step": 12906 + }, + { + "epoch": 2.70643740826169, + "grad_norm": 3.1542944862368487, + "learning_rate": 4.980765171758206e-07, + "loss": 0.1273, + "step": 12907 + }, + { + "epoch": 2.7066470958272175, + "grad_norm": 3.7359439573337854, + "learning_rate": 4.973711891580079e-07, + "loss": 0.1238, + "step": 12908 + }, + { + "epoch": 2.7068567833927446, + "grad_norm": 4.324340520305529, + "learning_rate": 4.966663481664058e-07, + "loss": 0.1166, + "step": 12909 + }, + { + "epoch": 2.707066470958272, + "grad_norm": 3.812328832626003, + "learning_rate": 4.959619942371352e-07, + "loss": 0.1153, + "step": 12910 + }, + { + "epoch": 2.7072761585237997, + "grad_norm": 5.194414708895172, + "learning_rate": 4.952581274062951e-07, + "loss": 0.1695, + "step": 12911 + }, + { + "epoch": 2.7074858460893267, + "grad_norm": 4.40246419500683, + "learning_rate": 4.945547477099633e-07, + "loss": 0.1276, + "step": 12912 + }, + { + "epoch": 2.7076955336548543, + "grad_norm": 3.311785189662271, + "learning_rate": 4.938518551841842e-07, + "loss": 0.1299, + "step": 12913 + }, + { + "epoch": 2.707905221220382, + "grad_norm": 3.3137349807363097, + "learning_rate": 4.93149449864988e-07, + "loss": 0.0855, + "step": 12914 + }, + { + "epoch": 2.708114908785909, + "grad_norm": 5.11668873569344, + "learning_rate": 4.924475317883704e-07, + "loss": 0.1189, + "step": 12915 + }, + { + "epoch": 2.7083245963514364, + "grad_norm": 3.1832860599040154, + "learning_rate": 4.917461009903057e-07, + "loss": 0.1022, + "step": 12916 + }, + { + "epoch": 2.708534283916964, + "grad_norm": 5.489876373463343, + "learning_rate": 4.910451575067443e-07, + "loss": 0.1563, + "step": 12917 + }, + { + "epoch": 2.708743971482491, + "grad_norm": 4.857369726719416, + "learning_rate": 4.903447013736129e-07, + "loss": 0.1705, + "step": 12918 + }, + { + "epoch": 2.7089536590480185, + "grad_norm": 4.027220215947041, + "learning_rate": 4.896447326268083e-07, + "loss": 0.1423, + "step": 12919 + }, + { + "epoch": 2.709163346613546, + "grad_norm": 4.472407174254757, + "learning_rate": 4.889452513022053e-07, + "loss": 0.1266, + "step": 12920 + }, + { + "epoch": 2.709373034179073, + "grad_norm": 4.981540546093964, + "learning_rate": 4.882462574356562e-07, + "loss": 0.1316, + "step": 12921 + }, + { + "epoch": 2.7095827217446007, + "grad_norm": 3.371009902419683, + "learning_rate": 4.875477510629823e-07, + "loss": 0.1176, + "step": 12922 + }, + { + "epoch": 2.7097924093101278, + "grad_norm": 5.501066025328657, + "learning_rate": 4.86849732219985e-07, + "loss": 0.178, + "step": 12923 + }, + { + "epoch": 2.7100020968756553, + "grad_norm": 5.155077674741697, + "learning_rate": 4.861522009424391e-07, + "loss": 0.1266, + "step": 12924 + }, + { + "epoch": 2.7102117844411824, + "grad_norm": 4.636654473753456, + "learning_rate": 4.854551572660915e-07, + "loss": 0.1465, + "step": 12925 + }, + { + "epoch": 2.71042147200671, + "grad_norm": 5.304478354853692, + "learning_rate": 4.847586012266725e-07, + "loss": 0.2087, + "step": 12926 + }, + { + "epoch": 2.7106311595722374, + "grad_norm": 3.467300868807927, + "learning_rate": 4.840625328598769e-07, + "loss": 0.1102, + "step": 12927 + }, + { + "epoch": 2.7108408471377645, + "grad_norm": 4.4499461528255635, + "learning_rate": 4.833669522013807e-07, + "loss": 0.1294, + "step": 12928 + }, + { + "epoch": 2.711050534703292, + "grad_norm": 4.017344033093376, + "learning_rate": 4.82671859286834e-07, + "loss": 0.1165, + "step": 12929 + }, + { + "epoch": 2.7112602222688196, + "grad_norm": 4.395317674417498, + "learning_rate": 4.81977254151862e-07, + "loss": 0.1614, + "step": 12930 + }, + { + "epoch": 2.7114699098343467, + "grad_norm": 5.442910382218076, + "learning_rate": 4.812831368320647e-07, + "loss": 0.1315, + "step": 12931 + }, + { + "epoch": 2.711679597399874, + "grad_norm": 4.53188196216915, + "learning_rate": 4.80589507363014e-07, + "loss": 0.1201, + "step": 12932 + }, + { + "epoch": 2.7118892849654017, + "grad_norm": 4.769564464365183, + "learning_rate": 4.798963657802624e-07, + "loss": 0.1668, + "step": 12933 + }, + { + "epoch": 2.712098972530929, + "grad_norm": 4.541470336707967, + "learning_rate": 4.792037121193349e-07, + "loss": 0.1466, + "step": 12934 + }, + { + "epoch": 2.7123086600964563, + "grad_norm": 4.174641683307479, + "learning_rate": 4.785115464157297e-07, + "loss": 0.1485, + "step": 12935 + }, + { + "epoch": 2.712518347661984, + "grad_norm": 4.6907800493546645, + "learning_rate": 4.778198687049207e-07, + "loss": 0.142, + "step": 12936 + }, + { + "epoch": 2.712728035227511, + "grad_norm": 3.9731316734444397, + "learning_rate": 4.771286790223606e-07, + "loss": 0.1324, + "step": 12937 + }, + { + "epoch": 2.7129377227930385, + "grad_norm": 3.5745997010911847, + "learning_rate": 4.7643797740346997e-07, + "loss": 0.1039, + "step": 12938 + }, + { + "epoch": 2.713147410358566, + "grad_norm": 3.6178250386276884, + "learning_rate": 4.7574776388365383e-07, + "loss": 0.1122, + "step": 12939 + }, + { + "epoch": 2.713357097924093, + "grad_norm": 3.7975845545905185, + "learning_rate": 4.750580384982828e-07, + "loss": 0.1273, + "step": 12940 + }, + { + "epoch": 2.7135667854896206, + "grad_norm": 5.987865909498283, + "learning_rate": 4.7436880128270526e-07, + "loss": 0.1548, + "step": 12941 + }, + { + "epoch": 2.7137764730551477, + "grad_norm": 3.6207442005610466, + "learning_rate": 4.7368005227224954e-07, + "loss": 0.1184, + "step": 12942 + }, + { + "epoch": 2.713986160620675, + "grad_norm": 4.321823345711403, + "learning_rate": 4.729917915022153e-07, + "loss": 0.1436, + "step": 12943 + }, + { + "epoch": 2.7141958481862023, + "grad_norm": 4.483822184350326, + "learning_rate": 4.7230401900787315e-07, + "loss": 0.1119, + "step": 12944 + }, + { + "epoch": 2.71440553575173, + "grad_norm": 5.099378431563094, + "learning_rate": 4.716167348244749e-07, + "loss": 0.1782, + "step": 12945 + }, + { + "epoch": 2.7146152233172574, + "grad_norm": 6.093865063693587, + "learning_rate": 4.709299389872468e-07, + "loss": 0.1995, + "step": 12946 + }, + { + "epoch": 2.7148249108827844, + "grad_norm": 4.033200298151314, + "learning_rate": 4.7024363153138633e-07, + "loss": 0.1486, + "step": 12947 + }, + { + "epoch": 2.715034598448312, + "grad_norm": 3.732926921607242, + "learning_rate": 4.6955781249206747e-07, + "loss": 0.1044, + "step": 12948 + }, + { + "epoch": 2.7152442860138395, + "grad_norm": 3.7122810034242995, + "learning_rate": 4.68872481904441e-07, + "loss": 0.1088, + "step": 12949 + }, + { + "epoch": 2.7154539735793666, + "grad_norm": 5.315774523405423, + "learning_rate": 4.6818763980363e-07, + "loss": 0.1681, + "step": 12950 + }, + { + "epoch": 2.715663661144894, + "grad_norm": 3.595935883284847, + "learning_rate": 4.675032862247364e-07, + "loss": 0.1232, + "step": 12951 + }, + { + "epoch": 2.7158733487104216, + "grad_norm": 5.176459058248521, + "learning_rate": 4.6681942120283094e-07, + "loss": 0.143, + "step": 12952 + }, + { + "epoch": 2.7160830362759487, + "grad_norm": 4.663626960206136, + "learning_rate": 4.661360447729646e-07, + "loss": 0.1663, + "step": 12953 + }, + { + "epoch": 2.7162927238414762, + "grad_norm": 4.295038124005719, + "learning_rate": 4.6545315697016146e-07, + "loss": 0.1573, + "step": 12954 + }, + { + "epoch": 2.7165024114070038, + "grad_norm": 3.9227550300988363, + "learning_rate": 4.647707578294225e-07, + "loss": 0.1322, + "step": 12955 + }, + { + "epoch": 2.716712098972531, + "grad_norm": 3.6938144716384564, + "learning_rate": 4.6408884738571966e-07, + "loss": 0.145, + "step": 12956 + }, + { + "epoch": 2.7169217865380584, + "grad_norm": 2.9999797074214003, + "learning_rate": 4.6340742567400065e-07, + "loss": 0.1018, + "step": 12957 + }, + { + "epoch": 2.717131474103586, + "grad_norm": 3.7452665704529684, + "learning_rate": 4.6272649272919414e-07, + "loss": 0.1174, + "step": 12958 + }, + { + "epoch": 2.717341161669113, + "grad_norm": 3.5840532657436315, + "learning_rate": 4.620460485861944e-07, + "loss": 0.1175, + "step": 12959 + }, + { + "epoch": 2.7175508492346405, + "grad_norm": 3.502673722676137, + "learning_rate": 4.61366093279878e-07, + "loss": 0.0967, + "step": 12960 + }, + { + "epoch": 2.7177605368001676, + "grad_norm": 5.068836699777926, + "learning_rate": 4.606866268450949e-07, + "loss": 0.122, + "step": 12961 + }, + { + "epoch": 2.717970224365695, + "grad_norm": 4.161801295578957, + "learning_rate": 4.6000764931666496e-07, + "loss": 0.1232, + "step": 12962 + }, + { + "epoch": 2.718179911931222, + "grad_norm": 4.573145471927253, + "learning_rate": 4.593291607293893e-07, + "loss": 0.1699, + "step": 12963 + }, + { + "epoch": 2.7183895994967497, + "grad_norm": 3.896022721718441, + "learning_rate": 4.5865116111804444e-07, + "loss": 0.1226, + "step": 12964 + }, + { + "epoch": 2.7185992870622773, + "grad_norm": 4.46992078456667, + "learning_rate": 4.579736505173726e-07, + "loss": 0.1021, + "step": 12965 + }, + { + "epoch": 2.7188089746278044, + "grad_norm": 5.699091132044439, + "learning_rate": 4.572966289621028e-07, + "loss": 0.1371, + "step": 12966 + }, + { + "epoch": 2.719018662193332, + "grad_norm": 4.583468468504497, + "learning_rate": 4.566200964869316e-07, + "loss": 0.1434, + "step": 12967 + }, + { + "epoch": 2.7192283497588594, + "grad_norm": 3.8987073071538085, + "learning_rate": 4.5594405312653357e-07, + "loss": 0.1134, + "step": 12968 + }, + { + "epoch": 2.7194380373243865, + "grad_norm": 4.154138249860281, + "learning_rate": 4.5526849891555424e-07, + "loss": 0.1391, + "step": 12969 + }, + { + "epoch": 2.719647724889914, + "grad_norm": 4.777505832992595, + "learning_rate": 4.545934338886204e-07, + "loss": 0.1832, + "step": 12970 + }, + { + "epoch": 2.7198574124554415, + "grad_norm": 4.293642384747073, + "learning_rate": 4.5391885808032664e-07, + "loss": 0.1306, + "step": 12971 + }, + { + "epoch": 2.7200671000209686, + "grad_norm": 3.7549427768051395, + "learning_rate": 4.532447715252508e-07, + "loss": 0.1271, + "step": 12972 + }, + { + "epoch": 2.720276787586496, + "grad_norm": 5.894509316397902, + "learning_rate": 4.525711742579375e-07, + "loss": 0.1577, + "step": 12973 + }, + { + "epoch": 2.7204864751520237, + "grad_norm": 4.241173070553654, + "learning_rate": 4.518980663129102e-07, + "loss": 0.1186, + "step": 12974 + }, + { + "epoch": 2.7206961627175508, + "grad_norm": 3.967952873027723, + "learning_rate": 4.512254477246669e-07, + "loss": 0.1154, + "step": 12975 + }, + { + "epoch": 2.7209058502830783, + "grad_norm": 3.4704061749036668, + "learning_rate": 4.5055331852768113e-07, + "loss": 0.117, + "step": 12976 + }, + { + "epoch": 2.721115537848606, + "grad_norm": 4.66058442829878, + "learning_rate": 4.49881678756402e-07, + "loss": 0.1511, + "step": 12977 + }, + { + "epoch": 2.721325225414133, + "grad_norm": 4.549925513913783, + "learning_rate": 4.492105284452486e-07, + "loss": 0.1344, + "step": 12978 + }, + { + "epoch": 2.7215349129796604, + "grad_norm": 4.603219116507876, + "learning_rate": 4.4853986762862234e-07, + "loss": 0.1226, + "step": 12979 + }, + { + "epoch": 2.7217446005451875, + "grad_norm": 5.4852833301333845, + "learning_rate": 4.478696963408924e-07, + "loss": 0.1511, + "step": 12980 + }, + { + "epoch": 2.721954288110715, + "grad_norm": 4.620012538146422, + "learning_rate": 4.4720001461641015e-07, + "loss": 0.1485, + "step": 12981 + }, + { + "epoch": 2.722163975676242, + "grad_norm": 4.113204601674995, + "learning_rate": 4.4653082248949487e-07, + "loss": 0.1229, + "step": 12982 + }, + { + "epoch": 2.7223736632417697, + "grad_norm": 4.59889031244395, + "learning_rate": 4.458621199944424e-07, + "loss": 0.1148, + "step": 12983 + }, + { + "epoch": 2.722583350807297, + "grad_norm": 5.381763701939643, + "learning_rate": 4.451939071655298e-07, + "loss": 0.1614, + "step": 12984 + }, + { + "epoch": 2.7227930383728243, + "grad_norm": 3.9638831508810317, + "learning_rate": 4.445261840370008e-07, + "loss": 0.1215, + "step": 12985 + }, + { + "epoch": 2.723002725938352, + "grad_norm": 4.190737809321449, + "learning_rate": 4.4385895064307684e-07, + "loss": 0.1286, + "step": 12986 + }, + { + "epoch": 2.7232124135038793, + "grad_norm": 5.706131868331729, + "learning_rate": 4.431922070179562e-07, + "loss": 0.1293, + "step": 12987 + }, + { + "epoch": 2.7234221010694064, + "grad_norm": 4.1399119658392145, + "learning_rate": 4.425259531958115e-07, + "loss": 0.1186, + "step": 12988 + }, + { + "epoch": 2.723631788634934, + "grad_norm": 3.722007400210006, + "learning_rate": 4.4186018921078765e-07, + "loss": 0.122, + "step": 12989 + }, + { + "epoch": 2.7238414762004615, + "grad_norm": 4.45556209010085, + "learning_rate": 4.4119491509700627e-07, + "loss": 0.1111, + "step": 12990 + }, + { + "epoch": 2.7240511637659885, + "grad_norm": 4.614646855155972, + "learning_rate": 4.4053013088856566e-07, + "loss": 0.1699, + "step": 12991 + }, + { + "epoch": 2.724260851331516, + "grad_norm": 3.730685125062525, + "learning_rate": 4.398658366195341e-07, + "loss": 0.1319, + "step": 12992 + }, + { + "epoch": 2.7244705388970436, + "grad_norm": 4.176612248536511, + "learning_rate": 4.3920203232396096e-07, + "loss": 0.1617, + "step": 12993 + }, + { + "epoch": 2.7246802264625707, + "grad_norm": 4.4207480177285685, + "learning_rate": 4.385387180358658e-07, + "loss": 0.1419, + "step": 12994 + }, + { + "epoch": 2.724889914028098, + "grad_norm": 4.1412969041830685, + "learning_rate": 4.378758937892447e-07, + "loss": 0.1101, + "step": 12995 + }, + { + "epoch": 2.7250996015936257, + "grad_norm": 5.035721551259342, + "learning_rate": 4.372135596180682e-07, + "loss": 0.1088, + "step": 12996 + }, + { + "epoch": 2.725309289159153, + "grad_norm": 6.026930671252997, + "learning_rate": 4.365517155562837e-07, + "loss": 0.1452, + "step": 12997 + }, + { + "epoch": 2.7255189767246804, + "grad_norm": 4.530969704085446, + "learning_rate": 4.3589036163781185e-07, + "loss": 0.1268, + "step": 12998 + }, + { + "epoch": 2.7257286642902074, + "grad_norm": 4.161639657181135, + "learning_rate": 4.3522949789654434e-07, + "loss": 0.1482, + "step": 12999 + }, + { + "epoch": 2.725938351855735, + "grad_norm": 3.515417560841747, + "learning_rate": 4.3456912436635414e-07, + "loss": 0.1391, + "step": 13000 + }, + { + "epoch": 2.726148039421262, + "grad_norm": 2.776970370880082, + "learning_rate": 4.339092410810908e-07, + "loss": 0.0796, + "step": 13001 + }, + { + "epoch": 2.7263577269867896, + "grad_norm": 4.28631678012816, + "learning_rate": 4.3324984807456613e-07, + "loss": 0.1286, + "step": 13002 + }, + { + "epoch": 2.726567414552317, + "grad_norm": 3.336565772482864, + "learning_rate": 4.32590945380581e-07, + "loss": 0.1048, + "step": 13003 + }, + { + "epoch": 2.726777102117844, + "grad_norm": 4.81039271802697, + "learning_rate": 4.3193253303290385e-07, + "loss": 0.1464, + "step": 13004 + }, + { + "epoch": 2.7269867896833717, + "grad_norm": 3.947673232245669, + "learning_rate": 4.312746110652799e-07, + "loss": 0.1222, + "step": 13005 + }, + { + "epoch": 2.7271964772488992, + "grad_norm": 4.288011507579844, + "learning_rate": 4.306171795114267e-07, + "loss": 0.0879, + "step": 13006 + }, + { + "epoch": 2.7274061648144263, + "grad_norm": 5.598255226914071, + "learning_rate": 4.2996023840504166e-07, + "loss": 0.1808, + "step": 13007 + }, + { + "epoch": 2.727615852379954, + "grad_norm": 3.9635573762059355, + "learning_rate": 4.293037877797912e-07, + "loss": 0.1116, + "step": 13008 + }, + { + "epoch": 2.7278255399454814, + "grad_norm": 4.1031806293641955, + "learning_rate": 4.286478276693229e-07, + "loss": 0.134, + "step": 13009 + }, + { + "epoch": 2.7280352275110085, + "grad_norm": 5.06334606984689, + "learning_rate": 4.2799235810725426e-07, + "loss": 0.1658, + "step": 13010 + }, + { + "epoch": 2.728244915076536, + "grad_norm": 3.719275434923693, + "learning_rate": 4.273373791271773e-07, + "loss": 0.1127, + "step": 13011 + }, + { + "epoch": 2.7284546026420635, + "grad_norm": 3.8670569350256696, + "learning_rate": 4.26682890762663e-07, + "loss": 0.1186, + "step": 13012 + }, + { + "epoch": 2.7286642902075906, + "grad_norm": 4.161937875338079, + "learning_rate": 4.2602889304725556e-07, + "loss": 0.1246, + "step": 13013 + }, + { + "epoch": 2.728873977773118, + "grad_norm": 3.3669231756630325, + "learning_rate": 4.253753860144727e-07, + "loss": 0.1011, + "step": 13014 + }, + { + "epoch": 2.7290836653386457, + "grad_norm": 3.6344981940066408, + "learning_rate": 4.2472236969780647e-07, + "loss": 0.1411, + "step": 13015 + }, + { + "epoch": 2.7292933529041727, + "grad_norm": 3.759262770552555, + "learning_rate": 4.240698441307278e-07, + "loss": 0.1518, + "step": 13016 + }, + { + "epoch": 2.7295030404697003, + "grad_norm": 4.196837596671605, + "learning_rate": 4.234178093466768e-07, + "loss": 0.1181, + "step": 13017 + }, + { + "epoch": 2.7297127280352274, + "grad_norm": 4.830146643294918, + "learning_rate": 4.2276626537907315e-07, + "loss": 0.1321, + "step": 13018 + }, + { + "epoch": 2.729922415600755, + "grad_norm": 3.9996514286075486, + "learning_rate": 4.2211521226131037e-07, + "loss": 0.1265, + "step": 13019 + }, + { + "epoch": 2.730132103166282, + "grad_norm": 4.86915354529976, + "learning_rate": 4.2146465002675274e-07, + "loss": 0.1527, + "step": 13020 + }, + { + "epoch": 2.7303417907318095, + "grad_norm": 4.009173848251095, + "learning_rate": 4.208145787087459e-07, + "loss": 0.1478, + "step": 13021 + }, + { + "epoch": 2.730551478297337, + "grad_norm": 2.928690850804952, + "learning_rate": 4.201649983406086e-07, + "loss": 0.1029, + "step": 13022 + }, + { + "epoch": 2.730761165862864, + "grad_norm": 3.122123351868127, + "learning_rate": 4.195159089556278e-07, + "loss": 0.0888, + "step": 13023 + }, + { + "epoch": 2.7309708534283916, + "grad_norm": 3.728001608132078, + "learning_rate": 4.188673105870733e-07, + "loss": 0.108, + "step": 13024 + }, + { + "epoch": 2.731180540993919, + "grad_norm": 4.0771304710357805, + "learning_rate": 4.1821920326818753e-07, + "loss": 0.1324, + "step": 13025 + }, + { + "epoch": 2.7313902285594462, + "grad_norm": 4.485887408354709, + "learning_rate": 4.175715870321861e-07, + "loss": 0.1598, + "step": 13026 + }, + { + "epoch": 2.7315999161249738, + "grad_norm": 3.3454733109642416, + "learning_rate": 4.1692446191226035e-07, + "loss": 0.1367, + "step": 13027 + }, + { + "epoch": 2.7318096036905013, + "grad_norm": 3.804249223834096, + "learning_rate": 4.16277827941578e-07, + "loss": 0.1286, + "step": 13028 + }, + { + "epoch": 2.7320192912560284, + "grad_norm": 3.686310386817152, + "learning_rate": 4.156316851532771e-07, + "loss": 0.1397, + "step": 13029 + }, + { + "epoch": 2.732228978821556, + "grad_norm": 4.350701934111513, + "learning_rate": 4.1498603358047787e-07, + "loss": 0.1476, + "step": 13030 + }, + { + "epoch": 2.7324386663870834, + "grad_norm": 3.9165162074522812, + "learning_rate": 4.1434087325626816e-07, + "loss": 0.1453, + "step": 13031 + }, + { + "epoch": 2.7326483539526105, + "grad_norm": 3.8880507342164403, + "learning_rate": 4.136962042137127e-07, + "loss": 0.1369, + "step": 13032 + }, + { + "epoch": 2.732858041518138, + "grad_norm": 4.557953933665127, + "learning_rate": 4.1305202648585397e-07, + "loss": 0.1279, + "step": 13033 + }, + { + "epoch": 2.7330677290836656, + "grad_norm": 3.930596345472188, + "learning_rate": 4.1240834010570774e-07, + "loss": 0.1439, + "step": 13034 + }, + { + "epoch": 2.7332774166491927, + "grad_norm": 4.608424277575705, + "learning_rate": 4.117651451062632e-07, + "loss": 0.1428, + "step": 13035 + }, + { + "epoch": 2.73348710421472, + "grad_norm": 4.064467420122084, + "learning_rate": 4.1112244152048285e-07, + "loss": 0.1503, + "step": 13036 + }, + { + "epoch": 2.7336967917802473, + "grad_norm": 5.437481104267108, + "learning_rate": 4.104802293813104e-07, + "loss": 0.146, + "step": 13037 + }, + { + "epoch": 2.733906479345775, + "grad_norm": 4.617233642135366, + "learning_rate": 4.098385087216561e-07, + "loss": 0.1382, + "step": 13038 + }, + { + "epoch": 2.7341161669113023, + "grad_norm": 4.295030715442799, + "learning_rate": 4.0919727957441366e-07, + "loss": 0.126, + "step": 13039 + }, + { + "epoch": 2.7343258544768294, + "grad_norm": 5.4199496086373316, + "learning_rate": 4.0855654197244464e-07, + "loss": 0.1482, + "step": 13040 + }, + { + "epoch": 2.734535542042357, + "grad_norm": 4.454260250891233, + "learning_rate": 4.079162959485883e-07, + "loss": 0.1479, + "step": 13041 + }, + { + "epoch": 2.734745229607884, + "grad_norm": 4.033597316713583, + "learning_rate": 4.072765415356572e-07, + "loss": 0.1303, + "step": 13042 + }, + { + "epoch": 2.7349549171734115, + "grad_norm": 4.6331033193228155, + "learning_rate": 4.066372787664441e-07, + "loss": 0.1487, + "step": 13043 + }, + { + "epoch": 2.735164604738939, + "grad_norm": 5.362642571081246, + "learning_rate": 4.0599850767370607e-07, + "loss": 0.1663, + "step": 13044 + }, + { + "epoch": 2.735374292304466, + "grad_norm": 2.8049229516249965, + "learning_rate": 4.0536022829018476e-07, + "loss": 0.0842, + "step": 13045 + }, + { + "epoch": 2.7355839798699937, + "grad_norm": 3.9739064855732606, + "learning_rate": 4.04722440648595e-07, + "loss": 0.1513, + "step": 13046 + }, + { + "epoch": 2.735793667435521, + "grad_norm": 4.188148948106658, + "learning_rate": 4.0408514478162075e-07, + "loss": 0.1345, + "step": 13047 + }, + { + "epoch": 2.7360033550010483, + "grad_norm": 4.483543983162668, + "learning_rate": 4.0344834072192585e-07, + "loss": 0.1226, + "step": 13048 + }, + { + "epoch": 2.736213042566576, + "grad_norm": 4.380791260737828, + "learning_rate": 4.028120285021486e-07, + "loss": 0.1314, + "step": 13049 + }, + { + "epoch": 2.7364227301321034, + "grad_norm": 5.135582382637659, + "learning_rate": 4.0217620815489964e-07, + "loss": 0.205, + "step": 13050 + }, + { + "epoch": 2.7366324176976304, + "grad_norm": 4.049195953770861, + "learning_rate": 4.015408797127684e-07, + "loss": 0.128, + "step": 13051 + }, + { + "epoch": 2.736842105263158, + "grad_norm": 3.517403727679536, + "learning_rate": 4.009060432083134e-07, + "loss": 0.1277, + "step": 13052 + }, + { + "epoch": 2.7370517928286855, + "grad_norm": 4.828535459679986, + "learning_rate": 4.0027169867407177e-07, + "loss": 0.1706, + "step": 13053 + }, + { + "epoch": 2.7372614803942126, + "grad_norm": 4.011334635681633, + "learning_rate": 3.996378461425554e-07, + "loss": 0.1162, + "step": 13054 + }, + { + "epoch": 2.73747116795974, + "grad_norm": 3.4721676360167564, + "learning_rate": 3.990044856462527e-07, + "loss": 0.1127, + "step": 13055 + }, + { + "epoch": 2.737680855525267, + "grad_norm": 4.407681619585082, + "learning_rate": 3.9837161721762106e-07, + "loss": 0.1211, + "step": 13056 + }, + { + "epoch": 2.7378905430907947, + "grad_norm": 4.229210439689979, + "learning_rate": 3.9773924088909566e-07, + "loss": 0.1221, + "step": 13057 + }, + { + "epoch": 2.7381002306563222, + "grad_norm": 3.8301973295774, + "learning_rate": 3.971073566930894e-07, + "loss": 0.1256, + "step": 13058 + }, + { + "epoch": 2.7383099182218493, + "grad_norm": 4.465339519646707, + "learning_rate": 3.9647596466198757e-07, + "loss": 0.1249, + "step": 13059 + }, + { + "epoch": 2.738519605787377, + "grad_norm": 3.1818519083640497, + "learning_rate": 3.9584506482814864e-07, + "loss": 0.1154, + "step": 13060 + }, + { + "epoch": 2.738729293352904, + "grad_norm": 4.987861233438751, + "learning_rate": 3.952146572239068e-07, + "loss": 0.1227, + "step": 13061 + }, + { + "epoch": 2.7389389809184315, + "grad_norm": 4.510606008019649, + "learning_rate": 3.945847418815729e-07, + "loss": 0.1, + "step": 13062 + }, + { + "epoch": 2.739148668483959, + "grad_norm": 3.7856506544172133, + "learning_rate": 3.9395531883343105e-07, + "loss": 0.1297, + "step": 13063 + }, + { + "epoch": 2.739358356049486, + "grad_norm": 3.5064900545571827, + "learning_rate": 3.9332638811173885e-07, + "loss": 0.1263, + "step": 13064 + }, + { + "epoch": 2.7395680436150136, + "grad_norm": 4.160328132812707, + "learning_rate": 3.9269794974873264e-07, + "loss": 0.125, + "step": 13065 + }, + { + "epoch": 2.739777731180541, + "grad_norm": 4.585377453459852, + "learning_rate": 3.920700037766179e-07, + "loss": 0.1321, + "step": 13066 + }, + { + "epoch": 2.739987418746068, + "grad_norm": 3.5077687544314538, + "learning_rate": 3.9144255022758203e-07, + "loss": 0.1344, + "step": 13067 + }, + { + "epoch": 2.7401971063115957, + "grad_norm": 6.688263692432872, + "learning_rate": 3.9081558913377947e-07, + "loss": 0.1662, + "step": 13068 + }, + { + "epoch": 2.7404067938771233, + "grad_norm": 3.761823956359265, + "learning_rate": 3.9018912052734335e-07, + "loss": 0.1213, + "step": 13069 + }, + { + "epoch": 2.7406164814426504, + "grad_norm": 3.784821829057151, + "learning_rate": 3.8956314444038135e-07, + "loss": 0.1189, + "step": 13070 + }, + { + "epoch": 2.740826169008178, + "grad_norm": 4.316685418555946, + "learning_rate": 3.8893766090497886e-07, + "loss": 0.1209, + "step": 13071 + }, + { + "epoch": 2.7410358565737054, + "grad_norm": 4.130071273999352, + "learning_rate": 3.8831266995319027e-07, + "loss": 0.1089, + "step": 13072 + }, + { + "epoch": 2.7412455441392325, + "grad_norm": 5.564957271679498, + "learning_rate": 3.8768817161704776e-07, + "loss": 0.1665, + "step": 13073 + }, + { + "epoch": 2.74145523170476, + "grad_norm": 6.426699029027333, + "learning_rate": 3.870641659285579e-07, + "loss": 0.1848, + "step": 13074 + }, + { + "epoch": 2.7416649192702875, + "grad_norm": 4.390466017640438, + "learning_rate": 3.864406529197018e-07, + "loss": 0.151, + "step": 13075 + }, + { + "epoch": 2.7418746068358146, + "grad_norm": 3.3542251218238226, + "learning_rate": 3.858176326224372e-07, + "loss": 0.0865, + "step": 13076 + }, + { + "epoch": 2.742084294401342, + "grad_norm": 3.791685912559164, + "learning_rate": 3.8519510506869417e-07, + "loss": 0.1176, + "step": 13077 + }, + { + "epoch": 2.7422939819668692, + "grad_norm": 5.21462192045731, + "learning_rate": 3.8457307029037716e-07, + "loss": 0.1649, + "step": 13078 + }, + { + "epoch": 2.7425036695323968, + "grad_norm": 4.529877945127559, + "learning_rate": 3.8395152831936734e-07, + "loss": 0.1316, + "step": 13079 + }, + { + "epoch": 2.742713357097924, + "grad_norm": 3.7712872183011563, + "learning_rate": 3.833304791875214e-07, + "loss": 0.1273, + "step": 13080 + }, + { + "epoch": 2.7429230446634514, + "grad_norm": 3.548834989663908, + "learning_rate": 3.827099229266662e-07, + "loss": 0.103, + "step": 13081 + }, + { + "epoch": 2.743132732228979, + "grad_norm": 5.361699992626051, + "learning_rate": 3.8208985956860736e-07, + "loss": 0.1374, + "step": 13082 + }, + { + "epoch": 2.743342419794506, + "grad_norm": 3.719129501816349, + "learning_rate": 3.8147028914512494e-07, + "loss": 0.1011, + "step": 13083 + }, + { + "epoch": 2.7435521073600335, + "grad_norm": 5.664017426870495, + "learning_rate": 3.808512116879737e-07, + "loss": 0.1705, + "step": 13084 + }, + { + "epoch": 2.743761794925561, + "grad_norm": 4.727731389146657, + "learning_rate": 3.8023262722888033e-07, + "loss": 0.1709, + "step": 13085 + }, + { + "epoch": 2.743971482491088, + "grad_norm": 5.1136299412198145, + "learning_rate": 3.7961453579955064e-07, + "loss": 0.1307, + "step": 13086 + }, + { + "epoch": 2.7441811700566157, + "grad_norm": 4.169705775287418, + "learning_rate": 3.7899693743166043e-07, + "loss": 0.1354, + "step": 13087 + }, + { + "epoch": 2.744390857622143, + "grad_norm": 4.510827347190437, + "learning_rate": 3.783798321568655e-07, + "loss": 0.1526, + "step": 13088 + }, + { + "epoch": 2.7446005451876703, + "grad_norm": 4.172586836498216, + "learning_rate": 3.7776322000679156e-07, + "loss": 0.1028, + "step": 13089 + }, + { + "epoch": 2.744810232753198, + "grad_norm": 3.943851788797182, + "learning_rate": 3.771471010130401e-07, + "loss": 0.1451, + "step": 13090 + }, + { + "epoch": 2.7450199203187253, + "grad_norm": 3.8749849758751007, + "learning_rate": 3.7653147520719024e-07, + "loss": 0.1375, + "step": 13091 + }, + { + "epoch": 2.7452296078842524, + "grad_norm": 4.9005040556261, + "learning_rate": 3.759163426207935e-07, + "loss": 0.1407, + "step": 13092 + }, + { + "epoch": 2.74543929544978, + "grad_norm": 3.898923187637183, + "learning_rate": 3.7530170328537674e-07, + "loss": 0.1416, + "step": 13093 + }, + { + "epoch": 2.7456489830153075, + "grad_norm": 3.9399313927751427, + "learning_rate": 3.7468755723244046e-07, + "loss": 0.12, + "step": 13094 + }, + { + "epoch": 2.7458586705808345, + "grad_norm": 6.3920146592992655, + "learning_rate": 3.740739044934616e-07, + "loss": 0.1696, + "step": 13095 + }, + { + "epoch": 2.746068358146362, + "grad_norm": 3.397969669856621, + "learning_rate": 3.734607450998895e-07, + "loss": 0.1066, + "step": 13096 + }, + { + "epoch": 2.746278045711889, + "grad_norm": 3.9531536301510926, + "learning_rate": 3.7284807908315235e-07, + "loss": 0.113, + "step": 13097 + }, + { + "epoch": 2.7464877332774167, + "grad_norm": 4.779758917615418, + "learning_rate": 3.7223590647464833e-07, + "loss": 0.1656, + "step": 13098 + }, + { + "epoch": 2.7466974208429438, + "grad_norm": 3.7435038922630723, + "learning_rate": 3.7162422730575133e-07, + "loss": 0.1249, + "step": 13099 + }, + { + "epoch": 2.7469071084084713, + "grad_norm": 3.7160965670180497, + "learning_rate": 3.710130416078117e-07, + "loss": 0.115, + "step": 13100 + }, + { + "epoch": 2.747116795973999, + "grad_norm": 4.142113055418561, + "learning_rate": 3.704023494121567e-07, + "loss": 0.1099, + "step": 13101 + }, + { + "epoch": 2.747326483539526, + "grad_norm": 3.197186020749749, + "learning_rate": 3.6979215075008123e-07, + "loss": 0.112, + "step": 13102 + }, + { + "epoch": 2.7475361711050534, + "grad_norm": 3.588662667123561, + "learning_rate": 3.6918244565286033e-07, + "loss": 0.1217, + "step": 13103 + }, + { + "epoch": 2.747745858670581, + "grad_norm": 3.5467705783460692, + "learning_rate": 3.685732341517445e-07, + "loss": 0.1136, + "step": 13104 + }, + { + "epoch": 2.747955546236108, + "grad_norm": 5.1426684179322315, + "learning_rate": 3.679645162779555e-07, + "loss": 0.1472, + "step": 13105 + }, + { + "epoch": 2.7481652338016356, + "grad_norm": 6.6765549065229575, + "learning_rate": 3.673562920626894e-07, + "loss": 0.1398, + "step": 13106 + }, + { + "epoch": 2.748374921367163, + "grad_norm": 3.082389406644875, + "learning_rate": 3.6674856153712137e-07, + "loss": 0.0969, + "step": 13107 + }, + { + "epoch": 2.74858460893269, + "grad_norm": 4.900001066246391, + "learning_rate": 3.6614132473239636e-07, + "loss": 0.1569, + "step": 13108 + }, + { + "epoch": 2.7487942964982177, + "grad_norm": 3.4354965307657737, + "learning_rate": 3.655345816796385e-07, + "loss": 0.1186, + "step": 13109 + }, + { + "epoch": 2.7490039840637452, + "grad_norm": 3.9565278578734144, + "learning_rate": 3.6492833240994395e-07, + "loss": 0.1305, + "step": 13110 + }, + { + "epoch": 2.7492136716292723, + "grad_norm": 5.237203355733942, + "learning_rate": 3.6432257695438235e-07, + "loss": 0.1631, + "step": 13111 + }, + { + "epoch": 2.7494233591948, + "grad_norm": 3.5204876778776804, + "learning_rate": 3.63717315344001e-07, + "loss": 0.1264, + "step": 13112 + }, + { + "epoch": 2.7496330467603274, + "grad_norm": 4.89761190286342, + "learning_rate": 3.6311254760982074e-07, + "loss": 0.1841, + "step": 13113 + }, + { + "epoch": 2.7498427343258545, + "grad_norm": 3.1440827744403914, + "learning_rate": 3.625082737828378e-07, + "loss": 0.1168, + "step": 13114 + }, + { + "epoch": 2.750052421891382, + "grad_norm": 4.355234658361031, + "learning_rate": 3.619044938940197e-07, + "loss": 0.1101, + "step": 13115 + }, + { + "epoch": 2.750262109456909, + "grad_norm": 5.427987352236284, + "learning_rate": 3.613012079743128e-07, + "loss": 0.1426, + "step": 13116 + }, + { + "epoch": 2.7504717970224366, + "grad_norm": 4.228338932327801, + "learning_rate": 3.606984160546378e-07, + "loss": 0.1167, + "step": 13117 + }, + { + "epoch": 2.7506814845879637, + "grad_norm": 3.7465332245080663, + "learning_rate": 3.6009611816588685e-07, + "loss": 0.125, + "step": 13118 + }, + { + "epoch": 2.750891172153491, + "grad_norm": 4.250225679873633, + "learning_rate": 3.5949431433892846e-07, + "loss": 0.1267, + "step": 13119 + }, + { + "epoch": 2.7511008597190187, + "grad_norm": 3.92420688975659, + "learning_rate": 3.588930046046091e-07, + "loss": 0.133, + "step": 13120 + }, + { + "epoch": 2.751310547284546, + "grad_norm": 4.063127554739173, + "learning_rate": 3.58292188993743e-07, + "loss": 0.1219, + "step": 13121 + }, + { + "epoch": 2.7515202348500734, + "grad_norm": 4.375136017281564, + "learning_rate": 3.576918675371255e-07, + "loss": 0.1196, + "step": 13122 + }, + { + "epoch": 2.751729922415601, + "grad_norm": 4.5861848087416615, + "learning_rate": 3.5709204026552425e-07, + "loss": 0.1233, + "step": 13123 + }, + { + "epoch": 2.751939609981128, + "grad_norm": 3.6517934292796053, + "learning_rate": 3.564927072096802e-07, + "loss": 0.122, + "step": 13124 + }, + { + "epoch": 2.7521492975466555, + "grad_norm": 2.9779729000042923, + "learning_rate": 3.5589386840031105e-07, + "loss": 0.112, + "step": 13125 + }, + { + "epoch": 2.752358985112183, + "grad_norm": 5.492115655330109, + "learning_rate": 3.5529552386810773e-07, + "loss": 0.1505, + "step": 13126 + }, + { + "epoch": 2.75256867267771, + "grad_norm": 3.271788870604396, + "learning_rate": 3.5469767364373577e-07, + "loss": 0.097, + "step": 13127 + }, + { + "epoch": 2.7527783602432376, + "grad_norm": 7.163102690944309, + "learning_rate": 3.5410031775783727e-07, + "loss": 0.1506, + "step": 13128 + }, + { + "epoch": 2.752988047808765, + "grad_norm": 4.274985455413665, + "learning_rate": 3.5350345624102886e-07, + "loss": 0.1164, + "step": 13129 + }, + { + "epoch": 2.7531977353742922, + "grad_norm": 3.102493697348989, + "learning_rate": 3.529070891238984e-07, + "loss": 0.0944, + "step": 13130 + }, + { + "epoch": 2.7534074229398198, + "grad_norm": 5.902044342160964, + "learning_rate": 3.5231121643701127e-07, + "loss": 0.1502, + "step": 13131 + }, + { + "epoch": 2.7536171105053473, + "grad_norm": 4.372388953363877, + "learning_rate": 3.5171583821090873e-07, + "loss": 0.1171, + "step": 13132 + }, + { + "epoch": 2.7538267980708744, + "grad_norm": 5.440613931656129, + "learning_rate": 3.511209544761018e-07, + "loss": 0.1643, + "step": 13133 + }, + { + "epoch": 2.754036485636402, + "grad_norm": 4.084115200852558, + "learning_rate": 3.505265652630818e-07, + "loss": 0.1381, + "step": 13134 + }, + { + "epoch": 2.754246173201929, + "grad_norm": 4.1777709926562725, + "learning_rate": 3.4993267060231206e-07, + "loss": 0.1338, + "step": 13135 + }, + { + "epoch": 2.7544558607674565, + "grad_norm": 4.6997194254327805, + "learning_rate": 3.4933927052422936e-07, + "loss": 0.1102, + "step": 13136 + }, + { + "epoch": 2.7546655483329836, + "grad_norm": 3.562132076821083, + "learning_rate": 3.487463650592471e-07, + "loss": 0.1091, + "step": 13137 + }, + { + "epoch": 2.754875235898511, + "grad_norm": 3.860834567962666, + "learning_rate": 3.481539542377532e-07, + "loss": 0.1294, + "step": 13138 + }, + { + "epoch": 2.7550849234640387, + "grad_norm": 5.435384765939473, + "learning_rate": 3.475620380901101e-07, + "loss": 0.1501, + "step": 13139 + }, + { + "epoch": 2.7552946110295657, + "grad_norm": 3.9516121408780593, + "learning_rate": 3.4697061664665354e-07, + "loss": 0.1336, + "step": 13140 + }, + { + "epoch": 2.7555042985950933, + "grad_norm": 4.011476632573279, + "learning_rate": 3.463796899376948e-07, + "loss": 0.1416, + "step": 13141 + }, + { + "epoch": 2.755713986160621, + "grad_norm": 4.517091721202496, + "learning_rate": 3.457892579935218e-07, + "loss": 0.1137, + "step": 13142 + }, + { + "epoch": 2.755923673726148, + "grad_norm": 4.630549011987817, + "learning_rate": 3.4519932084439154e-07, + "loss": 0.1625, + "step": 13143 + }, + { + "epoch": 2.7561333612916754, + "grad_norm": 3.5814864642386772, + "learning_rate": 3.446098785205432e-07, + "loss": 0.1199, + "step": 13144 + }, + { + "epoch": 2.756343048857203, + "grad_norm": 5.190948645736391, + "learning_rate": 3.440209310521825e-07, + "loss": 0.1736, + "step": 13145 + }, + { + "epoch": 2.75655273642273, + "grad_norm": 3.3138893189654897, + "learning_rate": 3.4343247846949976e-07, + "loss": 0.1143, + "step": 13146 + }, + { + "epoch": 2.7567624239882575, + "grad_norm": 3.583057535002732, + "learning_rate": 3.4284452080264985e-07, + "loss": 0.1269, + "step": 13147 + }, + { + "epoch": 2.756972111553785, + "grad_norm": 4.282476344919575, + "learning_rate": 3.4225705808176633e-07, + "loss": 0.1437, + "step": 13148 + }, + { + "epoch": 2.757181799119312, + "grad_norm": 3.794478606518034, + "learning_rate": 3.4167009033695963e-07, + "loss": 0.094, + "step": 13149 + }, + { + "epoch": 2.7573914866848397, + "grad_norm": 6.620890247824128, + "learning_rate": 3.4108361759831345e-07, + "loss": 0.1573, + "step": 13150 + }, + { + "epoch": 2.757601174250367, + "grad_norm": 3.8967377074189216, + "learning_rate": 3.404976398958837e-07, + "loss": 0.1208, + "step": 13151 + }, + { + "epoch": 2.7578108618158943, + "grad_norm": 4.651315657966747, + "learning_rate": 3.399121572597031e-07, + "loss": 0.128, + "step": 13152 + }, + { + "epoch": 2.758020549381422, + "grad_norm": 4.711174221559803, + "learning_rate": 3.3932716971977976e-07, + "loss": 0.1253, + "step": 13153 + }, + { + "epoch": 2.758230236946949, + "grad_norm": 4.4749655339351255, + "learning_rate": 3.3874267730609314e-07, + "loss": 0.163, + "step": 13154 + }, + { + "epoch": 2.7584399245124764, + "grad_norm": 5.135842414627989, + "learning_rate": 3.381586800486025e-07, + "loss": 0.1334, + "step": 13155 + }, + { + "epoch": 2.7586496120780035, + "grad_norm": 3.933703983546546, + "learning_rate": 3.375751779772374e-07, + "loss": 0.1396, + "step": 13156 + }, + { + "epoch": 2.758859299643531, + "grad_norm": 4.354213085627745, + "learning_rate": 3.3699217112190153e-07, + "loss": 0.1366, + "step": 13157 + }, + { + "epoch": 2.7590689872090586, + "grad_norm": 4.499426226080546, + "learning_rate": 3.364096595124766e-07, + "loss": 0.1251, + "step": 13158 + }, + { + "epoch": 2.7592786747745857, + "grad_norm": 4.374934836204621, + "learning_rate": 3.3582764317881986e-07, + "loss": 0.1322, + "step": 13159 + }, + { + "epoch": 2.759488362340113, + "grad_norm": 3.9063781644538143, + "learning_rate": 3.3524612215075524e-07, + "loss": 0.1419, + "step": 13160 + }, + { + "epoch": 2.7596980499056407, + "grad_norm": 5.475052792974205, + "learning_rate": 3.3466509645809e-07, + "loss": 0.1405, + "step": 13161 + }, + { + "epoch": 2.759907737471168, + "grad_norm": 4.366633109627006, + "learning_rate": 3.3408456613060245e-07, + "loss": 0.1328, + "step": 13162 + }, + { + "epoch": 2.7601174250366953, + "grad_norm": 6.7125982497515375, + "learning_rate": 3.335045311980467e-07, + "loss": 0.2084, + "step": 13163 + }, + { + "epoch": 2.760327112602223, + "grad_norm": 3.876246424795854, + "learning_rate": 3.329249916901478e-07, + "loss": 0.1327, + "step": 13164 + }, + { + "epoch": 2.76053680016775, + "grad_norm": 3.5167242501898626, + "learning_rate": 3.323459476366109e-07, + "loss": 0.1036, + "step": 13165 + }, + { + "epoch": 2.7607464877332775, + "grad_norm": 6.586141742223875, + "learning_rate": 3.317673990671111e-07, + "loss": 0.1746, + "step": 13166 + }, + { + "epoch": 2.760956175298805, + "grad_norm": 3.8549413323907205, + "learning_rate": 3.311893460113014e-07, + "loss": 0.1316, + "step": 13167 + }, + { + "epoch": 2.761165862864332, + "grad_norm": 4.710660994142496, + "learning_rate": 3.3061178849880695e-07, + "loss": 0.1774, + "step": 13168 + }, + { + "epoch": 2.7613755504298596, + "grad_norm": 4.45990056994574, + "learning_rate": 3.300347265592285e-07, + "loss": 0.1628, + "step": 13169 + }, + { + "epoch": 2.761585237995387, + "grad_norm": 4.370817919551644, + "learning_rate": 3.2945816022214247e-07, + "loss": 0.1406, + "step": 13170 + }, + { + "epoch": 2.761794925560914, + "grad_norm": 3.2183257385117368, + "learning_rate": 3.2888208951709966e-07, + "loss": 0.1176, + "step": 13171 + }, + { + "epoch": 2.7620046131264417, + "grad_norm": 3.627993035052207, + "learning_rate": 3.2830651447362304e-07, + "loss": 0.1095, + "step": 13172 + }, + { + "epoch": 2.762214300691969, + "grad_norm": 4.313726478601724, + "learning_rate": 3.277314351212102e-07, + "loss": 0.1335, + "step": 13173 + }, + { + "epoch": 2.7624239882574964, + "grad_norm": 3.8086468137810288, + "learning_rate": 3.2715685148933976e-07, + "loss": 0.1191, + "step": 13174 + }, + { + "epoch": 2.7626336758230234, + "grad_norm": 3.009829538989232, + "learning_rate": 3.265827636074548e-07, + "loss": 0.081, + "step": 13175 + }, + { + "epoch": 2.762843363388551, + "grad_norm": 4.338289492860503, + "learning_rate": 3.260091715049829e-07, + "loss": 0.1237, + "step": 13176 + }, + { + "epoch": 2.7630530509540785, + "grad_norm": 4.242690293121193, + "learning_rate": 3.2543607521131835e-07, + "loss": 0.1333, + "step": 13177 + }, + { + "epoch": 2.7632627385196056, + "grad_norm": 4.337602066338219, + "learning_rate": 3.248634747558366e-07, + "loss": 0.1307, + "step": 13178 + }, + { + "epoch": 2.763472426085133, + "grad_norm": 5.4042180155193495, + "learning_rate": 3.242913701678796e-07, + "loss": 0.1744, + "step": 13179 + }, + { + "epoch": 2.7636821136506606, + "grad_norm": 4.447769647449593, + "learning_rate": 3.2371976147677396e-07, + "loss": 0.1477, + "step": 13180 + }, + { + "epoch": 2.7638918012161877, + "grad_norm": 3.8935621359356585, + "learning_rate": 3.23148648711813e-07, + "loss": 0.1319, + "step": 13181 + }, + { + "epoch": 2.7641014887817152, + "grad_norm": 4.072471285733634, + "learning_rate": 3.2257803190226534e-07, + "loss": 0.1161, + "step": 13182 + }, + { + "epoch": 2.7643111763472428, + "grad_norm": 4.554949117515458, + "learning_rate": 3.2200791107737884e-07, + "loss": 0.1426, + "step": 13183 + }, + { + "epoch": 2.76452086391277, + "grad_norm": 3.8996627795410435, + "learning_rate": 3.2143828626637564e-07, + "loss": 0.1214, + "step": 13184 + }, + { + "epoch": 2.7647305514782974, + "grad_norm": 3.2681467738030805, + "learning_rate": 3.2086915749844347e-07, + "loss": 0.0944, + "step": 13185 + }, + { + "epoch": 2.764940239043825, + "grad_norm": 3.6782714882792775, + "learning_rate": 3.2030052480275464e-07, + "loss": 0.0849, + "step": 13186 + }, + { + "epoch": 2.765149926609352, + "grad_norm": 3.9787891303993788, + "learning_rate": 3.1973238820845465e-07, + "loss": 0.1022, + "step": 13187 + }, + { + "epoch": 2.7653596141748795, + "grad_norm": 5.513086441261112, + "learning_rate": 3.1916474774465797e-07, + "loss": 0.1556, + "step": 13188 + }, + { + "epoch": 2.765569301740407, + "grad_norm": 4.42266094987165, + "learning_rate": 3.185976034404581e-07, + "loss": 0.1327, + "step": 13189 + }, + { + "epoch": 2.765778989305934, + "grad_norm": 2.774821760792061, + "learning_rate": 3.1803095532492387e-07, + "loss": 0.0775, + "step": 13190 + }, + { + "epoch": 2.7659886768714617, + "grad_norm": 4.7093929105051355, + "learning_rate": 3.174648034270944e-07, + "loss": 0.1502, + "step": 13191 + }, + { + "epoch": 2.7661983644369887, + "grad_norm": 5.950973909952062, + "learning_rate": 3.168991477759886e-07, + "loss": 0.1964, + "step": 13192 + }, + { + "epoch": 2.7664080520025163, + "grad_norm": 4.124575489567756, + "learning_rate": 3.163339884005956e-07, + "loss": 0.1224, + "step": 13193 + }, + { + "epoch": 2.7666177395680434, + "grad_norm": 5.701105732450794, + "learning_rate": 3.157693253298799e-07, + "loss": 0.151, + "step": 13194 + }, + { + "epoch": 2.766827427133571, + "grad_norm": 4.971578767651743, + "learning_rate": 3.15205158592784e-07, + "loss": 0.1753, + "step": 13195 + }, + { + "epoch": 2.7670371146990984, + "grad_norm": 4.406002302237589, + "learning_rate": 3.146414882182214e-07, + "loss": 0.1239, + "step": 13196 + }, + { + "epoch": 2.7672468022646255, + "grad_norm": 5.1151282844597, + "learning_rate": 3.1407831423508006e-07, + "loss": 0.1418, + "step": 13197 + }, + { + "epoch": 2.767456489830153, + "grad_norm": 3.9696552442640556, + "learning_rate": 3.1351563667222474e-07, + "loss": 0.1536, + "step": 13198 + }, + { + "epoch": 2.7676661773956805, + "grad_norm": 4.228064521135329, + "learning_rate": 3.1295345555849344e-07, + "loss": 0.1322, + "step": 13199 + }, + { + "epoch": 2.7678758649612076, + "grad_norm": 4.404335508896068, + "learning_rate": 3.123917709226987e-07, + "loss": 0.1375, + "step": 13200 + }, + { + "epoch": 2.768085552526735, + "grad_norm": 4.354174268486226, + "learning_rate": 3.118305827936285e-07, + "loss": 0.1489, + "step": 13201 + }, + { + "epoch": 2.7682952400922627, + "grad_norm": 3.7798264144815943, + "learning_rate": 3.112698912000445e-07, + "loss": 0.1124, + "step": 13202 + }, + { + "epoch": 2.7685049276577898, + "grad_norm": 4.033037960351409, + "learning_rate": 3.1070969617068123e-07, + "loss": 0.108, + "step": 13203 + }, + { + "epoch": 2.7687146152233173, + "grad_norm": 4.606385702313986, + "learning_rate": 3.1014999773425257e-07, + "loss": 0.1498, + "step": 13204 + }, + { + "epoch": 2.768924302788845, + "grad_norm": 3.9903196210433194, + "learning_rate": 3.095907959194433e-07, + "loss": 0.1197, + "step": 13205 + }, + { + "epoch": 2.769133990354372, + "grad_norm": 5.000950281356826, + "learning_rate": 3.0903209075491046e-07, + "loss": 0.1409, + "step": 13206 + }, + { + "epoch": 2.7693436779198994, + "grad_norm": 4.097114424889244, + "learning_rate": 3.084738822692923e-07, + "loss": 0.1381, + "step": 13207 + }, + { + "epoch": 2.769553365485427, + "grad_norm": 4.2408223082803405, + "learning_rate": 3.0791617049119816e-07, + "loss": 0.1668, + "step": 13208 + }, + { + "epoch": 2.769763053050954, + "grad_norm": 4.701739691559669, + "learning_rate": 3.0735895544920957e-07, + "loss": 0.161, + "step": 13209 + }, + { + "epoch": 2.7699727406164816, + "grad_norm": 4.374463970052749, + "learning_rate": 3.0680223717188483e-07, + "loss": 0.1005, + "step": 13210 + }, + { + "epoch": 2.7701824281820087, + "grad_norm": 3.731434750810305, + "learning_rate": 3.062460156877578e-07, + "loss": 0.121, + "step": 13211 + }, + { + "epoch": 2.770392115747536, + "grad_norm": 3.6962792969628846, + "learning_rate": 3.056902910253345e-07, + "loss": 0.1008, + "step": 13212 + }, + { + "epoch": 2.7706018033130633, + "grad_norm": 3.575788771358574, + "learning_rate": 3.0513506321310007e-07, + "loss": 0.1198, + "step": 13213 + }, + { + "epoch": 2.770811490878591, + "grad_norm": 6.130992160026336, + "learning_rate": 3.0458033227950714e-07, + "loss": 0.1674, + "step": 13214 + }, + { + "epoch": 2.7710211784441183, + "grad_norm": 4.110630643355918, + "learning_rate": 3.0402609825298744e-07, + "loss": 0.1338, + "step": 13215 + }, + { + "epoch": 2.7712308660096454, + "grad_norm": 4.421286954444065, + "learning_rate": 3.0347236116194724e-07, + "loss": 0.1642, + "step": 13216 + }, + { + "epoch": 2.771440553575173, + "grad_norm": 3.642610598744105, + "learning_rate": 3.0291912103476705e-07, + "loss": 0.1415, + "step": 13217 + }, + { + "epoch": 2.7716502411407005, + "grad_norm": 4.492624602509829, + "learning_rate": 3.0236637789980094e-07, + "loss": 0.1837, + "step": 13218 + }, + { + "epoch": 2.7718599287062275, + "grad_norm": 4.084580975000896, + "learning_rate": 3.0181413178537623e-07, + "loss": 0.1328, + "step": 13219 + }, + { + "epoch": 2.772069616271755, + "grad_norm": 4.221260181438162, + "learning_rate": 3.012623827197991e-07, + "loss": 0.1304, + "step": 13220 + }, + { + "epoch": 2.7722793038372826, + "grad_norm": 2.6678786441456386, + "learning_rate": 3.0071113073134595e-07, + "loss": 0.0824, + "step": 13221 + }, + { + "epoch": 2.7724889914028097, + "grad_norm": 4.386237886871211, + "learning_rate": 3.001603758482696e-07, + "loss": 0.1636, + "step": 13222 + }, + { + "epoch": 2.772698678968337, + "grad_norm": 5.483496072159809, + "learning_rate": 2.996101180987987e-07, + "loss": 0.1284, + "step": 13223 + }, + { + "epoch": 2.7729083665338647, + "grad_norm": 4.901220317174622, + "learning_rate": 2.990603575111328e-07, + "loss": 0.1454, + "step": 13224 + }, + { + "epoch": 2.773118054099392, + "grad_norm": 4.096252670016384, + "learning_rate": 2.9851109411344946e-07, + "loss": 0.1577, + "step": 13225 + }, + { + "epoch": 2.7733277416649194, + "grad_norm": 3.1037193506652656, + "learning_rate": 2.979623279338983e-07, + "loss": 0.0851, + "step": 13226 + }, + { + "epoch": 2.773537429230447, + "grad_norm": 4.193902508319986, + "learning_rate": 2.974140590006058e-07, + "loss": 0.105, + "step": 13227 + }, + { + "epoch": 2.773747116795974, + "grad_norm": 5.239245632473883, + "learning_rate": 2.9686628734166944e-07, + "loss": 0.1914, + "step": 13228 + }, + { + "epoch": 2.7739568043615015, + "grad_norm": 3.965351670363004, + "learning_rate": 2.963190129851667e-07, + "loss": 0.1016, + "step": 13229 + }, + { + "epoch": 2.7741664919270286, + "grad_norm": 3.810717325074233, + "learning_rate": 2.957722359591453e-07, + "loss": 0.1267, + "step": 13230 + }, + { + "epoch": 2.774376179492556, + "grad_norm": 4.540066015976268, + "learning_rate": 2.95225956291626e-07, + "loss": 0.1536, + "step": 13231 + }, + { + "epoch": 2.774585867058083, + "grad_norm": 5.6886808235922155, + "learning_rate": 2.9468017401061087e-07, + "loss": 0.1417, + "step": 13232 + }, + { + "epoch": 2.7747955546236107, + "grad_norm": 3.6858062081808116, + "learning_rate": 2.941348891440676e-07, + "loss": 0.1103, + "step": 13233 + }, + { + "epoch": 2.7750052421891382, + "grad_norm": 2.8136004089160527, + "learning_rate": 2.9359010171994605e-07, + "loss": 0.0968, + "step": 13234 + }, + { + "epoch": 2.7752149297546653, + "grad_norm": 3.7777329352138147, + "learning_rate": 2.930458117661683e-07, + "loss": 0.1353, + "step": 13235 + }, + { + "epoch": 2.775424617320193, + "grad_norm": 3.324504828568484, + "learning_rate": 2.925020193106265e-07, + "loss": 0.116, + "step": 13236 + }, + { + "epoch": 2.7756343048857204, + "grad_norm": 4.8591723309736485, + "learning_rate": 2.919587243811928e-07, + "loss": 0.1537, + "step": 13237 + }, + { + "epoch": 2.7758439924512475, + "grad_norm": 5.177073868079106, + "learning_rate": 2.9141592700571265e-07, + "loss": 0.1665, + "step": 13238 + }, + { + "epoch": 2.776053680016775, + "grad_norm": 5.356782641086413, + "learning_rate": 2.908736272120061e-07, + "loss": 0.1497, + "step": 13239 + }, + { + "epoch": 2.7762633675823025, + "grad_norm": 5.354002925526554, + "learning_rate": 2.903318250278642e-07, + "loss": 0.1611, + "step": 13240 + }, + { + "epoch": 2.7764730551478296, + "grad_norm": 4.872058242263722, + "learning_rate": 2.8979052048105694e-07, + "loss": 0.1889, + "step": 13241 + }, + { + "epoch": 2.776682742713357, + "grad_norm": 4.386710451305927, + "learning_rate": 2.8924971359932996e-07, + "loss": 0.142, + "step": 13242 + }, + { + "epoch": 2.7768924302788847, + "grad_norm": 4.44115690340378, + "learning_rate": 2.887094044103944e-07, + "loss": 0.1553, + "step": 13243 + }, + { + "epoch": 2.7771021178444117, + "grad_norm": 3.9178146887532592, + "learning_rate": 2.881695929419459e-07, + "loss": 0.142, + "step": 13244 + }, + { + "epoch": 2.7773118054099393, + "grad_norm": 4.850911207776765, + "learning_rate": 2.876302792216501e-07, + "loss": 0.1315, + "step": 13245 + }, + { + "epoch": 2.777521492975467, + "grad_norm": 5.007647178907057, + "learning_rate": 2.870914632771482e-07, + "loss": 0.1467, + "step": 13246 + }, + { + "epoch": 2.777731180540994, + "grad_norm": 4.294105792587498, + "learning_rate": 2.8655314513605483e-07, + "loss": 0.1604, + "step": 13247 + }, + { + "epoch": 2.7779408681065214, + "grad_norm": 4.720307300617483, + "learning_rate": 2.860153248259601e-07, + "loss": 0.1859, + "step": 13248 + }, + { + "epoch": 2.7781505556720485, + "grad_norm": 4.2996407201859395, + "learning_rate": 2.8547800237442746e-07, + "loss": 0.1193, + "step": 13249 + }, + { + "epoch": 2.778360243237576, + "grad_norm": 6.042867224906332, + "learning_rate": 2.8494117780899613e-07, + "loss": 0.2001, + "step": 13250 + }, + { + "epoch": 2.778569930803103, + "grad_norm": 4.753267286995288, + "learning_rate": 2.844048511571806e-07, + "loss": 0.1591, + "step": 13251 + }, + { + "epoch": 2.7787796183686306, + "grad_norm": 3.3992199079327023, + "learning_rate": 2.8386902244646573e-07, + "loss": 0.1234, + "step": 13252 + }, + { + "epoch": 2.778989305934158, + "grad_norm": 3.3785206999929733, + "learning_rate": 2.83333691704315e-07, + "loss": 0.1131, + "step": 13253 + }, + { + "epoch": 2.7791989934996852, + "grad_norm": 3.7202110345582877, + "learning_rate": 2.827988589581665e-07, + "loss": 0.1032, + "step": 13254 + }, + { + "epoch": 2.7794086810652128, + "grad_norm": 4.387286658140256, + "learning_rate": 2.822645242354305e-07, + "loss": 0.1224, + "step": 13255 + }, + { + "epoch": 2.7796183686307403, + "grad_norm": 4.064823347330441, + "learning_rate": 2.8173068756349065e-07, + "loss": 0.1123, + "step": 13256 + }, + { + "epoch": 2.7798280561962674, + "grad_norm": 4.477784656540853, + "learning_rate": 2.811973489697106e-07, + "loss": 0.1306, + "step": 13257 + }, + { + "epoch": 2.780037743761795, + "grad_norm": 4.993774824728272, + "learning_rate": 2.806645084814208e-07, + "loss": 0.134, + "step": 13258 + }, + { + "epoch": 2.7802474313273224, + "grad_norm": 3.463738373333924, + "learning_rate": 2.8013216612593376e-07, + "loss": 0.1061, + "step": 13259 + }, + { + "epoch": 2.7804571188928495, + "grad_norm": 3.94592705805319, + "learning_rate": 2.79600321930531e-07, + "loss": 0.1192, + "step": 13260 + }, + { + "epoch": 2.780666806458377, + "grad_norm": 4.547791766194817, + "learning_rate": 2.790689759224696e-07, + "loss": 0.1759, + "step": 13261 + }, + { + "epoch": 2.7808764940239046, + "grad_norm": 4.670913251031462, + "learning_rate": 2.785381281289845e-07, + "loss": 0.1147, + "step": 13262 + }, + { + "epoch": 2.7810861815894317, + "grad_norm": 3.587542422623015, + "learning_rate": 2.780077785772828e-07, + "loss": 0.1112, + "step": 13263 + }, + { + "epoch": 2.781295869154959, + "grad_norm": 4.4830026088330905, + "learning_rate": 2.774779272945416e-07, + "loss": 0.151, + "step": 13264 + }, + { + "epoch": 2.7815055567204867, + "grad_norm": 4.199113395056815, + "learning_rate": 2.769485743079203e-07, + "loss": 0.127, + "step": 13265 + }, + { + "epoch": 2.781715244286014, + "grad_norm": 5.028060822858209, + "learning_rate": 2.764197196445495e-07, + "loss": 0.174, + "step": 13266 + }, + { + "epoch": 2.7819249318515413, + "grad_norm": 3.791738029566469, + "learning_rate": 2.7589136333153186e-07, + "loss": 0.1046, + "step": 13267 + }, + { + "epoch": 2.7821346194170684, + "grad_norm": 4.87676423634727, + "learning_rate": 2.7536350539594694e-07, + "loss": 0.1525, + "step": 13268 + }, + { + "epoch": 2.782344306982596, + "grad_norm": 3.853860273551908, + "learning_rate": 2.7483614586485075e-07, + "loss": 0.1057, + "step": 13269 + }, + { + "epoch": 2.782553994548123, + "grad_norm": 5.380177082396529, + "learning_rate": 2.743092847652673e-07, + "loss": 0.1391, + "step": 13270 + }, + { + "epoch": 2.7827636821136505, + "grad_norm": 3.7406023384844285, + "learning_rate": 2.737829221242028e-07, + "loss": 0.1207, + "step": 13271 + }, + { + "epoch": 2.782973369679178, + "grad_norm": 3.731899972741993, + "learning_rate": 2.7325705796863333e-07, + "loss": 0.1035, + "step": 13272 + }, + { + "epoch": 2.783183057244705, + "grad_norm": 3.7811428801923723, + "learning_rate": 2.7273169232550743e-07, + "loss": 0.109, + "step": 13273 + }, + { + "epoch": 2.7833927448102327, + "grad_norm": 3.634877454276499, + "learning_rate": 2.7220682522175465e-07, + "loss": 0.1139, + "step": 13274 + }, + { + "epoch": 2.78360243237576, + "grad_norm": 4.811346723262429, + "learning_rate": 2.7168245668427464e-07, + "loss": 0.1475, + "step": 13275 + }, + { + "epoch": 2.7838121199412873, + "grad_norm": 4.460029592229935, + "learning_rate": 2.711585867399413e-07, + "loss": 0.1412, + "step": 13276 + }, + { + "epoch": 2.784021807506815, + "grad_norm": 3.7279404908001363, + "learning_rate": 2.7063521541560444e-07, + "loss": 0.1025, + "step": 13277 + }, + { + "epoch": 2.7842314950723424, + "grad_norm": 2.6041155861722505, + "learning_rate": 2.7011234273808693e-07, + "loss": 0.0832, + "step": 13278 + }, + { + "epoch": 2.7844411826378694, + "grad_norm": 4.1101684924925665, + "learning_rate": 2.695899687341874e-07, + "loss": 0.1454, + "step": 13279 + }, + { + "epoch": 2.784650870203397, + "grad_norm": 4.486403356900338, + "learning_rate": 2.6906809343067997e-07, + "loss": 0.1305, + "step": 13280 + }, + { + "epoch": 2.7848605577689245, + "grad_norm": 4.548052337126846, + "learning_rate": 2.6854671685430875e-07, + "loss": 0.1334, + "step": 13281 + }, + { + "epoch": 2.7850702453344516, + "grad_norm": 4.348308977147785, + "learning_rate": 2.680258390317969e-07, + "loss": 0.1302, + "step": 13282 + }, + { + "epoch": 2.785279932899979, + "grad_norm": 3.812327835913294, + "learning_rate": 2.675054599898408e-07, + "loss": 0.1208, + "step": 13283 + }, + { + "epoch": 2.7854896204655066, + "grad_norm": 3.7762053949250074, + "learning_rate": 2.6698557975511017e-07, + "loss": 0.0939, + "step": 13284 + }, + { + "epoch": 2.7856993080310337, + "grad_norm": 4.065729623158698, + "learning_rate": 2.6646619835424824e-07, + "loss": 0.1398, + "step": 13285 + }, + { + "epoch": 2.7859089955965612, + "grad_norm": 3.97148992639932, + "learning_rate": 2.659473158138759e-07, + "loss": 0.1337, + "step": 13286 + }, + { + "epoch": 2.7861186831620883, + "grad_norm": 3.8726561537283373, + "learning_rate": 2.654289321605863e-07, + "loss": 0.0981, + "step": 13287 + }, + { + "epoch": 2.786328370727616, + "grad_norm": 4.023019213367914, + "learning_rate": 2.649110474209493e-07, + "loss": 0.1509, + "step": 13288 + }, + { + "epoch": 2.7865380582931434, + "grad_norm": 3.8401844888756225, + "learning_rate": 2.643936616215037e-07, + "loss": 0.1319, + "step": 13289 + }, + { + "epoch": 2.7867477458586705, + "grad_norm": 3.837441245795726, + "learning_rate": 2.638767747887705e-07, + "loss": 0.1347, + "step": 13290 + }, + { + "epoch": 2.786957433424198, + "grad_norm": 5.22801696706483, + "learning_rate": 2.633603869492374e-07, + "loss": 0.1563, + "step": 13291 + }, + { + "epoch": 2.787167120989725, + "grad_norm": 5.577282455268896, + "learning_rate": 2.6284449812937317e-07, + "loss": 0.1395, + "step": 13292 + }, + { + "epoch": 2.7873768085552526, + "grad_norm": 4.596783573410365, + "learning_rate": 2.6232910835561565e-07, + "loss": 0.1156, + "step": 13293 + }, + { + "epoch": 2.78758649612078, + "grad_norm": 4.47478025260815, + "learning_rate": 2.618142176543803e-07, + "loss": 0.1439, + "step": 13294 + }, + { + "epoch": 2.787796183686307, + "grad_norm": 3.825297221115946, + "learning_rate": 2.6129982605205605e-07, + "loss": 0.1649, + "step": 13295 + }, + { + "epoch": 2.7880058712518347, + "grad_norm": 4.996769482937837, + "learning_rate": 2.607859335750085e-07, + "loss": 0.1374, + "step": 13296 + }, + { + "epoch": 2.7882155588173623, + "grad_norm": 4.379278775838625, + "learning_rate": 2.6027254024957205e-07, + "loss": 0.1055, + "step": 13297 + }, + { + "epoch": 2.7884252463828894, + "grad_norm": 5.090274587988631, + "learning_rate": 2.5975964610206125e-07, + "loss": 0.1799, + "step": 13298 + }, + { + "epoch": 2.788634933948417, + "grad_norm": 3.465402424116034, + "learning_rate": 2.592472511587607e-07, + "loss": 0.0833, + "step": 13299 + }, + { + "epoch": 2.7888446215139444, + "grad_norm": 4.514405846601945, + "learning_rate": 2.58735355445936e-07, + "loss": 0.151, + "step": 13300 + }, + { + "epoch": 2.7890543090794715, + "grad_norm": 3.683925046413724, + "learning_rate": 2.5822395898981724e-07, + "loss": 0.1417, + "step": 13301 + }, + { + "epoch": 2.789263996644999, + "grad_norm": 4.396313897432646, + "learning_rate": 2.5771306181661794e-07, + "loss": 0.1254, + "step": 13302 + }, + { + "epoch": 2.7894736842105265, + "grad_norm": 3.449012913831787, + "learning_rate": 2.5720266395252044e-07, + "loss": 0.104, + "step": 13303 + }, + { + "epoch": 2.7896833717760536, + "grad_norm": 4.105720631955689, + "learning_rate": 2.566927654236862e-07, + "loss": 0.1275, + "step": 13304 + }, + { + "epoch": 2.789893059341581, + "grad_norm": 3.531049041639615, + "learning_rate": 2.561833662562452e-07, + "loss": 0.1012, + "step": 13305 + }, + { + "epoch": 2.7901027469071082, + "grad_norm": 3.0278918375427497, + "learning_rate": 2.5567446647630665e-07, + "loss": 0.1038, + "step": 13306 + }, + { + "epoch": 2.7903124344726358, + "grad_norm": 4.244885059852428, + "learning_rate": 2.5516606610995197e-07, + "loss": 0.1552, + "step": 13307 + }, + { + "epoch": 2.7905221220381633, + "grad_norm": 4.26462786422547, + "learning_rate": 2.5465816518323916e-07, + "loss": 0.1255, + "step": 13308 + }, + { + "epoch": 2.7907318096036904, + "grad_norm": 3.901507664181694, + "learning_rate": 2.541507637221974e-07, + "loss": 0.1547, + "step": 13309 + }, + { + "epoch": 2.790941497169218, + "grad_norm": 3.7429994203070915, + "learning_rate": 2.536438617528314e-07, + "loss": 0.1356, + "step": 13310 + }, + { + "epoch": 2.791151184734745, + "grad_norm": 5.248432174028611, + "learning_rate": 2.5313745930112153e-07, + "loss": 0.1585, + "step": 13311 + }, + { + "epoch": 2.7913608723002725, + "grad_norm": 3.638367645037781, + "learning_rate": 2.5263155639302375e-07, + "loss": 0.1187, + "step": 13312 + }, + { + "epoch": 2.7915705598658, + "grad_norm": 4.016157366733858, + "learning_rate": 2.5212615305446384e-07, + "loss": 0.1479, + "step": 13313 + }, + { + "epoch": 2.791780247431327, + "grad_norm": 3.2226092877723107, + "learning_rate": 2.516212493113446e-07, + "loss": 0.1437, + "step": 13314 + }, + { + "epoch": 2.7919899349968547, + "grad_norm": 3.6950571738641846, + "learning_rate": 2.5111684518954515e-07, + "loss": 0.1085, + "step": 13315 + }, + { + "epoch": 2.792199622562382, + "grad_norm": 4.5492402780003145, + "learning_rate": 2.5061294071491493e-07, + "loss": 0.1115, + "step": 13316 + }, + { + "epoch": 2.7924093101279093, + "grad_norm": 4.140942686642433, + "learning_rate": 2.501095359132821e-07, + "loss": 0.121, + "step": 13317 + }, + { + "epoch": 2.792618997693437, + "grad_norm": 4.391077385069247, + "learning_rate": 2.49606630810445e-07, + "loss": 0.1496, + "step": 13318 + }, + { + "epoch": 2.7928286852589643, + "grad_norm": 4.174312966372431, + "learning_rate": 2.491042254321796e-07, + "loss": 0.1259, + "step": 13319 + }, + { + "epoch": 2.7930383728244914, + "grad_norm": 4.698581190928134, + "learning_rate": 2.486023198042342e-07, + "loss": 0.1673, + "step": 13320 + }, + { + "epoch": 2.793248060390019, + "grad_norm": 3.865628368547563, + "learning_rate": 2.4810091395233384e-07, + "loss": 0.1602, + "step": 13321 + }, + { + "epoch": 2.7934577479555465, + "grad_norm": 5.551023702138432, + "learning_rate": 2.476000079021745e-07, + "loss": 0.1346, + "step": 13322 + }, + { + "epoch": 2.7936674355210735, + "grad_norm": 4.238893734277088, + "learning_rate": 2.470996016794291e-07, + "loss": 0.1393, + "step": 13323 + }, + { + "epoch": 2.793877123086601, + "grad_norm": 3.636960002009199, + "learning_rate": 2.4659969530974603e-07, + "loss": 0.1055, + "step": 13324 + }, + { + "epoch": 2.7940868106521286, + "grad_norm": 3.6920532101143513, + "learning_rate": 2.461002888187447e-07, + "loss": 0.1378, + "step": 13325 + }, + { + "epoch": 2.7942964982176557, + "grad_norm": 3.347715153476394, + "learning_rate": 2.4560138223202023e-07, + "loss": 0.11, + "step": 13326 + }, + { + "epoch": 2.794506185783183, + "grad_norm": 5.117964436612665, + "learning_rate": 2.451029755751444e-07, + "loss": 0.17, + "step": 13327 + }, + { + "epoch": 2.7947158733487103, + "grad_norm": 3.8774752157245924, + "learning_rate": 2.4460506887365786e-07, + "loss": 0.1284, + "step": 13328 + }, + { + "epoch": 2.794925560914238, + "grad_norm": 5.531526208772738, + "learning_rate": 2.4410766215308355e-07, + "loss": 0.1557, + "step": 13329 + }, + { + "epoch": 2.795135248479765, + "grad_norm": 4.775971273409095, + "learning_rate": 2.436107554389111e-07, + "loss": 0.14, + "step": 13330 + }, + { + "epoch": 2.7953449360452924, + "grad_norm": 3.501103370556629, + "learning_rate": 2.43114348756609e-07, + "loss": 0.1256, + "step": 13331 + }, + { + "epoch": 2.79555462361082, + "grad_norm": 3.6012625215727354, + "learning_rate": 2.426184421316191e-07, + "loss": 0.1224, + "step": 13332 + }, + { + "epoch": 2.795764311176347, + "grad_norm": 3.022861407191782, + "learning_rate": 2.421230355893578e-07, + "loss": 0.0864, + "step": 13333 + }, + { + "epoch": 2.7959739987418746, + "grad_norm": 4.342200000330737, + "learning_rate": 2.4162812915521473e-07, + "loss": 0.1639, + "step": 13334 + }, + { + "epoch": 2.796183686307402, + "grad_norm": 3.8348169688319476, + "learning_rate": 2.4113372285455517e-07, + "loss": 0.13, + "step": 13335 + }, + { + "epoch": 2.796393373872929, + "grad_norm": 4.248904874029018, + "learning_rate": 2.4063981671271886e-07, + "loss": 0.139, + "step": 13336 + }, + { + "epoch": 2.7966030614384567, + "grad_norm": 5.337087106996787, + "learning_rate": 2.401464107550167e-07, + "loss": 0.1676, + "step": 13337 + }, + { + "epoch": 2.7968127490039842, + "grad_norm": 3.9329469502031547, + "learning_rate": 2.3965350500674054e-07, + "loss": 0.1134, + "step": 13338 + }, + { + "epoch": 2.7970224365695113, + "grad_norm": 4.717925912477342, + "learning_rate": 2.3916109949315036e-07, + "loss": 0.139, + "step": 13339 + }, + { + "epoch": 2.797232124135039, + "grad_norm": 3.758103720231858, + "learning_rate": 2.386691942394825e-07, + "loss": 0.1475, + "step": 13340 + }, + { + "epoch": 2.7974418117005664, + "grad_norm": 4.221343060797928, + "learning_rate": 2.3817778927094914e-07, + "loss": 0.1142, + "step": 13341 + }, + { + "epoch": 2.7976514992660935, + "grad_norm": 5.398830119028365, + "learning_rate": 2.376868846127356e-07, + "loss": 0.1739, + "step": 13342 + }, + { + "epoch": 2.797861186831621, + "grad_norm": 3.816363564538564, + "learning_rate": 2.3719648029000065e-07, + "loss": 0.1201, + "step": 13343 + }, + { + "epoch": 2.7980708743971485, + "grad_norm": 4.510563421124062, + "learning_rate": 2.3670657632787753e-07, + "loss": 0.1547, + "step": 13344 + }, + { + "epoch": 2.7982805619626756, + "grad_norm": 4.6583565798660596, + "learning_rate": 2.3621717275147727e-07, + "loss": 0.1657, + "step": 13345 + }, + { + "epoch": 2.798490249528203, + "grad_norm": 4.078213667682109, + "learning_rate": 2.3572826958588202e-07, + "loss": 0.1321, + "step": 13346 + }, + { + "epoch": 2.79869993709373, + "grad_norm": 4.088288727620804, + "learning_rate": 2.352398668561462e-07, + "loss": 0.1069, + "step": 13347 + }, + { + "epoch": 2.7989096246592577, + "grad_norm": 4.5720136626806065, + "learning_rate": 2.3475196458730531e-07, + "loss": 0.1516, + "step": 13348 + }, + { + "epoch": 2.799119312224785, + "grad_norm": 4.707557919328625, + "learning_rate": 2.342645628043616e-07, + "loss": 0.117, + "step": 13349 + }, + { + "epoch": 2.7993289997903124, + "grad_norm": 4.166837922515698, + "learning_rate": 2.337776615322984e-07, + "loss": 0.1267, + "step": 13350 + }, + { + "epoch": 2.79953868735584, + "grad_norm": 3.985888703049242, + "learning_rate": 2.3329126079606802e-07, + "loss": 0.097, + "step": 13351 + }, + { + "epoch": 2.799748374921367, + "grad_norm": 4.8595647552521015, + "learning_rate": 2.3280536062059932e-07, + "loss": 0.118, + "step": 13352 + }, + { + "epoch": 2.7999580624868945, + "grad_norm": 3.822679603469462, + "learning_rate": 2.3231996103079578e-07, + "loss": 0.1343, + "step": 13353 + }, + { + "epoch": 2.800167750052422, + "grad_norm": 4.325511570692241, + "learning_rate": 2.318350620515375e-07, + "loss": 0.1219, + "step": 13354 + }, + { + "epoch": 2.800377437617949, + "grad_norm": 4.187537200583681, + "learning_rate": 2.3135066370767346e-07, + "loss": 0.1268, + "step": 13355 + }, + { + "epoch": 2.8005871251834766, + "grad_norm": 3.423713938821339, + "learning_rate": 2.3086676602403045e-07, + "loss": 0.1202, + "step": 13356 + }, + { + "epoch": 2.800796812749004, + "grad_norm": 4.178897016689882, + "learning_rate": 2.3038336902540982e-07, + "loss": 0.128, + "step": 13357 + }, + { + "epoch": 2.8010065003145312, + "grad_norm": 5.440551312740216, + "learning_rate": 2.29900472736585e-07, + "loss": 0.1098, + "step": 13358 + }, + { + "epoch": 2.8012161878800588, + "grad_norm": 4.876131371008478, + "learning_rate": 2.2941807718230736e-07, + "loss": 0.1226, + "step": 13359 + }, + { + "epoch": 2.8014258754455863, + "grad_norm": 3.391857696486738, + "learning_rate": 2.289361823872982e-07, + "loss": 0.12, + "step": 13360 + }, + { + "epoch": 2.8016355630111134, + "grad_norm": 3.9792692125491143, + "learning_rate": 2.284547883762589e-07, + "loss": 0.1006, + "step": 13361 + }, + { + "epoch": 2.801845250576641, + "grad_norm": 4.641618454122115, + "learning_rate": 2.2797389517385858e-07, + "loss": 0.2001, + "step": 13362 + }, + { + "epoch": 2.8020549381421684, + "grad_norm": 3.4121088944780333, + "learning_rate": 2.2749350280474425e-07, + "loss": 0.1185, + "step": 13363 + }, + { + "epoch": 2.8022646257076955, + "grad_norm": 4.710892306019308, + "learning_rate": 2.2701361129353838e-07, + "loss": 0.1793, + "step": 13364 + }, + { + "epoch": 2.802474313273223, + "grad_norm": 3.083631263512698, + "learning_rate": 2.2653422066483355e-07, + "loss": 0.0793, + "step": 13365 + }, + { + "epoch": 2.80268400083875, + "grad_norm": 5.497128376372013, + "learning_rate": 2.2605533094320343e-07, + "loss": 0.1497, + "step": 13366 + }, + { + "epoch": 2.8028936884042777, + "grad_norm": 4.879799166224526, + "learning_rate": 2.2557694215318838e-07, + "loss": 0.1598, + "step": 13367 + }, + { + "epoch": 2.8031033759698047, + "grad_norm": 5.877407969954646, + "learning_rate": 2.2509905431930768e-07, + "loss": 0.1611, + "step": 13368 + }, + { + "epoch": 2.8033130635353323, + "grad_norm": 3.599705482147894, + "learning_rate": 2.2462166746605396e-07, + "loss": 0.1113, + "step": 13369 + }, + { + "epoch": 2.80352275110086, + "grad_norm": 4.146496729286574, + "learning_rate": 2.2414478161789545e-07, + "loss": 0.1182, + "step": 13370 + }, + { + "epoch": 2.803732438666387, + "grad_norm": 4.385728670934805, + "learning_rate": 2.2366839679927254e-07, + "loss": 0.1419, + "step": 13371 + }, + { + "epoch": 2.8039421262319144, + "grad_norm": 4.956617048613931, + "learning_rate": 2.231925130345991e-07, + "loss": 0.1748, + "step": 13372 + }, + { + "epoch": 2.804151813797442, + "grad_norm": 5.182060125863725, + "learning_rate": 2.227171303482678e-07, + "loss": 0.1413, + "step": 13373 + }, + { + "epoch": 2.804361501362969, + "grad_norm": 3.987481742093236, + "learning_rate": 2.222422487646403e-07, + "loss": 0.1471, + "step": 13374 + }, + { + "epoch": 2.8045711889284966, + "grad_norm": 4.643603968233447, + "learning_rate": 2.2176786830805596e-07, + "loss": 0.1566, + "step": 13375 + }, + { + "epoch": 2.804780876494024, + "grad_norm": 3.9919272863346658, + "learning_rate": 2.2129398900282984e-07, + "loss": 0.1261, + "step": 13376 + }, + { + "epoch": 2.804990564059551, + "grad_norm": 3.5639047839085705, + "learning_rate": 2.2082061087324468e-07, + "loss": 0.1345, + "step": 13377 + }, + { + "epoch": 2.8052002516250787, + "grad_norm": 4.925106947407544, + "learning_rate": 2.2034773394356447e-07, + "loss": 0.1646, + "step": 13378 + }, + { + "epoch": 2.805409939190606, + "grad_norm": 4.197184426324162, + "learning_rate": 2.1987535823802754e-07, + "loss": 0.1185, + "step": 13379 + }, + { + "epoch": 2.8056196267561333, + "grad_norm": 4.343692399447504, + "learning_rate": 2.1940348378083896e-07, + "loss": 0.1401, + "step": 13380 + }, + { + "epoch": 2.805829314321661, + "grad_norm": 3.23282190610629, + "learning_rate": 2.1893211059618503e-07, + "loss": 0.1018, + "step": 13381 + }, + { + "epoch": 2.8060390018871884, + "grad_norm": 4.045202791922959, + "learning_rate": 2.1846123870822633e-07, + "loss": 0.1276, + "step": 13382 + }, + { + "epoch": 2.8062486894527154, + "grad_norm": 4.788440782226839, + "learning_rate": 2.179908681410936e-07, + "loss": 0.1395, + "step": 13383 + }, + { + "epoch": 2.806458377018243, + "grad_norm": 3.3259733894606747, + "learning_rate": 2.1752099891889533e-07, + "loss": 0.1045, + "step": 13384 + }, + { + "epoch": 2.80666806458377, + "grad_norm": 4.871621131166538, + "learning_rate": 2.1705163106571225e-07, + "loss": 0.1445, + "step": 13385 + }, + { + "epoch": 2.8068777521492976, + "grad_norm": 4.006487365183586, + "learning_rate": 2.1658276460560068e-07, + "loss": 0.1418, + "step": 13386 + }, + { + "epoch": 2.8070874397148247, + "grad_norm": 4.344836029063606, + "learning_rate": 2.1611439956259139e-07, + "loss": 0.1303, + "step": 13387 + }, + { + "epoch": 2.807297127280352, + "grad_norm": 3.511000874662097, + "learning_rate": 2.1564653596068852e-07, + "loss": 0.1036, + "step": 13388 + }, + { + "epoch": 2.8075068148458797, + "grad_norm": 4.093945265846804, + "learning_rate": 2.1517917382386955e-07, + "loss": 0.1181, + "step": 13389 + }, + { + "epoch": 2.807716502411407, + "grad_norm": 5.547162431251077, + "learning_rate": 2.1471231317608975e-07, + "loss": 0.1618, + "step": 13390 + }, + { + "epoch": 2.8079261899769343, + "grad_norm": 5.366522434937, + "learning_rate": 2.1424595404127556e-07, + "loss": 0.1507, + "step": 13391 + }, + { + "epoch": 2.808135877542462, + "grad_norm": 4.440964241269104, + "learning_rate": 2.1378009644332897e-07, + "loss": 0.1737, + "step": 13392 + }, + { + "epoch": 2.808345565107989, + "grad_norm": 3.5677402250983015, + "learning_rate": 2.133147404061253e-07, + "loss": 0.1071, + "step": 13393 + }, + { + "epoch": 2.8085552526735165, + "grad_norm": 3.605226177024849, + "learning_rate": 2.1284988595351553e-07, + "loss": 0.1086, + "step": 13394 + }, + { + "epoch": 2.808764940239044, + "grad_norm": 3.6027122635852007, + "learning_rate": 2.1238553310932385e-07, + "loss": 0.1293, + "step": 13395 + }, + { + "epoch": 2.808974627804571, + "grad_norm": 4.340822327230391, + "learning_rate": 2.119216818973502e-07, + "loss": 0.1062, + "step": 13396 + }, + { + "epoch": 2.8091843153700986, + "grad_norm": 6.030837404852382, + "learning_rate": 2.1145833234136659e-07, + "loss": 0.1862, + "step": 13397 + }, + { + "epoch": 2.809394002935626, + "grad_norm": 4.807526726194187, + "learning_rate": 2.1099548446511963e-07, + "loss": 0.1721, + "step": 13398 + }, + { + "epoch": 2.809603690501153, + "grad_norm": 4.319526975824163, + "learning_rate": 2.105331382923337e-07, + "loss": 0.1059, + "step": 13399 + }, + { + "epoch": 2.8098133780666807, + "grad_norm": 3.814477333747934, + "learning_rate": 2.1007129384670423e-07, + "loss": 0.1005, + "step": 13400 + }, + { + "epoch": 2.8100230656322083, + "grad_norm": 3.6704887588770476, + "learning_rate": 2.0960995115189785e-07, + "loss": 0.1166, + "step": 13401 + }, + { + "epoch": 2.8102327531977354, + "grad_norm": 3.6289756505723387, + "learning_rate": 2.0914911023156347e-07, + "loss": 0.1023, + "step": 13402 + }, + { + "epoch": 2.810442440763263, + "grad_norm": 3.410194110482676, + "learning_rate": 2.0868877110931885e-07, + "loss": 0.1111, + "step": 13403 + }, + { + "epoch": 2.81065212832879, + "grad_norm": 4.328741617231303, + "learning_rate": 2.082289338087573e-07, + "loss": 0.1339, + "step": 13404 + }, + { + "epoch": 2.8108618158943175, + "grad_norm": 3.5247446369169677, + "learning_rate": 2.0776959835344447e-07, + "loss": 0.1114, + "step": 13405 + }, + { + "epoch": 2.8110715034598446, + "grad_norm": 4.774572776704726, + "learning_rate": 2.073107647669248e-07, + "loss": 0.1395, + "step": 13406 + }, + { + "epoch": 2.811281191025372, + "grad_norm": 4.512397297700276, + "learning_rate": 2.0685243307271176e-07, + "loss": 0.1488, + "step": 13407 + }, + { + "epoch": 2.8114908785908996, + "grad_norm": 3.3179684025415326, + "learning_rate": 2.0639460329429872e-07, + "loss": 0.1108, + "step": 13408 + }, + { + "epoch": 2.8117005661564267, + "grad_norm": 4.2258377561081275, + "learning_rate": 2.0593727545514698e-07, + "loss": 0.1343, + "step": 13409 + }, + { + "epoch": 2.8119102537219542, + "grad_norm": 3.473784463380286, + "learning_rate": 2.0548044957869661e-07, + "loss": 0.1222, + "step": 13410 + }, + { + "epoch": 2.8121199412874818, + "grad_norm": 4.206296141960858, + "learning_rate": 2.0502412568836118e-07, + "loss": 0.1096, + "step": 13411 + }, + { + "epoch": 2.812329628853009, + "grad_norm": 4.0664203284667595, + "learning_rate": 2.0456830380752857e-07, + "loss": 0.1378, + "step": 13412 + }, + { + "epoch": 2.8125393164185364, + "grad_norm": 3.2432684241207714, + "learning_rate": 2.0411298395955904e-07, + "loss": 0.0881, + "step": 13413 + }, + { + "epoch": 2.812749003984064, + "grad_norm": 5.433889733033916, + "learning_rate": 2.0365816616778945e-07, + "loss": 0.1785, + "step": 13414 + }, + { + "epoch": 2.812958691549591, + "grad_norm": 3.9911981227287283, + "learning_rate": 2.0320385045553004e-07, + "loss": 0.1316, + "step": 13415 + }, + { + "epoch": 2.8131683791151185, + "grad_norm": 4.253736715467679, + "learning_rate": 2.0275003684606443e-07, + "loss": 0.1122, + "step": 13416 + }, + { + "epoch": 2.813378066680646, + "grad_norm": 3.9183504876836603, + "learning_rate": 2.0229672536265177e-07, + "loss": 0.108, + "step": 13417 + }, + { + "epoch": 2.813587754246173, + "grad_norm": 3.9947142758005874, + "learning_rate": 2.0184391602852572e-07, + "loss": 0.0991, + "step": 13418 + }, + { + "epoch": 2.8137974418117007, + "grad_norm": 4.945407354279828, + "learning_rate": 2.0139160886689213e-07, + "loss": 0.1568, + "step": 13419 + }, + { + "epoch": 2.814007129377228, + "grad_norm": 4.2994197695479395, + "learning_rate": 2.009398039009336e-07, + "loss": 0.1459, + "step": 13420 + }, + { + "epoch": 2.8142168169427553, + "grad_norm": 5.941397765208098, + "learning_rate": 2.0048850115380713e-07, + "loss": 0.1378, + "step": 13421 + }, + { + "epoch": 2.814426504508283, + "grad_norm": 4.090855495964533, + "learning_rate": 2.0003770064864092e-07, + "loss": 0.1277, + "step": 13422 + }, + { + "epoch": 2.81463619207381, + "grad_norm": 3.6856477686470397, + "learning_rate": 1.9958740240853868e-07, + "loss": 0.1156, + "step": 13423 + }, + { + "epoch": 2.8148458796393374, + "grad_norm": 3.6133898588559776, + "learning_rate": 1.9913760645658087e-07, + "loss": 0.1181, + "step": 13424 + }, + { + "epoch": 2.8150555672048645, + "grad_norm": 5.668607099434243, + "learning_rate": 1.9868831281581902e-07, + "loss": 0.1823, + "step": 13425 + }, + { + "epoch": 2.815265254770392, + "grad_norm": 4.876721713511406, + "learning_rate": 1.9823952150928027e-07, + "loss": 0.1394, + "step": 13426 + }, + { + "epoch": 2.8154749423359196, + "grad_norm": 4.731609195984963, + "learning_rate": 1.9779123255996625e-07, + "loss": 0.1926, + "step": 13427 + }, + { + "epoch": 2.8156846299014466, + "grad_norm": 3.238563367417499, + "learning_rate": 1.9734344599085408e-07, + "loss": 0.1237, + "step": 13428 + }, + { + "epoch": 2.815894317466974, + "grad_norm": 3.6109569587311143, + "learning_rate": 1.9689616182489102e-07, + "loss": 0.1229, + "step": 13429 + }, + { + "epoch": 2.8161040050325017, + "grad_norm": 4.674669086871531, + "learning_rate": 1.9644938008500313e-07, + "loss": 0.1334, + "step": 13430 + }, + { + "epoch": 2.8163136925980288, + "grad_norm": 5.58283762666598, + "learning_rate": 1.9600310079408767e-07, + "loss": 0.1531, + "step": 13431 + }, + { + "epoch": 2.8165233801635563, + "grad_norm": 5.608198001567224, + "learning_rate": 1.9555732397501636e-07, + "loss": 0.1644, + "step": 13432 + }, + { + "epoch": 2.816733067729084, + "grad_norm": 4.966514836845429, + "learning_rate": 1.9511204965063756e-07, + "loss": 0.1409, + "step": 13433 + }, + { + "epoch": 2.816942755294611, + "grad_norm": 3.643769678865067, + "learning_rate": 1.9466727784377304e-07, + "loss": 0.0919, + "step": 13434 + }, + { + "epoch": 2.8171524428601384, + "grad_norm": 3.0153687350004903, + "learning_rate": 1.9422300857721566e-07, + "loss": 0.1018, + "step": 13435 + }, + { + "epoch": 2.817362130425666, + "grad_norm": 3.2210536573591324, + "learning_rate": 1.937792418737372e-07, + "loss": 0.0821, + "step": 13436 + }, + { + "epoch": 2.817571817991193, + "grad_norm": 3.9368810303183195, + "learning_rate": 1.933359777560806e-07, + "loss": 0.1141, + "step": 13437 + }, + { + "epoch": 2.8177815055567206, + "grad_norm": 4.343287273930058, + "learning_rate": 1.9289321624696322e-07, + "loss": 0.1394, + "step": 13438 + }, + { + "epoch": 2.817991193122248, + "grad_norm": 3.950078195549755, + "learning_rate": 1.9245095736907805e-07, + "loss": 0.1274, + "step": 13439 + }, + { + "epoch": 2.818200880687775, + "grad_norm": 4.474596061212188, + "learning_rate": 1.9200920114509248e-07, + "loss": 0.1497, + "step": 13440 + }, + { + "epoch": 2.8184105682533027, + "grad_norm": 4.488894173186195, + "learning_rate": 1.915679475976473e-07, + "loss": 0.1663, + "step": 13441 + }, + { + "epoch": 2.81862025581883, + "grad_norm": 4.115302858837432, + "learning_rate": 1.9112719674935442e-07, + "loss": 0.1387, + "step": 13442 + }, + { + "epoch": 2.8188299433843573, + "grad_norm": 4.5602341278182825, + "learning_rate": 1.906869486228069e-07, + "loss": 0.1523, + "step": 13443 + }, + { + "epoch": 2.8190396309498844, + "grad_norm": 4.295693618321269, + "learning_rate": 1.9024720324056555e-07, + "loss": 0.1345, + "step": 13444 + }, + { + "epoch": 2.819249318515412, + "grad_norm": 4.037791341024337, + "learning_rate": 1.8980796062517015e-07, + "loss": 0.1189, + "step": 13445 + }, + { + "epoch": 2.8194590060809395, + "grad_norm": 4.729523782311469, + "learning_rate": 1.8936922079913155e-07, + "loss": 0.1345, + "step": 13446 + }, + { + "epoch": 2.8196686936464666, + "grad_norm": 3.2219591194457897, + "learning_rate": 1.8893098378493625e-07, + "loss": 0.0863, + "step": 13447 + }, + { + "epoch": 2.819878381211994, + "grad_norm": 3.9535016148577804, + "learning_rate": 1.8849324960504288e-07, + "loss": 0.1276, + "step": 13448 + }, + { + "epoch": 2.8200880687775216, + "grad_norm": 4.669735554991172, + "learning_rate": 1.8805601828188914e-07, + "loss": 0.1369, + "step": 13449 + }, + { + "epoch": 2.8202977563430487, + "grad_norm": 4.035105933095399, + "learning_rate": 1.8761928983788258e-07, + "loss": 0.1084, + "step": 13450 + }, + { + "epoch": 2.820507443908576, + "grad_norm": 4.492576791024945, + "learning_rate": 1.8718306429540534e-07, + "loss": 0.1534, + "step": 13451 + }, + { + "epoch": 2.8207171314741037, + "grad_norm": 4.959557698820072, + "learning_rate": 1.867473416768162e-07, + "loss": 0.1383, + "step": 13452 + }, + { + "epoch": 2.820926819039631, + "grad_norm": 4.469564353974139, + "learning_rate": 1.8631212200444504e-07, + "loss": 0.1082, + "step": 13453 + }, + { + "epoch": 2.8211365066051584, + "grad_norm": 3.8561423171316833, + "learning_rate": 1.858774053005996e-07, + "loss": 0.1277, + "step": 13454 + }, + { + "epoch": 2.821346194170686, + "grad_norm": 2.7173770401458146, + "learning_rate": 1.8544319158755876e-07, + "loss": 0.0854, + "step": 13455 + }, + { + "epoch": 2.821555881736213, + "grad_norm": 3.8595467412918096, + "learning_rate": 1.8500948088757687e-07, + "loss": 0.132, + "step": 13456 + }, + { + "epoch": 2.8217655693017405, + "grad_norm": 3.8034064217534502, + "learning_rate": 1.8457627322288064e-07, + "loss": 0.1086, + "step": 13457 + }, + { + "epoch": 2.821975256867268, + "grad_norm": 5.1509953222698694, + "learning_rate": 1.8414356861567784e-07, + "loss": 0.1249, + "step": 13458 + }, + { + "epoch": 2.822184944432795, + "grad_norm": 5.3051581073928284, + "learning_rate": 1.8371136708813852e-07, + "loss": 0.1858, + "step": 13459 + }, + { + "epoch": 2.8223946319983226, + "grad_norm": 3.7460790321609916, + "learning_rate": 1.832796686624183e-07, + "loss": 0.1171, + "step": 13460 + }, + { + "epoch": 2.8226043195638497, + "grad_norm": 3.256306776841788, + "learning_rate": 1.8284847336064171e-07, + "loss": 0.1264, + "step": 13461 + }, + { + "epoch": 2.8228140071293772, + "grad_norm": 4.4672830710706934, + "learning_rate": 1.824177812049077e-07, + "loss": 0.1145, + "step": 13462 + }, + { + "epoch": 2.8230236946949043, + "grad_norm": 4.142844164501183, + "learning_rate": 1.819875922172898e-07, + "loss": 0.147, + "step": 13463 + }, + { + "epoch": 2.823233382260432, + "grad_norm": 6.499598665238347, + "learning_rate": 1.8155790641983696e-07, + "loss": 0.1753, + "step": 13464 + }, + { + "epoch": 2.8234430698259594, + "grad_norm": 4.206203718005195, + "learning_rate": 1.8112872383456937e-07, + "loss": 0.1626, + "step": 13465 + }, + { + "epoch": 2.8236527573914865, + "grad_norm": 4.53781318891553, + "learning_rate": 1.807000444834861e-07, + "loss": 0.1755, + "step": 13466 + }, + { + "epoch": 2.823862444957014, + "grad_norm": 4.249738883313692, + "learning_rate": 1.8027186838855626e-07, + "loss": 0.1362, + "step": 13467 + }, + { + "epoch": 2.8240721325225415, + "grad_norm": 4.560572345384787, + "learning_rate": 1.798441955717234e-07, + "loss": 0.1466, + "step": 13468 + }, + { + "epoch": 2.8242818200880686, + "grad_norm": 4.182646673478448, + "learning_rate": 1.7941702605490773e-07, + "loss": 0.1522, + "step": 13469 + }, + { + "epoch": 2.824491507653596, + "grad_norm": 6.43631398921247, + "learning_rate": 1.789903598600029e-07, + "loss": 0.1756, + "step": 13470 + }, + { + "epoch": 2.8247011952191237, + "grad_norm": 4.109778781233847, + "learning_rate": 1.7856419700887694e-07, + "loss": 0.1193, + "step": 13471 + }, + { + "epoch": 2.8249108827846507, + "grad_norm": 4.8815366638382205, + "learning_rate": 1.7813853752336908e-07, + "loss": 0.1775, + "step": 13472 + }, + { + "epoch": 2.8251205703501783, + "grad_norm": 4.277927107201278, + "learning_rate": 1.7771338142529625e-07, + "loss": 0.1675, + "step": 13473 + }, + { + "epoch": 2.825330257915706, + "grad_norm": 3.198094000113271, + "learning_rate": 1.7728872873644886e-07, + "loss": 0.0756, + "step": 13474 + }, + { + "epoch": 2.825539945481233, + "grad_norm": 3.70242927349456, + "learning_rate": 1.7686457947859058e-07, + "loss": 0.1214, + "step": 13475 + }, + { + "epoch": 2.8257496330467604, + "grad_norm": 4.600217984464785, + "learning_rate": 1.764409336734607e-07, + "loss": 0.1765, + "step": 13476 + }, + { + "epoch": 2.825959320612288, + "grad_norm": 3.8869130889910912, + "learning_rate": 1.7601779134276963e-07, + "loss": 0.1319, + "step": 13477 + }, + { + "epoch": 2.826169008177815, + "grad_norm": 6.095516558269942, + "learning_rate": 1.7559515250820558e-07, + "loss": 0.1807, + "step": 13478 + }, + { + "epoch": 2.8263786957433426, + "grad_norm": 4.550213065818699, + "learning_rate": 1.751730171914312e-07, + "loss": 0.1463, + "step": 13479 + }, + { + "epoch": 2.8265883833088696, + "grad_norm": 3.786959131550005, + "learning_rate": 1.7475138541407922e-07, + "loss": 0.1104, + "step": 13480 + }, + { + "epoch": 2.826798070874397, + "grad_norm": 5.38653754157049, + "learning_rate": 1.7433025719775898e-07, + "loss": 0.1591, + "step": 13481 + }, + { + "epoch": 2.8270077584399242, + "grad_norm": 4.347449163323004, + "learning_rate": 1.7390963256405545e-07, + "loss": 0.1217, + "step": 13482 + }, + { + "epoch": 2.8272174460054518, + "grad_norm": 4.044976624560833, + "learning_rate": 1.7348951153452697e-07, + "loss": 0.1449, + "step": 13483 + }, + { + "epoch": 2.8274271335709793, + "grad_norm": 4.406117673376607, + "learning_rate": 1.7306989413070186e-07, + "loss": 0.13, + "step": 13484 + }, + { + "epoch": 2.8276368211365064, + "grad_norm": 4.16371531027683, + "learning_rate": 1.7265078037408956e-07, + "loss": 0.1264, + "step": 13485 + }, + { + "epoch": 2.827846508702034, + "grad_norm": 3.9000760317126826, + "learning_rate": 1.7223217028616956e-07, + "loss": 0.1234, + "step": 13486 + }, + { + "epoch": 2.8280561962675614, + "grad_norm": 5.126820395795172, + "learning_rate": 1.7181406388839695e-07, + "loss": 0.1408, + "step": 13487 + }, + { + "epoch": 2.8282658838330885, + "grad_norm": 4.732556929375315, + "learning_rate": 1.71396461202199e-07, + "loss": 0.1818, + "step": 13488 + }, + { + "epoch": 2.828475571398616, + "grad_norm": 4.501669049557819, + "learning_rate": 1.709793622489797e-07, + "loss": 0.1237, + "step": 13489 + }, + { + "epoch": 2.8286852589641436, + "grad_norm": 3.696165522573165, + "learning_rate": 1.7056276705011422e-07, + "loss": 0.0866, + "step": 13490 + }, + { + "epoch": 2.8288949465296707, + "grad_norm": 4.140680627404227, + "learning_rate": 1.7014667562695765e-07, + "loss": 0.1647, + "step": 13491 + }, + { + "epoch": 2.829104634095198, + "grad_norm": 3.7161421675275257, + "learning_rate": 1.6973108800083181e-07, + "loss": 0.1265, + "step": 13492 + }, + { + "epoch": 2.8293143216607257, + "grad_norm": 4.9794656184629105, + "learning_rate": 1.693160041930364e-07, + "loss": 0.137, + "step": 13493 + }, + { + "epoch": 2.829524009226253, + "grad_norm": 3.137659319844232, + "learning_rate": 1.6890142422484658e-07, + "loss": 0.0996, + "step": 13494 + }, + { + "epoch": 2.8297336967917803, + "grad_norm": 3.3701607995097835, + "learning_rate": 1.6848734811751088e-07, + "loss": 0.1156, + "step": 13495 + }, + { + "epoch": 2.829943384357308, + "grad_norm": 4.526917866353032, + "learning_rate": 1.6807377589225016e-07, + "loss": 0.1719, + "step": 13496 + }, + { + "epoch": 2.830153071922835, + "grad_norm": 5.4668251423924366, + "learning_rate": 1.6766070757026075e-07, + "loss": 0.1153, + "step": 13497 + }, + { + "epoch": 2.8303627594883625, + "grad_norm": 4.904087989765102, + "learning_rate": 1.6724814317271354e-07, + "loss": 0.1839, + "step": 13498 + }, + { + "epoch": 2.8305724470538896, + "grad_norm": 3.242284173558282, + "learning_rate": 1.668360827207527e-07, + "loss": 0.1211, + "step": 13499 + }, + { + "epoch": 2.830782134619417, + "grad_norm": 4.643365991129911, + "learning_rate": 1.664245262354991e-07, + "loss": 0.1358, + "step": 13500 + }, + { + "epoch": 2.830991822184944, + "grad_norm": 4.607655380221388, + "learning_rate": 1.6601347373804256e-07, + "loss": 0.1231, + "step": 13501 + }, + { + "epoch": 2.8312015097504717, + "grad_norm": 3.9491623004016327, + "learning_rate": 1.6560292524945176e-07, + "loss": 0.1272, + "step": 13502 + }, + { + "epoch": 2.831411197315999, + "grad_norm": 4.734050570244802, + "learning_rate": 1.6519288079076768e-07, + "loss": 0.1557, + "step": 13503 + }, + { + "epoch": 2.8316208848815263, + "grad_norm": 4.296665014316216, + "learning_rate": 1.647833403830068e-07, + "loss": 0.1257, + "step": 13504 + }, + { + "epoch": 2.831830572447054, + "grad_norm": 4.272249345063031, + "learning_rate": 1.6437430404715682e-07, + "loss": 0.1195, + "step": 13505 + }, + { + "epoch": 2.8320402600125814, + "grad_norm": 5.772249320373049, + "learning_rate": 1.6396577180418316e-07, + "loss": 0.1745, + "step": 13506 + }, + { + "epoch": 2.8322499475781084, + "grad_norm": 4.651866047745959, + "learning_rate": 1.6355774367502463e-07, + "loss": 0.1666, + "step": 13507 + }, + { + "epoch": 2.832459635143636, + "grad_norm": 3.0581453236408978, + "learning_rate": 1.631502196805912e-07, + "loss": 0.0895, + "step": 13508 + }, + { + "epoch": 2.8326693227091635, + "grad_norm": 3.5516137441847735, + "learning_rate": 1.6274319984177057e-07, + "loss": 0.1089, + "step": 13509 + }, + { + "epoch": 2.8328790102746906, + "grad_norm": 4.348700589667592, + "learning_rate": 1.6233668417942273e-07, + "loss": 0.1317, + "step": 13510 + }, + { + "epoch": 2.833088697840218, + "grad_norm": 3.923422756873794, + "learning_rate": 1.6193067271438213e-07, + "loss": 0.1152, + "step": 13511 + }, + { + "epoch": 2.8332983854057456, + "grad_norm": 3.379124584751275, + "learning_rate": 1.6152516546745766e-07, + "loss": 0.1048, + "step": 13512 + }, + { + "epoch": 2.8335080729712727, + "grad_norm": 3.8366215183652916, + "learning_rate": 1.6112016245943274e-07, + "loss": 0.1421, + "step": 13513 + }, + { + "epoch": 2.8337177605368002, + "grad_norm": 5.087542291677511, + "learning_rate": 1.6071566371106406e-07, + "loss": 0.1755, + "step": 13514 + }, + { + "epoch": 2.8339274481023278, + "grad_norm": 3.451368807985004, + "learning_rate": 1.6031166924308284e-07, + "loss": 0.0982, + "step": 13515 + }, + { + "epoch": 2.834137135667855, + "grad_norm": 4.593682387054715, + "learning_rate": 1.599081790761947e-07, + "loss": 0.1273, + "step": 13516 + }, + { + "epoch": 2.8343468232333824, + "grad_norm": 4.368376212679015, + "learning_rate": 1.5950519323107984e-07, + "loss": 0.1178, + "step": 13517 + }, + { + "epoch": 2.8345565107989095, + "grad_norm": 6.243624632763378, + "learning_rate": 1.5910271172839053e-07, + "loss": 0.2089, + "step": 13518 + }, + { + "epoch": 2.834766198364437, + "grad_norm": 4.027191688453906, + "learning_rate": 1.587007345887559e-07, + "loss": 0.156, + "step": 13519 + }, + { + "epoch": 2.834975885929964, + "grad_norm": 5.65532073269755, + "learning_rate": 1.5829926183277722e-07, + "loss": 0.1536, + "step": 13520 + }, + { + "epoch": 2.8351855734954916, + "grad_norm": 4.614701479719456, + "learning_rate": 1.5789829348103137e-07, + "loss": 0.1399, + "step": 13521 + }, + { + "epoch": 2.835395261061019, + "grad_norm": 5.035728326187406, + "learning_rate": 1.5749782955406745e-07, + "loss": 0.1695, + "step": 13522 + }, + { + "epoch": 2.835604948626546, + "grad_norm": 5.493209243927689, + "learning_rate": 1.5709787007241128e-07, + "loss": 0.1555, + "step": 13523 + }, + { + "epoch": 2.8358146361920737, + "grad_norm": 4.508705264071455, + "learning_rate": 1.566984150565609e-07, + "loss": 0.1106, + "step": 13524 + }, + { + "epoch": 2.8360243237576013, + "grad_norm": 5.494970843066494, + "learning_rate": 1.5629946452698886e-07, + "loss": 0.1618, + "step": 13525 + }, + { + "epoch": 2.8362340113231284, + "grad_norm": 4.099897825889496, + "learning_rate": 1.5590101850414207e-07, + "loss": 0.1428, + "step": 13526 + }, + { + "epoch": 2.836443698888656, + "grad_norm": 3.8661647393033602, + "learning_rate": 1.5550307700844203e-07, + "loss": 0.1224, + "step": 13527 + }, + { + "epoch": 2.8366533864541834, + "grad_norm": 5.323388608347447, + "learning_rate": 1.551056400602835e-07, + "loss": 0.1616, + "step": 13528 + }, + { + "epoch": 2.8368630740197105, + "grad_norm": 4.917644817297917, + "learning_rate": 1.5470870768003686e-07, + "loss": 0.1529, + "step": 13529 + }, + { + "epoch": 2.837072761585238, + "grad_norm": 5.104726337910391, + "learning_rate": 1.5431227988804365e-07, + "loss": 0.19, + "step": 13530 + }, + { + "epoch": 2.8372824491507656, + "grad_norm": 5.622229177454503, + "learning_rate": 1.5391635670462424e-07, + "loss": 0.1559, + "step": 13531 + }, + { + "epoch": 2.8374921367162926, + "grad_norm": 3.9099025297449517, + "learning_rate": 1.5352093815006684e-07, + "loss": 0.1036, + "step": 13532 + }, + { + "epoch": 2.83770182428182, + "grad_norm": 4.44382294992835, + "learning_rate": 1.5312602424463974e-07, + "loss": 0.0961, + "step": 13533 + }, + { + "epoch": 2.8379115118473477, + "grad_norm": 3.9443254029146764, + "learning_rate": 1.5273161500858336e-07, + "loss": 0.1097, + "step": 13534 + }, + { + "epoch": 2.8381211994128748, + "grad_norm": 3.6514105217803907, + "learning_rate": 1.523377104621093e-07, + "loss": 0.1311, + "step": 13535 + }, + { + "epoch": 2.8383308869784023, + "grad_norm": 5.2831244240714454, + "learning_rate": 1.5194431062540817e-07, + "loss": 0.1514, + "step": 13536 + }, + { + "epoch": 2.8385405745439294, + "grad_norm": 4.198605410894636, + "learning_rate": 1.5155141551864373e-07, + "loss": 0.1618, + "step": 13537 + }, + { + "epoch": 2.838750262109457, + "grad_norm": 5.559591065808132, + "learning_rate": 1.511590251619477e-07, + "loss": 0.1532, + "step": 13538 + }, + { + "epoch": 2.8389599496749844, + "grad_norm": 4.978930894028863, + "learning_rate": 1.50767139575434e-07, + "loss": 0.15, + "step": 13539 + }, + { + "epoch": 2.8391696372405115, + "grad_norm": 4.295804930222321, + "learning_rate": 1.5037575877918875e-07, + "loss": 0.1349, + "step": 13540 + }, + { + "epoch": 2.839379324806039, + "grad_norm": 3.9601735192046714, + "learning_rate": 1.4998488279326816e-07, + "loss": 0.105, + "step": 13541 + }, + { + "epoch": 2.839589012371566, + "grad_norm": 4.8306204935383334, + "learning_rate": 1.495945116377051e-07, + "loss": 0.1292, + "step": 13542 + }, + { + "epoch": 2.8397986999370937, + "grad_norm": 3.841762238212064, + "learning_rate": 1.492046453325091e-07, + "loss": 0.1346, + "step": 13543 + }, + { + "epoch": 2.840008387502621, + "grad_norm": 4.523805638306732, + "learning_rate": 1.4881528389766087e-07, + "loss": 0.085, + "step": 13544 + }, + { + "epoch": 2.8402180750681483, + "grad_norm": 4.741838921582771, + "learning_rate": 1.484264273531144e-07, + "loss": 0.1553, + "step": 13545 + }, + { + "epoch": 2.840427762633676, + "grad_norm": 4.47256299540073, + "learning_rate": 1.4803807571880047e-07, + "loss": 0.142, + "step": 13546 + }, + { + "epoch": 2.8406374501992033, + "grad_norm": 3.4233333964104173, + "learning_rate": 1.476502290146231e-07, + "loss": 0.0947, + "step": 13547 + }, + { + "epoch": 2.8408471377647304, + "grad_norm": 4.485976764046395, + "learning_rate": 1.472628872604587e-07, + "loss": 0.1325, + "step": 13548 + }, + { + "epoch": 2.841056825330258, + "grad_norm": 4.352710032384308, + "learning_rate": 1.4687605047616126e-07, + "loss": 0.1256, + "step": 13549 + }, + { + "epoch": 2.8412665128957855, + "grad_norm": 4.786726859081019, + "learning_rate": 1.4648971868155505e-07, + "loss": 0.1572, + "step": 13550 + }, + { + "epoch": 2.8414762004613126, + "grad_norm": 4.592737482641132, + "learning_rate": 1.461038918964397e-07, + "loss": 0.124, + "step": 13551 + }, + { + "epoch": 2.84168588802684, + "grad_norm": 5.078285260239264, + "learning_rate": 1.4571857014059165e-07, + "loss": 0.1786, + "step": 13552 + }, + { + "epoch": 2.8418955755923676, + "grad_norm": 4.8586291165436375, + "learning_rate": 1.4533375343375845e-07, + "loss": 0.1516, + "step": 13553 + }, + { + "epoch": 2.8421052631578947, + "grad_norm": 4.483687482484891, + "learning_rate": 1.4494944179566206e-07, + "loss": 0.1576, + "step": 13554 + }, + { + "epoch": 2.842314950723422, + "grad_norm": 4.422660326661012, + "learning_rate": 1.4456563524599899e-07, + "loss": 0.1603, + "step": 13555 + }, + { + "epoch": 2.8425246382889493, + "grad_norm": 3.159919696835004, + "learning_rate": 1.4418233380444125e-07, + "loss": 0.1169, + "step": 13556 + }, + { + "epoch": 2.842734325854477, + "grad_norm": 4.573223903512745, + "learning_rate": 1.4379953749063202e-07, + "loss": 0.1542, + "step": 13557 + }, + { + "epoch": 2.8429440134200044, + "grad_norm": 2.7437999341616037, + "learning_rate": 1.434172463241923e-07, + "loss": 0.0744, + "step": 13558 + }, + { + "epoch": 2.8431537009855314, + "grad_norm": 3.3932337019077132, + "learning_rate": 1.4303546032471417e-07, + "loss": 0.1102, + "step": 13559 + }, + { + "epoch": 2.843363388551059, + "grad_norm": 3.6212030201899, + "learning_rate": 1.426541795117642e-07, + "loss": 0.1502, + "step": 13560 + }, + { + "epoch": 2.843573076116586, + "grad_norm": 4.456743997131811, + "learning_rate": 1.4227340390488342e-07, + "loss": 0.1525, + "step": 13561 + }, + { + "epoch": 2.8437827636821136, + "grad_norm": 3.1878796132562086, + "learning_rate": 1.4189313352358957e-07, + "loss": 0.1002, + "step": 13562 + }, + { + "epoch": 2.843992451247641, + "grad_norm": 5.415522014487006, + "learning_rate": 1.415133683873704e-07, + "loss": 0.1632, + "step": 13563 + }, + { + "epoch": 2.844202138813168, + "grad_norm": 4.380361604196579, + "learning_rate": 1.411341085156881e-07, + "loss": 0.1792, + "step": 13564 + }, + { + "epoch": 2.8444118263786957, + "grad_norm": 3.835866387843376, + "learning_rate": 1.4075535392798378e-07, + "loss": 0.1075, + "step": 13565 + }, + { + "epoch": 2.8446215139442232, + "grad_norm": 3.9882846191496637, + "learning_rate": 1.403771046436675e-07, + "loss": 0.1188, + "step": 13566 + }, + { + "epoch": 2.8448312015097503, + "grad_norm": 7.257783162906677, + "learning_rate": 1.3999936068212483e-07, + "loss": 0.1883, + "step": 13567 + }, + { + "epoch": 2.845040889075278, + "grad_norm": 3.657899516699932, + "learning_rate": 1.3962212206271696e-07, + "loss": 0.0999, + "step": 13568 + }, + { + "epoch": 2.8452505766408054, + "grad_norm": 4.10619393383268, + "learning_rate": 1.3924538880477623e-07, + "loss": 0.1325, + "step": 13569 + }, + { + "epoch": 2.8454602642063325, + "grad_norm": 4.222522438815761, + "learning_rate": 1.388691609276116e-07, + "loss": 0.131, + "step": 13570 + }, + { + "epoch": 2.84566995177186, + "grad_norm": 5.461991239664398, + "learning_rate": 1.3849343845050768e-07, + "loss": 0.1327, + "step": 13571 + }, + { + "epoch": 2.8458796393373875, + "grad_norm": 3.965255898866399, + "learning_rate": 1.381182213927168e-07, + "loss": 0.0976, + "step": 13572 + }, + { + "epoch": 2.8460893269029146, + "grad_norm": 3.839500547731919, + "learning_rate": 1.3774350977347251e-07, + "loss": 0.1045, + "step": 13573 + }, + { + "epoch": 2.846299014468442, + "grad_norm": 5.968003001890681, + "learning_rate": 1.3736930361197832e-07, + "loss": 0.1571, + "step": 13574 + }, + { + "epoch": 2.8465087020339697, + "grad_norm": 3.3105197807626765, + "learning_rate": 1.3699560292741442e-07, + "loss": 0.0998, + "step": 13575 + }, + { + "epoch": 2.8467183895994967, + "grad_norm": 4.267024282311256, + "learning_rate": 1.366224077389311e-07, + "loss": 0.1403, + "step": 13576 + }, + { + "epoch": 2.8469280771650243, + "grad_norm": 4.823953508506322, + "learning_rate": 1.3624971806565633e-07, + "loss": 0.1453, + "step": 13577 + }, + { + "epoch": 2.8471377647305514, + "grad_norm": 4.29776446033827, + "learning_rate": 1.3587753392669157e-07, + "loss": 0.1312, + "step": 13578 + }, + { + "epoch": 2.847347452296079, + "grad_norm": 4.534087073953428, + "learning_rate": 1.355058553411126e-07, + "loss": 0.176, + "step": 13579 + }, + { + "epoch": 2.847557139861606, + "grad_norm": 7.797080729346616, + "learning_rate": 1.3513468232796757e-07, + "loss": 0.1706, + "step": 13580 + }, + { + "epoch": 2.8477668274271335, + "grad_norm": 3.9273316841090007, + "learning_rate": 1.3476401490627898e-07, + "loss": 0.1291, + "step": 13581 + }, + { + "epoch": 2.847976514992661, + "grad_norm": 4.446979082184453, + "learning_rate": 1.3439385309504504e-07, + "loss": 0.1889, + "step": 13582 + }, + { + "epoch": 2.848186202558188, + "grad_norm": 4.727092161727425, + "learning_rate": 1.3402419691323832e-07, + "loss": 0.12, + "step": 13583 + }, + { + "epoch": 2.8483958901237156, + "grad_norm": 4.305504846235536, + "learning_rate": 1.3365504637980142e-07, + "loss": 0.1261, + "step": 13584 + }, + { + "epoch": 2.848605577689243, + "grad_norm": 4.92008365737643, + "learning_rate": 1.332864015136559e-07, + "loss": 0.1519, + "step": 13585 + }, + { + "epoch": 2.8488152652547702, + "grad_norm": 3.7348946808043566, + "learning_rate": 1.3291826233369665e-07, + "loss": 0.1283, + "step": 13586 + }, + { + "epoch": 2.8490249528202978, + "grad_norm": 3.54063830124864, + "learning_rate": 1.3255062885878967e-07, + "loss": 0.1102, + "step": 13587 + }, + { + "epoch": 2.8492346403858253, + "grad_norm": 5.733590426744869, + "learning_rate": 1.3218350110777656e-07, + "loss": 0.1542, + "step": 13588 + }, + { + "epoch": 2.8494443279513524, + "grad_norm": 3.999563133469065, + "learning_rate": 1.3181687909947338e-07, + "loss": 0.1402, + "step": 13589 + }, + { + "epoch": 2.84965401551688, + "grad_norm": 4.774945477740314, + "learning_rate": 1.3145076285267177e-07, + "loss": 0.1453, + "step": 13590 + }, + { + "epoch": 2.8498637030824074, + "grad_norm": 3.3548180552347384, + "learning_rate": 1.310851523861345e-07, + "loss": 0.1035, + "step": 13591 + }, + { + "epoch": 2.8500733906479345, + "grad_norm": 4.133150867012464, + "learning_rate": 1.3072004771859992e-07, + "loss": 0.1038, + "step": 13592 + }, + { + "epoch": 2.850283078213462, + "grad_norm": 4.094446470054902, + "learning_rate": 1.303554488687797e-07, + "loss": 0.1518, + "step": 13593 + }, + { + "epoch": 2.8504927657789896, + "grad_norm": 3.290340668771646, + "learning_rate": 1.2999135585536006e-07, + "loss": 0.107, + "step": 13594 + }, + { + "epoch": 2.8507024533445167, + "grad_norm": 4.553580807567586, + "learning_rate": 1.2962776869700378e-07, + "loss": 0.1287, + "step": 13595 + }, + { + "epoch": 2.850912140910044, + "grad_norm": 4.028074819887243, + "learning_rate": 1.2926468741234378e-07, + "loss": 0.1574, + "step": 13596 + }, + { + "epoch": 2.8511218284755713, + "grad_norm": 3.457077367186731, + "learning_rate": 1.2890211201998736e-07, + "loss": 0.1066, + "step": 13597 + }, + { + "epoch": 2.851331516041099, + "grad_norm": 5.371522879016928, + "learning_rate": 1.285400425385186e-07, + "loss": 0.1499, + "step": 13598 + }, + { + "epoch": 2.851541203606626, + "grad_norm": 4.239696298104947, + "learning_rate": 1.2817847898649482e-07, + "loss": 0.1297, + "step": 13599 + }, + { + "epoch": 2.8517508911721534, + "grad_norm": 3.9951595769445616, + "learning_rate": 1.2781742138244345e-07, + "loss": 0.1426, + "step": 13600 + }, + { + "epoch": 2.851960578737681, + "grad_norm": 4.466207240802085, + "learning_rate": 1.2745686974487413e-07, + "loss": 0.1571, + "step": 13601 + }, + { + "epoch": 2.852170266303208, + "grad_norm": 4.169059567774628, + "learning_rate": 1.2709682409226099e-07, + "loss": 0.1251, + "step": 13602 + }, + { + "epoch": 2.8523799538687356, + "grad_norm": 5.604564169424631, + "learning_rate": 1.2673728444306144e-07, + "loss": 0.1819, + "step": 13603 + }, + { + "epoch": 2.852589641434263, + "grad_norm": 4.982648487967696, + "learning_rate": 1.2637825081569855e-07, + "loss": 0.1782, + "step": 13604 + }, + { + "epoch": 2.85279932899979, + "grad_norm": 3.7567597521064466, + "learning_rate": 1.260197232285765e-07, + "loss": 0.099, + "step": 13605 + }, + { + "epoch": 2.8530090165653177, + "grad_norm": 5.773398639899188, + "learning_rate": 1.2566170170006831e-07, + "loss": 0.1778, + "step": 13606 + }, + { + "epoch": 2.853218704130845, + "grad_norm": 3.6153543505579013, + "learning_rate": 1.2530418624852493e-07, + "loss": 0.1174, + "step": 13607 + }, + { + "epoch": 2.8534283916963723, + "grad_norm": 3.0286179382909086, + "learning_rate": 1.2494717689226942e-07, + "loss": 0.0961, + "step": 13608 + }, + { + "epoch": 2.8536380792619, + "grad_norm": 3.7187505126499976, + "learning_rate": 1.245906736495972e-07, + "loss": 0.1099, + "step": 13609 + }, + { + "epoch": 2.8538477668274274, + "grad_norm": 3.848745846292349, + "learning_rate": 1.242346765387803e-07, + "loss": 0.1015, + "step": 13610 + }, + { + "epoch": 2.8540574543929544, + "grad_norm": 5.346604607808305, + "learning_rate": 1.2387918557806632e-07, + "loss": 0.1211, + "step": 13611 + }, + { + "epoch": 2.854267141958482, + "grad_norm": 4.931097679466003, + "learning_rate": 1.2352420078567295e-07, + "loss": 0.1491, + "step": 13612 + }, + { + "epoch": 2.8544768295240095, + "grad_norm": 4.234824569570955, + "learning_rate": 1.231697221797945e-07, + "loss": 0.118, + "step": 13613 + }, + { + "epoch": 2.8546865170895366, + "grad_norm": 4.376378728016561, + "learning_rate": 1.2281574977859755e-07, + "loss": 0.1345, + "step": 13614 + }, + { + "epoch": 2.854896204655064, + "grad_norm": 5.14643806422206, + "learning_rate": 1.2246228360022538e-07, + "loss": 0.1657, + "step": 13615 + }, + { + "epoch": 2.855105892220591, + "grad_norm": 4.027745349310622, + "learning_rate": 1.2210932366279238e-07, + "loss": 0.1296, + "step": 13616 + }, + { + "epoch": 2.8553155797861187, + "grad_norm": 4.428884014545528, + "learning_rate": 1.2175686998438963e-07, + "loss": 0.1324, + "step": 13617 + }, + { + "epoch": 2.855525267351646, + "grad_norm": 3.2877188365182843, + "learning_rate": 1.2140492258307824e-07, + "loss": 0.1233, + "step": 13618 + }, + { + "epoch": 2.8557349549171733, + "grad_norm": 4.282642228408265, + "learning_rate": 1.2105348147689934e-07, + "loss": 0.1451, + "step": 13619 + }, + { + "epoch": 2.855944642482701, + "grad_norm": 4.714871118872217, + "learning_rate": 1.2070254668386517e-07, + "loss": 0.1614, + "step": 13620 + }, + { + "epoch": 2.856154330048228, + "grad_norm": 4.1193856952778996, + "learning_rate": 1.2035211822195802e-07, + "loss": 0.1295, + "step": 13621 + }, + { + "epoch": 2.8563640176137555, + "grad_norm": 3.142993769562701, + "learning_rate": 1.200021961091402e-07, + "loss": 0.0862, + "step": 13622 + }, + { + "epoch": 2.856573705179283, + "grad_norm": 4.5550732570915295, + "learning_rate": 1.1965278036334737e-07, + "loss": 0.1141, + "step": 13623 + }, + { + "epoch": 2.85678339274481, + "grad_norm": 4.487123890452544, + "learning_rate": 1.193038710024852e-07, + "loss": 0.1231, + "step": 13624 + }, + { + "epoch": 2.8569930803103376, + "grad_norm": 4.608936735912965, + "learning_rate": 1.1895546804443713e-07, + "loss": 0.1337, + "step": 13625 + }, + { + "epoch": 2.857202767875865, + "grad_norm": 4.131497070113812, + "learning_rate": 1.1860757150705893e-07, + "loss": 0.1281, + "step": 13626 + }, + { + "epoch": 2.857412455441392, + "grad_norm": 4.512176705212993, + "learning_rate": 1.1826018140818074e-07, + "loss": 0.1844, + "step": 13627 + }, + { + "epoch": 2.8576221430069197, + "grad_norm": 3.952044341528827, + "learning_rate": 1.1791329776560834e-07, + "loss": 0.1415, + "step": 13628 + }, + { + "epoch": 2.8578318305724473, + "grad_norm": 4.370192464467855, + "learning_rate": 1.175669205971186e-07, + "loss": 0.1167, + "step": 13629 + }, + { + "epoch": 2.8580415181379744, + "grad_norm": 4.6871348205020045, + "learning_rate": 1.1722104992046401e-07, + "loss": 0.1291, + "step": 13630 + }, + { + "epoch": 2.858251205703502, + "grad_norm": 3.950759548623538, + "learning_rate": 1.1687568575337149e-07, + "loss": 0.1158, + "step": 13631 + }, + { + "epoch": 2.8584608932690294, + "grad_norm": 3.9022129223537605, + "learning_rate": 1.1653082811354132e-07, + "loss": 0.1351, + "step": 13632 + }, + { + "epoch": 2.8586705808345565, + "grad_norm": 4.150530887589807, + "learning_rate": 1.1618647701864938e-07, + "loss": 0.1423, + "step": 13633 + }, + { + "epoch": 2.858880268400084, + "grad_norm": 3.8147957289538628, + "learning_rate": 1.1584263248634153e-07, + "loss": 0.1223, + "step": 13634 + }, + { + "epoch": 2.859089955965611, + "grad_norm": 4.744537730792014, + "learning_rate": 1.1549929453424258e-07, + "loss": 0.1795, + "step": 13635 + }, + { + "epoch": 2.8592996435311386, + "grad_norm": 3.6599870359446127, + "learning_rate": 1.1515646317994844e-07, + "loss": 0.1182, + "step": 13636 + }, + { + "epoch": 2.8595093310966657, + "grad_norm": 4.4819778434380195, + "learning_rate": 1.1481413844103062e-07, + "loss": 0.1361, + "step": 13637 + }, + { + "epoch": 2.8597190186621932, + "grad_norm": 4.825349831900849, + "learning_rate": 1.1447232033503286e-07, + "loss": 0.1793, + "step": 13638 + }, + { + "epoch": 2.8599287062277208, + "grad_norm": 5.164136180734384, + "learning_rate": 1.1413100887947448e-07, + "loss": 0.1636, + "step": 13639 + }, + { + "epoch": 2.860138393793248, + "grad_norm": 5.2535938873817996, + "learning_rate": 1.1379020409184704e-07, + "loss": 0.1819, + "step": 13640 + }, + { + "epoch": 2.8603480813587754, + "grad_norm": 3.2605737583106236, + "learning_rate": 1.1344990598961992e-07, + "loss": 0.1109, + "step": 13641 + }, + { + "epoch": 2.860557768924303, + "grad_norm": 5.3740688320670955, + "learning_rate": 1.1311011459023135e-07, + "loss": 0.1797, + "step": 13642 + }, + { + "epoch": 2.86076745648983, + "grad_norm": 4.577988669941811, + "learning_rate": 1.1277082991109745e-07, + "loss": 0.1261, + "step": 13643 + }, + { + "epoch": 2.8609771440553575, + "grad_norm": 3.1814478824021264, + "learning_rate": 1.124320519696076e-07, + "loss": 0.095, + "step": 13644 + }, + { + "epoch": 2.861186831620885, + "grad_norm": 5.147039954673129, + "learning_rate": 1.120937807831235e-07, + "loss": 0.1086, + "step": 13645 + }, + { + "epoch": 2.861396519186412, + "grad_norm": 5.340453266931193, + "learning_rate": 1.1175601636898348e-07, + "loss": 0.1562, + "step": 13646 + }, + { + "epoch": 2.8616062067519397, + "grad_norm": 3.6952859300944545, + "learning_rate": 1.1141875874449815e-07, + "loss": 0.1158, + "step": 13647 + }, + { + "epoch": 2.861815894317467, + "grad_norm": 5.301684438619656, + "learning_rate": 1.1108200792695035e-07, + "loss": 0.1422, + "step": 13648 + }, + { + "epoch": 2.8620255818829943, + "grad_norm": 4.8367563037495795, + "learning_rate": 1.1074576393360292e-07, + "loss": 0.1494, + "step": 13649 + }, + { + "epoch": 2.862235269448522, + "grad_norm": 4.364164084450006, + "learning_rate": 1.1041002678168766e-07, + "loss": 0.1626, + "step": 13650 + }, + { + "epoch": 2.8624449570140493, + "grad_norm": 3.877871500441854, + "learning_rate": 1.1007479648840969e-07, + "loss": 0.1289, + "step": 13651 + }, + { + "epoch": 2.8626546445795764, + "grad_norm": 3.969067891692186, + "learning_rate": 1.0974007307095192e-07, + "loss": 0.1265, + "step": 13652 + }, + { + "epoch": 2.862864332145104, + "grad_norm": 4.057390512293068, + "learning_rate": 1.0940585654646951e-07, + "loss": 0.1592, + "step": 13653 + }, + { + "epoch": 2.863074019710631, + "grad_norm": 3.7477973147476034, + "learning_rate": 1.09072146932091e-07, + "loss": 0.1117, + "step": 13654 + }, + { + "epoch": 2.8632837072761586, + "grad_norm": 3.2109220481475242, + "learning_rate": 1.0873894424492049e-07, + "loss": 0.1131, + "step": 13655 + }, + { + "epoch": 2.8634933948416856, + "grad_norm": 3.4511645709102434, + "learning_rate": 1.084062485020343e-07, + "loss": 0.1062, + "step": 13656 + }, + { + "epoch": 2.863703082407213, + "grad_norm": 4.179331793199333, + "learning_rate": 1.0807405972048324e-07, + "loss": 0.1392, + "step": 13657 + }, + { + "epoch": 2.8639127699727407, + "grad_norm": 4.738768710333367, + "learning_rate": 1.0774237791729481e-07, + "loss": 0.182, + "step": 13658 + }, + { + "epoch": 2.8641224575382678, + "grad_norm": 3.8175376001693713, + "learning_rate": 1.074112031094654e-07, + "loss": 0.113, + "step": 13659 + }, + { + "epoch": 2.8643321451037953, + "grad_norm": 4.626430669270818, + "learning_rate": 1.0708053531397034e-07, + "loss": 0.1467, + "step": 13660 + }, + { + "epoch": 2.864541832669323, + "grad_norm": 5.131946922938686, + "learning_rate": 1.0675037454775606e-07, + "loss": 0.149, + "step": 13661 + }, + { + "epoch": 2.86475152023485, + "grad_norm": 5.36406272708449, + "learning_rate": 1.0642072082774458e-07, + "loss": 0.1221, + "step": 13662 + }, + { + "epoch": 2.8649612078003774, + "grad_norm": 3.4170416277557294, + "learning_rate": 1.0609157417083016e-07, + "loss": 0.1027, + "step": 13663 + }, + { + "epoch": 2.865170895365905, + "grad_norm": 3.38670810976359, + "learning_rate": 1.0576293459388265e-07, + "loss": 0.0981, + "step": 13664 + }, + { + "epoch": 2.865380582931432, + "grad_norm": 3.7759771735066, + "learning_rate": 1.0543480211374524e-07, + "loss": 0.12, + "step": 13665 + }, + { + "epoch": 2.8655902704969596, + "grad_norm": 3.614647847213322, + "learning_rate": 1.0510717674723558e-07, + "loss": 0.1175, + "step": 13666 + }, + { + "epoch": 2.865799958062487, + "grad_norm": 4.336401271825973, + "learning_rate": 1.047800585111447e-07, + "loss": 0.1509, + "step": 13667 + }, + { + "epoch": 2.866009645628014, + "grad_norm": 5.227772158608982, + "learning_rate": 1.0445344742223917e-07, + "loss": 0.1494, + "step": 13668 + }, + { + "epoch": 2.8662193331935417, + "grad_norm": 4.68598737073222, + "learning_rate": 1.0412734349725673e-07, + "loss": 0.1401, + "step": 13669 + }, + { + "epoch": 2.8664290207590692, + "grad_norm": 5.020608384899008, + "learning_rate": 1.0380174675291177e-07, + "loss": 0.1454, + "step": 13670 + }, + { + "epoch": 2.8666387083245963, + "grad_norm": 3.462378392602713, + "learning_rate": 1.0347665720589095e-07, + "loss": 0.1056, + "step": 13671 + }, + { + "epoch": 2.866848395890124, + "grad_norm": 3.6725823120002308, + "learning_rate": 1.0315207487285651e-07, + "loss": 0.1149, + "step": 13672 + }, + { + "epoch": 2.867058083455651, + "grad_norm": 4.507916037553875, + "learning_rate": 1.0282799977044289e-07, + "loss": 0.1566, + "step": 13673 + }, + { + "epoch": 2.8672677710211785, + "grad_norm": 5.161568807846467, + "learning_rate": 1.0250443191526016e-07, + "loss": 0.1412, + "step": 13674 + }, + { + "epoch": 2.8674774585867056, + "grad_norm": 4.41848591076201, + "learning_rate": 1.0218137132389283e-07, + "loss": 0.1496, + "step": 13675 + }, + { + "epoch": 2.867687146152233, + "grad_norm": 7.149009844531486, + "learning_rate": 1.0185881801289543e-07, + "loss": 0.1928, + "step": 13676 + }, + { + "epoch": 2.8678968337177606, + "grad_norm": 3.4195707639008748, + "learning_rate": 1.0153677199880141e-07, + "loss": 0.1018, + "step": 13677 + }, + { + "epoch": 2.8681065212832877, + "grad_norm": 3.5456014790915598, + "learning_rate": 1.0121523329811756e-07, + "loss": 0.1182, + "step": 13678 + }, + { + "epoch": 2.868316208848815, + "grad_norm": 3.9180992896039615, + "learning_rate": 1.0089420192731958e-07, + "loss": 0.1486, + "step": 13679 + }, + { + "epoch": 2.8685258964143427, + "grad_norm": 4.356664955517061, + "learning_rate": 1.0057367790286321e-07, + "loss": 0.138, + "step": 13680 + }, + { + "epoch": 2.86873558397987, + "grad_norm": 4.064940342487734, + "learning_rate": 1.0025366124117641e-07, + "loss": 0.1364, + "step": 13681 + }, + { + "epoch": 2.8689452715453974, + "grad_norm": 4.3236333787680525, + "learning_rate": 9.993415195865941e-08, + "loss": 0.1442, + "step": 13682 + }, + { + "epoch": 2.869154959110925, + "grad_norm": 4.45191416256865, + "learning_rate": 9.961515007168687e-08, + "loss": 0.151, + "step": 13683 + }, + { + "epoch": 2.869364646676452, + "grad_norm": 3.294083362729306, + "learning_rate": 9.929665559661017e-08, + "loss": 0.0995, + "step": 13684 + }, + { + "epoch": 2.8695743342419795, + "grad_norm": 4.2225052346037035, + "learning_rate": 9.89786685497518e-08, + "loss": 0.1357, + "step": 13685 + }, + { + "epoch": 2.869784021807507, + "grad_norm": 3.9305347718127757, + "learning_rate": 9.866118894740873e-08, + "loss": 0.1286, + "step": 13686 + }, + { + "epoch": 2.869993709373034, + "grad_norm": 5.737158323653031, + "learning_rate": 9.834421680585349e-08, + "loss": 0.1399, + "step": 13687 + }, + { + "epoch": 2.8702033969385616, + "grad_norm": 4.756270010999447, + "learning_rate": 9.802775214132976e-08, + "loss": 0.1572, + "step": 13688 + }, + { + "epoch": 2.870413084504089, + "grad_norm": 4.233421918282514, + "learning_rate": 9.77117949700579e-08, + "loss": 0.141, + "step": 13689 + }, + { + "epoch": 2.8706227720696162, + "grad_norm": 4.952355570721876, + "learning_rate": 9.739634530823161e-08, + "loss": 0.1496, + "step": 13690 + }, + { + "epoch": 2.8708324596351438, + "grad_norm": 4.018547560313546, + "learning_rate": 9.708140317201798e-08, + "loss": 0.1279, + "step": 13691 + }, + { + "epoch": 2.871042147200671, + "grad_norm": 4.731739098056344, + "learning_rate": 9.676696857755741e-08, + "loss": 0.1572, + "step": 13692 + }, + { + "epoch": 2.8712518347661984, + "grad_norm": 4.966048422968939, + "learning_rate": 9.645304154096702e-08, + "loss": 0.1618, + "step": 13693 + }, + { + "epoch": 2.8714615223317255, + "grad_norm": 5.2444004576972905, + "learning_rate": 9.613962207833394e-08, + "loss": 0.1775, + "step": 13694 + }, + { + "epoch": 2.871671209897253, + "grad_norm": 3.616721146755147, + "learning_rate": 9.582671020572421e-08, + "loss": 0.1033, + "step": 13695 + }, + { + "epoch": 2.8718808974627805, + "grad_norm": 3.8408486694755655, + "learning_rate": 9.551430593917277e-08, + "loss": 0.104, + "step": 13696 + }, + { + "epoch": 2.8720905850283076, + "grad_norm": 4.105991190431302, + "learning_rate": 9.520240929469127e-08, + "loss": 0.1328, + "step": 13697 + }, + { + "epoch": 2.872300272593835, + "grad_norm": 4.4469561334081, + "learning_rate": 9.489102028826469e-08, + "loss": 0.1413, + "step": 13698 + }, + { + "epoch": 2.8725099601593627, + "grad_norm": 4.618809256452602, + "learning_rate": 9.458013893585472e-08, + "loss": 0.1706, + "step": 13699 + }, + { + "epoch": 2.8727196477248897, + "grad_norm": 3.7284777983080004, + "learning_rate": 9.426976525339082e-08, + "loss": 0.1176, + "step": 13700 + }, + { + "epoch": 2.8729293352904173, + "grad_norm": 3.3685397630278966, + "learning_rate": 9.39598992567825e-08, + "loss": 0.115, + "step": 13701 + }, + { + "epoch": 2.873139022855945, + "grad_norm": 3.5640782762719314, + "learning_rate": 9.365054096191151e-08, + "loss": 0.1076, + "step": 13702 + }, + { + "epoch": 2.873348710421472, + "grad_norm": 4.23863695515395, + "learning_rate": 9.33416903846307e-08, + "loss": 0.13, + "step": 13703 + }, + { + "epoch": 2.8735583979869994, + "grad_norm": 4.047976180621325, + "learning_rate": 9.303334754077075e-08, + "loss": 0.1148, + "step": 13704 + }, + { + "epoch": 2.873768085552527, + "grad_norm": 3.5901009762709357, + "learning_rate": 9.27255124461357e-08, + "loss": 0.1155, + "step": 13705 + }, + { + "epoch": 2.873977773118054, + "grad_norm": 4.966260848858032, + "learning_rate": 9.241818511649958e-08, + "loss": 0.1441, + "step": 13706 + }, + { + "epoch": 2.8741874606835816, + "grad_norm": 3.886238328796598, + "learning_rate": 9.211136556761646e-08, + "loss": 0.1092, + "step": 13707 + }, + { + "epoch": 2.874397148249109, + "grad_norm": 3.6683485106978684, + "learning_rate": 9.180505381521044e-08, + "loss": 0.1103, + "step": 13708 + }, + { + "epoch": 2.874606835814636, + "grad_norm": 3.872272491107465, + "learning_rate": 9.149924987498004e-08, + "loss": 0.1304, + "step": 13709 + }, + { + "epoch": 2.8748165233801637, + "grad_norm": 3.8082056448113555, + "learning_rate": 9.119395376259832e-08, + "loss": 0.1475, + "step": 13710 + }, + { + "epoch": 2.8750262109456908, + "grad_norm": 4.7427498974542095, + "learning_rate": 9.088916549371163e-08, + "loss": 0.1498, + "step": 13711 + }, + { + "epoch": 2.8752358985112183, + "grad_norm": 4.803805907711026, + "learning_rate": 9.058488508394304e-08, + "loss": 0.1642, + "step": 13712 + }, + { + "epoch": 2.8754455860767454, + "grad_norm": 5.149907046021405, + "learning_rate": 9.02811125488856e-08, + "loss": 0.1628, + "step": 13713 + }, + { + "epoch": 2.875655273642273, + "grad_norm": 4.874141858966118, + "learning_rate": 8.997784790410913e-08, + "loss": 0.1387, + "step": 13714 + }, + { + "epoch": 2.8758649612078004, + "grad_norm": 4.594539811031173, + "learning_rate": 8.96750911651545e-08, + "loss": 0.1265, + "step": 13715 + }, + { + "epoch": 2.8760746487733275, + "grad_norm": 4.651032619188963, + "learning_rate": 8.937284234754151e-08, + "loss": 0.1228, + "step": 13716 + }, + { + "epoch": 2.876284336338855, + "grad_norm": 5.052123448656687, + "learning_rate": 8.907110146676001e-08, + "loss": 0.1518, + "step": 13717 + }, + { + "epoch": 2.8764940239043826, + "grad_norm": 4.707168220470122, + "learning_rate": 8.876986853827208e-08, + "loss": 0.1506, + "step": 13718 + }, + { + "epoch": 2.8767037114699097, + "grad_norm": 3.56967054969513, + "learning_rate": 8.84691435775198e-08, + "loss": 0.118, + "step": 13719 + }, + { + "epoch": 2.876913399035437, + "grad_norm": 4.519674345350076, + "learning_rate": 8.816892659991528e-08, + "loss": 0.1423, + "step": 13720 + }, + { + "epoch": 2.8771230866009647, + "grad_norm": 4.455753854022332, + "learning_rate": 8.786921762084288e-08, + "loss": 0.1247, + "step": 13721 + }, + { + "epoch": 2.877332774166492, + "grad_norm": 3.527412898433023, + "learning_rate": 8.757001665566589e-08, + "loss": 0.12, + "step": 13722 + }, + { + "epoch": 2.8775424617320193, + "grad_norm": 5.534074890372293, + "learning_rate": 8.727132371971758e-08, + "loss": 0.1695, + "step": 13723 + }, + { + "epoch": 2.877752149297547, + "grad_norm": 4.990068682814447, + "learning_rate": 8.697313882830683e-08, + "loss": 0.1522, + "step": 13724 + }, + { + "epoch": 2.877961836863074, + "grad_norm": 4.334415369019435, + "learning_rate": 8.667546199671473e-08, + "loss": 0.1588, + "step": 13725 + }, + { + "epoch": 2.8781715244286015, + "grad_norm": 2.98322187906449, + "learning_rate": 8.63782932402002e-08, + "loss": 0.102, + "step": 13726 + }, + { + "epoch": 2.878381211994129, + "grad_norm": 4.484368980134519, + "learning_rate": 8.608163257399216e-08, + "loss": 0.1533, + "step": 13727 + }, + { + "epoch": 2.878590899559656, + "grad_norm": 4.7900535118592815, + "learning_rate": 8.578548001329622e-08, + "loss": 0.1409, + "step": 13728 + }, + { + "epoch": 2.8788005871251836, + "grad_norm": 4.01954994588884, + "learning_rate": 8.548983557328805e-08, + "loss": 0.099, + "step": 13729 + }, + { + "epoch": 2.8790102746907107, + "grad_norm": 5.4917598870929325, + "learning_rate": 8.519469926912327e-08, + "loss": 0.1627, + "step": 13730 + }, + { + "epoch": 2.879219962256238, + "grad_norm": 4.984366279029392, + "learning_rate": 8.490007111592536e-08, + "loss": 0.1383, + "step": 13731 + }, + { + "epoch": 2.8794296498217653, + "grad_norm": 3.6714831612389447, + "learning_rate": 8.460595112879555e-08, + "loss": 0.0931, + "step": 13732 + }, + { + "epoch": 2.879639337387293, + "grad_norm": 3.946642438472358, + "learning_rate": 8.431233932280958e-08, + "loss": 0.1052, + "step": 13733 + }, + { + "epoch": 2.8798490249528204, + "grad_norm": 4.045919873937914, + "learning_rate": 8.401923571301207e-08, + "loss": 0.1302, + "step": 13734 + }, + { + "epoch": 2.8800587125183474, + "grad_norm": 3.8497475079643655, + "learning_rate": 8.372664031442768e-08, + "loss": 0.0944, + "step": 13735 + }, + { + "epoch": 2.880268400083875, + "grad_norm": 3.6240896143379615, + "learning_rate": 8.343455314205329e-08, + "loss": 0.1475, + "step": 13736 + }, + { + "epoch": 2.8804780876494025, + "grad_norm": 3.891431118500464, + "learning_rate": 8.314297421085693e-08, + "loss": 0.125, + "step": 13737 + }, + { + "epoch": 2.8806877752149296, + "grad_norm": 4.9170641642873845, + "learning_rate": 8.28519035357822e-08, + "loss": 0.1436, + "step": 13738 + }, + { + "epoch": 2.880897462780457, + "grad_norm": 4.474136144980559, + "learning_rate": 8.25613411317483e-08, + "loss": 0.1383, + "step": 13739 + }, + { + "epoch": 2.8811071503459846, + "grad_norm": 5.154780107564171, + "learning_rate": 8.227128701364662e-08, + "loss": 0.1351, + "step": 13740 + }, + { + "epoch": 2.8813168379115117, + "grad_norm": 5.117974511806987, + "learning_rate": 8.198174119634306e-08, + "loss": 0.1823, + "step": 13741 + }, + { + "epoch": 2.8815265254770392, + "grad_norm": 3.690876287181319, + "learning_rate": 8.169270369467685e-08, + "loss": 0.1077, + "step": 13742 + }, + { + "epoch": 2.8817362130425668, + "grad_norm": 4.823497337386282, + "learning_rate": 8.140417452346283e-08, + "loss": 0.1821, + "step": 13743 + }, + { + "epoch": 2.881945900608094, + "grad_norm": 6.941536931131761, + "learning_rate": 8.111615369748693e-08, + "loss": 0.1342, + "step": 13744 + }, + { + "epoch": 2.8821555881736214, + "grad_norm": 3.437278536143222, + "learning_rate": 8.08286412315118e-08, + "loss": 0.1154, + "step": 13745 + }, + { + "epoch": 2.882365275739149, + "grad_norm": 4.598256447624964, + "learning_rate": 8.05416371402723e-08, + "loss": 0.1275, + "step": 13746 + }, + { + "epoch": 2.882574963304676, + "grad_norm": 3.8816519812415597, + "learning_rate": 8.02551414384789e-08, + "loss": 0.1171, + "step": 13747 + }, + { + "epoch": 2.8827846508702035, + "grad_norm": 3.6155420120416086, + "learning_rate": 7.99691541408143e-08, + "loss": 0.1164, + "step": 13748 + }, + { + "epoch": 2.8829943384357306, + "grad_norm": 4.289405757122124, + "learning_rate": 7.968367526193676e-08, + "loss": 0.1515, + "step": 13749 + }, + { + "epoch": 2.883204026001258, + "grad_norm": 3.8146134816238515, + "learning_rate": 7.939870481647571e-08, + "loss": 0.1102, + "step": 13750 + }, + { + "epoch": 2.883413713566785, + "grad_norm": 3.686923433923327, + "learning_rate": 7.911424281903723e-08, + "loss": 0.1196, + "step": 13751 + }, + { + "epoch": 2.8836234011323127, + "grad_norm": 3.4201529018802317, + "learning_rate": 7.883028928420188e-08, + "loss": 0.1105, + "step": 13752 + }, + { + "epoch": 2.8838330886978403, + "grad_norm": 4.684602793377559, + "learning_rate": 7.854684422652026e-08, + "loss": 0.135, + "step": 13753 + }, + { + "epoch": 2.8840427762633674, + "grad_norm": 3.3743559909457534, + "learning_rate": 7.826390766052184e-08, + "loss": 0.1093, + "step": 13754 + }, + { + "epoch": 2.884252463828895, + "grad_norm": 5.07677136829841, + "learning_rate": 7.798147960070612e-08, + "loss": 0.1603, + "step": 13755 + }, + { + "epoch": 2.8844621513944224, + "grad_norm": 4.319795751580264, + "learning_rate": 7.769956006154821e-08, + "loss": 0.1548, + "step": 13756 + }, + { + "epoch": 2.8846718389599495, + "grad_norm": 4.139001292683012, + "learning_rate": 7.741814905749767e-08, + "loss": 0.1317, + "step": 13757 + }, + { + "epoch": 2.884881526525477, + "grad_norm": 3.6525670994330754, + "learning_rate": 7.713724660297628e-08, + "loss": 0.12, + "step": 13758 + }, + { + "epoch": 2.8850912140910046, + "grad_norm": 5.127382731285871, + "learning_rate": 7.685685271238141e-08, + "loss": 0.1588, + "step": 13759 + }, + { + "epoch": 2.8853009016565316, + "grad_norm": 3.766361395381949, + "learning_rate": 7.65769674000849e-08, + "loss": 0.1171, + "step": 13760 + }, + { + "epoch": 2.885510589222059, + "grad_norm": 4.290458624274155, + "learning_rate": 7.629759068042863e-08, + "loss": 0.1138, + "step": 13761 + }, + { + "epoch": 2.8857202767875867, + "grad_norm": 3.971951778673911, + "learning_rate": 7.601872256773335e-08, + "loss": 0.1314, + "step": 13762 + }, + { + "epoch": 2.8859299643531138, + "grad_norm": 4.6176523666054425, + "learning_rate": 7.574036307629096e-08, + "loss": 0.146, + "step": 13763 + }, + { + "epoch": 2.8861396519186413, + "grad_norm": 3.9580045566408972, + "learning_rate": 7.546251222036671e-08, + "loss": 0.1262, + "step": 13764 + }, + { + "epoch": 2.886349339484169, + "grad_norm": 4.600645487202965, + "learning_rate": 7.518517001420256e-08, + "loss": 0.1318, + "step": 13765 + }, + { + "epoch": 2.886559027049696, + "grad_norm": 4.560487759266229, + "learning_rate": 7.490833647201268e-08, + "loss": 0.1682, + "step": 13766 + }, + { + "epoch": 2.8867687146152234, + "grad_norm": 4.061911558898941, + "learning_rate": 7.463201160798461e-08, + "loss": 0.112, + "step": 13767 + }, + { + "epoch": 2.8869784021807505, + "grad_norm": 4.184090925820003, + "learning_rate": 7.435619543627925e-08, + "loss": 0.1418, + "step": 13768 + }, + { + "epoch": 2.887188089746278, + "grad_norm": 5.73613067416364, + "learning_rate": 7.40808879710353e-08, + "loss": 0.1338, + "step": 13769 + }, + { + "epoch": 2.887397777311805, + "grad_norm": 3.5627548911905484, + "learning_rate": 7.380608922636145e-08, + "loss": 0.1053, + "step": 13770 + }, + { + "epoch": 2.8876074648773327, + "grad_norm": 4.208702908196676, + "learning_rate": 7.353179921634091e-08, + "loss": 0.1197, + "step": 13771 + }, + { + "epoch": 2.88781715244286, + "grad_norm": 5.590035603354481, + "learning_rate": 7.325801795503351e-08, + "loss": 0.1418, + "step": 13772 + }, + { + "epoch": 2.8880268400083873, + "grad_norm": 4.084587563667208, + "learning_rate": 7.298474545646917e-08, + "loss": 0.116, + "step": 13773 + }, + { + "epoch": 2.888236527573915, + "grad_norm": 4.686438539645786, + "learning_rate": 7.271198173465332e-08, + "loss": 0.0872, + "step": 13774 + }, + { + "epoch": 2.8884462151394423, + "grad_norm": 3.822956589497696, + "learning_rate": 7.243972680356814e-08, + "loss": 0.1371, + "step": 13775 + }, + { + "epoch": 2.8886559027049694, + "grad_norm": 5.197993450817628, + "learning_rate": 7.216798067716357e-08, + "loss": 0.1737, + "step": 13776 + }, + { + "epoch": 2.888865590270497, + "grad_norm": 4.259208436369405, + "learning_rate": 7.189674336936958e-08, + "loss": 0.1197, + "step": 13777 + }, + { + "epoch": 2.8890752778360245, + "grad_norm": 4.649931417656924, + "learning_rate": 7.162601489408728e-08, + "loss": 0.1443, + "step": 13778 + }, + { + "epoch": 2.8892849654015516, + "grad_norm": 3.85257100949017, + "learning_rate": 7.135579526519109e-08, + "loss": 0.1632, + "step": 13779 + }, + { + "epoch": 2.889494652967079, + "grad_norm": 3.811742704204405, + "learning_rate": 7.108608449653109e-08, + "loss": 0.1138, + "step": 13780 + }, + { + "epoch": 2.8897043405326066, + "grad_norm": 3.778131224519266, + "learning_rate": 7.081688260193065e-08, + "loss": 0.1406, + "step": 13781 + }, + { + "epoch": 2.8899140280981337, + "grad_norm": 3.79256460308423, + "learning_rate": 7.054818959518538e-08, + "loss": 0.1309, + "step": 13782 + }, + { + "epoch": 2.890123715663661, + "grad_norm": 4.793540105986759, + "learning_rate": 7.028000549006764e-08, + "loss": 0.1595, + "step": 13783 + }, + { + "epoch": 2.8903334032291887, + "grad_norm": 3.659232194122678, + "learning_rate": 7.001233030032195e-08, + "loss": 0.1216, + "step": 13784 + }, + { + "epoch": 2.890543090794716, + "grad_norm": 5.305010152163205, + "learning_rate": 6.974516403966624e-08, + "loss": 0.1375, + "step": 13785 + }, + { + "epoch": 2.8907527783602434, + "grad_norm": 4.532380999215023, + "learning_rate": 6.947850672179512e-08, + "loss": 0.117, + "step": 13786 + }, + { + "epoch": 2.8909624659257704, + "grad_norm": 5.073945022028327, + "learning_rate": 6.921235836037321e-08, + "loss": 0.1645, + "step": 13787 + }, + { + "epoch": 2.891172153491298, + "grad_norm": 5.683442423631372, + "learning_rate": 6.894671896904292e-08, + "loss": 0.1768, + "step": 13788 + }, + { + "epoch": 2.891381841056825, + "grad_norm": 4.370577394239539, + "learning_rate": 6.868158856141671e-08, + "loss": 0.1251, + "step": 13789 + }, + { + "epoch": 2.8915915286223526, + "grad_norm": 3.681890435938822, + "learning_rate": 6.841696715108481e-08, + "loss": 0.1079, + "step": 13790 + }, + { + "epoch": 2.89180121618788, + "grad_norm": 4.356078192171075, + "learning_rate": 6.81528547516086e-08, + "loss": 0.184, + "step": 13791 + }, + { + "epoch": 2.892010903753407, + "grad_norm": 5.938306152051643, + "learning_rate": 6.788925137652391e-08, + "loss": 0.1676, + "step": 13792 + }, + { + "epoch": 2.8922205913189347, + "grad_norm": 4.354426777975696, + "learning_rate": 6.762615703934217e-08, + "loss": 0.1464, + "step": 13793 + }, + { + "epoch": 2.8924302788844622, + "grad_norm": 4.491786503723854, + "learning_rate": 6.736357175354702e-08, + "loss": 0.1375, + "step": 13794 + }, + { + "epoch": 2.8926399664499893, + "grad_norm": 4.523607806926432, + "learning_rate": 6.710149553259548e-08, + "loss": 0.131, + "step": 13795 + }, + { + "epoch": 2.892849654015517, + "grad_norm": 4.52541757002651, + "learning_rate": 6.683992838992015e-08, + "loss": 0.1218, + "step": 13796 + }, + { + "epoch": 2.8930593415810444, + "grad_norm": 4.748950922398838, + "learning_rate": 6.657887033892695e-08, + "loss": 0.1455, + "step": 13797 + }, + { + "epoch": 2.8932690291465715, + "grad_norm": 4.290681213145596, + "learning_rate": 6.631832139299521e-08, + "loss": 0.1551, + "step": 13798 + }, + { + "epoch": 2.893478716712099, + "grad_norm": 5.030716401845468, + "learning_rate": 6.605828156547866e-08, + "loss": 0.1692, + "step": 13799 + }, + { + "epoch": 2.8936884042776265, + "grad_norm": 4.357220335680682, + "learning_rate": 6.579875086970444e-08, + "loss": 0.1397, + "step": 13800 + }, + { + "epoch": 2.8938980918431536, + "grad_norm": 3.429329552577863, + "learning_rate": 6.553972931897412e-08, + "loss": 0.1189, + "step": 13801 + }, + { + "epoch": 2.894107779408681, + "grad_norm": 3.297131085647814, + "learning_rate": 6.528121692656375e-08, + "loss": 0.0964, + "step": 13802 + }, + { + "epoch": 2.8943174669742087, + "grad_norm": 4.453298450216547, + "learning_rate": 6.502321370572162e-08, + "loss": 0.1459, + "step": 13803 + }, + { + "epoch": 2.8945271545397357, + "grad_norm": 4.725714378085602, + "learning_rate": 6.47657196696705e-08, + "loss": 0.142, + "step": 13804 + }, + { + "epoch": 2.8947368421052633, + "grad_norm": 4.80071577620342, + "learning_rate": 6.450873483160869e-08, + "loss": 0.1317, + "step": 13805 + }, + { + "epoch": 2.8949465296707904, + "grad_norm": 4.1107128031699895, + "learning_rate": 6.425225920470679e-08, + "loss": 0.0985, + "step": 13806 + }, + { + "epoch": 2.895156217236318, + "grad_norm": 3.6194322138437522, + "learning_rate": 6.399629280210873e-08, + "loss": 0.1375, + "step": 13807 + }, + { + "epoch": 2.8953659048018454, + "grad_norm": 4.882275750413403, + "learning_rate": 6.374083563693401e-08, + "loss": 0.1952, + "step": 13808 + }, + { + "epoch": 2.8955755923673725, + "grad_norm": 5.317718561990381, + "learning_rate": 6.348588772227548e-08, + "loss": 0.1493, + "step": 13809 + }, + { + "epoch": 2.8957852799329, + "grad_norm": 3.6934150767753158, + "learning_rate": 6.323144907119827e-08, + "loss": 0.1394, + "step": 13810 + }, + { + "epoch": 2.895994967498427, + "grad_norm": 3.5514128854193987, + "learning_rate": 6.297751969674415e-08, + "loss": 0.128, + "step": 13811 + }, + { + "epoch": 2.8962046550639546, + "grad_norm": 4.516790239362552, + "learning_rate": 6.272409961192716e-08, + "loss": 0.1338, + "step": 13812 + }, + { + "epoch": 2.896414342629482, + "grad_norm": 3.5341204487815614, + "learning_rate": 6.247118882973468e-08, + "loss": 0.1209, + "step": 13813 + }, + { + "epoch": 2.8966240301950092, + "grad_norm": 3.9490917623555966, + "learning_rate": 6.22187873631297e-08, + "loss": 0.1036, + "step": 13814 + }, + { + "epoch": 2.8968337177605368, + "grad_norm": 4.233294170353973, + "learning_rate": 6.196689522504962e-08, + "loss": 0.1395, + "step": 13815 + }, + { + "epoch": 2.8970434053260643, + "grad_norm": 4.446203336134346, + "learning_rate": 6.17155124284019e-08, + "loss": 0.151, + "step": 13816 + }, + { + "epoch": 2.8972530928915914, + "grad_norm": 3.5325822238506888, + "learning_rate": 6.146463898607069e-08, + "loss": 0.1092, + "step": 13817 + }, + { + "epoch": 2.897462780457119, + "grad_norm": 4.846831844204268, + "learning_rate": 6.121427491091458e-08, + "loss": 0.1551, + "step": 13818 + }, + { + "epoch": 2.8976724680226464, + "grad_norm": 4.000238338136815, + "learning_rate": 6.096442021576443e-08, + "loss": 0.1474, + "step": 13819 + }, + { + "epoch": 2.8978821555881735, + "grad_norm": 4.955699476281586, + "learning_rate": 6.071507491342665e-08, + "loss": 0.1372, + "step": 13820 + }, + { + "epoch": 2.898091843153701, + "grad_norm": 3.542579338749896, + "learning_rate": 6.046623901667992e-08, + "loss": 0.1156, + "step": 13821 + }, + { + "epoch": 2.8983015307192286, + "grad_norm": 3.7799674950957827, + "learning_rate": 6.021791253827736e-08, + "loss": 0.1193, + "step": 13822 + }, + { + "epoch": 2.8985112182847557, + "grad_norm": 4.202857556972337, + "learning_rate": 5.997009549094768e-08, + "loss": 0.1248, + "step": 13823 + }, + { + "epoch": 2.898720905850283, + "grad_norm": 4.364355450884744, + "learning_rate": 5.972278788738961e-08, + "loss": 0.1512, + "step": 13824 + }, + { + "epoch": 2.8989305934158107, + "grad_norm": 2.853207299829324, + "learning_rate": 5.947598974027968e-08, + "loss": 0.0813, + "step": 13825 + }, + { + "epoch": 2.899140280981338, + "grad_norm": 4.566298793849188, + "learning_rate": 5.922970106226555e-08, + "loss": 0.1471, + "step": 13826 + }, + { + "epoch": 2.8993499685468653, + "grad_norm": 5.040366639444228, + "learning_rate": 5.898392186597157e-08, + "loss": 0.1305, + "step": 13827 + }, + { + "epoch": 2.8995596561123924, + "grad_norm": 4.525005780057943, + "learning_rate": 5.873865216399322e-08, + "loss": 0.1175, + "step": 13828 + }, + { + "epoch": 2.89976934367792, + "grad_norm": 4.590534552067123, + "learning_rate": 5.849389196890043e-08, + "loss": 0.1329, + "step": 13829 + }, + { + "epoch": 2.899979031243447, + "grad_norm": 4.094531859106882, + "learning_rate": 5.8249641293238733e-08, + "loss": 0.1175, + "step": 13830 + }, + { + "epoch": 2.9001887188089746, + "grad_norm": 3.823303721530756, + "learning_rate": 5.80059001495259e-08, + "loss": 0.1231, + "step": 13831 + }, + { + "epoch": 2.900398406374502, + "grad_norm": 4.307276652204052, + "learning_rate": 5.776266855025525e-08, + "loss": 0.1367, + "step": 13832 + }, + { + "epoch": 2.900608093940029, + "grad_norm": 3.6817124414143034, + "learning_rate": 5.751994650789017e-08, + "loss": 0.1153, + "step": 13833 + }, + { + "epoch": 2.9008177815055567, + "grad_norm": 4.0263924498439785, + "learning_rate": 5.727773403487291e-08, + "loss": 0.1505, + "step": 13834 + }, + { + "epoch": 2.901027469071084, + "grad_norm": 4.945908552617482, + "learning_rate": 5.7036031143616887e-08, + "loss": 0.1921, + "step": 13835 + }, + { + "epoch": 2.9012371566366113, + "grad_norm": 4.108744952475568, + "learning_rate": 5.679483784650996e-08, + "loss": 0.1076, + "step": 13836 + }, + { + "epoch": 2.901446844202139, + "grad_norm": 4.4400271410893835, + "learning_rate": 5.655415415591226e-08, + "loss": 0.1371, + "step": 13837 + }, + { + "epoch": 2.9016565317676664, + "grad_norm": 4.797076049437198, + "learning_rate": 5.631398008415945e-08, + "loss": 0.1434, + "step": 13838 + }, + { + "epoch": 2.9018662193331934, + "grad_norm": 3.7751024552391326, + "learning_rate": 5.60743156435628e-08, + "loss": 0.1315, + "step": 13839 + }, + { + "epoch": 2.902075906898721, + "grad_norm": 4.186904641720187, + "learning_rate": 5.583516084640472e-08, + "loss": 0.1533, + "step": 13840 + }, + { + "epoch": 2.9022855944642485, + "grad_norm": 5.739363903527498, + "learning_rate": 5.559651570494096e-08, + "loss": 0.1479, + "step": 13841 + }, + { + "epoch": 2.9024952820297756, + "grad_norm": 4.2788848719969135, + "learning_rate": 5.535838023140394e-08, + "loss": 0.1403, + "step": 13842 + }, + { + "epoch": 2.902704969595303, + "grad_norm": 3.430011620513265, + "learning_rate": 5.5120754437997245e-08, + "loss": 0.092, + "step": 13843 + }, + { + "epoch": 2.9029146571608306, + "grad_norm": 4.518674634722797, + "learning_rate": 5.4883638336901135e-08, + "loss": 0.1527, + "step": 13844 + }, + { + "epoch": 2.9031243447263577, + "grad_norm": 4.162951439508755, + "learning_rate": 5.46470319402681e-08, + "loss": 0.1324, + "step": 13845 + }, + { + "epoch": 2.9033340322918852, + "grad_norm": 4.386123205141205, + "learning_rate": 5.4410935260222894e-08, + "loss": 0.1002, + "step": 13846 + }, + { + "epoch": 2.9035437198574123, + "grad_norm": 4.0480535192231715, + "learning_rate": 5.417534830886695e-08, + "loss": 0.14, + "step": 13847 + }, + { + "epoch": 2.90375340742294, + "grad_norm": 3.7826651089442933, + "learning_rate": 5.394027109827504e-08, + "loss": 0.1208, + "step": 13848 + }, + { + "epoch": 2.903963094988467, + "grad_norm": 3.678151201637555, + "learning_rate": 5.3705703640495324e-08, + "loss": 0.1318, + "step": 13849 + }, + { + "epoch": 2.9041727825539945, + "grad_norm": 3.5182865330739075, + "learning_rate": 5.347164594754817e-08, + "loss": 0.1136, + "step": 13850 + }, + { + "epoch": 2.904382470119522, + "grad_norm": 4.064808322572299, + "learning_rate": 5.3238098031430655e-08, + "loss": 0.1105, + "step": 13851 + }, + { + "epoch": 2.904592157685049, + "grad_norm": 5.281592986083384, + "learning_rate": 5.30050599041132e-08, + "loss": 0.1616, + "step": 13852 + }, + { + "epoch": 2.9048018452505766, + "grad_norm": 5.943235721674502, + "learning_rate": 5.2772531577538475e-08, + "loss": 0.1454, + "step": 13853 + }, + { + "epoch": 2.905011532816104, + "grad_norm": 3.5064856709648153, + "learning_rate": 5.254051306362362e-08, + "loss": 0.1006, + "step": 13854 + }, + { + "epoch": 2.905221220381631, + "grad_norm": 5.404028522927519, + "learning_rate": 5.230900437426134e-08, + "loss": 0.18, + "step": 13855 + }, + { + "epoch": 2.9054309079471587, + "grad_norm": 6.1948843858583835, + "learning_rate": 5.2078005521315477e-08, + "loss": 0.1746, + "step": 13856 + }, + { + "epoch": 2.9056405955126863, + "grad_norm": 3.4122949095594595, + "learning_rate": 5.184751651662545e-08, + "loss": 0.0912, + "step": 13857 + }, + { + "epoch": 2.9058502830782134, + "grad_norm": 5.30846854473473, + "learning_rate": 5.161753737200403e-08, + "loss": 0.1461, + "step": 13858 + }, + { + "epoch": 2.906059970643741, + "grad_norm": 3.9785285506134485, + "learning_rate": 5.1388068099238465e-08, + "loss": 0.1301, + "step": 13859 + }, + { + "epoch": 2.9062696582092684, + "grad_norm": 4.20944891432764, + "learning_rate": 5.115910871009045e-08, + "loss": 0.1384, + "step": 13860 + }, + { + "epoch": 2.9064793457747955, + "grad_norm": 4.480922277547946, + "learning_rate": 5.093065921629281e-08, + "loss": 0.1462, + "step": 13861 + }, + { + "epoch": 2.906689033340323, + "grad_norm": 4.278201349934148, + "learning_rate": 5.070271962955509e-08, + "loss": 0.1201, + "step": 13862 + }, + { + "epoch": 2.9068987209058506, + "grad_norm": 4.651361254412583, + "learning_rate": 5.0475289961557926e-08, + "loss": 0.141, + "step": 13863 + }, + { + "epoch": 2.9071084084713776, + "grad_norm": 4.22654048225129, + "learning_rate": 5.0248370223959786e-08, + "loss": 0.1356, + "step": 13864 + }, + { + "epoch": 2.907318096036905, + "grad_norm": 3.24158329116449, + "learning_rate": 5.0021960428390245e-08, + "loss": 0.1053, + "step": 13865 + }, + { + "epoch": 2.9075277836024322, + "grad_norm": 4.4276829740959185, + "learning_rate": 4.9796060586451146e-08, + "loss": 0.147, + "step": 13866 + }, + { + "epoch": 2.9077374711679598, + "grad_norm": 4.670308421373421, + "learning_rate": 4.95706707097221e-08, + "loss": 0.1478, + "step": 13867 + }, + { + "epoch": 2.907947158733487, + "grad_norm": 4.05677502817684, + "learning_rate": 4.934579080975499e-08, + "loss": 0.1283, + "step": 13868 + }, + { + "epoch": 2.9081568462990144, + "grad_norm": 4.501895783570169, + "learning_rate": 4.9121420898073924e-08, + "loss": 0.1794, + "step": 13869 + }, + { + "epoch": 2.908366533864542, + "grad_norm": 4.106850386948959, + "learning_rate": 4.8897560986179706e-08, + "loss": 0.1453, + "step": 13870 + }, + { + "epoch": 2.908576221430069, + "grad_norm": 4.2627163870116815, + "learning_rate": 4.867421108554427e-08, + "loss": 0.1363, + "step": 13871 + }, + { + "epoch": 2.9087859089955965, + "grad_norm": 3.5400293298784837, + "learning_rate": 4.845137120761512e-08, + "loss": 0.0897, + "step": 13872 + }, + { + "epoch": 2.908995596561124, + "grad_norm": 5.8029248200901025, + "learning_rate": 4.822904136381312e-08, + "loss": 0.2023, + "step": 13873 + }, + { + "epoch": 2.909205284126651, + "grad_norm": 3.275673715138521, + "learning_rate": 4.800722156553361e-08, + "loss": 0.0994, + "step": 13874 + }, + { + "epoch": 2.9094149716921787, + "grad_norm": 4.386439585549887, + "learning_rate": 4.778591182414527e-08, + "loss": 0.1242, + "step": 13875 + }, + { + "epoch": 2.909624659257706, + "grad_norm": 4.7506408851180035, + "learning_rate": 4.756511215098902e-08, + "loss": 0.163, + "step": 13876 + }, + { + "epoch": 2.9098343468232333, + "grad_norm": 4.985320959921978, + "learning_rate": 4.734482255738359e-08, + "loss": 0.1488, + "step": 13877 + }, + { + "epoch": 2.910044034388761, + "grad_norm": 5.124547729074882, + "learning_rate": 4.7125043054617734e-08, + "loss": 0.1471, + "step": 13878 + }, + { + "epoch": 2.9102537219542883, + "grad_norm": 5.114608704899934, + "learning_rate": 4.6905773653955764e-08, + "loss": 0.136, + "step": 13879 + }, + { + "epoch": 2.9104634095198154, + "grad_norm": 4.506309649445351, + "learning_rate": 4.6687014366635363e-08, + "loss": 0.1436, + "step": 13880 + }, + { + "epoch": 2.910673097085343, + "grad_norm": 4.053470722084693, + "learning_rate": 4.646876520386867e-08, + "loss": 0.1426, + "step": 13881 + }, + { + "epoch": 2.9108827846508705, + "grad_norm": 4.134292669077508, + "learning_rate": 4.6251026176841186e-08, + "loss": 0.1392, + "step": 13882 + }, + { + "epoch": 2.9110924722163976, + "grad_norm": 3.9967663684574695, + "learning_rate": 4.603379729671287e-08, + "loss": 0.1261, + "step": 13883 + }, + { + "epoch": 2.911302159781925, + "grad_norm": 3.758979553684162, + "learning_rate": 4.5817078574615926e-08, + "loss": 0.1299, + "step": 13884 + }, + { + "epoch": 2.911511847347452, + "grad_norm": 3.982747293125947, + "learning_rate": 4.560087002165925e-08, + "loss": 0.1208, + "step": 13885 + }, + { + "epoch": 2.9117215349129797, + "grad_norm": 4.438562871389678, + "learning_rate": 4.5385171648922864e-08, + "loss": 0.1231, + "step": 13886 + }, + { + "epoch": 2.9119312224785068, + "grad_norm": 3.768071921602696, + "learning_rate": 4.516998346746126e-08, + "loss": 0.1314, + "step": 13887 + }, + { + "epoch": 2.9121409100440343, + "grad_norm": 5.303100985996429, + "learning_rate": 4.495530548830451e-08, + "loss": 0.1632, + "step": 13888 + }, + { + "epoch": 2.912350597609562, + "grad_norm": 3.6550054292674794, + "learning_rate": 4.474113772245381e-08, + "loss": 0.1212, + "step": 13889 + }, + { + "epoch": 2.912560285175089, + "grad_norm": 3.974327290387412, + "learning_rate": 4.452748018088704e-08, + "loss": 0.1219, + "step": 13890 + }, + { + "epoch": 2.9127699727406164, + "grad_norm": 5.213483999082169, + "learning_rate": 4.431433287455322e-08, + "loss": 0.1525, + "step": 13891 + }, + { + "epoch": 2.912979660306144, + "grad_norm": 3.425745334716992, + "learning_rate": 4.4101695814378066e-08, + "loss": 0.1329, + "step": 13892 + }, + { + "epoch": 2.913189347871671, + "grad_norm": 3.5104345243880095, + "learning_rate": 4.3889569011257296e-08, + "loss": 0.0966, + "step": 13893 + }, + { + "epoch": 2.9133990354371986, + "grad_norm": 4.969545049795125, + "learning_rate": 4.367795247606443e-08, + "loss": 0.1629, + "step": 13894 + }, + { + "epoch": 2.913608723002726, + "grad_norm": 3.4012278642741545, + "learning_rate": 4.346684621964636e-08, + "loss": 0.1196, + "step": 13895 + }, + { + "epoch": 2.913818410568253, + "grad_norm": 4.59595447939931, + "learning_rate": 4.3256250252818874e-08, + "loss": 0.1269, + "step": 13896 + }, + { + "epoch": 2.9140280981337807, + "grad_norm": 4.255845890378312, + "learning_rate": 4.304616458637889e-08, + "loss": 0.1475, + "step": 13897 + }, + { + "epoch": 2.9142377856993082, + "grad_norm": 4.566365290259029, + "learning_rate": 4.283658923109335e-08, + "loss": 0.1582, + "step": 13898 + }, + { + "epoch": 2.9144474732648353, + "grad_norm": 4.368190747496577, + "learning_rate": 4.262752419770033e-08, + "loss": 0.1482, + "step": 13899 + }, + { + "epoch": 2.914657160830363, + "grad_norm": 4.859932985690303, + "learning_rate": 4.241896949691793e-08, + "loss": 0.1771, + "step": 13900 + }, + { + "epoch": 2.9148668483958904, + "grad_norm": 3.5907133471378576, + "learning_rate": 4.2210925139434255e-08, + "loss": 0.1185, + "step": 13901 + }, + { + "epoch": 2.9150765359614175, + "grad_norm": 7.081136643831514, + "learning_rate": 4.200339113591079e-08, + "loss": 0.1642, + "step": 13902 + }, + { + "epoch": 2.915286223526945, + "grad_norm": 4.187403115532427, + "learning_rate": 4.1796367496985677e-08, + "loss": 0.1366, + "step": 13903 + }, + { + "epoch": 2.915495911092472, + "grad_norm": 5.160077260832875, + "learning_rate": 4.1589854233268225e-08, + "loss": 0.1804, + "step": 13904 + }, + { + "epoch": 2.9157055986579996, + "grad_norm": 4.262358913060092, + "learning_rate": 4.138385135534218e-08, + "loss": 0.1318, + "step": 13905 + }, + { + "epoch": 2.9159152862235267, + "grad_norm": 3.5770074583303066, + "learning_rate": 4.117835887376576e-08, + "loss": 0.1054, + "step": 13906 + }, + { + "epoch": 2.916124973789054, + "grad_norm": 3.050250517123977, + "learning_rate": 4.0973376799072764e-08, + "loss": 0.0912, + "step": 13907 + }, + { + "epoch": 2.9163346613545817, + "grad_norm": 3.860480659267379, + "learning_rate": 4.0768905141765905e-08, + "loss": 0.1206, + "step": 13908 + }, + { + "epoch": 2.916544348920109, + "grad_norm": 4.584263218329184, + "learning_rate": 4.0564943912325686e-08, + "loss": 0.1423, + "step": 13909 + }, + { + "epoch": 2.9167540364856364, + "grad_norm": 3.5705741797122394, + "learning_rate": 4.0361493121207074e-08, + "loss": 0.1154, + "step": 13910 + }, + { + "epoch": 2.916963724051164, + "grad_norm": 5.100120065401551, + "learning_rate": 4.015855277883507e-08, + "loss": 0.2026, + "step": 13911 + }, + { + "epoch": 2.917173411616691, + "grad_norm": 4.196606669747442, + "learning_rate": 3.995612289561246e-08, + "loss": 0.126, + "step": 13912 + }, + { + "epoch": 2.9173830991822185, + "grad_norm": 3.589288420317563, + "learning_rate": 3.975420348191317e-08, + "loss": 0.131, + "step": 13913 + }, + { + "epoch": 2.917592786747746, + "grad_norm": 3.7852268865655376, + "learning_rate": 3.955279454808558e-08, + "loss": 0.1126, + "step": 13914 + }, + { + "epoch": 2.917802474313273, + "grad_norm": 4.663914447100712, + "learning_rate": 3.935189610445367e-08, + "loss": 0.1307, + "step": 13915 + }, + { + "epoch": 2.9180121618788006, + "grad_norm": 4.7435244379552755, + "learning_rate": 3.915150816131252e-08, + "loss": 0.1392, + "step": 13916 + }, + { + "epoch": 2.918221849444328, + "grad_norm": 3.76151980783133, + "learning_rate": 3.895163072893282e-08, + "loss": 0.0946, + "step": 13917 + }, + { + "epoch": 2.9184315370098552, + "grad_norm": 4.2346506394165715, + "learning_rate": 3.8752263817558586e-08, + "loss": 0.1546, + "step": 13918 + }, + { + "epoch": 2.9186412245753828, + "grad_norm": 3.7144545055107625, + "learning_rate": 3.855340743740832e-08, + "loss": 0.1204, + "step": 13919 + }, + { + "epoch": 2.9188509121409103, + "grad_norm": 3.5046941695673888, + "learning_rate": 3.835506159867164e-08, + "loss": 0.1198, + "step": 13920 + }, + { + "epoch": 2.9190605997064374, + "grad_norm": 4.6708525694880585, + "learning_rate": 3.815722631151708e-08, + "loss": 0.1518, + "step": 13921 + }, + { + "epoch": 2.919270287271965, + "grad_norm": 4.127380332846929, + "learning_rate": 3.7959901586082096e-08, + "loss": 0.1363, + "step": 13922 + }, + { + "epoch": 2.919479974837492, + "grad_norm": 4.66279199582513, + "learning_rate": 3.7763087432480805e-08, + "loss": 0.1692, + "step": 13923 + }, + { + "epoch": 2.9196896624030195, + "grad_norm": 4.947903623887908, + "learning_rate": 3.7566783860798485e-08, + "loss": 0.1626, + "step": 13924 + }, + { + "epoch": 2.9198993499685466, + "grad_norm": 4.0221018937736, + "learning_rate": 3.73709908810993e-08, + "loss": 0.137, + "step": 13925 + }, + { + "epoch": 2.920109037534074, + "grad_norm": 3.910044626830158, + "learning_rate": 3.717570850341523e-08, + "loss": 0.1048, + "step": 13926 + }, + { + "epoch": 2.9203187250996017, + "grad_norm": 4.458567824605001, + "learning_rate": 3.698093673775605e-08, + "loss": 0.1181, + "step": 13927 + }, + { + "epoch": 2.9205284126651287, + "grad_norm": 5.213427102620994, + "learning_rate": 3.678667559410376e-08, + "loss": 0.1819, + "step": 13928 + }, + { + "epoch": 2.9207381002306563, + "grad_norm": 5.659916239836797, + "learning_rate": 3.659292508241485e-08, + "loss": 0.1268, + "step": 13929 + }, + { + "epoch": 2.920947787796184, + "grad_norm": 4.477291676708218, + "learning_rate": 3.639968521261805e-08, + "loss": 0.1506, + "step": 13930 + }, + { + "epoch": 2.921157475361711, + "grad_norm": 5.449949705498544, + "learning_rate": 3.620695599461987e-08, + "loss": 0.1412, + "step": 13931 + }, + { + "epoch": 2.9213671629272384, + "grad_norm": 4.394465920280594, + "learning_rate": 3.601473743829576e-08, + "loss": 0.1508, + "step": 13932 + }, + { + "epoch": 2.921576850492766, + "grad_norm": 2.816249277826936, + "learning_rate": 3.582302955349781e-08, + "loss": 0.1038, + "step": 13933 + }, + { + "epoch": 2.921786538058293, + "grad_norm": 4.175153740127488, + "learning_rate": 3.563183235005263e-08, + "loss": 0.1137, + "step": 13934 + }, + { + "epoch": 2.9219962256238206, + "grad_norm": 3.33408377772412, + "learning_rate": 3.544114583775682e-08, + "loss": 0.1333, + "step": 13935 + }, + { + "epoch": 2.922205913189348, + "grad_norm": 4.340920796641978, + "learning_rate": 3.525097002638589e-08, + "loss": 0.1344, + "step": 13936 + }, + { + "epoch": 2.922415600754875, + "grad_norm": 3.6627366153789804, + "learning_rate": 3.506130492568649e-08, + "loss": 0.1045, + "step": 13937 + }, + { + "epoch": 2.9226252883204027, + "grad_norm": 2.814032885182091, + "learning_rate": 3.48721505453764e-08, + "loss": 0.0935, + "step": 13938 + }, + { + "epoch": 2.92283497588593, + "grad_norm": 6.412692403706989, + "learning_rate": 3.468350689515343e-08, + "loss": 0.1779, + "step": 13939 + }, + { + "epoch": 2.9230446634514573, + "grad_norm": 4.920794573151626, + "learning_rate": 3.4495373984685385e-08, + "loss": 0.1331, + "step": 13940 + }, + { + "epoch": 2.923254351016985, + "grad_norm": 4.904411245461035, + "learning_rate": 3.4307751823613454e-08, + "loss": 0.1597, + "step": 13941 + }, + { + "epoch": 2.923464038582512, + "grad_norm": 4.414442687097795, + "learning_rate": 3.412064042155327e-08, + "loss": 0.1225, + "step": 13942 + }, + { + "epoch": 2.9236737261480394, + "grad_norm": 5.052710193565847, + "learning_rate": 3.393403978809495e-08, + "loss": 0.1619, + "step": 13943 + }, + { + "epoch": 2.9238834137135665, + "grad_norm": 3.4007471547207544, + "learning_rate": 3.374794993280306e-08, + "loss": 0.1331, + "step": 13944 + }, + { + "epoch": 2.924093101279094, + "grad_norm": 3.748072517271522, + "learning_rate": 3.356237086521441e-08, + "loss": 0.1373, + "step": 13945 + }, + { + "epoch": 2.9243027888446216, + "grad_norm": 3.7353635029820103, + "learning_rate": 3.3377302594839176e-08, + "loss": 0.1249, + "step": 13946 + }, + { + "epoch": 2.9245124764101487, + "grad_norm": 3.416482306283054, + "learning_rate": 3.3192745131164216e-08, + "loss": 0.1291, + "step": 13947 + }, + { + "epoch": 2.924722163975676, + "grad_norm": 6.730602859460566, + "learning_rate": 3.300869848364752e-08, + "loss": 0.1552, + "step": 13948 + }, + { + "epoch": 2.9249318515412037, + "grad_norm": 4.316560544410731, + "learning_rate": 3.282516266172264e-08, + "loss": 0.1118, + "step": 13949 + }, + { + "epoch": 2.925141539106731, + "grad_norm": 3.5773092827211177, + "learning_rate": 3.264213767479429e-08, + "loss": 0.1204, + "step": 13950 + }, + { + "epoch": 2.9253512266722583, + "grad_norm": 2.937889099626319, + "learning_rate": 3.2459623532243856e-08, + "loss": 0.0917, + "step": 13951 + }, + { + "epoch": 2.925560914237786, + "grad_norm": 3.825910843634259, + "learning_rate": 3.227762024342607e-08, + "loss": 0.1049, + "step": 13952 + }, + { + "epoch": 2.925770601803313, + "grad_norm": 4.3420650901909905, + "learning_rate": 3.209612781766791e-08, + "loss": 0.142, + "step": 13953 + }, + { + "epoch": 2.9259802893688405, + "grad_norm": 3.7042870581956504, + "learning_rate": 3.191514626427195e-08, + "loss": 0.0924, + "step": 13954 + }, + { + "epoch": 2.926189976934368, + "grad_norm": 4.123512173707741, + "learning_rate": 3.173467559251298e-08, + "loss": 0.113, + "step": 13955 + }, + { + "epoch": 2.926399664499895, + "grad_norm": 3.389649384580066, + "learning_rate": 3.155471581164138e-08, + "loss": 0.1012, + "step": 13956 + }, + { + "epoch": 2.9266093520654226, + "grad_norm": 5.268738340406482, + "learning_rate": 3.1375266930880884e-08, + "loss": 0.1404, + "step": 13957 + }, + { + "epoch": 2.92681903963095, + "grad_norm": 3.2101388296053615, + "learning_rate": 3.1196328959426367e-08, + "loss": 0.1056, + "step": 13958 + }, + { + "epoch": 2.927028727196477, + "grad_norm": 4.122237853021114, + "learning_rate": 3.101790190645049e-08, + "loss": 0.1306, + "step": 13959 + }, + { + "epoch": 2.9272384147620047, + "grad_norm": 3.9863838285827797, + "learning_rate": 3.0839985781098144e-08, + "loss": 0.1207, + "step": 13960 + }, + { + "epoch": 2.927448102327532, + "grad_norm": 4.4089049057234, + "learning_rate": 3.0662580592486503e-08, + "loss": 0.148, + "step": 13961 + }, + { + "epoch": 2.9276577898930594, + "grad_norm": 3.9135444398438355, + "learning_rate": 3.048568634970828e-08, + "loss": 0.1343, + "step": 13962 + }, + { + "epoch": 2.9278674774585864, + "grad_norm": 5.186688028911975, + "learning_rate": 3.030930306182955e-08, + "loss": 0.1549, + "step": 13963 + }, + { + "epoch": 2.928077165024114, + "grad_norm": 6.640647039937862, + "learning_rate": 3.013343073789088e-08, + "loss": 0.1689, + "step": 13964 + }, + { + "epoch": 2.9282868525896415, + "grad_norm": 3.464229211126237, + "learning_rate": 2.995806938690615e-08, + "loss": 0.1127, + "step": 13965 + }, + { + "epoch": 2.9284965401551686, + "grad_norm": 3.4243503063960214, + "learning_rate": 2.978321901786263e-08, + "loss": 0.0996, + "step": 13966 + }, + { + "epoch": 2.928706227720696, + "grad_norm": 4.351702447743387, + "learning_rate": 2.9608879639722032e-08, + "loss": 0.1344, + "step": 13967 + }, + { + "epoch": 2.9289159152862236, + "grad_norm": 3.3542825491752444, + "learning_rate": 2.9435051261417213e-08, + "loss": 0.1062, + "step": 13968 + }, + { + "epoch": 2.9291256028517507, + "grad_norm": 4.08233877682267, + "learning_rate": 2.9261733891861043e-08, + "loss": 0.1421, + "step": 13969 + }, + { + "epoch": 2.9293352904172782, + "grad_norm": 3.474413169464223, + "learning_rate": 2.9088927539934197e-08, + "loss": 0.0908, + "step": 13970 + }, + { + "epoch": 2.9295449779828058, + "grad_norm": 3.3475353472019425, + "learning_rate": 2.8916632214491813e-08, + "loss": 0.1253, + "step": 13971 + }, + { + "epoch": 2.929754665548333, + "grad_norm": 4.933767218081028, + "learning_rate": 2.8744847924366825e-08, + "loss": 0.1778, + "step": 13972 + }, + { + "epoch": 2.9299643531138604, + "grad_norm": 4.478609169159007, + "learning_rate": 2.8573574678363304e-08, + "loss": 0.1525, + "step": 13973 + }, + { + "epoch": 2.930174040679388, + "grad_norm": 3.3138384669875895, + "learning_rate": 2.840281248525867e-08, + "loss": 0.099, + "step": 13974 + }, + { + "epoch": 2.930383728244915, + "grad_norm": 3.3888198157895295, + "learning_rate": 2.82325613538037e-08, + "loss": 0.1241, + "step": 13975 + }, + { + "epoch": 2.9305934158104425, + "grad_norm": 3.7459067474753205, + "learning_rate": 2.8062821292725862e-08, + "loss": 0.1008, + "step": 13976 + }, + { + "epoch": 2.93080310337597, + "grad_norm": 5.009107959812513, + "learning_rate": 2.789359231072486e-08, + "loss": 0.1735, + "step": 13977 + }, + { + "epoch": 2.931012790941497, + "grad_norm": 3.656686547140454, + "learning_rate": 2.7724874416471536e-08, + "loss": 0.1044, + "step": 13978 + }, + { + "epoch": 2.9312224785070247, + "grad_norm": 4.434217021526881, + "learning_rate": 2.755666761861453e-08, + "loss": 0.1709, + "step": 13979 + }, + { + "epoch": 2.9314321660725517, + "grad_norm": 4.5129939733528825, + "learning_rate": 2.738897192577472e-08, + "loss": 0.1376, + "step": 13980 + }, + { + "epoch": 2.9316418536380793, + "grad_norm": 5.075825057230401, + "learning_rate": 2.7221787346547455e-08, + "loss": 0.1238, + "step": 13981 + }, + { + "epoch": 2.9318515412036064, + "grad_norm": 3.3560917976551896, + "learning_rate": 2.7055113889499218e-08, + "loss": 0.0918, + "step": 13982 + }, + { + "epoch": 2.932061228769134, + "grad_norm": 4.188881502857614, + "learning_rate": 2.6888951563175393e-08, + "loss": 0.1188, + "step": 13983 + }, + { + "epoch": 2.9322709163346614, + "grad_norm": 4.896688883086082, + "learning_rate": 2.6723300376088058e-08, + "loss": 0.1726, + "step": 13984 + }, + { + "epoch": 2.9324806039001885, + "grad_norm": 3.5600686145757066, + "learning_rate": 2.6558160336731533e-08, + "loss": 0.1198, + "step": 13985 + }, + { + "epoch": 2.932690291465716, + "grad_norm": 3.838743330537526, + "learning_rate": 2.6393531453565713e-08, + "loss": 0.128, + "step": 13986 + }, + { + "epoch": 2.9328999790312436, + "grad_norm": 4.078330470646543, + "learning_rate": 2.6229413735030517e-08, + "loss": 0.1148, + "step": 13987 + }, + { + "epoch": 2.9331096665967706, + "grad_norm": 3.642321848227449, + "learning_rate": 2.6065807189535885e-08, + "loss": 0.1263, + "step": 13988 + }, + { + "epoch": 2.933319354162298, + "grad_norm": 3.1172119729884744, + "learning_rate": 2.5902711825468442e-08, + "loss": 0.1291, + "step": 13989 + }, + { + "epoch": 2.9335290417278257, + "grad_norm": 4.684465519876334, + "learning_rate": 2.574012765118594e-08, + "loss": 0.1429, + "step": 13990 + }, + { + "epoch": 2.9337387292933528, + "grad_norm": 3.01121543863825, + "learning_rate": 2.5578054675020613e-08, + "loss": 0.0866, + "step": 13991 + }, + { + "epoch": 2.9339484168588803, + "grad_norm": 5.473591564101314, + "learning_rate": 2.5416492905280255e-08, + "loss": 0.1556, + "step": 13992 + }, + { + "epoch": 2.934158104424408, + "grad_norm": 3.653811067465383, + "learning_rate": 2.5255442350244906e-08, + "loss": 0.1204, + "step": 13993 + }, + { + "epoch": 2.934367791989935, + "grad_norm": 3.698297934846077, + "learning_rate": 2.509490301816908e-08, + "loss": 0.1095, + "step": 13994 + }, + { + "epoch": 2.9345774795554624, + "grad_norm": 3.5545899276913646, + "learning_rate": 2.4934874917279528e-08, + "loss": 0.1111, + "step": 13995 + }, + { + "epoch": 2.93478716712099, + "grad_norm": 4.037017916389461, + "learning_rate": 2.4775358055777464e-08, + "loss": 0.1256, + "step": 13996 + }, + { + "epoch": 2.934996854686517, + "grad_norm": 5.076727774124706, + "learning_rate": 2.4616352441839685e-08, + "loss": 0.1411, + "step": 13997 + }, + { + "epoch": 2.9352065422520446, + "grad_norm": 3.7402751409933503, + "learning_rate": 2.4457858083616337e-08, + "loss": 0.1196, + "step": 13998 + }, + { + "epoch": 2.9354162298175717, + "grad_norm": 4.331800431877805, + "learning_rate": 2.42998749892287e-08, + "loss": 0.1427, + "step": 13999 + }, + { + "epoch": 2.935625917383099, + "grad_norm": 3.1987137567685977, + "learning_rate": 2.414240316677474e-08, + "loss": 0.1106, + "step": 14000 + }, + { + "epoch": 2.9358356049486263, + "grad_norm": 5.136016044489025, + "learning_rate": 2.3985442624325782e-08, + "loss": 0.121, + "step": 14001 + }, + { + "epoch": 2.936045292514154, + "grad_norm": 5.107095871544669, + "learning_rate": 2.382899336992428e-08, + "loss": 0.1619, + "step": 14002 + }, + { + "epoch": 2.9362549800796813, + "grad_norm": 6.264521096526768, + "learning_rate": 2.3673055411589374e-08, + "loss": 0.1813, + "step": 14003 + }, + { + "epoch": 2.9364646676452084, + "grad_norm": 4.228474019707086, + "learning_rate": 2.351762875731467e-08, + "loss": 0.1397, + "step": 14004 + }, + { + "epoch": 2.936674355210736, + "grad_norm": 4.310940557719353, + "learning_rate": 2.33627134150638e-08, + "loss": 0.1753, + "step": 14005 + }, + { + "epoch": 2.9368840427762635, + "grad_norm": 4.057385504818411, + "learning_rate": 2.3208309392777073e-08, + "loss": 0.1197, + "step": 14006 + }, + { + "epoch": 2.9370937303417906, + "grad_norm": 4.243241990474552, + "learning_rate": 2.3054416698369274e-08, + "loss": 0.1796, + "step": 14007 + }, + { + "epoch": 2.937303417907318, + "grad_norm": 4.631136099005787, + "learning_rate": 2.290103533972632e-08, + "loss": 0.1454, + "step": 14008 + }, + { + "epoch": 2.9375131054728456, + "grad_norm": 4.2891667512738, + "learning_rate": 2.2748165324709692e-08, + "loss": 0.1203, + "step": 14009 + }, + { + "epoch": 2.9377227930383727, + "grad_norm": 4.203436378862493, + "learning_rate": 2.259580666115535e-08, + "loss": 0.1324, + "step": 14010 + }, + { + "epoch": 2.9379324806039, + "grad_norm": 4.369069255835128, + "learning_rate": 2.244395935686927e-08, + "loss": 0.1396, + "step": 14011 + }, + { + "epoch": 2.9381421681694277, + "grad_norm": 4.0064812653629565, + "learning_rate": 2.229262341963634e-08, + "loss": 0.1403, + "step": 14012 + }, + { + "epoch": 2.938351855734955, + "grad_norm": 5.066171107022914, + "learning_rate": 2.214179885721146e-08, + "loss": 0.1717, + "step": 14013 + }, + { + "epoch": 2.9385615433004824, + "grad_norm": 3.3948684547943353, + "learning_rate": 2.199148567732512e-08, + "loss": 0.11, + "step": 14014 + }, + { + "epoch": 2.93877123086601, + "grad_norm": 3.706344215985088, + "learning_rate": 2.1841683887682263e-08, + "loss": 0.0959, + "step": 14015 + }, + { + "epoch": 2.938980918431537, + "grad_norm": 4.378779376299231, + "learning_rate": 2.1692393495957864e-08, + "loss": 0.1173, + "step": 14016 + }, + { + "epoch": 2.9391906059970645, + "grad_norm": 4.003900627969656, + "learning_rate": 2.1543614509805798e-08, + "loss": 0.1193, + "step": 14017 + }, + { + "epoch": 2.9394002935625916, + "grad_norm": 3.541209324869028, + "learning_rate": 2.139534693684886e-08, + "loss": 0.0994, + "step": 14018 + }, + { + "epoch": 2.939609981128119, + "grad_norm": 4.313493241728477, + "learning_rate": 2.1247590784688743e-08, + "loss": 0.1296, + "step": 14019 + }, + { + "epoch": 2.939819668693646, + "grad_norm": 3.5703947216391145, + "learning_rate": 2.110034606089717e-08, + "loss": 0.1053, + "step": 14020 + }, + { + "epoch": 2.9400293562591737, + "grad_norm": 3.8103718959416963, + "learning_rate": 2.0953612773019217e-08, + "loss": 0.1283, + "step": 14021 + }, + { + "epoch": 2.9402390438247012, + "grad_norm": 4.448374640631488, + "learning_rate": 2.0807390928576645e-08, + "loss": 0.151, + "step": 14022 + }, + { + "epoch": 2.9404487313902283, + "grad_norm": 4.0347241627781205, + "learning_rate": 2.0661680535063454e-08, + "loss": 0.1411, + "step": 14023 + }, + { + "epoch": 2.940658418955756, + "grad_norm": 3.6302195825880235, + "learning_rate": 2.0516481599948124e-08, + "loss": 0.0954, + "step": 14024 + }, + { + "epoch": 2.9408681065212834, + "grad_norm": 5.134454722476279, + "learning_rate": 2.0371794130671362e-08, + "loss": 0.1561, + "step": 14025 + }, + { + "epoch": 2.9410777940868105, + "grad_norm": 3.8021449726776955, + "learning_rate": 2.0227618134648352e-08, + "loss": 0.1075, + "step": 14026 + }, + { + "epoch": 2.941287481652338, + "grad_norm": 4.238122889201493, + "learning_rate": 2.0083953619268737e-08, + "loss": 0.1622, + "step": 14027 + }, + { + "epoch": 2.9414971692178655, + "grad_norm": 5.570521877234014, + "learning_rate": 1.9940800591896627e-08, + "loss": 0.1787, + "step": 14028 + }, + { + "epoch": 2.9417068567833926, + "grad_norm": 4.8336469262760104, + "learning_rate": 1.979815905986726e-08, + "loss": 0.1358, + "step": 14029 + }, + { + "epoch": 2.94191654434892, + "grad_norm": 3.48769105268091, + "learning_rate": 1.9656029030492574e-08, + "loss": 0.102, + "step": 14030 + }, + { + "epoch": 2.9421262319144477, + "grad_norm": 2.743218031227074, + "learning_rate": 1.9514410511056737e-08, + "loss": 0.1179, + "step": 14031 + }, + { + "epoch": 2.9423359194799747, + "grad_norm": 3.9727427171713825, + "learning_rate": 1.9373303508817275e-08, + "loss": 0.1436, + "step": 14032 + }, + { + "epoch": 2.9425456070455023, + "grad_norm": 4.040769724649306, + "learning_rate": 1.9232708031006187e-08, + "loss": 0.0859, + "step": 14033 + }, + { + "epoch": 2.94275529461103, + "grad_norm": 4.462170073294156, + "learning_rate": 1.9092624084828815e-08, + "loss": 0.1528, + "step": 14034 + }, + { + "epoch": 2.942964982176557, + "grad_norm": 6.2021401053450615, + "learning_rate": 1.8953051677467192e-08, + "loss": 0.1729, + "step": 14035 + }, + { + "epoch": 2.9431746697420844, + "grad_norm": 4.065509747590375, + "learning_rate": 1.881399081607227e-08, + "loss": 0.112, + "step": 14036 + }, + { + "epoch": 2.9433843573076115, + "grad_norm": 3.945715062688238, + "learning_rate": 1.8675441507771673e-08, + "loss": 0.1386, + "step": 14037 + }, + { + "epoch": 2.943594044873139, + "grad_norm": 4.69755821704635, + "learning_rate": 1.8537403759666394e-08, + "loss": 0.1485, + "step": 14038 + }, + { + "epoch": 2.943803732438666, + "grad_norm": 4.869885747025418, + "learning_rate": 1.839987757883077e-08, + "loss": 0.1437, + "step": 14039 + }, + { + "epoch": 2.9440134200041936, + "grad_norm": 4.028640331619228, + "learning_rate": 1.826286297231472e-08, + "loss": 0.1249, + "step": 14040 + }, + { + "epoch": 2.944223107569721, + "grad_norm": 3.782761214663777, + "learning_rate": 1.8126359947138184e-08, + "loss": 0.1034, + "step": 14041 + }, + { + "epoch": 2.9444327951352482, + "grad_norm": 3.5617392100415697, + "learning_rate": 1.7990368510298896e-08, + "loss": 0.1192, + "step": 14042 + }, + { + "epoch": 2.9446424827007758, + "grad_norm": 4.643180751184854, + "learning_rate": 1.7854888668766834e-08, + "loss": 0.127, + "step": 14043 + }, + { + "epoch": 2.9448521702663033, + "grad_norm": 3.8499578217733665, + "learning_rate": 1.7719920429484228e-08, + "loss": 0.1208, + "step": 14044 + }, + { + "epoch": 2.9450618578318304, + "grad_norm": 4.272185846791371, + "learning_rate": 1.7585463799368875e-08, + "loss": 0.1415, + "step": 14045 + }, + { + "epoch": 2.945271545397358, + "grad_norm": 4.159644292393944, + "learning_rate": 1.7451518785311928e-08, + "loss": 0.1657, + "step": 14046 + }, + { + "epoch": 2.9454812329628854, + "grad_norm": 4.192730621289511, + "learning_rate": 1.7318085394179008e-08, + "loss": 0.1391, + "step": 14047 + }, + { + "epoch": 2.9456909205284125, + "grad_norm": 3.7863515557501097, + "learning_rate": 1.7185163632809087e-08, + "loss": 0.1228, + "step": 14048 + }, + { + "epoch": 2.94590060809394, + "grad_norm": 3.9476606396926823, + "learning_rate": 1.7052753508012276e-08, + "loss": 0.1083, + "step": 14049 + }, + { + "epoch": 2.9461102956594676, + "grad_norm": 3.785811694255531, + "learning_rate": 1.6920855026577588e-08, + "loss": 0.1083, + "step": 14050 + }, + { + "epoch": 2.9463199832249947, + "grad_norm": 4.2238728495838584, + "learning_rate": 1.6789468195262948e-08, + "loss": 0.142, + "step": 14051 + }, + { + "epoch": 2.946529670790522, + "grad_norm": 4.182371135740404, + "learning_rate": 1.665859302080297e-08, + "loss": 0.1362, + "step": 14052 + }, + { + "epoch": 2.9467393583560497, + "grad_norm": 3.6252832551261953, + "learning_rate": 1.6528229509906735e-08, + "loss": 0.128, + "step": 14053 + }, + { + "epoch": 2.946949045921577, + "grad_norm": 4.5244449688114985, + "learning_rate": 1.639837766925334e-08, + "loss": 0.1448, + "step": 14054 + }, + { + "epoch": 2.9471587334871043, + "grad_norm": 4.734922864621065, + "learning_rate": 1.6269037505497464e-08, + "loss": 0.1081, + "step": 14055 + }, + { + "epoch": 2.9473684210526314, + "grad_norm": 4.06027416831395, + "learning_rate": 1.614020902527158e-08, + "loss": 0.1326, + "step": 14056 + }, + { + "epoch": 2.947578108618159, + "grad_norm": 3.8486219024600357, + "learning_rate": 1.6011892235174853e-08, + "loss": 0.1228, + "step": 14057 + }, + { + "epoch": 2.9477877961836865, + "grad_norm": 4.277160497554599, + "learning_rate": 1.5884087141785354e-08, + "loss": 0.1505, + "step": 14058 + }, + { + "epoch": 2.9479974837492136, + "grad_norm": 4.148364684822753, + "learning_rate": 1.575679375165229e-08, + "loss": 0.1122, + "step": 14059 + }, + { + "epoch": 2.948207171314741, + "grad_norm": 3.5426302058610206, + "learning_rate": 1.5630012071301548e-08, + "loss": 0.1369, + "step": 14060 + }, + { + "epoch": 2.948416858880268, + "grad_norm": 3.528340764967035, + "learning_rate": 1.550374210722905e-08, + "loss": 0.1037, + "step": 14061 + }, + { + "epoch": 2.9486265464457957, + "grad_norm": 4.41991677865507, + "learning_rate": 1.5377983865906276e-08, + "loss": 0.1427, + "step": 14062 + }, + { + "epoch": 2.948836234011323, + "grad_norm": 4.275846745295044, + "learning_rate": 1.52527373537803e-08, + "loss": 0.1592, + "step": 14063 + }, + { + "epoch": 2.9490459215768503, + "grad_norm": 5.542345568275729, + "learning_rate": 1.5128002577269317e-08, + "loss": 0.1605, + "step": 14064 + }, + { + "epoch": 2.949255609142378, + "grad_norm": 4.757814127431632, + "learning_rate": 1.5003779542764884e-08, + "loss": 0.1538, + "step": 14065 + }, + { + "epoch": 2.9494652967079054, + "grad_norm": 4.428855808668618, + "learning_rate": 1.4880068256636349e-08, + "loss": 0.1107, + "step": 14066 + }, + { + "epoch": 2.9496749842734324, + "grad_norm": 5.027763117273203, + "learning_rate": 1.4756868725220863e-08, + "loss": 0.1609, + "step": 14067 + }, + { + "epoch": 2.94988467183896, + "grad_norm": 4.988007408196683, + "learning_rate": 1.4634180954835598e-08, + "loss": 0.1483, + "step": 14068 + }, + { + "epoch": 2.9500943594044875, + "grad_norm": 4.980686909348065, + "learning_rate": 1.4512004951767745e-08, + "loss": 0.1595, + "step": 14069 + }, + { + "epoch": 2.9503040469700146, + "grad_norm": 4.178873859518184, + "learning_rate": 1.4390340722277851e-08, + "loss": 0.1319, + "step": 14070 + }, + { + "epoch": 2.950513734535542, + "grad_norm": 4.488116823658961, + "learning_rate": 1.4269188272602042e-08, + "loss": 0.1454, + "step": 14071 + }, + { + "epoch": 2.9507234221010696, + "grad_norm": 4.355603178584133, + "learning_rate": 1.4148547608949793e-08, + "loss": 0.1493, + "step": 14072 + }, + { + "epoch": 2.9509331096665967, + "grad_norm": 3.304392985046987, + "learning_rate": 1.4028418737505045e-08, + "loss": 0.0967, + "step": 14073 + }, + { + "epoch": 2.9511427972321242, + "grad_norm": 5.439950625921383, + "learning_rate": 1.3908801664422878e-08, + "loss": 0.1273, + "step": 14074 + }, + { + "epoch": 2.9513524847976518, + "grad_norm": 5.422996009175124, + "learning_rate": 1.378969639583394e-08, + "loss": 0.1905, + "step": 14075 + }, + { + "epoch": 2.951562172363179, + "grad_norm": 3.6524470416038537, + "learning_rate": 1.367110293784335e-08, + "loss": 0.1201, + "step": 14076 + }, + { + "epoch": 2.9517718599287064, + "grad_norm": 3.8917884818864064, + "learning_rate": 1.3553021296530688e-08, + "loss": 0.1563, + "step": 14077 + }, + { + "epoch": 2.9519815474942335, + "grad_norm": 6.399820970154888, + "learning_rate": 1.343545147794445e-08, + "loss": 0.191, + "step": 14078 + }, + { + "epoch": 2.952191235059761, + "grad_norm": 4.182213950976563, + "learning_rate": 1.3318393488113147e-08, + "loss": 0.1704, + "step": 14079 + }, + { + "epoch": 2.952400922625288, + "grad_norm": 4.504965990511345, + "learning_rate": 1.3201847333035311e-08, + "loss": 0.1658, + "step": 14080 + }, + { + "epoch": 2.9526106101908156, + "grad_norm": 4.25142980376893, + "learning_rate": 1.3085813018683946e-08, + "loss": 0.1263, + "step": 14081 + }, + { + "epoch": 2.952820297756343, + "grad_norm": 3.7597530610010845, + "learning_rate": 1.2970290551005404e-08, + "loss": 0.1263, + "step": 14082 + }, + { + "epoch": 2.95302998532187, + "grad_norm": 4.213781412407541, + "learning_rate": 1.2855279935921617e-08, + "loss": 0.1641, + "step": 14083 + }, + { + "epoch": 2.9532396728873977, + "grad_norm": 3.51960655122851, + "learning_rate": 1.2740781179325646e-08, + "loss": 0.0979, + "step": 14084 + }, + { + "epoch": 2.9534493604529253, + "grad_norm": 4.7330294194293225, + "learning_rate": 1.2626794287088351e-08, + "loss": 0.1282, + "step": 14085 + }, + { + "epoch": 2.9536590480184524, + "grad_norm": 4.515445262019607, + "learning_rate": 1.2513319265049507e-08, + "loss": 0.1276, + "step": 14086 + }, + { + "epoch": 2.95386873558398, + "grad_norm": 3.5936529415956, + "learning_rate": 1.2400356119024459e-08, + "loss": 0.1234, + "step": 14087 + }, + { + "epoch": 2.9540784231495074, + "grad_norm": 6.248758685403473, + "learning_rate": 1.2287904854804134e-08, + "loss": 0.1204, + "step": 14088 + }, + { + "epoch": 2.9542881107150345, + "grad_norm": 3.4180275791418784, + "learning_rate": 1.2175965478152807e-08, + "loss": 0.1245, + "step": 14089 + }, + { + "epoch": 2.954497798280562, + "grad_norm": 4.7399059175177864, + "learning_rate": 1.2064537994804782e-08, + "loss": 0.1285, + "step": 14090 + }, + { + "epoch": 2.9547074858460896, + "grad_norm": 3.0976946437219386, + "learning_rate": 1.1953622410473264e-08, + "loss": 0.0932, + "step": 14091 + }, + { + "epoch": 2.9549171734116166, + "grad_norm": 3.702908563299613, + "learning_rate": 1.1843218730841488e-08, + "loss": 0.1505, + "step": 14092 + }, + { + "epoch": 2.955126860977144, + "grad_norm": 3.87931292743932, + "learning_rate": 1.1733326961568259e-08, + "loss": 0.1378, + "step": 14093 + }, + { + "epoch": 2.9553365485426717, + "grad_norm": 3.215499501005404, + "learning_rate": 1.1623947108285738e-08, + "loss": 0.1046, + "step": 14094 + }, + { + "epoch": 2.9555462361081988, + "grad_norm": 3.3500192644438074, + "learning_rate": 1.1515079176600552e-08, + "loss": 0.0771, + "step": 14095 + }, + { + "epoch": 2.9557559236737263, + "grad_norm": 2.8545588453916455, + "learning_rate": 1.1406723172091572e-08, + "loss": 0.0918, + "step": 14096 + }, + { + "epoch": 2.9559656112392534, + "grad_norm": 4.2568969985816425, + "learning_rate": 1.1298879100311022e-08, + "loss": 0.1514, + "step": 14097 + }, + { + "epoch": 2.956175298804781, + "grad_norm": 4.316764277393882, + "learning_rate": 1.1191546966788925e-08, + "loss": 0.1323, + "step": 14098 + }, + { + "epoch": 2.956384986370308, + "grad_norm": 4.420689405349377, + "learning_rate": 1.1084726777024213e-08, + "loss": 0.1353, + "step": 14099 + }, + { + "epoch": 2.9565946739358355, + "grad_norm": 4.3683750946763595, + "learning_rate": 1.0978418536492507e-08, + "loss": 0.1445, + "step": 14100 + }, + { + "epoch": 2.956804361501363, + "grad_norm": 4.703686206228155, + "learning_rate": 1.0872622250641674e-08, + "loss": 0.1331, + "step": 14101 + }, + { + "epoch": 2.95701404906689, + "grad_norm": 3.125036009342627, + "learning_rate": 1.0767337924894038e-08, + "loss": 0.0897, + "step": 14102 + }, + { + "epoch": 2.9572237366324177, + "grad_norm": 3.130697772858216, + "learning_rate": 1.0662565564646398e-08, + "loss": 0.1158, + "step": 14103 + }, + { + "epoch": 2.957433424197945, + "grad_norm": 3.875383946649768, + "learning_rate": 1.0558305175266682e-08, + "loss": 0.1131, + "step": 14104 + }, + { + "epoch": 2.9576431117634723, + "grad_norm": 4.237132588760174, + "learning_rate": 1.045455676210061e-08, + "loss": 0.1401, + "step": 14105 + }, + { + "epoch": 2.957852799329, + "grad_norm": 5.153738027330773, + "learning_rate": 1.0351320330465043e-08, + "loss": 0.15, + "step": 14106 + }, + { + "epoch": 2.9580624868945273, + "grad_norm": 3.967829597323568, + "learning_rate": 1.0248595885649081e-08, + "loss": 0.1511, + "step": 14107 + }, + { + "epoch": 2.9582721744600544, + "grad_norm": 4.6359349701507755, + "learning_rate": 1.0146383432919626e-08, + "loss": 0.1511, + "step": 14108 + }, + { + "epoch": 2.958481862025582, + "grad_norm": 3.881871876700195, + "learning_rate": 1.0044682977514708e-08, + "loss": 0.1357, + "step": 14109 + }, + { + "epoch": 2.9586915495911095, + "grad_norm": 4.798127728287746, + "learning_rate": 9.943494524646825e-09, + "loss": 0.1509, + "step": 14110 + }, + { + "epoch": 2.9589012371566366, + "grad_norm": 4.273614643947953, + "learning_rate": 9.842818079500715e-09, + "loss": 0.1326, + "step": 14111 + }, + { + "epoch": 2.959110924722164, + "grad_norm": 5.6311653223125155, + "learning_rate": 9.74265364723781e-09, + "loss": 0.1109, + "step": 14112 + }, + { + "epoch": 2.9593206122876916, + "grad_norm": 5.534786551727605, + "learning_rate": 9.643001232990667e-09, + "loss": 0.1524, + "step": 14113 + }, + { + "epoch": 2.9595302998532187, + "grad_norm": 4.833773908175576, + "learning_rate": 9.543860841867425e-09, + "loss": 0.128, + "step": 14114 + }, + { + "epoch": 2.959739987418746, + "grad_norm": 4.801457682181429, + "learning_rate": 9.445232478949573e-09, + "loss": 0.1644, + "step": 14115 + }, + { + "epoch": 2.9599496749842733, + "grad_norm": 4.4077670662743245, + "learning_rate": 9.347116149289737e-09, + "loss": 0.1331, + "step": 14116 + }, + { + "epoch": 2.960159362549801, + "grad_norm": 3.9503032056599396, + "learning_rate": 9.249511857919446e-09, + "loss": 0.1171, + "step": 14117 + }, + { + "epoch": 2.960369050115328, + "grad_norm": 4.159254517792313, + "learning_rate": 9.152419609839148e-09, + "loss": 0.1507, + "step": 14118 + }, + { + "epoch": 2.9605787376808554, + "grad_norm": 3.2806793475210876, + "learning_rate": 9.055839410025969e-09, + "loss": 0.1176, + "step": 14119 + }, + { + "epoch": 2.960788425246383, + "grad_norm": 3.478096097793844, + "learning_rate": 8.959771263429285e-09, + "loss": 0.0945, + "step": 14120 + }, + { + "epoch": 2.96099811281191, + "grad_norm": 3.7308225943232443, + "learning_rate": 8.864215174972934e-09, + "loss": 0.1114, + "step": 14121 + }, + { + "epoch": 2.9612078003774376, + "grad_norm": 4.610521197152587, + "learning_rate": 8.769171149554111e-09, + "loss": 0.1355, + "step": 14122 + }, + { + "epoch": 2.961417487942965, + "grad_norm": 4.0872662001732865, + "learning_rate": 8.674639192044475e-09, + "loss": 0.1507, + "step": 14123 + }, + { + "epoch": 2.961627175508492, + "grad_norm": 4.38588082377484, + "learning_rate": 8.58061930728793e-09, + "loss": 0.1438, + "step": 14124 + }, + { + "epoch": 2.9618368630740197, + "grad_norm": 4.133623764095485, + "learning_rate": 8.487111500103951e-09, + "loss": 0.1331, + "step": 14125 + }, + { + "epoch": 2.9620465506395472, + "grad_norm": 4.2474664480122994, + "learning_rate": 8.394115775286481e-09, + "loss": 0.1216, + "step": 14126 + }, + { + "epoch": 2.9622562382050743, + "grad_norm": 4.422076686776864, + "learning_rate": 8.30163213759838e-09, + "loss": 0.1319, + "step": 14127 + }, + { + "epoch": 2.962465925770602, + "grad_norm": 5.634129090279814, + "learning_rate": 8.20966059178252e-09, + "loss": 0.1585, + "step": 14128 + }, + { + "epoch": 2.9626756133361294, + "grad_norm": 3.956225010042866, + "learning_rate": 8.118201142550685e-09, + "loss": 0.13, + "step": 14129 + }, + { + "epoch": 2.9628853009016565, + "grad_norm": 3.469172799246103, + "learning_rate": 8.02725379459135e-09, + "loss": 0.0961, + "step": 14130 + }, + { + "epoch": 2.963094988467184, + "grad_norm": 5.339445172687056, + "learning_rate": 7.936818552565229e-09, + "loss": 0.118, + "step": 14131 + }, + { + "epoch": 2.9633046760327115, + "grad_norm": 3.402977064657105, + "learning_rate": 7.846895421107504e-09, + "loss": 0.1091, + "step": 14132 + }, + { + "epoch": 2.9635143635982386, + "grad_norm": 5.4545633818913615, + "learning_rate": 7.75748440482671e-09, + "loss": 0.1335, + "step": 14133 + }, + { + "epoch": 2.963724051163766, + "grad_norm": 4.749579349435152, + "learning_rate": 7.668585508305849e-09, + "loss": 0.136, + "step": 14134 + }, + { + "epoch": 2.963933738729293, + "grad_norm": 6.624755674368858, + "learning_rate": 7.580198736101273e-09, + "loss": 0.2214, + "step": 14135 + }, + { + "epoch": 2.9641434262948207, + "grad_norm": 4.889040306502076, + "learning_rate": 7.492324092742697e-09, + "loss": 0.1428, + "step": 14136 + }, + { + "epoch": 2.964353113860348, + "grad_norm": 4.254509176834944, + "learning_rate": 7.404961582732073e-09, + "loss": 0.1112, + "step": 14137 + }, + { + "epoch": 2.9645628014258754, + "grad_norm": 4.240742374038316, + "learning_rate": 7.318111210550261e-09, + "loss": 0.1358, + "step": 14138 + }, + { + "epoch": 2.964772488991403, + "grad_norm": 4.075283158601231, + "learning_rate": 7.2317729806448135e-09, + "loss": 0.0999, + "step": 14139 + }, + { + "epoch": 2.96498217655693, + "grad_norm": 5.028783092365597, + "learning_rate": 7.145946897444411e-09, + "loss": 0.1376, + "step": 14140 + }, + { + "epoch": 2.9651918641224575, + "grad_norm": 4.71427835216695, + "learning_rate": 7.060632965344427e-09, + "loss": 0.1208, + "step": 14141 + }, + { + "epoch": 2.965401551687985, + "grad_norm": 5.4726197349070755, + "learning_rate": 6.97583118872025e-09, + "loss": 0.1841, + "step": 14142 + }, + { + "epoch": 2.965611239253512, + "grad_norm": 4.362963347051351, + "learning_rate": 6.8915415719150726e-09, + "loss": 0.1277, + "step": 14143 + }, + { + "epoch": 2.9658209268190396, + "grad_norm": 3.810859246137735, + "learning_rate": 6.807764119252102e-09, + "loss": 0.1462, + "step": 14144 + }, + { + "epoch": 2.966030614384567, + "grad_norm": 4.489727302955854, + "learning_rate": 6.724498835022353e-09, + "loss": 0.138, + "step": 14145 + }, + { + "epoch": 2.9662403019500942, + "grad_norm": 3.7570083717167995, + "learning_rate": 6.64174572349574e-09, + "loss": 0.0769, + "step": 14146 + }, + { + "epoch": 2.9664499895156218, + "grad_norm": 5.596001719452043, + "learning_rate": 6.559504788911098e-09, + "loss": 0.1526, + "step": 14147 + }, + { + "epoch": 2.9666596770811493, + "grad_norm": 3.282781314020852, + "learning_rate": 6.477776035485051e-09, + "loss": 0.1118, + "step": 14148 + }, + { + "epoch": 2.9668693646466764, + "grad_norm": 3.8763042191558386, + "learning_rate": 6.396559467406471e-09, + "loss": 0.1519, + "step": 14149 + }, + { + "epoch": 2.967079052212204, + "grad_norm": 3.966179237688947, + "learning_rate": 6.315855088836476e-09, + "loss": 0.1213, + "step": 14150 + }, + { + "epoch": 2.9672887397777314, + "grad_norm": 4.5402958041005625, + "learning_rate": 6.235662903911754e-09, + "loss": 0.1317, + "step": 14151 + }, + { + "epoch": 2.9674984273432585, + "grad_norm": 4.6955379480158355, + "learning_rate": 6.155982916743463e-09, + "loss": 0.1793, + "step": 14152 + }, + { + "epoch": 2.967708114908786, + "grad_norm": 3.0117341915061315, + "learning_rate": 6.076815131415004e-09, + "loss": 0.0911, + "step": 14153 + }, + { + "epoch": 2.967917802474313, + "grad_norm": 3.3552523549860522, + "learning_rate": 5.998159551982019e-09, + "loss": 0.11, + "step": 14154 + }, + { + "epoch": 2.9681274900398407, + "grad_norm": 4.620456743351178, + "learning_rate": 5.92001618247684e-09, + "loss": 0.1245, + "step": 14155 + }, + { + "epoch": 2.9683371776053677, + "grad_norm": 4.004282909447476, + "learning_rate": 5.842385026906261e-09, + "loss": 0.1238, + "step": 14156 + }, + { + "epoch": 2.9685468651708953, + "grad_norm": 4.604404931440074, + "learning_rate": 5.765266089245991e-09, + "loss": 0.1603, + "step": 14157 + }, + { + "epoch": 2.968756552736423, + "grad_norm": 4.744417334818594, + "learning_rate": 5.688659373450644e-09, + "loss": 0.1156, + "step": 14158 + }, + { + "epoch": 2.96896624030195, + "grad_norm": 3.9765035255092926, + "learning_rate": 5.61256488344597e-09, + "loss": 0.1157, + "step": 14159 + }, + { + "epoch": 2.9691759278674774, + "grad_norm": 3.626478872507043, + "learning_rate": 5.536982623131071e-09, + "loss": 0.1292, + "step": 14160 + }, + { + "epoch": 2.969385615433005, + "grad_norm": 3.034996859619405, + "learning_rate": 5.4619125963806255e-09, + "loss": 0.0909, + "step": 14161 + }, + { + "epoch": 2.969595302998532, + "grad_norm": 3.5489429510148462, + "learning_rate": 5.3873548070426665e-09, + "loss": 0.087, + "step": 14162 + }, + { + "epoch": 2.9698049905640596, + "grad_norm": 4.456468645690029, + "learning_rate": 5.313309258936361e-09, + "loss": 0.1391, + "step": 14163 + }, + { + "epoch": 2.970014678129587, + "grad_norm": 4.3669368723374715, + "learning_rate": 5.239775955858673e-09, + "loss": 0.1327, + "step": 14164 + }, + { + "epoch": 2.970224365695114, + "grad_norm": 3.453636289223555, + "learning_rate": 5.166754901576587e-09, + "loss": 0.1083, + "step": 14165 + }, + { + "epoch": 2.9704340532606417, + "grad_norm": 3.774700351581176, + "learning_rate": 5.094246099833777e-09, + "loss": 0.1323, + "step": 14166 + }, + { + "epoch": 2.970643740826169, + "grad_norm": 4.406581763555014, + "learning_rate": 5.0222495543461595e-09, + "loss": 0.1447, + "step": 14167 + }, + { + "epoch": 2.9708534283916963, + "grad_norm": 4.036319666846888, + "learning_rate": 4.9507652688041145e-09, + "loss": 0.1455, + "step": 14168 + }, + { + "epoch": 2.971063115957224, + "grad_norm": 3.9097460877956274, + "learning_rate": 4.879793246870268e-09, + "loss": 0.1304, + "step": 14169 + }, + { + "epoch": 2.9712728035227514, + "grad_norm": 3.4211235322165545, + "learning_rate": 4.809333492183932e-09, + "loss": 0.1398, + "step": 14170 + }, + { + "epoch": 2.9714824910882784, + "grad_norm": 3.707036201767182, + "learning_rate": 4.739386008353331e-09, + "loss": 0.0915, + "step": 14171 + }, + { + "epoch": 2.971692178653806, + "grad_norm": 3.4758162801310193, + "learning_rate": 4.669950798965595e-09, + "loss": 0.119, + "step": 14172 + }, + { + "epoch": 2.971901866219333, + "grad_norm": 3.667608740590732, + "learning_rate": 4.601027867580099e-09, + "loss": 0.1155, + "step": 14173 + }, + { + "epoch": 2.9721115537848606, + "grad_norm": 4.607440536294672, + "learning_rate": 4.532617217726243e-09, + "loss": 0.1543, + "step": 14174 + }, + { + "epoch": 2.9723212413503877, + "grad_norm": 3.3147481178439575, + "learning_rate": 4.464718852913441e-09, + "loss": 0.1157, + "step": 14175 + }, + { + "epoch": 2.972530928915915, + "grad_norm": 3.61358622771724, + "learning_rate": 4.397332776620023e-09, + "loss": 0.1133, + "step": 14176 + }, + { + "epoch": 2.9727406164814427, + "grad_norm": 5.71801667384248, + "learning_rate": 4.33045899229878e-09, + "loss": 0.1635, + "step": 14177 + }, + { + "epoch": 2.97295030404697, + "grad_norm": 4.847679761357739, + "learning_rate": 4.264097503379194e-09, + "loss": 0.1331, + "step": 14178 + }, + { + "epoch": 2.9731599916124973, + "grad_norm": 3.3976399018744523, + "learning_rate": 4.198248313260767e-09, + "loss": 0.0992, + "step": 14179 + }, + { + "epoch": 2.973369679178025, + "grad_norm": 3.613015178730115, + "learning_rate": 4.132911425318575e-09, + "loss": 0.0826, + "step": 14180 + }, + { + "epoch": 2.973579366743552, + "grad_norm": 4.159919588986853, + "learning_rate": 4.068086842902163e-09, + "loss": 0.1382, + "step": 14181 + }, + { + "epoch": 2.9737890543090795, + "grad_norm": 4.817047176948955, + "learning_rate": 4.0037745693333185e-09, + "loss": 0.1166, + "step": 14182 + }, + { + "epoch": 2.973998741874607, + "grad_norm": 6.062466634057586, + "learning_rate": 3.939974607908293e-09, + "loss": 0.1548, + "step": 14183 + }, + { + "epoch": 2.974208429440134, + "grad_norm": 3.5185610690126508, + "learning_rate": 3.876686961896692e-09, + "loss": 0.1116, + "step": 14184 + }, + { + "epoch": 2.9744181170056616, + "grad_norm": 4.937806368366158, + "learning_rate": 3.813911634542589e-09, + "loss": 0.1549, + "step": 14185 + }, + { + "epoch": 2.974627804571189, + "grad_norm": 3.997290833991194, + "learning_rate": 3.751648629063409e-09, + "loss": 0.1069, + "step": 14186 + }, + { + "epoch": 2.974837492136716, + "grad_norm": 4.13437208317382, + "learning_rate": 3.689897948648824e-09, + "loss": 0.1478, + "step": 14187 + }, + { + "epoch": 2.9750471797022437, + "grad_norm": 3.4103122089167477, + "learning_rate": 3.6286595964651893e-09, + "loss": 0.1006, + "step": 14188 + }, + { + "epoch": 2.9752568672677713, + "grad_norm": 4.0256495984750575, + "learning_rate": 3.5679335756511057e-09, + "loss": 0.1235, + "step": 14189 + }, + { + "epoch": 2.9754665548332984, + "grad_norm": 4.039987075504377, + "learning_rate": 3.507719889317418e-09, + "loss": 0.1416, + "step": 14190 + }, + { + "epoch": 2.975676242398826, + "grad_norm": 4.676280311929736, + "learning_rate": 3.4480185405516562e-09, + "loss": 0.1797, + "step": 14191 + }, + { + "epoch": 2.975885929964353, + "grad_norm": 4.245659538942573, + "learning_rate": 3.3888295324124854e-09, + "loss": 0.1625, + "step": 14192 + }, + { + "epoch": 2.9760956175298805, + "grad_norm": 4.885486084064281, + "learning_rate": 3.3301528679341444e-09, + "loss": 0.1249, + "step": 14193 + }, + { + "epoch": 2.9763053050954076, + "grad_norm": 3.753814168439786, + "learning_rate": 3.2719885501242276e-09, + "loss": 0.1369, + "step": 14194 + }, + { + "epoch": 2.976514992660935, + "grad_norm": 3.46367356200396, + "learning_rate": 3.214336581963684e-09, + "loss": 0.0988, + "step": 14195 + }, + { + "epoch": 2.9767246802264626, + "grad_norm": 3.2551930545368846, + "learning_rate": 3.157196966406817e-09, + "loss": 0.0713, + "step": 14196 + }, + { + "epoch": 2.9769343677919897, + "grad_norm": 3.6762343714189556, + "learning_rate": 3.1005697063812845e-09, + "loss": 0.117, + "step": 14197 + }, + { + "epoch": 2.9771440553575172, + "grad_norm": 4.195869216396358, + "learning_rate": 3.04445480479143e-09, + "loss": 0.1307, + "step": 14198 + }, + { + "epoch": 2.9773537429230448, + "grad_norm": 4.830250833372161, + "learning_rate": 2.9888522645116215e-09, + "loss": 0.1115, + "step": 14199 + }, + { + "epoch": 2.977563430488572, + "grad_norm": 4.068902606082635, + "learning_rate": 2.9337620883918004e-09, + "loss": 0.1408, + "step": 14200 + }, + { + "epoch": 2.9777731180540994, + "grad_norm": 3.928014834035326, + "learning_rate": 2.879184279256375e-09, + "loss": 0.1511, + "step": 14201 + }, + { + "epoch": 2.977982805619627, + "grad_norm": 4.174579938672508, + "learning_rate": 2.8251188399008865e-09, + "loss": 0.1077, + "step": 14202 + }, + { + "epoch": 2.978192493185154, + "grad_norm": 4.472218819325068, + "learning_rate": 2.771565773097562e-09, + "loss": 0.1558, + "step": 14203 + }, + { + "epoch": 2.9784021807506815, + "grad_norm": 3.378865315142501, + "learning_rate": 2.7185250815908728e-09, + "loss": 0.1181, + "step": 14204 + }, + { + "epoch": 2.978611868316209, + "grad_norm": 3.8621367295032343, + "learning_rate": 2.6659967680997544e-09, + "loss": 0.0962, + "step": 14205 + }, + { + "epoch": 2.978821555881736, + "grad_norm": 5.053110291358464, + "learning_rate": 2.6139808353153885e-09, + "loss": 0.1412, + "step": 14206 + }, + { + "epoch": 2.9790312434472637, + "grad_norm": 4.463334170656582, + "learning_rate": 2.56247728590453e-09, + "loss": 0.1328, + "step": 14207 + }, + { + "epoch": 2.979240931012791, + "grad_norm": 3.6619533364431383, + "learning_rate": 2.5114861225061795e-09, + "loss": 0.0833, + "step": 14208 + }, + { + "epoch": 2.9794506185783183, + "grad_norm": 5.613834569852398, + "learning_rate": 2.4610073477338013e-09, + "loss": 0.1355, + "step": 14209 + }, + { + "epoch": 2.979660306143846, + "grad_norm": 6.845096687523262, + "learning_rate": 2.411040964174216e-09, + "loss": 0.1855, + "step": 14210 + }, + { + "epoch": 2.979869993709373, + "grad_norm": 4.105777188877517, + "learning_rate": 2.3615869743887075e-09, + "loss": 0.1076, + "step": 14211 + }, + { + "epoch": 2.9800796812749004, + "grad_norm": 7.2156246461822775, + "learning_rate": 2.312645380913026e-09, + "loss": 0.1738, + "step": 14212 + }, + { + "epoch": 2.9802893688404275, + "grad_norm": 4.228085455036401, + "learning_rate": 2.2642161862529434e-09, + "loss": 0.1382, + "step": 14213 + }, + { + "epoch": 2.980499056405955, + "grad_norm": 5.241006719368144, + "learning_rate": 2.2162993928931397e-09, + "loss": 0.1529, + "step": 14214 + }, + { + "epoch": 2.9807087439714826, + "grad_norm": 5.091989618735093, + "learning_rate": 2.1688950032883183e-09, + "loss": 0.1724, + "step": 14215 + }, + { + "epoch": 2.9809184315370096, + "grad_norm": 4.055609026644988, + "learning_rate": 2.1220030198665366e-09, + "loss": 0.1268, + "step": 14216 + }, + { + "epoch": 2.981128119102537, + "grad_norm": 4.4428377679534385, + "learning_rate": 2.0756234450336475e-09, + "loss": 0.1443, + "step": 14217 + }, + { + "epoch": 2.9813378066680647, + "grad_norm": 4.169969193504995, + "learning_rate": 2.029756281165529e-09, + "loss": 0.1151, + "step": 14218 + }, + { + "epoch": 2.9815474942335918, + "grad_norm": 3.324104290161633, + "learning_rate": 1.984401530612523e-09, + "loss": 0.096, + "step": 14219 + }, + { + "epoch": 2.9817571817991193, + "grad_norm": 4.090981989088206, + "learning_rate": 1.939559195700547e-09, + "loss": 0.1211, + "step": 14220 + }, + { + "epoch": 2.981966869364647, + "grad_norm": 4.746819267430773, + "learning_rate": 1.8952292787266514e-09, + "loss": 0.1231, + "step": 14221 + }, + { + "epoch": 2.982176556930174, + "grad_norm": 5.209049922443455, + "learning_rate": 1.8514117819623534e-09, + "loss": 0.1256, + "step": 14222 + }, + { + "epoch": 2.9823862444957014, + "grad_norm": 6.032526163186445, + "learning_rate": 1.8081067076547443e-09, + "loss": 0.1765, + "step": 14223 + }, + { + "epoch": 2.982595932061229, + "grad_norm": 4.949502494411583, + "learning_rate": 1.7653140580220495e-09, + "loss": 0.1297, + "step": 14224 + }, + { + "epoch": 2.982805619626756, + "grad_norm": 4.369051221594472, + "learning_rate": 1.7230338352591803e-09, + "loss": 0.1643, + "step": 14225 + }, + { + "epoch": 2.9830153071922836, + "grad_norm": 3.6989507289178816, + "learning_rate": 1.681266041531071e-09, + "loss": 0.1491, + "step": 14226 + }, + { + "epoch": 2.983224994757811, + "grad_norm": 4.150111420832969, + "learning_rate": 1.6400106789793424e-09, + "loss": 0.1247, + "step": 14227 + }, + { + "epoch": 2.983434682323338, + "grad_norm": 3.1265917227548012, + "learning_rate": 1.5992677497178587e-09, + "loss": 0.0898, + "step": 14228 + }, + { + "epoch": 2.9836443698888657, + "grad_norm": 3.944158507349355, + "learning_rate": 1.55903725583606e-09, + "loss": 0.1542, + "step": 14229 + }, + { + "epoch": 2.983854057454393, + "grad_norm": 3.3946792700038815, + "learning_rate": 1.51931919939452e-09, + "loss": 0.1065, + "step": 14230 + }, + { + "epoch": 2.9840637450199203, + "grad_norm": 4.300415961632463, + "learning_rate": 1.480113582430498e-09, + "loss": 0.1214, + "step": 14231 + }, + { + "epoch": 2.9842734325854474, + "grad_norm": 4.162462353786712, + "learning_rate": 1.441420406951277e-09, + "loss": 0.1317, + "step": 14232 + }, + { + "epoch": 2.984483120150975, + "grad_norm": 4.793681544454836, + "learning_rate": 1.403239674940826e-09, + "loss": 0.1837, + "step": 14233 + }, + { + "epoch": 2.9846928077165025, + "grad_norm": 3.9930645499280746, + "learning_rate": 1.365571388356468e-09, + "loss": 0.1218, + "step": 14234 + }, + { + "epoch": 2.9849024952820296, + "grad_norm": 4.503888404654512, + "learning_rate": 1.3284155491277706e-09, + "loss": 0.154, + "step": 14235 + }, + { + "epoch": 2.985112182847557, + "grad_norm": 3.5728393507715244, + "learning_rate": 1.2917721591598764e-09, + "loss": 0.1022, + "step": 14236 + }, + { + "epoch": 2.9853218704130846, + "grad_norm": 3.774709005663481, + "learning_rate": 1.2556412203301727e-09, + "loss": 0.1492, + "step": 14237 + }, + { + "epoch": 2.9855315579786117, + "grad_norm": 4.30946456616066, + "learning_rate": 1.2200227344916216e-09, + "loss": 0.1868, + "step": 14238 + }, + { + "epoch": 2.985741245544139, + "grad_norm": 4.5471076876092855, + "learning_rate": 1.1849167034683196e-09, + "loss": 0.1598, + "step": 14239 + }, + { + "epoch": 2.9859509331096667, + "grad_norm": 4.80132463610375, + "learning_rate": 1.150323129059938e-09, + "loss": 0.1626, + "step": 14240 + }, + { + "epoch": 2.986160620675194, + "grad_norm": 3.6972411754906074, + "learning_rate": 1.1162420130395036e-09, + "loss": 0.1489, + "step": 14241 + }, + { + "epoch": 2.9863703082407214, + "grad_norm": 4.252336012160244, + "learning_rate": 1.0826733571545068e-09, + "loss": 0.1735, + "step": 14242 + }, + { + "epoch": 2.986579995806249, + "grad_norm": 4.622739605589684, + "learning_rate": 1.0496171631246831e-09, + "loss": 0.1614, + "step": 14243 + }, + { + "epoch": 2.986789683371776, + "grad_norm": 4.575671271683195, + "learning_rate": 1.0170734326442332e-09, + "loss": 0.1383, + "step": 14244 + }, + { + "epoch": 2.9869993709373035, + "grad_norm": 4.055892033166719, + "learning_rate": 9.85042167380712e-10, + "loss": 0.136, + "step": 14245 + }, + { + "epoch": 2.987209058502831, + "grad_norm": 3.718926650698118, + "learning_rate": 9.535233689761392e-10, + "loss": 0.1122, + "step": 14246 + }, + { + "epoch": 2.987418746068358, + "grad_norm": 5.665129866823334, + "learning_rate": 9.225170390469995e-10, + "loss": 0.1379, + "step": 14247 + }, + { + "epoch": 2.9876284336338856, + "grad_norm": 3.9556185589182222, + "learning_rate": 8.920231791798017e-10, + "loss": 0.1131, + "step": 14248 + }, + { + "epoch": 2.9878381211994127, + "grad_norm": 4.8565627229061015, + "learning_rate": 8.620417909399603e-10, + "loss": 0.1106, + "step": 14249 + }, + { + "epoch": 2.9880478087649402, + "grad_norm": 5.022450264050266, + "learning_rate": 8.325728758629137e-10, + "loss": 0.1611, + "step": 14250 + }, + { + "epoch": 2.9882574963304673, + "grad_norm": 5.32275588463611, + "learning_rate": 8.036164354596753e-10, + "loss": 0.1456, + "step": 14251 + }, + { + "epoch": 2.988467183895995, + "grad_norm": 4.717961748688537, + "learning_rate": 7.751724712135034e-10, + "loss": 0.1316, + "step": 14252 + }, + { + "epoch": 2.9886768714615224, + "grad_norm": 5.079328021821132, + "learning_rate": 7.472409845821205e-10, + "loss": 0.1571, + "step": 14253 + }, + { + "epoch": 2.9888865590270495, + "grad_norm": 4.578899448642258, + "learning_rate": 7.198219769977144e-10, + "loss": 0.1626, + "step": 14254 + }, + { + "epoch": 2.989096246592577, + "grad_norm": 5.765370571071565, + "learning_rate": 6.929154498658275e-10, + "loss": 0.185, + "step": 14255 + }, + { + "epoch": 2.9893059341581045, + "grad_norm": 4.101698138628173, + "learning_rate": 6.665214045642466e-10, + "loss": 0.1459, + "step": 14256 + }, + { + "epoch": 2.9895156217236316, + "grad_norm": 6.011572093054696, + "learning_rate": 6.406398424463334e-10, + "loss": 0.1642, + "step": 14257 + }, + { + "epoch": 2.989725309289159, + "grad_norm": 3.406181106997159, + "learning_rate": 6.152707648388045e-10, + "loss": 0.1187, + "step": 14258 + }, + { + "epoch": 2.9899349968546867, + "grad_norm": 5.184306824398564, + "learning_rate": 5.904141730417312e-10, + "loss": 0.1352, + "step": 14259 + }, + { + "epoch": 2.9901446844202137, + "grad_norm": 3.4342081212772215, + "learning_rate": 5.660700683296494e-10, + "loss": 0.1024, + "step": 14260 + }, + { + "epoch": 2.9903543719857413, + "grad_norm": 4.716916031842858, + "learning_rate": 5.422384519493395e-10, + "loss": 0.1497, + "step": 14261 + }, + { + "epoch": 2.990564059551269, + "grad_norm": 4.631812140620607, + "learning_rate": 5.18919325122047e-10, + "loss": 0.112, + "step": 14262 + }, + { + "epoch": 2.990773747116796, + "grad_norm": 3.108250847427697, + "learning_rate": 4.96112689043482e-10, + "loss": 0.1021, + "step": 14263 + }, + { + "epoch": 2.9909834346823234, + "grad_norm": 4.221239962168933, + "learning_rate": 4.738185448827093e-10, + "loss": 0.1362, + "step": 14264 + }, + { + "epoch": 2.991193122247851, + "grad_norm": 4.145358739628082, + "learning_rate": 4.520368937810382e-10, + "loss": 0.1159, + "step": 14265 + }, + { + "epoch": 2.991402809813378, + "grad_norm": 3.9226733597650285, + "learning_rate": 4.307677368564633e-10, + "loss": 0.1304, + "step": 14266 + }, + { + "epoch": 2.9916124973789056, + "grad_norm": 3.6770658974418744, + "learning_rate": 4.1001107519922365e-10, + "loss": 0.1336, + "step": 14267 + }, + { + "epoch": 2.9918221849444326, + "grad_norm": 3.8167229222572634, + "learning_rate": 3.8976690987180265e-10, + "loss": 0.1163, + "step": 14268 + }, + { + "epoch": 2.99203187250996, + "grad_norm": 4.153900105342872, + "learning_rate": 3.700352419111486e-10, + "loss": 0.1235, + "step": 14269 + }, + { + "epoch": 2.9922415600754873, + "grad_norm": 4.695739705732879, + "learning_rate": 3.508160723308951e-10, + "loss": 0.1417, + "step": 14270 + }, + { + "epoch": 2.9924512476410148, + "grad_norm": 4.184957444078295, + "learning_rate": 3.3210940211469977e-10, + "loss": 0.144, + "step": 14271 + }, + { + "epoch": 2.9926609352065423, + "grad_norm": 3.5827991661900027, + "learning_rate": 3.139152322217953e-10, + "loss": 0.1222, + "step": 14272 + }, + { + "epoch": 2.9928706227720694, + "grad_norm": 5.0263091575559, + "learning_rate": 2.9623356358365884e-10, + "loss": 0.1634, + "step": 14273 + }, + { + "epoch": 2.993080310337597, + "grad_norm": 3.332992455458001, + "learning_rate": 2.7906439710734255e-10, + "loss": 0.1123, + "step": 14274 + }, + { + "epoch": 2.9932899979031244, + "grad_norm": 3.6597966219254774, + "learning_rate": 2.624077336721431e-10, + "loss": 0.0896, + "step": 14275 + }, + { + "epoch": 2.9934996854686515, + "grad_norm": 4.049995356358593, + "learning_rate": 2.462635741329322e-10, + "loss": 0.1267, + "step": 14276 + }, + { + "epoch": 2.993709373034179, + "grad_norm": 4.861911612798047, + "learning_rate": 2.3063191931571583e-10, + "loss": 0.167, + "step": 14277 + }, + { + "epoch": 2.9939190605997066, + "grad_norm": 3.785223173552741, + "learning_rate": 2.1551277002318515e-10, + "loss": 0.1303, + "step": 14278 + }, + { + "epoch": 2.9941287481652337, + "grad_norm": 5.943901446434463, + "learning_rate": 2.0090612702805546e-10, + "loss": 0.1445, + "step": 14279 + }, + { + "epoch": 2.994338435730761, + "grad_norm": 4.778420101108463, + "learning_rate": 1.868119910819477e-10, + "loss": 0.1707, + "step": 14280 + }, + { + "epoch": 2.9945481232962887, + "grad_norm": 4.351730393628573, + "learning_rate": 1.732303629042864e-10, + "loss": 0.1436, + "step": 14281 + }, + { + "epoch": 2.994757810861816, + "grad_norm": 4.887286616434012, + "learning_rate": 1.6016124319229165e-10, + "loss": 0.1628, + "step": 14282 + }, + { + "epoch": 2.9949674984273433, + "grad_norm": 4.680603516026108, + "learning_rate": 1.476046326165381e-10, + "loss": 0.165, + "step": 14283 + }, + { + "epoch": 2.995177185992871, + "grad_norm": 3.499897399743875, + "learning_rate": 1.3556053181984497e-10, + "loss": 0.1173, + "step": 14284 + }, + { + "epoch": 2.995386873558398, + "grad_norm": 3.3287776000213656, + "learning_rate": 1.2402894141949618e-10, + "loss": 0.1009, + "step": 14285 + }, + { + "epoch": 2.9955965611239255, + "grad_norm": 3.924063056275675, + "learning_rate": 1.1300986200724063e-10, + "loss": 0.1432, + "step": 14286 + }, + { + "epoch": 2.9958062486894526, + "grad_norm": 3.53551096117981, + "learning_rate": 1.0250329414596138e-10, + "loss": 0.1062, + "step": 14287 + }, + { + "epoch": 2.99601593625498, + "grad_norm": 4.012627932488079, + "learning_rate": 9.250923837633708e-11, + "loss": 0.1499, + "step": 14288 + }, + { + "epoch": 2.996225623820507, + "grad_norm": 3.0910890372762747, + "learning_rate": 8.302769520907028e-11, + "loss": 0.0994, + "step": 14289 + }, + { + "epoch": 2.9964353113860347, + "grad_norm": 4.383922935382639, + "learning_rate": 7.405866513043869e-11, + "loss": 0.1428, + "step": 14290 + }, + { + "epoch": 2.9966449989515622, + "grad_norm": 4.086305329462803, + "learning_rate": 6.560214860007463e-11, + "loss": 0.1404, + "step": 14291 + }, + { + "epoch": 2.9968546865170893, + "grad_norm": 3.7436921038095323, + "learning_rate": 5.765814605207532e-11, + "loss": 0.15, + "step": 14292 + }, + { + "epoch": 2.997064374082617, + "grad_norm": 3.412975188056128, + "learning_rate": 5.022665789278236e-11, + "loss": 0.1071, + "step": 14293 + }, + { + "epoch": 2.9972740616481444, + "grad_norm": 4.283350806075584, + "learning_rate": 4.330768450411249e-11, + "loss": 0.1298, + "step": 14294 + }, + { + "epoch": 2.9974837492136714, + "grad_norm": 4.452789328915619, + "learning_rate": 3.690122623911663e-11, + "loss": 0.1436, + "step": 14295 + }, + { + "epoch": 2.997693436779199, + "grad_norm": 4.88171505458831, + "learning_rate": 3.100728342753101e-11, + "loss": 0.1689, + "step": 14296 + }, + { + "epoch": 2.9979031243447265, + "grad_norm": 4.569187621717733, + "learning_rate": 2.5625856370226075e-11, + "loss": 0.171, + "step": 14297 + }, + { + "epoch": 2.9981128119102536, + "grad_norm": 4.181243597012907, + "learning_rate": 2.0756945344757585e-11, + "loss": 0.0922, + "step": 14298 + }, + { + "epoch": 2.998322499475781, + "grad_norm": 3.5021110538682705, + "learning_rate": 1.6400550598705266e-11, + "loss": 0.098, + "step": 14299 + }, + { + "epoch": 2.9985321870413086, + "grad_norm": 4.318872608265839, + "learning_rate": 1.2556672356334176e-11, + "loss": 0.1184, + "step": 14300 + }, + { + "epoch": 2.9987418746068357, + "grad_norm": 4.4613017870445635, + "learning_rate": 9.22531081526401e-12, + "loss": 0.1387, + "step": 14301 + }, + { + "epoch": 2.9989515621723632, + "grad_norm": 3.871335515730445, + "learning_rate": 6.4064661453588916e-12, + "loss": 0.1044, + "step": 14302 + }, + { + "epoch": 2.9991612497378908, + "grad_norm": 4.582158006035162, + "learning_rate": 4.100138489837591e-12, + "loss": 0.1301, + "step": 14303 + }, + { + "epoch": 2.999370937303418, + "grad_norm": 6.069782410370943, + "learning_rate": 2.306327969714417e-12, + "loss": 0.1158, + "step": 14304 + }, + { + "epoch": 2.9995806248689454, + "grad_norm": 4.739417643890966, + "learning_rate": 1.0250346749174356e-12, + "loss": 0.1418, + "step": 14305 + }, + { + "epoch": 2.9997903124344725, + "grad_norm": 4.02449458850959, + "learning_rate": 2.5625867206002795e-13, + "loss": 0.1291, + "step": 14306 + }, + { + "epoch": 3.0, + "grad_norm": 3.351427500888645, + "learning_rate": 0.0, + "loss": 0.1252, + "step": 14307 + }, + { + "epoch": 3.0, + "step": 14307, + "total_flos": 5.786700751120105e+18, + "train_loss": 1.0319068685825985, + "train_runtime": 97307.4105, + "train_samples_per_second": 9.41, + "train_steps_per_second": 0.147 + } + ], + "logging_steps": 1.0, + "max_steps": 14307, + "num_input_tokens_seen": 0, + "num_train_epochs": 3, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 5.786700751120105e+18, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}