diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4662 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9977298524404086,
+  "eval_steps": 500,
+  "global_step": 660,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003026863412788498,
+      "grad_norm": 3.162303897505498,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.6712,
+      "step": 1
+    },
+    {
+      "epoch": 0.006053726825576996,
+      "grad_norm": 3.477619096038061,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.7363,
+      "step": 2
+    },
+    {
+      "epoch": 0.009080590238365494,
+      "grad_norm": 3.0248475084484228,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.687,
+      "step": 3
+    },
+    {
+      "epoch": 0.012107453651153992,
+      "grad_norm": 3.033130199661253,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.7019,
+      "step": 4
+    },
+    {
+      "epoch": 0.01513431706394249,
+      "grad_norm": 3.3646155807952245,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.7094,
+      "step": 5
+    },
+    {
+      "epoch": 0.018161180476730987,
+      "grad_norm": 2.9404958154054035,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.6818,
+      "step": 6
+    },
+    {
+      "epoch": 0.021188043889519486,
+      "grad_norm": 2.8650780042759862,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.6813,
+      "step": 7
+    },
+    {
+      "epoch": 0.024214907302307985,
+      "grad_norm": 2.4703666904101054,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.6663,
+      "step": 8
+    },
+    {
+      "epoch": 0.02724177071509648,
+      "grad_norm": 1.9244437005019943,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.6616,
+      "step": 9
+    },
+    {
+      "epoch": 0.03026863412788498,
+      "grad_norm": 1.6092680212698305,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.6197,
+      "step": 10
+    },
+    {
+      "epoch": 0.03329549754067348,
+      "grad_norm": 1.4819151969579991,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.6023,
+      "step": 11
+    },
+    {
+      "epoch": 0.036322360953461974,
+      "grad_norm": 1.8513992114922788,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.5918,
+      "step": 12
+    },
+    {
+      "epoch": 0.03934922436625047,
+      "grad_norm": 2.871791816677462,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.6159,
+      "step": 13
+    },
+    {
+      "epoch": 0.04237608777903897,
+      "grad_norm": 2.369497741908694,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.5771,
+      "step": 14
+    },
+    {
+      "epoch": 0.04540295119182747,
+      "grad_norm": 1.7448680450540297,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.5693,
+      "step": 15
+    },
+    {
+      "epoch": 0.04842981460461597,
+      "grad_norm": 1.5369742432701625,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.5863,
+      "step": 16
+    },
+    {
+      "epoch": 0.051456678017404466,
+      "grad_norm": 1.414243083421236,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.5812,
+      "step": 17
+    },
+    {
+      "epoch": 0.05448354143019296,
+      "grad_norm": 1.4762539058139061,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.5636,
+      "step": 18
+    },
+    {
+      "epoch": 0.057510404842981463,
+      "grad_norm": 1.3550924339648431,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.553,
+      "step": 19
+    },
+    {
+      "epoch": 0.06053726825576996,
+      "grad_norm": 1.2129648873821985,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.5684,
+      "step": 20
+    },
+    {
+      "epoch": 0.06356413166855845,
+      "grad_norm": 1.0588640571770458,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.5535,
+      "step": 21
+    },
+    {
+      "epoch": 0.06659099508134696,
+      "grad_norm": 1.15031571250409,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.547,
+      "step": 22
+    },
+    {
+      "epoch": 0.06961785849413545,
+      "grad_norm": 1.0449936972053457,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.5122,
+      "step": 23
+    },
+    {
+      "epoch": 0.07264472190692395,
+      "grad_norm": 1.0157216157098372,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.5288,
+      "step": 24
+    },
+    {
+      "epoch": 0.07567158531971245,
+      "grad_norm": 0.9735496252115815,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.5123,
+      "step": 25
+    },
+    {
+      "epoch": 0.07869844873250094,
+      "grad_norm": 0.9863580607296011,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.4996,
+      "step": 26
+    },
+    {
+      "epoch": 0.08172531214528944,
+      "grad_norm": 0.9577383832312116,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.4912,
+      "step": 27
+    },
+    {
+      "epoch": 0.08475217555807794,
+      "grad_norm": 0.9307072505103615,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.5111,
+      "step": 28
+    },
+    {
+      "epoch": 0.08777903897086645,
+      "grad_norm": 0.8755091725616875,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 0.4955,
+      "step": 29
+    },
+    {
+      "epoch": 0.09080590238365494,
+      "grad_norm": 0.9857320816463918,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.5143,
+      "step": 30
+    },
+    {
+      "epoch": 0.09383276579644344,
+      "grad_norm": 0.9491374417664864,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 0.4962,
+      "step": 31
+    },
+    {
+      "epoch": 0.09685962920923194,
+      "grad_norm": 0.9168983960014782,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.5066,
+      "step": 32
+    },
+    {
+      "epoch": 0.09988649262202043,
+      "grad_norm": 0.8169961478108528,
+      "learning_rate": 1e-05,
+      "loss": 0.4847,
+      "step": 33
+    },
+    {
+      "epoch": 0.10291335603480893,
+      "grad_norm": 0.8852788220790511,
+      "learning_rate": 1.0303030303030304e-05,
+      "loss": 0.4827,
+      "step": 34
+    },
+    {
+      "epoch": 0.10594021944759743,
+      "grad_norm": 0.8625799732177928,
+      "learning_rate": 1.0606060606060606e-05,
+      "loss": 0.4839,
+      "step": 35
+    },
+    {
+      "epoch": 0.10896708286038592,
+      "grad_norm": 0.8921236901400129,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.5344,
+      "step": 36
+    },
+    {
+      "epoch": 0.11199394627317442,
+      "grad_norm": 0.9435694954111703,
+      "learning_rate": 1.1212121212121212e-05,
+      "loss": 0.5,
+      "step": 37
+    },
+    {
+      "epoch": 0.11502080968596293,
+      "grad_norm": 0.843047318465379,
+      "learning_rate": 1.1515151515151517e-05,
+      "loss": 0.4946,
+      "step": 38
+    },
+    {
+      "epoch": 0.11804767309875142,
+      "grad_norm": 0.8903147012048684,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.5086,
+      "step": 39
+    },
+    {
+      "epoch": 0.12107453651153992,
+      "grad_norm": 0.918274180490875,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.4784,
+      "step": 40
+    },
+    {
+      "epoch": 0.12410139992432842,
+      "grad_norm": 0.8432185141265864,
+      "learning_rate": 1.2424242424242425e-05,
+      "loss": 0.4754,
+      "step": 41
+    },
+    {
+      "epoch": 0.1271282633371169,
+      "grad_norm": 1.011495606718716,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.4734,
+      "step": 42
+    },
+    {
+      "epoch": 0.1301551267499054,
+      "grad_norm": 0.9693476290838801,
+      "learning_rate": 1.3030303030303032e-05,
+      "loss": 0.5061,
+      "step": 43
+    },
+    {
+      "epoch": 0.13318199016269391,
+      "grad_norm": 1.077209246229062,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.4914,
+      "step": 44
+    },
+    {
+      "epoch": 0.1362088535754824,
+      "grad_norm": 0.9519744865503152,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.4723,
+      "step": 45
+    },
+    {
+      "epoch": 0.1392357169882709,
+      "grad_norm": 1.0298364500086112,
+      "learning_rate": 1.3939393939393942e-05,
+      "loss": 0.4817,
+      "step": 46
+    },
+    {
+      "epoch": 0.1422625804010594,
+      "grad_norm": 0.9462687808138835,
+      "learning_rate": 1.4242424242424245e-05,
+      "loss": 0.4868,
+      "step": 47
+    },
+    {
+      "epoch": 0.1452894438138479,
+      "grad_norm": 0.8714639634329036,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.4834,
+      "step": 48
+    },
+    {
+      "epoch": 0.14831630722663638,
+      "grad_norm": 0.9702867483854449,
+      "learning_rate": 1.484848484848485e-05,
+      "loss": 0.4829,
+      "step": 49
+    },
+    {
+      "epoch": 0.1513431706394249,
+      "grad_norm": 0.8420826586340022,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.4896,
+      "step": 50
+    },
+    {
+      "epoch": 0.1543700340522134,
+      "grad_norm": 0.9489985723848989,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.5006,
+      "step": 51
+    },
+    {
+      "epoch": 0.15739689746500188,
+      "grad_norm": 0.928702761660508,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 0.4502,
+      "step": 52
+    },
+    {
+      "epoch": 0.1604237608777904,
+      "grad_norm": 0.8618303250523711,
+      "learning_rate": 1.606060606060606e-05,
+      "loss": 0.4792,
+      "step": 53
+    },
+    {
+      "epoch": 0.16345062429057888,
+      "grad_norm": 1.038824265747017,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.4786,
+      "step": 54
+    },
+    {
+      "epoch": 0.16647748770336737,
+      "grad_norm": 0.8348378254977165,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.4673,
+      "step": 55
+    },
+    {
+      "epoch": 0.1695043511161559,
+      "grad_norm": 1.0244052192139341,
+      "learning_rate": 1.6969696969696972e-05,
+      "loss": 0.4782,
+      "step": 56
+    },
+    {
+      "epoch": 0.17253121452894438,
+      "grad_norm": 0.8531946621240053,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.5047,
+      "step": 57
+    },
+    {
+      "epoch": 0.1755580779417329,
+      "grad_norm": 0.9908318880290045,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 0.5158,
+      "step": 58
+    },
+    {
+      "epoch": 0.17858494135452138,
+      "grad_norm": 0.8635400816337314,
+      "learning_rate": 1.787878787878788e-05,
+      "loss": 0.4981,
+      "step": 59
+    },
+    {
+      "epoch": 0.18161180476730987,
+      "grad_norm": 0.9556857704027323,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.4851,
+      "step": 60
+    },
+    {
+      "epoch": 0.1846386681800984,
+      "grad_norm": 0.9539408196871916,
+      "learning_rate": 1.8484848484848487e-05,
+      "loss": 0.4996,
+      "step": 61
+    },
+    {
+      "epoch": 0.18766553159288688,
+      "grad_norm": 0.9363118742956849,
+      "learning_rate": 1.8787878787878792e-05,
+      "loss": 0.4995,
+      "step": 62
+    },
+    {
+      "epoch": 0.19069239500567536,
+      "grad_norm": 0.9268946074627714,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.48,
+      "step": 63
+    },
+    {
+      "epoch": 0.19371925841846388,
+      "grad_norm": 0.8877679375426357,
+      "learning_rate": 1.9393939393939395e-05,
+      "loss": 0.4801,
+      "step": 64
+    },
+    {
+      "epoch": 0.19674612183125237,
+      "grad_norm": 0.9067043534302868,
+      "learning_rate": 1.96969696969697e-05,
+      "loss": 0.4638,
+      "step": 65
+    },
+    {
+      "epoch": 0.19977298524404086,
+      "grad_norm": 0.8578004879749674,
+      "learning_rate": 2e-05,
+      "loss": 0.4575,
+      "step": 66
+    },
+    {
+      "epoch": 0.20279984865682937,
+      "grad_norm": 1.2145449622871916,
+      "learning_rate": 1.9999860139251737e-05,
+      "loss": 0.4806,
+      "step": 67
+    },
+    {
+      "epoch": 0.20582671206961786,
+      "grad_norm": 0.846277156571831,
+      "learning_rate": 1.9999440560919153e-05,
+      "loss": 0.4738,
+      "step": 68
+    },
+    {
+      "epoch": 0.20885357548240635,
+      "grad_norm": 1.0772803975348055,
+      "learning_rate": 1.9998741276738753e-05,
+      "loss": 0.4868,
+      "step": 69
+    },
+    {
+      "epoch": 0.21188043889519487,
+      "grad_norm": 0.9202041448347507,
+      "learning_rate": 1.999776230627102e-05,
+      "loss": 0.4712,
+      "step": 70
+    },
+    {
+      "epoch": 0.21490730230798336,
+      "grad_norm": 0.9796153953340392,
+      "learning_rate": 1.9996503676899863e-05,
+      "loss": 0.4947,
+      "step": 71
+    },
+    {
+      "epoch": 0.21793416572077184,
+      "grad_norm": 0.8602897884508506,
+      "learning_rate": 1.9994965423831853e-05,
+      "loss": 0.4601,
+      "step": 72
+    },
+    {
+      "epoch": 0.22096102913356036,
+      "grad_norm": 1.0530215791895698,
+      "learning_rate": 1.9993147590095232e-05,
+      "loss": 0.4946,
+      "step": 73
+    },
+    {
+      "epoch": 0.22398789254634885,
+      "grad_norm": 0.8082950351691883,
+      "learning_rate": 1.999105022653872e-05,
+      "loss": 0.4551,
+      "step": 74
+    },
+    {
+      "epoch": 0.22701475595913734,
+      "grad_norm": 0.97267364684008,
+      "learning_rate": 1.9988673391830082e-05,
+      "loss": 0.487,
+      "step": 75
+    },
+    {
+      "epoch": 0.23004161937192585,
+      "grad_norm": 0.8239531783912563,
+      "learning_rate": 1.9986017152454497e-05,
+      "loss": 0.4682,
+      "step": 76
+    },
+    {
+      "epoch": 0.23306848278471434,
+      "grad_norm": 0.8325852999423533,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.4812,
+      "step": 77
+    },
+    {
+      "epoch": 0.23609534619750283,
+      "grad_norm": 0.7988804443746536,
+      "learning_rate": 1.9979866764718846e-05,
+      "loss": 0.4521,
+      "step": 78
+    },
+    {
+      "epoch": 0.23912220961029135,
+      "grad_norm": 0.9180978623984122,
+      "learning_rate": 1.997637278839835e-05,
+      "loss": 0.4778,
+      "step": 79
+    },
+    {
+      "epoch": 0.24214907302307984,
+      "grad_norm": 0.8383184142745751,
+      "learning_rate": 1.9972599751485225e-05,
+      "loss": 0.4694,
+      "step": 80
+    },
+    {
+      "epoch": 0.24517593643586832,
+      "grad_norm": 1.0297332574737117,
+      "learning_rate": 1.9968547759519426e-05,
+      "loss": 0.4604,
+      "step": 81
+    },
+    {
+      "epoch": 0.24820279984865684,
+      "grad_norm": 0.9255465235384883,
+      "learning_rate": 1.9964216925843876e-05,
+      "loss": 0.481,
+      "step": 82
+    },
+    {
+      "epoch": 0.2512296632614453,
+      "grad_norm": 0.9036368567155374,
+      "learning_rate": 1.9959607371601303e-05,
+      "loss": 0.4598,
+      "step": 83
+    },
+    {
+      "epoch": 0.2542565266742338,
+      "grad_norm": 0.8561269228567857,
+      "learning_rate": 1.9954719225730847e-05,
+      "loss": 0.4669,
+      "step": 84
+    },
+    {
+      "epoch": 0.25728339008702233,
+      "grad_norm": 1.1420539910264331,
+      "learning_rate": 1.994955262496446e-05,
+      "loss": 0.4919,
+      "step": 85
+    },
+    {
+      "epoch": 0.2603102534998108,
+      "grad_norm": 0.9128263980613053,
+      "learning_rate": 1.9944107713823068e-05,
+      "loss": 0.4782,
+      "step": 86
+    },
+    {
+      "epoch": 0.2633371169125993,
+      "grad_norm": 0.927014187405091,
+      "learning_rate": 1.9938384644612542e-05,
+      "loss": 0.4729,
+      "step": 87
+    },
+    {
+      "epoch": 0.26636398032538783,
+      "grad_norm": 1.007735162287187,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.4763,
+      "step": 88
+    },
+    {
+      "epoch": 0.2693908437381763,
+      "grad_norm": 0.846512682275217,
+      "learning_rate": 1.9926104680106484e-05,
+      "loss": 0.4909,
+      "step": 89
+    },
+    {
+      "epoch": 0.2724177071509648,
+      "grad_norm": 1.0062637527054976,
+      "learning_rate": 1.9919548128307954e-05,
+      "loss": 0.4788,
+      "step": 90
+    },
+    {
+      "epoch": 0.2754445705637533,
+      "grad_norm": 0.9030725255053689,
+      "learning_rate": 1.9912714105424694e-05,
+      "loss": 0.4676,
+      "step": 91
+    },
+    {
+      "epoch": 0.2784714339765418,
+      "grad_norm": 1.0198128119120289,
+      "learning_rate": 1.990560280261901e-05,
+      "loss": 0.4907,
+      "step": 92
+    },
+    {
+      "epoch": 0.2814982973893303,
+      "grad_norm": 0.8944201660226052,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4396,
+      "step": 93
+    },
+    {
+      "epoch": 0.2845251608021188,
+      "grad_norm": 0.9199835920722563,
+      "learning_rate": 1.9890549160664633e-05,
+      "loss": 0.4683,
+      "step": 94
+    },
+    {
+      "epoch": 0.2875520242149073,
+      "grad_norm": 0.9669168126928122,
+      "learning_rate": 1.9882607242598663e-05,
+      "loss": 0.4444,
+      "step": 95
+    },
+    {
+      "epoch": 0.2905788876276958,
+      "grad_norm": 0.8645963095184214,
+      "learning_rate": 1.9874388886763944e-05,
+      "loss": 0.4685,
+      "step": 96
+    },
+    {
+      "epoch": 0.2936057510404843,
+      "grad_norm": 0.9438347282413809,
+      "learning_rate": 1.9865894323045558e-05,
+      "loss": 0.4811,
+      "step": 97
+    },
+    {
+      "epoch": 0.29663261445327277,
+      "grad_norm": 0.8630787473105261,
+      "learning_rate": 1.9857123789054707e-05,
+      "loss": 0.5033,
+      "step": 98
+    },
+    {
+      "epoch": 0.2996594778660613,
+      "grad_norm": 0.8605619048523878,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.4695,
+      "step": 99
+    },
+    {
+      "epoch": 0.3026863412788498,
+      "grad_norm": 0.8186446718750028,
+      "learning_rate": 1.9838755799290993e-05,
+      "loss": 0.4757,
+      "step": 100
+    },
+    {
+      "epoch": 0.30571320469163826,
+      "grad_norm": 0.8322972402032934,
+      "learning_rate": 1.9829158857310288e-05,
+      "loss": 0.472,
+      "step": 101
+    },
+    {
+      "epoch": 0.3087400681044268,
+      "grad_norm": 0.8543323933645511,
+      "learning_rate": 1.9819286972627066e-05,
+      "loss": 0.4747,
+      "step": 102
+    },
+    {
+      "epoch": 0.3117669315172153,
+      "grad_norm": 0.9507409563228003,
+      "learning_rate": 1.9809140421379168e-05,
+      "loss": 0.508,
+      "step": 103
+    },
+    {
+      "epoch": 0.31479379493000376,
+      "grad_norm": 0.9980989261088086,
+      "learning_rate": 1.979871948738743e-05,
+      "loss": 0.4564,
+      "step": 104
+    },
+    {
+      "epoch": 0.3178206583427923,
+      "grad_norm": 0.8679886185359897,
+      "learning_rate": 1.978802446214779e-05,
+      "loss": 0.4864,
+      "step": 105
+    },
+    {
+      "epoch": 0.3208475217555808,
+      "grad_norm": 0.8787910036687884,
+      "learning_rate": 1.9777055644823087e-05,
+      "loss": 0.4629,
+      "step": 106
+    },
+    {
+      "epoch": 0.32387438516836925,
+      "grad_norm": 0.9010179518920617,
+      "learning_rate": 1.9765813342234726e-05,
+      "loss": 0.4751,
+      "step": 107
+    },
+    {
+      "epoch": 0.32690124858115777,
+      "grad_norm": 0.8798798390573066,
+      "learning_rate": 1.9754297868854075e-05,
+      "loss": 0.4822,
+      "step": 108
+    },
+    {
+      "epoch": 0.3299281119939463,
+      "grad_norm": 1.2453215735173986,
+      "learning_rate": 1.9742509546793673e-05,
+      "loss": 0.4536,
+      "step": 109
+    },
+    {
+      "epoch": 0.33295497540673474,
+      "grad_norm": 0.8995259856659721,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.4787,
+      "step": 110
+    },
+    {
+      "epoch": 0.33598183881952326,
+      "grad_norm": 1.0438663428426223,
+      "learning_rate": 1.9718115683235418e-05,
+      "loss": 0.4525,
+      "step": 111
+    },
+    {
+      "epoch": 0.3390087022323118,
+      "grad_norm": 0.8523579171515343,
+      "learning_rate": 1.970551082408636e-05,
+      "loss": 0.4607,
+      "step": 112
+    },
+    {
+      "epoch": 0.3420355656451003,
+      "grad_norm": 1.1630024642078751,
+      "learning_rate": 1.969263448093608e-05,
+      "loss": 0.4531,
+      "step": 113
+    },
+    {
+      "epoch": 0.34506242905788875,
+      "grad_norm": 0.8966891678664323,
+      "learning_rate": 1.9679487013963566e-05,
+      "loss": 0.4763,
+      "step": 114
+    },
+    {
+      "epoch": 0.34808929247067727,
+      "grad_norm": 1.0489966077543964,
+      "learning_rate": 1.9666068790931733e-05,
+      "loss": 0.4723,
+      "step": 115
+    },
+    {
+      "epoch": 0.3511161558834658,
+      "grad_norm": 0.8239524274910004,
+      "learning_rate": 1.9652380187177128e-05,
+      "loss": 0.4639,
+      "step": 116
+    },
+    {
+      "epoch": 0.35414301929625425,
+      "grad_norm": 0.7966715957847416,
+      "learning_rate": 1.9638421585599422e-05,
+      "loss": 0.4743,
+      "step": 117
+    },
+    {
+      "epoch": 0.35716988270904276,
+      "grad_norm": 0.8463328902156634,
+      "learning_rate": 1.9624193376650708e-05,
+      "loss": 0.4731,
+      "step": 118
+    },
+    {
+      "epoch": 0.3601967461218313,
+      "grad_norm": 0.814468179746004,
+      "learning_rate": 1.960969595832457e-05,
+      "loss": 0.4629,
+      "step": 119
+    },
+    {
+      "epoch": 0.36322360953461974,
+      "grad_norm": 0.8856975105597623,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4856,
+      "step": 120
+    },
+    {
+      "epoch": 0.36625047294740826,
+      "grad_norm": 0.7421164492375186,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.4473,
+      "step": 121
+    },
+    {
+      "epoch": 0.3692773363601968,
+      "grad_norm": 0.7761121924569291,
+      "learning_rate": 1.956459253990476e-05,
+      "loss": 0.4636,
+      "step": 122
+    },
+    {
+      "epoch": 0.37230419977298523,
+      "grad_norm": 0.8167225097857224,
+      "learning_rate": 1.9549022414440738e-05,
+      "loss": 0.4768,
+      "step": 123
+    },
+    {
+      "epoch": 0.37533106318577375,
+      "grad_norm": 0.8786562518609927,
+      "learning_rate": 1.9533185182292705e-05,
+      "loss": 0.4864,
+      "step": 124
+    },
+    {
+      "epoch": 0.37835792659856227,
+      "grad_norm": 0.779688363002822,
+      "learning_rate": 1.9517081286462082e-05,
+      "loss": 0.4756,
+      "step": 125
+    },
+    {
+      "epoch": 0.3813847900113507,
+      "grad_norm": 0.8206498545957859,
+      "learning_rate": 1.9500711177409456e-05,
+      "loss": 0.4846,
+      "step": 126
+    },
+    {
+      "epoch": 0.38441165342413924,
+      "grad_norm": 0.8324827802134679,
+      "learning_rate": 1.9484075313041968e-05,
+      "loss": 0.4536,
+      "step": 127
+    },
+    {
+      "epoch": 0.38743851683692776,
+      "grad_norm": 0.7478616093586664,
+      "learning_rate": 1.9467174158700507e-05,
+      "loss": 0.4523,
+      "step": 128
+    },
+    {
+      "epoch": 0.3904653802497162,
+      "grad_norm": 0.7748039603935429,
+      "learning_rate": 1.9450008187146685e-05,
+      "loss": 0.4376,
+      "step": 129
+    },
+    {
+      "epoch": 0.39349224366250474,
+      "grad_norm": 0.7859357092191007,
+      "learning_rate": 1.9432577878549635e-05,
+      "loss": 0.4832,
+      "step": 130
+    },
+    {
+      "epoch": 0.39651910707529325,
+      "grad_norm": 0.7567057335709949,
+      "learning_rate": 1.9414883720472557e-05,
+      "loss": 0.4594,
+      "step": 131
+    },
+    {
+      "epoch": 0.3995459704880817,
+      "grad_norm": 0.7938273707653927,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.4837,
+      "step": 132
+    },
+    {
+      "epoch": 0.40257283390087023,
+      "grad_norm": 0.7589483204428902,
+      "learning_rate": 1.937870584301945e-05,
+      "loss": 0.4632,
+      "step": 133
+    },
+    {
+      "epoch": 0.40559969731365875,
+      "grad_norm": 0.8120167737050139,
+      "learning_rate": 1.9360223135616423e-05,
+      "loss": 0.4695,
+      "step": 134
+    },
+    {
+      "epoch": 0.4086265607264472,
+      "grad_norm": 0.7779022154966455,
+      "learning_rate": 1.9341478602651068e-05,
+      "loss": 0.4828,
+      "step": 135
+    },
+    {
+      "epoch": 0.4116534241392357,
+      "grad_norm": 0.77798904742415,
+      "learning_rate": 1.932247276844826e-05,
+      "loss": 0.4458,
+      "step": 136
+    },
+    {
+      "epoch": 0.41468028755202424,
+      "grad_norm": 0.8457764141144063,
+      "learning_rate": 1.9303206164642037e-05,
+      "loss": 0.4692,
+      "step": 137
+    },
+    {
+      "epoch": 0.4177071509648127,
+      "grad_norm": 0.8064406254345843,
+      "learning_rate": 1.9283679330160726e-05,
+      "loss": 0.4809,
+      "step": 138
+    },
+    {
+      "epoch": 0.4207340143776012,
+      "grad_norm": 0.8247528026084574,
+      "learning_rate": 1.9263892811211865e-05,
+      "loss": 0.4838,
+      "step": 139
+    },
+    {
+      "epoch": 0.42376087779038973,
+      "grad_norm": 0.77324047002901,
+      "learning_rate": 1.9243847161266924e-05,
+      "loss": 0.4731,
+      "step": 140
+    },
+    {
+      "epoch": 0.4267877412031782,
+      "grad_norm": 0.7750589825274077,
+      "learning_rate": 1.9223542941045817e-05,
+      "loss": 0.4927,
+      "step": 141
+    },
+    {
+      "epoch": 0.4298146046159667,
+      "grad_norm": 0.829510661119134,
+      "learning_rate": 1.920298071850123e-05,
+      "loss": 0.4617,
+      "step": 142
+    },
+    {
+      "epoch": 0.4328414680287552,
+      "grad_norm": 0.763904599202375,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.4443,
+      "step": 143
+    },
+    {
+      "epoch": 0.4358683314415437,
+      "grad_norm": 0.8013752883055623,
+      "learning_rate": 1.9161084574320696e-05,
+      "loss": 0.4654,
+      "step": 144
+    },
+    {
+      "epoch": 0.4388951948543322,
+      "grad_norm": 0.8660041096208889,
+      "learning_rate": 1.913975182460996e-05,
+      "loss": 0.4724,
+      "step": 145
+    },
+    {
+      "epoch": 0.4419220582671207,
+      "grad_norm": 0.7538088274285658,
+      "learning_rate": 1.9118163416393392e-05,
+      "loss": 0.4568,
+      "step": 146
+    },
+    {
+      "epoch": 0.4449489216799092,
+      "grad_norm": 0.9063120964139122,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4943,
+      "step": 147
+    },
+    {
+      "epoch": 0.4479757850926977,
+      "grad_norm": 0.8244084440984972,
+      "learning_rate": 1.9074222047073945e-05,
+      "loss": 0.4767,
+      "step": 148
+    },
+    {
+      "epoch": 0.4510026485054862,
+      "grad_norm": 0.8427793996169352,
+      "learning_rate": 1.9051870315105626e-05,
+      "loss": 0.4636,
+      "step": 149
+    },
+    {
+      "epoch": 0.4540295119182747,
+      "grad_norm": 0.7672388909659206,
+      "learning_rate": 1.9029265382866216e-05,
+      "loss": 0.4523,
+      "step": 150
+    },
+    {
+      "epoch": 0.4570563753310632,
+      "grad_norm": 0.8388859822000623,
+      "learning_rate": 1.9006407882664256e-05,
+      "loss": 0.4376,
+      "step": 151
+    },
+    {
+      "epoch": 0.4600832387438517,
+      "grad_norm": 0.8310061822482814,
+      "learning_rate": 1.8983298453873172e-05,
+      "loss": 0.4623,
+      "step": 152
+    },
+    {
+      "epoch": 0.46311010215664017,
+      "grad_norm": 0.8313203200903586,
+      "learning_rate": 1.895993774291336e-05,
+      "loss": 0.4725,
+      "step": 153
+    },
+    {
+      "epoch": 0.4661369655694287,
+      "grad_norm": 0.6941943179807292,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.4349,
+      "step": 154
+    },
+    {
+      "epoch": 0.4691638289822172,
+      "grad_norm": 0.919659872228411,
+      "learning_rate": 1.891246509529539e-05,
+      "loss": 0.464,
+      "step": 155
+    },
+    {
+      "epoch": 0.47219069239500566,
+      "grad_norm": 0.7616237459318409,
+      "learning_rate": 1.8888354486549238e-05,
+      "loss": 0.4789,
+      "step": 156
+    },
+    {
+      "epoch": 0.4752175558077942,
+      "grad_norm": 0.7640584160047689,
+      "learning_rate": 1.886399525142122e-05,
+      "loss": 0.452,
+      "step": 157
+    },
+    {
+      "epoch": 0.4782444192205827,
+      "grad_norm": 0.7637732284533377,
+      "learning_rate": 1.8839388071291506e-05,
+      "loss": 0.458,
+      "step": 158
+    },
+    {
+      "epoch": 0.48127128263337116,
+      "grad_norm": 0.8049455348443784,
+      "learning_rate": 1.881453363447582e-05,
+      "loss": 0.4831,
+      "step": 159
+    },
+    {
+      "epoch": 0.48429814604615967,
+      "grad_norm": 0.8984650989053296,
+      "learning_rate": 1.8789432636206197e-05,
+      "loss": 0.4359,
+      "step": 160
+    },
+    {
+      "epoch": 0.4873250094589482,
+      "grad_norm": 0.8726512958851851,
+      "learning_rate": 1.8764085778611507e-05,
+      "loss": 0.4749,
+      "step": 161
+    },
+    {
+      "epoch": 0.49035187287173665,
+      "grad_norm": 0.876900644497976,
+      "learning_rate": 1.873849377069785e-05,
+      "loss": 0.4343,
+      "step": 162
+    },
+    {
+      "epoch": 0.49337873628452517,
+      "grad_norm": 0.8013923318694746,
+      "learning_rate": 1.87126573283287e-05,
+      "loss": 0.4535,
+      "step": 163
+    },
+    {
+      "epoch": 0.4964055996973137,
+      "grad_norm": 1.0174158100520057,
+      "learning_rate": 1.8686577174204887e-05,
+      "loss": 0.4673,
+      "step": 164
+    },
+    {
+      "epoch": 0.49943246311010214,
+      "grad_norm": 0.7775836364649137,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.4541,
+      "step": 165
+    },
+    {
+      "epoch": 0.5024593265228906,
+      "grad_norm": 0.9717206968796297,
+      "learning_rate": 1.863368865556191e-05,
+      "loss": 0.4493,
+      "step": 166
+    },
+    {
+      "epoch": 0.5054861899356792,
+      "grad_norm": 0.8631593419820175,
+      "learning_rate": 1.8606881770448305e-05,
+      "loss": 0.4428,
+      "step": 167
+    },
+    {
+      "epoch": 0.5085130533484676,
+      "grad_norm": 0.7793776266519721,
+      "learning_rate": 1.8579834132349773e-05,
+      "loss": 0.4576,
+      "step": 168
+    },
+    {
+      "epoch": 0.5115399167612561,
+      "grad_norm": 0.7482928709019844,
+      "learning_rate": 1.8552546497846893e-05,
+      "loss": 0.4491,
+      "step": 169
+    },
+    {
+      "epoch": 0.5145667801740447,
+      "grad_norm": 0.8268090879099498,
+      "learning_rate": 1.8525019630233463e-05,
+      "loss": 0.4876,
+      "step": 170
+    },
+    {
+      "epoch": 0.5175936435868331,
+      "grad_norm": 0.9516987012090243,
+      "learning_rate": 1.8497254299495147e-05,
+      "loss": 0.4567,
+      "step": 171
+    },
+    {
+      "epoch": 0.5206205069996216,
+      "grad_norm": 0.7572730016294951,
+      "learning_rate": 1.8469251282287925e-05,
+      "loss": 0.4668,
+      "step": 172
+    },
+    {
+      "epoch": 0.5236473704124102,
+      "grad_norm": 0.7863582941827981,
+      "learning_rate": 1.8441011361916387e-05,
+      "loss": 0.427,
+      "step": 173
+    },
+    {
+      "epoch": 0.5266742338251986,
+      "grad_norm": 0.8431233349582884,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4671,
+      "step": 174
+    },
+    {
+      "epoch": 0.5297010972379871,
+      "grad_norm": 0.7398126931258614,
+      "learning_rate": 1.8383823978010077e-05,
+      "loss": 0.4607,
+      "step": 175
+    },
+    {
+      "epoch": 0.5327279606507757,
+      "grad_norm": 0.8209924483862278,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.4208,
+      "step": 176
+    },
+    {
+      "epoch": 0.5357548240635641,
+      "grad_norm": 0.7953203563029778,
+      "learning_rate": 1.8325698546347714e-05,
+      "loss": 0.4719,
+      "step": 177
+    },
+    {
+      "epoch": 0.5387816874763526,
+      "grad_norm": 0.8162322592329131,
+      "learning_rate": 1.8296286090880362e-05,
+      "loss": 0.4442,
+      "step": 178
+    },
+    {
+      "epoch": 0.5418085508891411,
+      "grad_norm": 0.756757836008388,
+      "learning_rate": 1.8266641570456915e-05,
+      "loss": 0.4378,
+      "step": 179
+    },
+    {
+      "epoch": 0.5448354143019296,
+      "grad_norm": 0.7527936508224499,
+      "learning_rate": 1.8236765814298328e-05,
+      "loss": 0.4523,
+      "step": 180
+    },
+    {
+      "epoch": 0.5478622777147181,
+      "grad_norm": 0.7493626770062112,
+      "learning_rate": 1.820665965809373e-05,
+      "loss": 0.4516,
+      "step": 181
+    },
+    {
+      "epoch": 0.5508891411275066,
+      "grad_norm": 0.8085655095255699,
+      "learning_rate": 1.8176323943977034e-05,
+      "loss": 0.4461,
+      "step": 182
+    },
+    {
+      "epoch": 0.5539160045402951,
+      "grad_norm": 0.7689198462906255,
+      "learning_rate": 1.814575952050336e-05,
+      "loss": 0.4553,
+      "step": 183
+    },
+    {
+      "epoch": 0.5569428679530836,
+      "grad_norm": 0.7505589477376717,
+      "learning_rate": 1.8114967242625342e-05,
+      "loss": 0.4593,
+      "step": 184
+    },
+    {
+      "epoch": 0.5599697313658721,
+      "grad_norm": 0.7538416623482549,
+      "learning_rate": 1.808394797166919e-05,
+      "loss": 0.4496,
+      "step": 185
+    },
+    {
+      "epoch": 0.5629965947786606,
+      "grad_norm": 0.82567162277707,
+      "learning_rate": 1.8052702575310588e-05,
+      "loss": 0.4232,
+      "step": 186
+    },
+    {
+      "epoch": 0.5660234581914491,
+      "grad_norm": 0.7512467770844982,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.4487,
+      "step": 187
+    },
+    {
+      "epoch": 0.5690503216042376,
+      "grad_norm": 0.7089577320214763,
+      "learning_rate": 1.7989536908690413e-05,
+      "loss": 0.4252,
+      "step": 188
+    },
+    {
+      "epoch": 0.5720771850170261,
+      "grad_norm": 0.9590232796041346,
+      "learning_rate": 1.7957618405308323e-05,
+      "loss": 0.4669,
+      "step": 189
+    },
+    {
+      "epoch": 0.5751040484298146,
+      "grad_norm": 0.6544513145406187,
+      "learning_rate": 1.792547731023332e-05,
+      "loss": 0.4219,
+      "step": 190
+    },
+    {
+      "epoch": 0.5781309118426031,
+      "grad_norm": 1.0213854332076853,
+      "learning_rate": 1.789311452252092e-05,
+      "loss": 0.4326,
+      "step": 191
+    },
+    {
+      "epoch": 0.5811577752553916,
+      "grad_norm": 0.7331906821716333,
+      "learning_rate": 1.7860530947427878e-05,
+      "loss": 0.4288,
+      "step": 192
+    },
+    {
+      "epoch": 0.58418463866818,
+      "grad_norm": 0.9237485901277386,
+      "learning_rate": 1.782772749638682e-05,
+      "loss": 0.4379,
+      "step": 193
+    },
+    {
+      "epoch": 0.5872115020809686,
+      "grad_norm": 0.8002601541058559,
+      "learning_rate": 1.779470508698079e-05,
+      "loss": 0.4306,
+      "step": 194
+    },
+    {
+      "epoch": 0.5902383654937571,
+      "grad_norm": 0.8127815717823276,
+      "learning_rate": 1.776146464291757e-05,
+      "loss": 0.4378,
+      "step": 195
+    },
+    {
+      "epoch": 0.5932652289065455,
+      "grad_norm": 0.8594663461983456,
+      "learning_rate": 1.772800709400383e-05,
+      "loss": 0.4689,
+      "step": 196
+    },
+    {
+      "epoch": 0.5962920923193341,
+      "grad_norm": 0.7502752205664198,
+      "learning_rate": 1.7694333376119144e-05,
+      "loss": 0.4394,
+      "step": 197
+    },
+    {
+      "epoch": 0.5993189557321226,
+      "grad_norm": 0.9363880136401701,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.4396,
+      "step": 198
+    },
+    {
+      "epoch": 0.602345819144911,
+      "grad_norm": 0.6969681819802825,
+      "learning_rate": 1.762634120716238e-05,
+      "loss": 0.4662,
+      "step": 199
+    },
+    {
+      "epoch": 0.6053726825576996,
+      "grad_norm": 0.8262365462081662,
+      "learning_rate": 1.7592024657977432e-05,
+      "loss": 0.4499,
+      "step": 200
+    },
+    {
+      "epoch": 0.6083995459704881,
+      "grad_norm": 0.8404567350166624,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4574,
+      "step": 201
+    },
+    {
+      "epoch": 0.6114264093832765,
+      "grad_norm": 0.7548638861960166,
+      "learning_rate": 1.75227554297058e-05,
+      "loss": 0.4443,
+      "step": 202
+    },
+    {
+      "epoch": 0.6144532727960651,
+      "grad_norm": 0.8769374586359174,
+      "learning_rate": 1.7487804688228327e-05,
+      "loss": 0.442,
+      "step": 203
+    },
+    {
+      "epoch": 0.6174801362088536,
+      "grad_norm": 0.7359239818293573,
+      "learning_rate": 1.745264449675755e-05,
+      "loss": 0.4443,
+      "step": 204
+    },
+    {
+      "epoch": 0.620506999621642,
+      "grad_norm": 0.9185699319569386,
+      "learning_rate": 1.7417275838799596e-05,
+      "loss": 0.4416,
+      "step": 205
+    },
+    {
+      "epoch": 0.6235338630344306,
+      "grad_norm": 0.8769484809856751,
+      "learning_rate": 1.7381699703691866e-05,
+      "loss": 0.4471,
+      "step": 206
+    },
+    {
+      "epoch": 0.626560726447219,
+      "grad_norm": 0.8038975190789405,
+      "learning_rate": 1.734591708657533e-05,
+      "loss": 0.4201,
+      "step": 207
+    },
+    {
+      "epoch": 0.6295875898600075,
+      "grad_norm": 0.8223317500692707,
+      "learning_rate": 1.730992898836672e-05,
+      "loss": 0.4334,
+      "step": 208
+    },
+    {
+      "epoch": 0.6326144532727961,
+      "grad_norm": 0.7451475080445484,
+      "learning_rate": 1.7273736415730488e-05,
+      "loss": 0.4339,
+      "step": 209
+    },
+    {
+      "epoch": 0.6356413166855845,
+      "grad_norm": 0.8383419191899345,
+      "learning_rate": 1.72373403810507e-05,
+      "loss": 0.4678,
+      "step": 210
+    },
+    {
+      "epoch": 0.638668180098373,
+      "grad_norm": 0.7405080919955892,
+      "learning_rate": 1.720074190240269e-05,
+      "loss": 0.447,
+      "step": 211
+    },
+    {
+      "epoch": 0.6416950435111616,
+      "grad_norm": 0.8404781920720572,
+      "learning_rate": 1.7163942003524574e-05,
+      "loss": 0.4555,
+      "step": 212
+    },
+    {
+      "epoch": 0.64472190692395,
+      "grad_norm": 0.8037078444524368,
+      "learning_rate": 1.7126941713788633e-05,
+      "loss": 0.4669,
+      "step": 213
+    },
+    {
+      "epoch": 0.6477487703367385,
+      "grad_norm": 0.8052619489797984,
+      "learning_rate": 1.70897420681725e-05,
+      "loss": 0.4424,
+      "step": 214
+    },
+    {
+      "epoch": 0.6507756337495271,
+      "grad_norm": 0.8115253394702173,
+      "learning_rate": 1.7052344107230244e-05,
+      "loss": 0.4372,
+      "step": 215
+    },
+    {
+      "epoch": 0.6538024971623155,
+      "grad_norm": 0.7806855630239484,
+      "learning_rate": 1.7014748877063212e-05,
+      "loss": 0.4559,
+      "step": 216
+    },
+    {
+      "epoch": 0.656829360575104,
+      "grad_norm": 0.9359582566856662,
+      "learning_rate": 1.697695742929082e-05,
+      "loss": 0.4363,
+      "step": 217
+    },
+    {
+      "epoch": 0.6598562239878926,
+      "grad_norm": 0.8734447155852043,
+      "learning_rate": 1.693897082102109e-05,
+      "loss": 0.4475,
+      "step": 218
+    },
+    {
+      "epoch": 0.662883087400681,
+      "grad_norm": 0.8575929659814436,
+      "learning_rate": 1.6900790114821122e-05,
+      "loss": 0.4296,
+      "step": 219
+    },
+    {
+      "epoch": 0.6659099508134695,
+      "grad_norm": 0.7679802880580873,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.4379,
+      "step": 220
+    },
+    {
+      "epoch": 0.6689368142262581,
+      "grad_norm": 0.9503444920008771,
+      "learning_rate": 1.682385068601563e-05,
+      "loss": 0.4229,
+      "step": 221
+    },
+    {
+      "epoch": 0.6719636776390465,
+      "grad_norm": 0.9818965681698484,
+      "learning_rate": 1.6785094115571323e-05,
+      "loss": 0.4397,
+      "step": 222
+    },
+    {
+      "epoch": 0.6749905410518351,
+      "grad_norm": 0.949739940696302,
+      "learning_rate": 1.674614775145901e-05,
+      "loss": 0.4395,
+      "step": 223
+    },
+    {
+      "epoch": 0.6780174044646236,
+      "grad_norm": 0.9720478553088886,
+      "learning_rate": 1.670701268309221e-05,
+      "loss": 0.4338,
+      "step": 224
+    },
+    {
+      "epoch": 0.681044267877412,
+      "grad_norm": 0.6646901380911848,
+      "learning_rate": 1.666769000516292e-05,
+      "loss": 0.4447,
+      "step": 225
+    },
+    {
+      "epoch": 0.6840711312902006,
+      "grad_norm": 0.8620172646547367,
+      "learning_rate": 1.6628180817610963e-05,
+      "loss": 0.4407,
+      "step": 226
+    },
+    {
+      "epoch": 0.687097994702989,
+      "grad_norm": 0.8410825814209901,
+      "learning_rate": 1.658848622559325e-05,
+      "loss": 0.4314,
+      "step": 227
+    },
+    {
+      "epoch": 0.6901248581157775,
+      "grad_norm": 0.6750036021043792,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.4285,
+      "step": 228
+    },
+    {
+      "epoch": 0.6931517215285661,
+      "grad_norm": 0.8774881260425775,
+      "learning_rate": 1.6508545274687936e-05,
+      "loss": 0.4611,
+      "step": 229
+    },
+    {
+      "epoch": 0.6961785849413545,
+      "grad_norm": 0.799467547469604,
+      "learning_rate": 1.6468301151920576e-05,
+      "loss": 0.4538,
+      "step": 230
+    },
+    {
+      "epoch": 0.699205448354143,
+      "grad_norm": 0.8007649852269014,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.4469,
+      "step": 231
+    },
+    {
+      "epoch": 0.7022323117669316,
+      "grad_norm": 0.7928687230805521,
+      "learning_rate": 1.6387271240298082e-05,
+      "loss": 0.4316,
+      "step": 232
+    },
+    {
+      "epoch": 0.70525917517972,
+      "grad_norm": 0.7227786749570734,
+      "learning_rate": 1.6346487718023762e-05,
+      "loss": 0.4498,
+      "step": 233
+    },
+    {
+      "epoch": 0.7082860385925085,
+      "grad_norm": 0.8155972822029076,
+      "learning_rate": 1.6305526670845225e-05,
+      "loss": 0.4402,
+      "step": 234
+    },
+    {
+      "epoch": 0.7113129020052971,
+      "grad_norm": 0.835747470957178,
+      "learning_rate": 1.6264389244531015e-05,
+      "loss": 0.4499,
+      "step": 235
+    },
+    {
+      "epoch": 0.7143397654180855,
+      "grad_norm": 0.7789421214073318,
+      "learning_rate": 1.6223076589783368e-05,
+      "loss": 0.4608,
+      "step": 236
+    },
+    {
+      "epoch": 0.717366628830874,
+      "grad_norm": 0.8004779735160987,
+      "learning_rate": 1.6181589862206053e-05,
+      "loss": 0.4551,
+      "step": 237
+    },
+    {
+      "epoch": 0.7203934922436626,
+      "grad_norm": 0.8754865915069941,
+      "learning_rate": 1.613993022227202e-05,
+      "loss": 0.461,
+      "step": 238
+    },
+    {
+      "epoch": 0.723420355656451,
+      "grad_norm": 0.7913686558524112,
+      "learning_rate": 1.6098098835290955e-05,
+      "loss": 0.4195,
+      "step": 239
+    },
+    {
+      "epoch": 0.7264472190692395,
+      "grad_norm": 0.7833138215555906,
+      "learning_rate": 1.6056096871376667e-05,
+      "loss": 0.4197,
+      "step": 240
+    },
+    {
+      "epoch": 0.729474082482028,
+      "grad_norm": 0.7619727424964072,
+      "learning_rate": 1.6013925505414386e-05,
+      "loss": 0.4522,
+      "step": 241
+    },
+    {
+      "epoch": 0.7325009458948165,
+      "grad_norm": 0.8500538554542452,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.445,
+      "step": 242
+    },
+    {
+      "epoch": 0.735527809307605,
+      "grad_norm": 0.7163021394447447,
+      "learning_rate": 1.5929079290546408e-05,
+      "loss": 0.4411,
+      "step": 243
+    },
+    {
+      "epoch": 0.7385546727203935,
+      "grad_norm": 0.7325483025095844,
+      "learning_rate": 1.5886406814971728e-05,
+      "loss": 0.442,
+      "step": 244
+    },
+    {
+      "epoch": 0.741581536133182,
+      "grad_norm": 0.7423505630201604,
+      "learning_rate": 1.584356968394471e-05,
+      "loss": 0.4453,
+      "step": 245
+    },
+    {
+      "epoch": 0.7446083995459705,
+      "grad_norm": 0.9030628630736712,
+      "learning_rate": 1.5800569095711983e-05,
+      "loss": 0.4661,
+      "step": 246
+    },
+    {
+      "epoch": 0.747635262958759,
+      "grad_norm": 0.7355698554927667,
+      "learning_rate": 1.575740625309244e-05,
+      "loss": 0.479,
+      "step": 247
+    },
+    {
+      "epoch": 0.7506621263715475,
+      "grad_norm": 0.8091244523318958,
+      "learning_rate": 1.5714082363443576e-05,
+      "loss": 0.4433,
+      "step": 248
+    },
+    {
+      "epoch": 0.753688989784336,
+      "grad_norm": 0.8022265619078223,
+      "learning_rate": 1.5670598638627707e-05,
+      "loss": 0.4425,
+      "step": 249
+    },
+    {
+      "epoch": 0.7567158531971245,
+      "grad_norm": 0.6530655530355023,
+      "learning_rate": 1.5626956294978103e-05,
+      "loss": 0.4145,
+      "step": 250
+    },
+    {
+      "epoch": 0.759742716609913,
+      "grad_norm": 0.7280088596554589,
+      "learning_rate": 1.5583156553264923e-05,
+      "loss": 0.435,
+      "step": 251
+    },
+    {
+      "epoch": 0.7627695800227015,
+      "grad_norm": 0.8085717035300192,
+      "learning_rate": 1.5539200638661106e-05,
+      "loss": 0.4362,
+      "step": 252
+    },
+    {
+      "epoch": 0.76579644343549,
+      "grad_norm": 0.6901263962546973,
+      "learning_rate": 1.5495089780708062e-05,
+      "loss": 0.4654,
+      "step": 253
+    },
+    {
+      "epoch": 0.7688233068482785,
+      "grad_norm": 0.7098947827193745,
+      "learning_rate": 1.5450825213281317e-05,
+      "loss": 0.4346,
+      "step": 254
+    },
+    {
+      "epoch": 0.771850170261067,
+      "grad_norm": 0.7873272830702647,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4385,
+      "step": 255
+    },
+    {
+      "epoch": 0.7748770336738555,
+      "grad_norm": 0.7362688222696735,
+      "learning_rate": 1.5361839906972095e-05,
+      "loss": 0.417,
+      "step": 256
+    },
+    {
+      "epoch": 0.777903897086644,
+      "grad_norm": 0.7089500521002634,
+      "learning_rate": 1.531712165719992e-05,
+      "loss": 0.4319,
+      "step": 257
+    },
+    {
+      "epoch": 0.7809307604994324,
+      "grad_norm": 0.7664979494260672,
+      "learning_rate": 1.5272254676105026e-05,
+      "loss": 0.4364,
+      "step": 258
+    },
+    {
+      "epoch": 0.783957623912221,
+      "grad_norm": 0.7620633538637864,
+      "learning_rate": 1.5227240218713326e-05,
+      "loss": 0.429,
+      "step": 259
+    },
+    {
+      "epoch": 0.7869844873250095,
+      "grad_norm": 0.8208667703803756,
+      "learning_rate": 1.5182079544175957e-05,
+      "loss": 0.4522,
+      "step": 260
+    },
+    {
+      "epoch": 0.7900113507377979,
+      "grad_norm": 0.7167228193026779,
+      "learning_rate": 1.5136773915734067e-05,
+      "loss": 0.4399,
+      "step": 261
+    },
+    {
+      "epoch": 0.7930382141505865,
+      "grad_norm": 0.7025883626195679,
+      "learning_rate": 1.5091324600683472e-05,
+      "loss": 0.428,
+      "step": 262
+    },
+    {
+      "epoch": 0.796065077563375,
+      "grad_norm": 0.7483075988786887,
+      "learning_rate": 1.5045732870339213e-05,
+      "loss": 0.4071,
+      "step": 263
+    },
+    {
+      "epoch": 0.7990919409761634,
+      "grad_norm": 0.7973987482105404,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4523,
+      "step": 264
+    },
+    {
+      "epoch": 0.802118804388952,
+      "grad_norm": 0.7195636013634443,
+      "learning_rate": 1.4954127268912525e-05,
+      "loss": 0.4298,
+      "step": 265
+    },
+    {
+      "epoch": 0.8051456678017405,
+      "grad_norm": 0.8091464284418212,
+      "learning_rate": 1.4908115960235683e-05,
+      "loss": 0.4464,
+      "step": 266
+    },
+    {
+      "epoch": 0.8081725312145289,
+      "grad_norm": 0.7908091350019503,
+      "learning_rate": 1.4861967361004687e-05,
+      "loss": 0.4514,
+      "step": 267
+    },
+    {
+      "epoch": 0.8111993946273175,
+      "grad_norm": 0.723894131818355,
+      "learning_rate": 1.4815682762095065e-05,
+      "loss": 0.4167,
+      "step": 268
+    },
+    {
+      "epoch": 0.814226258040106,
+      "grad_norm": 0.7221637987629108,
+      "learning_rate": 1.476926345818654e-05,
+      "loss": 0.4517,
+      "step": 269
+    },
+    {
+      "epoch": 0.8172531214528944,
+      "grad_norm": 0.8062700782068588,
+      "learning_rate": 1.472271074772683e-05,
+      "loss": 0.437,
+      "step": 270
+    },
+    {
+      "epoch": 0.820279984865683,
+      "grad_norm": 0.6943138117225822,
+      "learning_rate": 1.4676025932895315e-05,
+      "loss": 0.4151,
+      "step": 271
+    },
+    {
+      "epoch": 0.8233068482784714,
+      "grad_norm": 0.7605336542921753,
+      "learning_rate": 1.4629210319566626e-05,
+      "loss": 0.4225,
+      "step": 272
+    },
+    {
+      "epoch": 0.8263337116912599,
+      "grad_norm": 0.7145524552095768,
+      "learning_rate": 1.4582265217274105e-05,
+      "loss": 0.4442,
+      "step": 273
+    },
+    {
+      "epoch": 0.8293605751040485,
+      "grad_norm": 0.8012818689165345,
+      "learning_rate": 1.4535191939173179e-05,
+      "loss": 0.435,
+      "step": 274
+    },
+    {
+      "epoch": 0.8323874385168369,
+      "grad_norm": 0.7795914380983968,
+      "learning_rate": 1.4487991802004625e-05,
+      "loss": 0.4156,
+      "step": 275
+    },
+    {
+      "epoch": 0.8354143019296254,
+      "grad_norm": 0.7255582567806893,
+      "learning_rate": 1.4440666126057743e-05,
+      "loss": 0.4437,
+      "step": 276
+    },
+    {
+      "epoch": 0.838441165342414,
+      "grad_norm": 0.7769738426696209,
+      "learning_rate": 1.4393216235133427e-05,
+      "loss": 0.417,
+      "step": 277
+    },
+    {
+      "epoch": 0.8414680287552024,
+      "grad_norm": 0.7027102720040505,
+      "learning_rate": 1.4345643456507126e-05,
+      "loss": 0.4235,
+      "step": 278
+    },
+    {
+      "epoch": 0.8444948921679909,
+      "grad_norm": 0.7789235539214701,
+      "learning_rate": 1.4297949120891718e-05,
+      "loss": 0.4183,
+      "step": 279
+    },
+    {
+      "epoch": 0.8475217555807795,
+      "grad_norm": 0.7175810397100073,
+      "learning_rate": 1.4250134562400301e-05,
+      "loss": 0.4367,
+      "step": 280
+    },
+    {
+      "epoch": 0.8505486189935679,
+      "grad_norm": 0.7084814524264332,
+      "learning_rate": 1.4202201118508863e-05,
+      "loss": 0.4249,
+      "step": 281
+    },
+    {
+      "epoch": 0.8535754824063564,
+      "grad_norm": 0.7891658287575073,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.4378,
+      "step": 282
+    },
+    {
+      "epoch": 0.856602345819145,
+      "grad_norm": 0.8237665106721381,
+      "learning_rate": 1.4105982941019751e-05,
+      "loss": 0.4679,
+      "step": 283
+    },
+    {
+      "epoch": 0.8596292092319334,
+      "grad_norm": 0.7660031680570911,
+      "learning_rate": 1.405770089885134e-05,
+      "loss": 0.4393,
+      "step": 284
+    },
+    {
+      "epoch": 0.8626560726447219,
+      "grad_norm": 0.6871374224955339,
+      "learning_rate": 1.4009305354066138e-05,
+      "loss": 0.419,
+      "step": 285
+    },
+    {
+      "epoch": 0.8656829360575105,
+      "grad_norm": 0.7150016514605124,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 0.4258,
+      "step": 286
+    },
+    {
+      "epoch": 0.8687097994702989,
+      "grad_norm": 0.7421544996311775,
+      "learning_rate": 1.39121791746921e-05,
+      "loss": 0.4386,
+      "step": 287
+    },
+    {
+      "epoch": 0.8717366628830874,
+      "grad_norm": 0.70004433222392,
+      "learning_rate": 1.3863451256931286e-05,
+      "loss": 0.4177,
+      "step": 288
+    },
+    {
+      "epoch": 0.874763526295876,
+      "grad_norm": 0.7586150361094046,
+      "learning_rate": 1.381461527013374e-05,
+      "loss": 0.4304,
+      "step": 289
+    },
+    {
+      "epoch": 0.8777903897086644,
+      "grad_norm": 0.7635109869891171,
+      "learning_rate": 1.3765672580346986e-05,
+      "loss": 0.459,
+      "step": 290
+    },
+    {
+      "epoch": 0.8808172531214529,
+      "grad_norm": 0.6917749468946599,
+      "learning_rate": 1.3716624556603275e-05,
+      "loss": 0.4358,
+      "step": 291
+    },
+    {
+      "epoch": 0.8838441165342414,
+      "grad_norm": 0.7771212534338365,
+      "learning_rate": 1.3667472570881264e-05,
+      "loss": 0.4375,
+      "step": 292
+    },
+    {
+      "epoch": 0.8868709799470299,
+      "grad_norm": 0.7740653275119871,
+      "learning_rate": 1.361821799806765e-05,
+      "loss": 0.4209,
+      "step": 293
+    },
+    {
+      "epoch": 0.8898978433598184,
+      "grad_norm": 0.7704664366953056,
+      "learning_rate": 1.356886221591872e-05,
+      "loss": 0.4254,
+      "step": 294
+    },
+    {
+      "epoch": 0.8929247067726069,
+      "grad_norm": 0.7725867598321083,
+      "learning_rate": 1.3519406605021797e-05,
+      "loss": 0.4197,
+      "step": 295
+    },
+    {
+      "epoch": 0.8959515701853954,
+      "grad_norm": 0.6921472011564272,
+      "learning_rate": 1.3469852548756626e-05,
+      "loss": 0.4083,
+      "step": 296
+    },
+    {
+      "epoch": 0.8989784335981839,
+      "grad_norm": 0.8067935628780358,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.4418,
+      "step": 297
+    },
+    {
+      "epoch": 0.9020052970109724,
+      "grad_norm": 0.7698609485942266,
+      "learning_rate": 1.3370454647370418e-05,
+      "loss": 0.4356,
+      "step": 298
+    },
+    {
+      "epoch": 0.9050321604237609,
+      "grad_norm": 0.7157149981965328,
+      "learning_rate": 1.3320613582622354e-05,
+      "loss": 0.4391,
+      "step": 299
+    },
+    {
+      "epoch": 0.9080590238365494,
+      "grad_norm": 0.7256418661667707,
+      "learning_rate": 1.3270679633174219e-05,
+      "loss": 0.4516,
+      "step": 300
+    },
+    {
+      "epoch": 0.9110858872493379,
+      "grad_norm": 0.8779309356974607,
+      "learning_rate": 1.3220654195785917e-05,
+      "loss": 0.4674,
+      "step": 301
+    },
+    {
+      "epoch": 0.9141127506621264,
+      "grad_norm": 0.7238912013338223,
+      "learning_rate": 1.3170538669776469e-05,
+      "loss": 0.4345,
+      "step": 302
+    },
+    {
+      "epoch": 0.9171396140749148,
+      "grad_norm": 0.7123343816766206,
+      "learning_rate": 1.3120334456984871e-05,
+      "loss": 0.4283,
+      "step": 303
+    },
+    {
+      "epoch": 0.9201664774877034,
+      "grad_norm": 0.7646726163595352,
+      "learning_rate": 1.3070042961730878e-05,
+      "loss": 0.4462,
+      "step": 304
+    },
+    {
+      "epoch": 0.9231933409004919,
+      "grad_norm": 0.6978429470697927,
+      "learning_rate": 1.3019665590775717e-05,
+      "loss": 0.3872,
+      "step": 305
+    },
+    {
+      "epoch": 0.9262202043132803,
+      "grad_norm": 0.8215814125195987,
+      "learning_rate": 1.296920375328275e-05,
+      "loss": 0.4478,
+      "step": 306
+    },
+    {
+      "epoch": 0.9292470677260689,
+      "grad_norm": 0.7827918744205924,
+      "learning_rate": 1.2918658860778046e-05,
+      "loss": 0.4244,
+      "step": 307
+    },
+    {
+      "epoch": 0.9322739311388574,
+      "grad_norm": 0.6776247304651067,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 0.4192,
+      "step": 308
+    },
+    {
+      "epoch": 0.9353007945516458,
+      "grad_norm": 0.7967104116779378,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.452,
+      "step": 309
+    },
+    {
+      "epoch": 0.9383276579644344,
+      "grad_norm": 0.8053807766212533,
+      "learning_rate": 1.2766540003065272e-05,
+      "loss": 0.4072,
+      "step": 310
+    },
+    {
+      "epoch": 0.9413545213772229,
+      "grad_norm": 0.6994544274711724,
+      "learning_rate": 1.2715677051645259e-05,
+      "loss": 0.4286,
+      "step": 311
+    },
+    {
+      "epoch": 0.9443813847900113,
+      "grad_norm": 0.7044953227460141,
+      "learning_rate": 1.266473813690035e-05,
+      "loss": 0.4416,
+      "step": 312
+    },
+    {
+      "epoch": 0.9474082482027999,
+      "grad_norm": 0.7527938302384102,
+      "learning_rate": 1.2613724683701491e-05,
+      "loss": 0.4421,
+      "step": 313
+    },
+    {
+      "epoch": 0.9504351116155884,
+      "grad_norm": 0.7109760891717083,
+      "learning_rate": 1.2562638119004627e-05,
+      "loss": 0.4349,
+      "step": 314
+    },
+    {
+      "epoch": 0.9534619750283768,
+      "grad_norm": 0.7056197486602497,
+      "learning_rate": 1.2511479871810792e-05,
+      "loss": 0.4326,
+      "step": 315
+    },
+    {
+      "epoch": 0.9564888384411654,
+      "grad_norm": 0.7191711844493504,
+      "learning_rate": 1.2460251373126136e-05,
+      "loss": 0.3988,
+      "step": 316
+    },
+    {
+      "epoch": 0.9595157018539539,
+      "grad_norm": 0.6688807025324857,
+      "learning_rate": 1.2408954055921884e-05,
+      "loss": 0.4172,
+      "step": 317
+    },
+    {
+      "epoch": 0.9625425652667423,
+      "grad_norm": 0.6470039185633302,
+      "learning_rate": 1.2357589355094275e-05,
+      "loss": 0.4096,
+      "step": 318
+    },
+    {
+      "epoch": 0.9655694286795309,
+      "grad_norm": 0.6561772531350585,
+      "learning_rate": 1.2306158707424402e-05,
+      "loss": 0.4185,
+      "step": 319
+    },
+    {
+      "epoch": 0.9685962920923193,
+      "grad_norm": 0.7233636477196307,
+      "learning_rate": 1.2254663551538047e-05,
+      "loss": 0.4068,
+      "step": 320
+    },
+    {
+      "epoch": 0.9716231555051078,
+      "grad_norm": 0.7403443632198313,
+      "learning_rate": 1.2203105327865407e-05,
+      "loss": 0.4287,
+      "step": 321
+    },
+    {
+      "epoch": 0.9746500189178964,
+      "grad_norm": 0.7803435460606253,
+      "learning_rate": 1.215148547860084e-05,
+      "loss": 0.428,
+      "step": 322
+    },
+    {
+      "epoch": 0.9776768823306848,
+      "grad_norm": 0.6792455690045772,
+      "learning_rate": 1.2099805447662485e-05,
+      "loss": 0.4039,
+      "step": 323
+    },
+    {
+      "epoch": 0.9807037457434733,
+      "grad_norm": 0.7055233352079842,
+      "learning_rate": 1.2048066680651908e-05,
+      "loss": 0.4387,
+      "step": 324
+    },
+    {
+      "epoch": 0.9837306091562619,
+      "grad_norm": 0.6606753657946727,
+      "learning_rate": 1.1996270624813642e-05,
+      "loss": 0.4205,
+      "step": 325
+    },
+    {
+      "epoch": 0.9867574725690503,
+      "grad_norm": 0.7813281916840149,
+      "learning_rate": 1.194441872899471e-05,
+      "loss": 0.4199,
+      "step": 326
+    },
+    {
+      "epoch": 0.9897843359818388,
+      "grad_norm": 0.6669814472420743,
+      "learning_rate": 1.1892512443604103e-05,
+      "loss": 0.4175,
+      "step": 327
+    },
+    {
+      "epoch": 0.9928111993946274,
+      "grad_norm": 0.7047789400127767,
+      "learning_rate": 1.1840553220572204e-05,
+      "loss": 0.4359,
+      "step": 328
+    },
+    {
+      "epoch": 0.9958380628074158,
+      "grad_norm": 0.9777661179524167,
+      "learning_rate": 1.1788542513310178e-05,
+      "loss": 0.4267,
+      "step": 329
+    },
+    {
+      "epoch": 0.9988649262202043,
+      "grad_norm": 0.6828108263463339,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.4445,
+      "step": 330
+    },
+    {
+      "epoch": 1.0018917896329929,
+      "grad_norm": 0.7806720093737719,
+      "learning_rate": 1.1684372466900306e-05,
+      "loss": 0.3617,
+      "step": 331
+    },
+    {
+      "epoch": 1.0049186530457812,
+      "grad_norm": 1.0759906196950666,
+      "learning_rate": 1.1632216041612595e-05,
+      "loss": 0.3141,
+      "step": 332
+    },
+    {
+      "epoch": 1.0079455164585698,
+      "grad_norm": 0.73376718792793,
+      "learning_rate": 1.15800139597335e-05,
+      "loss": 0.3328,
+      "step": 333
+    },
+    {
+      "epoch": 1.0109723798713584,
+      "grad_norm": 0.6791481985601627,
+      "learning_rate": 1.1527767681467472e-05,
+      "loss": 0.3298,
+      "step": 334
+    },
+    {
+      "epoch": 1.0139992432841467,
+      "grad_norm": 0.8683398096679862,
+      "learning_rate": 1.1475478668255223e-05,
+      "loss": 0.3273,
+      "step": 335
+    },
+    {
+      "epoch": 1.0170261066969353,
+      "grad_norm": 0.7658960614441493,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3104,
+      "step": 336
+    },
+    {
+      "epoch": 1.0200529701097238,
+      "grad_norm": 0.8172544759185724,
+      "learning_rate": 1.1370778288690947e-05,
+      "loss": 0.3181,
+      "step": 337
+    },
+    {
+      "epoch": 1.0230798335225122,
+      "grad_norm": 0.8294974569303851,
+      "learning_rate": 1.1318369851033604e-05,
+      "loss": 0.3131,
+      "step": 338
+    },
+    {
+      "epoch": 1.0261066969353008,
+      "grad_norm": 0.765202946957603,
+      "learning_rate": 1.1265924535737494e-05,
+      "loss": 0.3091,
+      "step": 339
+    },
+    {
+      "epoch": 1.0291335603480893,
+      "grad_norm": 0.7041478915966126,
+      "learning_rate": 1.121344380981082e-05,
+      "loss": 0.3167,
+      "step": 340
+    },
+    {
+      "epoch": 1.0321604237608777,
+      "grad_norm": 0.768354029357346,
+      "learning_rate": 1.1160929141252303e-05,
+      "loss": 0.3111,
+      "step": 341
+    },
+    {
+      "epoch": 1.0351872871736663,
+      "grad_norm": 0.8300836169159135,
+      "learning_rate": 1.1108381999010111e-05,
+      "loss": 0.3037,
+      "step": 342
+    },
+    {
+      "epoch": 1.0382141505864548,
+      "grad_norm": 0.7525580555877907,
+      "learning_rate": 1.1055803852940772e-05,
+      "loss": 0.3145,
+      "step": 343
+    },
+    {
+      "epoch": 1.0412410139992432,
+      "grad_norm": 0.7757851376888233,
+      "learning_rate": 1.1003196173768051e-05,
+      "loss": 0.3044,
+      "step": 344
+    },
+    {
+      "epoch": 1.0442678774120318,
+      "grad_norm": 0.8237855581212018,
+      "learning_rate": 1.0950560433041825e-05,
+      "loss": 0.3212,
+      "step": 345
+    },
+    {
+      "epoch": 1.0472947408248203,
+      "grad_norm": 0.7688327557728135,
+      "learning_rate": 1.0897898103096917e-05,
+      "loss": 0.3275,
+      "step": 346
+    },
+    {
+      "epoch": 1.0503216042376087,
+      "grad_norm": 0.7871148926174769,
+      "learning_rate": 1.0845210657011893e-05,
+      "loss": 0.3166,
+      "step": 347
+    },
+    {
+      "epoch": 1.0533484676503972,
+      "grad_norm": 0.8657333501041573,
+      "learning_rate": 1.0792499568567885e-05,
+      "loss": 0.3232,
+      "step": 348
+    },
+    {
+      "epoch": 1.0563753310631858,
+      "grad_norm": 0.7934936334258926,
+      "learning_rate": 1.0739766312207344e-05,
+      "loss": 0.2997,
+      "step": 349
+    },
+    {
+      "epoch": 1.0594021944759742,
+      "grad_norm": 0.7999960310314015,
+      "learning_rate": 1.068701236299281e-05,
+      "loss": 0.3169,
+      "step": 350
+    },
+    {
+      "epoch": 1.0624290578887627,
+      "grad_norm": 0.7289611954616597,
+      "learning_rate": 1.0634239196565646e-05,
+      "loss": 0.3091,
+      "step": 351
+    },
+    {
+      "epoch": 1.0654559213015513,
+      "grad_norm": 0.794627491960225,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 0.3217,
+      "step": 352
+    },
+    {
+      "epoch": 1.0684827847143397,
+      "grad_norm": 0.8020921857581723,
+      "learning_rate": 1.0528641117285315e-05,
+      "loss": 0.3156,
+      "step": 353
+    },
+    {
+      "epoch": 1.0715096481271282,
+      "grad_norm": 0.6952720408036114,
+      "learning_rate": 1.0475819158237426e-05,
+      "loss": 0.2863,
+      "step": 354
+    },
+    {
+      "epoch": 1.0745365115399168,
+      "grad_norm": 0.7705514503735514,
+      "learning_rate": 1.0422983889504831e-05,
+      "loss": 0.3304,
+      "step": 355
+    },
+    {
+      "epoch": 1.0775633749527052,
+      "grad_norm": 0.7183429621577557,
+      "learning_rate": 1.0370136789003582e-05,
+      "loss": 0.3054,
+      "step": 356
+    },
+    {
+      "epoch": 1.0805902383654937,
+      "grad_norm": 0.8339816376861587,
+      "learning_rate": 1.031727933498068e-05,
+      "loss": 0.2927,
+      "step": 357
+    },
+    {
+      "epoch": 1.0836171017782823,
+      "grad_norm": 0.7134626118237305,
+      "learning_rate": 1.0264413005972736e-05,
+      "loss": 0.3083,
+      "step": 358
+    },
+    {
+      "epoch": 1.0866439651910706,
+      "grad_norm": 0.7739194914553884,
+      "learning_rate": 1.0211539280764617e-05,
+      "loss": 0.3147,
+      "step": 359
+    },
+    {
+      "epoch": 1.0896708286038592,
+      "grad_norm": 0.8059578220481518,
+      "learning_rate": 1.015865963834808e-05,
+      "loss": 0.3058,
+      "step": 360
+    },
+    {
+      "epoch": 1.0926976920166478,
+      "grad_norm": 0.6966807733974929,
+      "learning_rate": 1.0105775557880398e-05,
+      "loss": 0.3114,
+      "step": 361
+    },
+    {
+      "epoch": 1.0957245554294364,
+      "grad_norm": 0.7259517784999577,
+      "learning_rate": 1.0052888518642978e-05,
+      "loss": 0.309,
+      "step": 362
+    },
+    {
+      "epoch": 1.0987514188422247,
+      "grad_norm": 0.70152001024236,
+      "learning_rate": 1e-05,
+      "loss": 0.3114,
+      "step": 363
+    },
+    {
+      "epoch": 1.1017782822550133,
+      "grad_norm": 0.7239071251160694,
+      "learning_rate": 9.947111481357023e-06,
+      "loss": 0.306,
+      "step": 364
+    },
+    {
+      "epoch": 1.1048051456678016,
+      "grad_norm": 0.7270720689325801,
+      "learning_rate": 9.894224442119606e-06,
+      "loss": 0.3034,
+      "step": 365
+    },
+    {
+      "epoch": 1.1078320090805902,
+      "grad_norm": 0.7075182994637685,
+      "learning_rate": 9.841340361651921e-06,
+      "loss": 0.304,
+      "step": 366
+    },
+    {
+      "epoch": 1.1108588724933788,
+      "grad_norm": 0.7667766876116868,
+      "learning_rate": 9.788460719235386e-06,
+      "loss": 0.3083,
+      "step": 367
+    },
+    {
+      "epoch": 1.1138857359061674,
+      "grad_norm": 0.7291972316247977,
+      "learning_rate": 9.735586994027267e-06,
+      "loss": 0.3021,
+      "step": 368
+    },
+    {
+      "epoch": 1.1169125993189557,
+      "grad_norm": 0.8051578846658292,
+      "learning_rate": 9.682720665019325e-06,
+      "loss": 0.3189,
+      "step": 369
+    },
+    {
+      "epoch": 1.1199394627317443,
+      "grad_norm": 0.7773044179566808,
+      "learning_rate": 9.62986321099642e-06,
+      "loss": 0.3012,
+      "step": 370
+    },
+    {
+      "epoch": 1.1229663261445326,
+      "grad_norm": 0.7813770410769499,
+      "learning_rate": 9.57701611049517e-06,
+      "loss": 0.3109,
+      "step": 371
+    },
+    {
+      "epoch": 1.1259931895573212,
+      "grad_norm": 0.7350996663815793,
+      "learning_rate": 9.524180841762577e-06,
+      "loss": 0.3129,
+      "step": 372
+    },
+    {
+      "epoch": 1.1290200529701098,
+      "grad_norm": 0.7415395614630593,
+      "learning_rate": 9.471358882714687e-06,
+      "loss": 0.3063,
+      "step": 373
+    },
+    {
+      "epoch": 1.1320469163828983,
+      "grad_norm": 0.7481524697041051,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 0.3166,
+      "step": 374
+    },
+    {
+      "epoch": 1.1350737797956867,
+      "grad_norm": 0.7875631631529152,
+      "learning_rate": 9.365760803434356e-06,
+      "loss": 0.3074,
+      "step": 375
+    },
+    {
+      "epoch": 1.1381006432084753,
+      "grad_norm": 0.7244454348708046,
+      "learning_rate": 9.312987637007191e-06,
+      "loss": 0.3147,
+      "step": 376
+    },
+    {
+      "epoch": 1.1411275066212636,
+      "grad_norm": 0.6806836196922613,
+      "learning_rate": 9.260233687792657e-06,
+      "loss": 0.306,
+      "step": 377
+    },
+    {
+      "epoch": 1.1441543700340522,
+      "grad_norm": 0.7244837806326596,
+      "learning_rate": 9.207500431432115e-06,
+      "loss": 0.3081,
+      "step": 378
+    },
+    {
+      "epoch": 1.1471812334468408,
+      "grad_norm": 0.7226113533607956,
+      "learning_rate": 9.154789342988108e-06,
+      "loss": 0.304,
+      "step": 379
+    },
+    {
+      "epoch": 1.1502080968596293,
+      "grad_norm": 0.8163526870441586,
+      "learning_rate": 9.102101896903084e-06,
+      "loss": 0.3366,
+      "step": 380
+    },
+    {
+      "epoch": 1.1532349602724177,
+      "grad_norm": 0.728443961988738,
+      "learning_rate": 9.049439566958176e-06,
+      "loss": 0.3162,
+      "step": 381
+    },
+    {
+      "epoch": 1.1562618236852062,
+      "grad_norm": 0.7334490478700961,
+      "learning_rate": 8.99680382623195e-06,
+      "loss": 0.3027,
+      "step": 382
+    },
+    {
+      "epoch": 1.1592886870979946,
+      "grad_norm": 0.7109888918743166,
+      "learning_rate": 8.944196147059233e-06,
+      "loss": 0.295,
+      "step": 383
+    },
+    {
+      "epoch": 1.1623155505107832,
+      "grad_norm": 0.7350918156249301,
+      "learning_rate": 8.89161800098989e-06,
+      "loss": 0.3216,
+      "step": 384
+    },
+    {
+      "epoch": 1.1653424139235717,
+      "grad_norm": 0.752134552079796,
+      "learning_rate": 8.839070858747697e-06,
+      "loss": 0.3165,
+      "step": 385
+    },
+    {
+      "epoch": 1.1683692773363603,
+      "grad_norm": 0.6864901484027058,
+      "learning_rate": 8.786556190189183e-06,
+      "loss": 0.2911,
+      "step": 386
+    },
+    {
+      "epoch": 1.1713961407491487,
+      "grad_norm": 0.7153415334943132,
+      "learning_rate": 8.734075464262507e-06,
+      "loss": 0.3113,
+      "step": 387
+    },
+    {
+      "epoch": 1.1744230041619372,
+      "grad_norm": 0.7326392093738365,
+      "learning_rate": 8.681630148966397e-06,
+      "loss": 0.3106,
+      "step": 388
+    },
+    {
+      "epoch": 1.1774498675747256,
+      "grad_norm": 0.6807092675215786,
+      "learning_rate": 8.629221711309056e-06,
+      "loss": 0.3116,
+      "step": 389
+    },
+    {
+      "epoch": 1.1804767309875142,
+      "grad_norm": 0.7279305685194482,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.2985,
+      "step": 390
+    },
+    {
+      "epoch": 1.1835035944003027,
+      "grad_norm": 0.6858441147300602,
+      "learning_rate": 8.52452133174478e-06,
+      "loss": 0.3011,
+      "step": 391
+    },
+    {
+      "epoch": 1.1865304578130913,
+      "grad_norm": 0.6905702343943123,
+      "learning_rate": 8.472232318532531e-06,
+      "loss": 0.2953,
+      "step": 392
+    },
+    {
+      "epoch": 1.1895573212258796,
+      "grad_norm": 0.7303909827222848,
+      "learning_rate": 8.419986040266502e-06,
+      "loss": 0.3072,
+      "step": 393
+    },
+    {
+      "epoch": 1.1925841846386682,
+      "grad_norm": 0.7129815044440446,
+      "learning_rate": 8.367783958387407e-06,
+      "loss": 0.2998,
+      "step": 394
+    },
+    {
+      "epoch": 1.1956110480514566,
+      "grad_norm": 0.7246704222275712,
+      "learning_rate": 8.315627533099697e-06,
+      "loss": 0.2992,
+      "step": 395
+    },
+    {
+      "epoch": 1.1986379114642451,
+      "grad_norm": 0.7363441494790591,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.3159,
+      "step": 396
+    },
+    {
+      "epoch": 1.2016647748770337,
+      "grad_norm": 0.7126685363543432,
+      "learning_rate": 8.211457486689829e-06,
+      "loss": 0.3033,
+      "step": 397
+    },
+    {
+      "epoch": 1.2046916382898223,
+      "grad_norm": 0.7717241454630607,
+      "learning_rate": 8.159446779427798e-06,
+      "loss": 0.3146,
+      "step": 398
+    },
+    {
+      "epoch": 1.2077185017026106,
+      "grad_norm": 0.8197333887387661,
+      "learning_rate": 8.107487556395902e-06,
+      "loss": 0.2899,
+      "step": 399
+    },
+    {
+      "epoch": 1.2107453651153992,
+      "grad_norm": 0.7348256512573579,
+      "learning_rate": 8.055581271005292e-06,
+      "loss": 0.3153,
+      "step": 400
+    },
+    {
+      "epoch": 1.2137722285281876,
+      "grad_norm": 0.794786568321028,
+      "learning_rate": 8.00372937518636e-06,
+      "loss": 0.3144,
+      "step": 401
+    },
+    {
+      "epoch": 1.2167990919409761,
+      "grad_norm": 0.6895299818724617,
+      "learning_rate": 7.951933319348095e-06,
+      "loss": 0.3031,
+      "step": 402
+    },
+    {
+      "epoch": 1.2198259553537647,
+      "grad_norm": 0.7263283159277609,
+      "learning_rate": 7.900194552337516e-06,
+      "loss": 0.3054,
+      "step": 403
+    },
+    {
+      "epoch": 1.2228528187665533,
+      "grad_norm": 0.7011561429110952,
+      "learning_rate": 7.848514521399167e-06,
+      "loss": 0.3089,
+      "step": 404
+    },
+    {
+      "epoch": 1.2258796821793416,
+      "grad_norm": 0.7142827938969311,
+      "learning_rate": 7.796894672134594e-06,
+      "loss": 0.3045,
+      "step": 405
+    },
+    {
+      "epoch": 1.2289065455921302,
+      "grad_norm": 0.6413910775489793,
+      "learning_rate": 7.745336448461958e-06,
+      "loss": 0.2905,
+      "step": 406
+    },
+    {
+      "epoch": 1.2319334090049185,
+      "grad_norm": 0.7248781267659022,
+      "learning_rate": 7.6938412925756e-06,
+      "loss": 0.3012,
+      "step": 407
+    },
+    {
+      "epoch": 1.2349602724177071,
+      "grad_norm": 0.7130970569758894,
+      "learning_rate": 7.642410644905726e-06,
+      "loss": 0.3068,
+      "step": 408
+    },
+    {
+      "epoch": 1.2379871358304957,
+      "grad_norm": 0.7068523166895104,
+      "learning_rate": 7.591045944078119e-06,
+      "loss": 0.2969,
+      "step": 409
+    },
+    {
+      "epoch": 1.2410139992432843,
+      "grad_norm": 0.7490440389632759,
+      "learning_rate": 7.539748626873866e-06,
+      "loss": 0.3154,
+      "step": 410
+    },
+    {
+      "epoch": 1.2440408626560726,
+      "grad_norm": 0.7629338274331703,
+      "learning_rate": 7.488520128189209e-06,
+      "loss": 0.3248,
+      "step": 411
+    },
+    {
+      "epoch": 1.2470677260688612,
+      "grad_norm": 0.7259415486385469,
+      "learning_rate": 7.4373618809953755e-06,
+      "loss": 0.3023,
+      "step": 412
+    },
+    {
+      "epoch": 1.2500945894816495,
+      "grad_norm": 0.7851448028937558,
+      "learning_rate": 7.386275316298513e-06,
+      "loss": 0.314,
+      "step": 413
+    },
+    {
+      "epoch": 1.253121452894438,
+      "grad_norm": 0.7103863390767462,
+      "learning_rate": 7.335261863099652e-06,
+      "loss": 0.2986,
+      "step": 414
+    },
+    {
+      "epoch": 1.2561483163072267,
+      "grad_norm": 0.7546164402331856,
+      "learning_rate": 7.2843229483547405e-06,
+      "loss": 0.315,
+      "step": 415
+    },
+    {
+      "epoch": 1.2591751797200152,
+      "grad_norm": 0.6816094066847943,
+      "learning_rate": 7.233459996934731e-06,
+      "loss": 0.3066,
+      "step": 416
+    },
+    {
+      "epoch": 1.2622020431328036,
+      "grad_norm": 0.7282618586230907,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.3069,
+      "step": 417
+    },
+    {
+      "epoch": 1.2652289065455922,
+      "grad_norm": 0.7454483673453621,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 0.2977,
+      "step": 418
+    },
+    {
+      "epoch": 1.2682557699583805,
+      "grad_norm": 0.791772260494691,
+      "learning_rate": 7.081341139221955e-06,
+      "loss": 0.3229,
+      "step": 419
+    },
+    {
+      "epoch": 1.271282633371169,
+      "grad_norm": 0.6961441591891994,
+      "learning_rate": 7.0307962467172555e-06,
+      "loss": 0.3034,
+      "step": 420
+    },
+    {
+      "epoch": 1.2743094967839577,
+      "grad_norm": 0.7539210186158104,
+      "learning_rate": 6.9803344092242855e-06,
+      "loss": 0.3113,
+      "step": 421
+    },
+    {
+      "epoch": 1.2773363601967462,
+      "grad_norm": 0.7140869464698587,
+      "learning_rate": 6.929957038269123e-06,
+      "loss": 0.3093,
+      "step": 422
+    },
+    {
+      "epoch": 1.2803632236095346,
+      "grad_norm": 0.7152554055811224,
+      "learning_rate": 6.87966554301513e-06,
+      "loss": 0.3038,
+      "step": 423
+    },
+    {
+      "epoch": 1.2833900870223232,
+      "grad_norm": 0.7090326055170555,
+      "learning_rate": 6.8294613302235325e-06,
+      "loss": 0.2897,
+      "step": 424
+    },
+    {
+      "epoch": 1.2864169504351115,
+      "grad_norm": 0.6920747684029929,
+      "learning_rate": 6.779345804214088e-06,
+      "loss": 0.3086,
+      "step": 425
+    },
+    {
+      "epoch": 1.2894438138479,
+      "grad_norm": 0.6910458786418171,
+      "learning_rate": 6.729320366825785e-06,
+      "loss": 0.2993,
+      "step": 426
+    },
+    {
+      "epoch": 1.2924706772606886,
+      "grad_norm": 0.7079605989793586,
+      "learning_rate": 6.679386417377649e-06,
+      "loss": 0.314,
+      "step": 427
+    },
+    {
+      "epoch": 1.2954975406734772,
+      "grad_norm": 0.6828401955243103,
+      "learning_rate": 6.629545352629583e-06,
+      "loss": 0.3051,
+      "step": 428
+    },
+    {
+      "epoch": 1.2985244040862656,
+      "grad_norm": 0.6858357132492148,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 0.2817,
+      "step": 429
+    },
+    {
+      "epoch": 1.3015512674990541,
+      "grad_norm": 0.7520651219869522,
+      "learning_rate": 6.530147451243377e-06,
+      "loss": 0.315,
+      "step": 430
+    },
+    {
+      "epoch": 1.3045781309118425,
+      "grad_norm": 0.787181077916253,
+      "learning_rate": 6.480593394978208e-06,
+      "loss": 0.3266,
+      "step": 431
+    },
+    {
+      "epoch": 1.307604994324631,
+      "grad_norm": 0.738439945164543,
+      "learning_rate": 6.431137784081283e-06,
+      "loss": 0.3036,
+      "step": 432
+    },
+    {
+      "epoch": 1.3106318577374196,
+      "grad_norm": 0.7267743762658395,
+      "learning_rate": 6.381782001932352e-06,
+      "loss": 0.2945,
+      "step": 433
+    },
+    {
+      "epoch": 1.3136587211502082,
+      "grad_norm": 0.6916869736761744,
+      "learning_rate": 6.33252742911874e-06,
+      "loss": 0.3096,
+      "step": 434
+    },
+    {
+      "epoch": 1.3166855845629966,
+      "grad_norm": 0.7369562751330266,
+      "learning_rate": 6.283375443396726e-06,
+      "loss": 0.3083,
+      "step": 435
+    },
+    {
+      "epoch": 1.3197124479757851,
+      "grad_norm": 0.772169666233919,
+      "learning_rate": 6.234327419653013e-06,
+      "loss": 0.3147,
+      "step": 436
+    },
+    {
+      "epoch": 1.3227393113885735,
+      "grad_norm": 0.6827348488273219,
+      "learning_rate": 6.185384729866264e-06,
+      "loss": 0.3041,
+      "step": 437
+    },
+    {
+      "epoch": 1.325766174801362,
+      "grad_norm": 0.7148168241171187,
+      "learning_rate": 6.136548743068713e-06,
+      "loss": 0.3059,
+      "step": 438
+    },
+    {
+      "epoch": 1.3287930382141506,
+      "grad_norm": 0.7023513265782648,
+      "learning_rate": 6.087820825307904e-06,
+      "loss": 0.293,
+      "step": 439
+    },
+    {
+      "epoch": 1.3318199016269392,
+      "grad_norm": 0.705539900140726,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 0.3002,
+      "step": 440
+    },
+    {
+      "epoch": 1.3348467650397275,
+      "grad_norm": 0.6976014155797954,
+      "learning_rate": 5.990694645933866e-06,
+      "loss": 0.2894,
+      "step": 441
+    },
+    {
+      "epoch": 1.3378736284525161,
+      "grad_norm": 0.6977973750673745,
+      "learning_rate": 5.9422991011486635e-06,
+      "loss": 0.2844,
+      "step": 442
+    },
+    {
+      "epoch": 1.3409004918653045,
+      "grad_norm": 0.6843186341500641,
+      "learning_rate": 5.894017058980249e-06,
+      "loss": 0.2909,
+      "step": 443
+    },
+    {
+      "epoch": 1.343927355278093,
+      "grad_norm": 0.7048872754004082,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.296,
+      "step": 444
+    },
+    {
+      "epoch": 1.3469542186908816,
+      "grad_norm": 0.7140854899947078,
+      "learning_rate": 5.797798881491138e-06,
+      "loss": 0.2981,
+      "step": 445
+    },
+    {
+      "epoch": 1.3499810821036702,
+      "grad_norm": 0.7365995160924693,
+      "learning_rate": 5.749865437599703e-06,
+      "loss": 0.3229,
+      "step": 446
+    },
+    {
+      "epoch": 1.3530079455164585,
+      "grad_norm": 0.6951202483867199,
+      "learning_rate": 5.702050879108284e-06,
+      "loss": 0.2888,
+      "step": 447
+    },
+    {
+      "epoch": 1.356034808929247,
+      "grad_norm": 0.6870270379078338,
+      "learning_rate": 5.654356543492883e-06,
+      "loss": 0.2953,
+      "step": 448
+    },
+    {
+      "epoch": 1.3590616723420355,
+      "grad_norm": 0.7059530258124969,
+      "learning_rate": 5.606783764866576e-06,
+      "loss": 0.2872,
+      "step": 449
+    },
+    {
+      "epoch": 1.362088535754824,
+      "grad_norm": 0.7349612184977304,
+      "learning_rate": 5.559333873942259e-06,
+      "loss": 0.287,
+      "step": 450
+    },
+    {
+      "epoch": 1.3651153991676126,
+      "grad_norm": 0.7339064167968615,
+      "learning_rate": 5.512008197995379e-06,
+      "loss": 0.3021,
+      "step": 451
+    },
+    {
+      "epoch": 1.3681422625804012,
+      "grad_norm": 0.7270910621695029,
+      "learning_rate": 5.464808060826825e-06,
+      "loss": 0.2947,
+      "step": 452
+    },
+    {
+      "epoch": 1.3711691259931895,
+      "grad_norm": 0.7408478744530012,
+      "learning_rate": 5.417734782725896e-06,
+      "loss": 0.3027,
+      "step": 453
+    },
+    {
+      "epoch": 1.374195989405978,
+      "grad_norm": 0.7417290855590758,
+      "learning_rate": 5.370789680433376e-06,
+      "loss": 0.3059,
+      "step": 454
+    },
+    {
+      "epoch": 1.3772228528187664,
+      "grad_norm": 0.7344927670851353,
+      "learning_rate": 5.323974067104687e-06,
+      "loss": 0.2838,
+      "step": 455
+    },
+    {
+      "epoch": 1.380249716231555,
+      "grad_norm": 0.703527577694739,
+      "learning_rate": 5.277289252273175e-06,
+      "loss": 0.2931,
+      "step": 456
+    },
+    {
+      "epoch": 1.3832765796443436,
+      "grad_norm": 0.718970883177768,
+      "learning_rate": 5.230736541813463e-06,
+      "loss": 0.3049,
+      "step": 457
+    },
+    {
+      "epoch": 1.3863034430571322,
+      "grad_norm": 0.7063767857541002,
+      "learning_rate": 5.184317237904939e-06,
+      "loss": 0.2982,
+      "step": 458
+    },
+    {
+      "epoch": 1.3893303064699205,
+      "grad_norm": 0.7379015905933879,
+      "learning_rate": 5.138032638995315e-06,
+      "loss": 0.3045,
+      "step": 459
+    },
+    {
+      "epoch": 1.392357169882709,
+      "grad_norm": 0.7343973803064894,
+      "learning_rate": 5.091884039764321e-06,
+      "loss": 0.2822,
+      "step": 460
+    },
+    {
+      "epoch": 1.3953840332954974,
+      "grad_norm": 0.727407837033042,
+      "learning_rate": 5.045872731087479e-06,
+      "loss": 0.3032,
+      "step": 461
+    },
+    {
+      "epoch": 1.398410896708286,
+      "grad_norm": 0.6948764390702157,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.303,
+      "step": 462
+    },
+    {
+      "epoch": 1.4014377601210746,
+      "grad_norm": 0.6974689047078668,
+      "learning_rate": 4.954267129660789e-06,
+      "loss": 0.2924,
+      "step": 463
+    },
+    {
+      "epoch": 1.4044646235338631,
+      "grad_norm": 0.7311130555148413,
+      "learning_rate": 4.908675399316534e-06,
+      "loss": 0.3121,
+      "step": 464
+    },
+    {
+      "epoch": 1.4074914869466515,
+      "grad_norm": 0.8268450669703035,
+      "learning_rate": 4.863226084265939e-06,
+      "loss": 0.3066,
+      "step": 465
+    },
+    {
+      "epoch": 1.41051835035944,
+      "grad_norm": 0.6972950595943486,
+      "learning_rate": 4.817920455824045e-06,
+      "loss": 0.2982,
+      "step": 466
+    },
+    {
+      "epoch": 1.4135452137722284,
+      "grad_norm": 0.7865362451362247,
+      "learning_rate": 4.772759781286679e-06,
+      "loss": 0.3203,
+      "step": 467
+    },
+    {
+      "epoch": 1.416572077185017,
+      "grad_norm": 0.7256738227002579,
+      "learning_rate": 4.727745323894976e-06,
+      "loss": 0.2871,
+      "step": 468
+    },
+    {
+      "epoch": 1.4195989405978056,
+      "grad_norm": 0.6891980442278347,
+      "learning_rate": 4.682878342800087e-06,
+      "loss": 0.2839,
+      "step": 469
+    },
+    {
+      "epoch": 1.4226258040105941,
+      "grad_norm": 0.7640904437955583,
+      "learning_rate": 4.638160093027908e-06,
+      "loss": 0.3051,
+      "step": 470
+    },
+    {
+      "epoch": 1.4256526674233825,
+      "grad_norm": 0.7503203938844507,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.3039,
+      "step": 471
+    },
+    {
+      "epoch": 1.428679530836171,
+      "grad_norm": 0.7656499859133818,
+      "learning_rate": 4.549174786718684e-06,
+      "loss": 0.2992,
+      "step": 472
+    },
+    {
+      "epoch": 1.4317063942489594,
+      "grad_norm": 0.7457892762990587,
+      "learning_rate": 4.504910219291941e-06,
+      "loss": 0.2992,
+      "step": 473
+    },
+    {
+      "epoch": 1.434733257661748,
+      "grad_norm": 0.7212585193362251,
+      "learning_rate": 4.460799361338898e-06,
+      "loss": 0.3032,
+      "step": 474
+    },
+    {
+      "epoch": 1.4377601210745365,
+      "grad_norm": 0.7162271467071009,
+      "learning_rate": 4.416843446735077e-06,
+      "loss": 0.3055,
+      "step": 475
+    },
+    {
+      "epoch": 1.4407869844873251,
+      "grad_norm": 0.715348310242356,
+      "learning_rate": 4.373043705021899e-06,
+      "loss": 0.2854,
+      "step": 476
+    },
+    {
+      "epoch": 1.4438138479001135,
+      "grad_norm": 0.7377369119658413,
+      "learning_rate": 4.3294013613722944e-06,
+      "loss": 0.291,
+      "step": 477
+    },
+    {
+      "epoch": 1.446840711312902,
+      "grad_norm": 0.7552225916372017,
+      "learning_rate": 4.2859176365564294e-06,
+      "loss": 0.3005,
+      "step": 478
+    },
+    {
+      "epoch": 1.4498675747256904,
+      "grad_norm": 0.6815984225008468,
+      "learning_rate": 4.2425937469075626e-06,
+      "loss": 0.2852,
+      "step": 479
+    },
+    {
+      "epoch": 1.452894438138479,
+      "grad_norm": 0.7594709726342207,
+      "learning_rate": 4.19943090428802e-06,
+      "loss": 0.3044,
+      "step": 480
+    },
+    {
+      "epoch": 1.4559213015512675,
+      "grad_norm": 0.7377317483549033,
+      "learning_rate": 4.1564303160552935e-06,
+      "loss": 0.3196,
+      "step": 481
+    },
+    {
+      "epoch": 1.458948164964056,
+      "grad_norm": 0.7316084471879414,
+      "learning_rate": 4.113593185028273e-06,
+      "loss": 0.2997,
+      "step": 482
+    },
+    {
+      "epoch": 1.4619750283768445,
+      "grad_norm": 0.6971728584769332,
+      "learning_rate": 4.070920709453597e-06,
+      "loss": 0.2972,
+      "step": 483
+    },
+    {
+      "epoch": 1.465001891789633,
+      "grad_norm": 0.7290188155613687,
+      "learning_rate": 4.028414082972141e-06,
+      "loss": 0.2922,
+      "step": 484
+    },
+    {
+      "epoch": 1.4680287552024214,
+      "grad_norm": 0.7902751771608626,
+      "learning_rate": 3.986074494585619e-06,
+      "loss": 0.3078,
+      "step": 485
+    },
+    {
+      "epoch": 1.47105561861521,
+      "grad_norm": 0.6762316383057356,
+      "learning_rate": 3.943903128623336e-06,
+      "loss": 0.2861,
+      "step": 486
+    },
+    {
+      "epoch": 1.4740824820279985,
+      "grad_norm": 0.7608264408842211,
+      "learning_rate": 3.9019011647090465e-06,
+      "loss": 0.3107,
+      "step": 487
+    },
+    {
+      "epoch": 1.477109345440787,
+      "grad_norm": 0.7669867910376652,
+      "learning_rate": 3.860069777727983e-06,
+      "loss": 0.3049,
+      "step": 488
+    },
+    {
+      "epoch": 1.4801362088535754,
+      "grad_norm": 0.6783574006762981,
+      "learning_rate": 3.818410137793947e-06,
+      "loss": 0.2967,
+      "step": 489
+    },
+    {
+      "epoch": 1.483163072266364,
+      "grad_norm": 0.6952496931356794,
+      "learning_rate": 3.7769234102166365e-06,
+      "loss": 0.2916,
+      "step": 490
+    },
+    {
+      "epoch": 1.4861899356791524,
+      "grad_norm": 0.6824315096918989,
+      "learning_rate": 3.735610755468988e-06,
+      "loss": 0.2929,
+      "step": 491
+    },
+    {
+      "epoch": 1.489216799091941,
+      "grad_norm": 0.7345315106094218,
+      "learning_rate": 3.6944733291547784e-06,
+      "loss": 0.311,
+      "step": 492
+    },
+    {
+      "epoch": 1.4922436625047295,
+      "grad_norm": 0.693607778293158,
+      "learning_rate": 3.653512281976238e-06,
+      "loss": 0.291,
+      "step": 493
+    },
+    {
+      "epoch": 1.495270525917518,
+      "grad_norm": 0.7308018624015185,
+      "learning_rate": 3.612728759701919e-06,
+      "loss": 0.3018,
+      "step": 494
+    },
+    {
+      "epoch": 1.4982973893303064,
+      "grad_norm": 0.7034929360426182,
+      "learning_rate": 3.5721239031346067e-06,
+      "loss": 0.297,
+      "step": 495
+    },
+    {
+      "epoch": 1.501324252743095,
+      "grad_norm": 0.7127132950737849,
+      "learning_rate": 3.5316988480794255e-06,
+      "loss": 0.2864,
+      "step": 496
+    },
+    {
+      "epoch": 1.5043511161558833,
+      "grad_norm": 0.7759506885922928,
+      "learning_rate": 3.4914547253120655e-06,
+      "loss": 0.3189,
+      "step": 497
+    },
+    {
+      "epoch": 1.507377979568672,
+      "grad_norm": 0.7132748155740757,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2993,
+      "step": 498
+    },
+    {
+      "epoch": 1.5104048429814605,
+      "grad_norm": 0.7110575633216043,
+      "learning_rate": 3.4115137744067516e-06,
+      "loss": 0.3014,
+      "step": 499
+    },
+    {
+      "epoch": 1.513431706394249,
+      "grad_norm": 0.754876398202877,
+      "learning_rate": 3.37181918238904e-06,
+      "loss": 0.3001,
+      "step": 500
+    },
+    {
+      "epoch": 1.5164585698070374,
+      "grad_norm": 0.719266706010558,
+      "learning_rate": 3.3323099948370853e-06,
+      "loss": 0.3135,
+      "step": 501
+    },
+    {
+      "epoch": 1.519485433219826,
+      "grad_norm": 0.7307656103097783,
+      "learning_rate": 3.292987316907792e-06,
+      "loss": 0.3015,
+      "step": 502
+    },
+    {
+      "epoch": 1.5225122966326143,
+      "grad_norm": 0.7140081855076122,
+      "learning_rate": 3.253852248540994e-06,
+      "loss": 0.2911,
+      "step": 503
+    },
+    {
+      "epoch": 1.525539160045403,
+      "grad_norm": 0.7049401517146187,
+      "learning_rate": 3.2149058844286796e-06,
+      "loss": 0.2973,
+      "step": 504
+    },
+    {
+      "epoch": 1.5285660234581915,
+      "grad_norm": 0.715126839619227,
+      "learning_rate": 3.1761493139843734e-06,
+      "loss": 0.2906,
+      "step": 505
+    },
+    {
+      "epoch": 1.53159288687098,
+      "grad_norm": 0.6994396506456242,
+      "learning_rate": 3.1375836213126653e-06,
+      "loss": 0.2737,
+      "step": 506
+    },
+    {
+      "epoch": 1.5346197502837684,
+      "grad_norm": 0.7078361088125315,
+      "learning_rate": 3.099209885178882e-06,
+      "loss": 0.307,
+      "step": 507
+    },
+    {
+      "epoch": 1.537646613696557,
+      "grad_norm": 0.7046668422771416,
+      "learning_rate": 3.0610291789789094e-06,
+      "loss": 0.3035,
+      "step": 508
+    },
+    {
+      "epoch": 1.5406734771093453,
+      "grad_norm": 0.7197751692874702,
+      "learning_rate": 3.023042570709185e-06,
+      "loss": 0.2998,
+      "step": 509
+    },
+    {
+      "epoch": 1.543700340522134,
+      "grad_norm": 0.7361920502511936,
+      "learning_rate": 2.9852511229367862e-06,
+      "loss": 0.2848,
+      "step": 510
+    },
+    {
+      "epoch": 1.5467272039349225,
+      "grad_norm": 0.7403071513659568,
+      "learning_rate": 2.9476558927697605e-06,
+      "loss": 0.3053,
+      "step": 511
+    },
+    {
+      "epoch": 1.549754067347711,
+      "grad_norm": 0.7292022507100604,
+      "learning_rate": 2.9102579318274994e-06,
+      "loss": 0.3025,
+      "step": 512
+    },
+    {
+      "epoch": 1.5527809307604994,
+      "grad_norm": 0.7211437332106418,
+      "learning_rate": 2.8730582862113743e-06,
+      "loss": 0.3002,
+      "step": 513
+    },
+    {
+      "epoch": 1.555807794173288,
+      "grad_norm": 0.7281769418859074,
+      "learning_rate": 2.8360579964754277e-06,
+      "loss": 0.2896,
+      "step": 514
+    },
+    {
+      "epoch": 1.5588346575860763,
+      "grad_norm": 0.7745925532544838,
+      "learning_rate": 2.7992580975973136e-06,
+      "loss": 0.315,
+      "step": 515
+    },
+    {
+      "epoch": 1.5618615209988649,
+      "grad_norm": 0.7650894273516721,
+      "learning_rate": 2.7626596189492983e-06,
+      "loss": 0.3007,
+      "step": 516
+    },
+    {
+      "epoch": 1.5648883844116535,
+      "grad_norm": 0.7565642482329465,
+      "learning_rate": 2.726263584269513e-06,
+      "loss": 0.2951,
+      "step": 517
+    },
+    {
+      "epoch": 1.567915247824442,
+      "grad_norm": 0.7263158043069446,
+      "learning_rate": 2.690071011633284e-06,
+      "loss": 0.2856,
+      "step": 518
+    },
+    {
+      "epoch": 1.5709421112372304,
+      "grad_norm": 0.7215010198325115,
+      "learning_rate": 2.6540829134246683e-06,
+      "loss": 0.2948,
+      "step": 519
+    },
+    {
+      "epoch": 1.573968974650019,
+      "grad_norm": 0.7293900204605486,
+      "learning_rate": 2.618300296308135e-06,
+      "loss": 0.2996,
+      "step": 520
+    },
+    {
+      "epoch": 1.5769958380628073,
+      "grad_norm": 0.7498511851116723,
+      "learning_rate": 2.582724161200405e-06,
+      "loss": 0.2963,
+      "step": 521
+    },
+    {
+      "epoch": 1.5800227014755959,
+      "grad_norm": 0.766235911861552,
+      "learning_rate": 2.5473555032424534e-06,
+      "loss": 0.3071,
+      "step": 522
+    },
+    {
+      "epoch": 1.5830495648883844,
+      "grad_norm": 0.7602513400266425,
+      "learning_rate": 2.5121953117716744e-06,
+      "loss": 0.3077,
+      "step": 523
+    },
+    {
+      "epoch": 1.586076428301173,
+      "grad_norm": 0.7313684645925771,
+      "learning_rate": 2.477244570294206e-06,
+      "loss": 0.3002,
+      "step": 524
+    },
+    {
+      "epoch": 1.5891032917139614,
+      "grad_norm": 0.7338457683356104,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.3083,
+      "step": 525
+    },
+    {
+      "epoch": 1.59213015512675,
+      "grad_norm": 0.7357718121542034,
+      "learning_rate": 2.4079753420225694e-06,
+      "loss": 0.2969,
+      "step": 526
+    },
+    {
+      "epoch": 1.5951570185395383,
+      "grad_norm": 0.6539416749565655,
+      "learning_rate": 2.3736587928376197e-06,
+      "loss": 0.2911,
+      "step": 527
+    },
+    {
+      "epoch": 1.5981838819523269,
+      "grad_norm": 0.6911013727733428,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.2817,
+      "step": 528
+    },
+    {
+      "epoch": 1.6012107453651154,
+      "grad_norm": 0.7257528957528827,
+      "learning_rate": 2.305666623880858e-06,
+      "loss": 0.2977,
+      "step": 529
+    },
+    {
+      "epoch": 1.604237608777904,
+      "grad_norm": 0.7277750683209819,
+      "learning_rate": 2.27199290599617e-06,
+      "loss": 0.2983,
+      "step": 530
+    },
+    {
+      "epoch": 1.6072644721906924,
+      "grad_norm": 0.7330171349131269,
+      "learning_rate": 2.2385353570824308e-06,
+      "loss": 0.3165,
+      "step": 531
+    },
+    {
+      "epoch": 1.610291335603481,
+      "grad_norm": 0.72839013329869,
+      "learning_rate": 2.2052949130192136e-06,
+      "loss": 0.3149,
+      "step": 532
+    },
+    {
+      "epoch": 1.6133181990162693,
+      "grad_norm": 0.7003290761499322,
+      "learning_rate": 2.172272503613183e-06,
+      "loss": 0.2916,
+      "step": 533
+    },
+    {
+      "epoch": 1.6163450624290578,
+      "grad_norm": 0.7778235605512103,
+      "learning_rate": 2.1394690525721275e-06,
+      "loss": 0.3041,
+      "step": 534
+    },
+    {
+      "epoch": 1.6193719258418464,
+      "grad_norm": 0.7015034164559416,
+      "learning_rate": 2.1068854774790783e-06,
+      "loss": 0.2932,
+      "step": 535
+    },
+    {
+      "epoch": 1.622398789254635,
+      "grad_norm": 0.6774106167712252,
+      "learning_rate": 2.0745226897666858e-06,
+      "loss": 0.3107,
+      "step": 536
+    },
+    {
+      "epoch": 1.6254256526674233,
+      "grad_norm": 0.7526836803648397,
+      "learning_rate": 2.0423815946916783e-06,
+      "loss": 0.3036,
+      "step": 537
+    },
+    {
+      "epoch": 1.628452516080212,
+      "grad_norm": 0.6855905129399048,
+      "learning_rate": 2.010463091309587e-06,
+      "loss": 0.2853,
+      "step": 538
+    },
+    {
+      "epoch": 1.6314793794930003,
+      "grad_norm": 0.7268700549990801,
+      "learning_rate": 1.9787680724495617e-06,
+      "loss": 0.2808,
+      "step": 539
+    },
+    {
+      "epoch": 1.6345062429057888,
+      "grad_norm": 0.7067953862625688,
+      "learning_rate": 1.947297424689414e-06,
+      "loss": 0.2931,
+      "step": 540
+    },
+    {
+      "epoch": 1.6375331063185774,
+      "grad_norm": 0.703170282490726,
+      "learning_rate": 1.9160520283308115e-06,
+      "loss": 0.3088,
+      "step": 541
+    },
+    {
+      "epoch": 1.640559969731366,
+      "grad_norm": 0.7264798156337304,
+      "learning_rate": 1.8850327573746584e-06,
+      "loss": 0.3047,
+      "step": 542
+    },
+    {
+      "epoch": 1.6435868331441543,
+      "grad_norm": 0.7227549430160181,
+      "learning_rate": 1.854240479496643e-06,
+      "loss": 0.2898,
+      "step": 543
+    },
+    {
+      "epoch": 1.646613696556943,
+      "grad_norm": 0.7015295971224901,
+      "learning_rate": 1.8236760560229715e-06,
+      "loss": 0.2984,
+      "step": 544
+    },
+    {
+      "epoch": 1.6496405599697312,
+      "grad_norm": 0.6866761580725703,
+      "learning_rate": 1.7933403419062689e-06,
+      "loss": 0.3051,
+      "step": 545
+    },
+    {
+      "epoch": 1.6526674233825198,
+      "grad_norm": 0.7251984488706571,
+      "learning_rate": 1.7632341857016733e-06,
+      "loss": 0.3024,
+      "step": 546
+    },
+    {
+      "epoch": 1.6556942867953084,
+      "grad_norm": 0.7376618283367711,
+      "learning_rate": 1.7333584295430894e-06,
+      "loss": 0.3053,
+      "step": 547
+    },
+    {
+      "epoch": 1.658721150208097,
+      "grad_norm": 0.748770474084299,
+      "learning_rate": 1.7037139091196396e-06,
+      "loss": 0.2996,
+      "step": 548
+    },
+    {
+      "epoch": 1.6617480136208853,
+      "grad_norm": 0.7192409165846234,
+      "learning_rate": 1.6743014536522872e-06,
+      "loss": 0.3071,
+      "step": 549
+    },
+    {
+      "epoch": 1.6647748770336739,
+      "grad_norm": 0.7345508241609225,
+      "learning_rate": 1.6451218858706374e-06,
+      "loss": 0.2926,
+      "step": 550
+    },
+    {
+      "epoch": 1.6678017404464622,
+      "grad_norm": 0.7065333425120881,
+      "learning_rate": 1.616176021989926e-06,
+      "loss": 0.2787,
+      "step": 551
+    },
+    {
+      "epoch": 1.6708286038592508,
+      "grad_norm": 0.7192763973802473,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2799,
+      "step": 552
+    },
+    {
+      "epoch": 1.6738554672720394,
+      "grad_norm": 0.7141377198088835,
+      "learning_rate": 1.558988638083616e-06,
+      "loss": 0.299,
+      "step": 553
+    },
+    {
+      "epoch": 1.676882330684828,
+      "grad_norm": 0.7926571608828284,
+      "learning_rate": 1.5307487177120773e-06,
+      "loss": 0.2957,
+      "step": 554
+    },
+    {
+      "epoch": 1.6799091940976163,
+      "grad_norm": 0.9140707338446863,
+      "learning_rate": 1.5027457005048573e-06,
+      "loss": 0.3096,
+      "step": 555
+    },
+    {
+      "epoch": 1.6829360575104049,
+      "grad_norm": 0.7323605405418113,
+      "learning_rate": 1.4749803697665366e-06,
+      "loss": 0.2959,
+      "step": 556
+    },
+    {
+      "epoch": 1.6859629209231932,
+      "grad_norm": 0.7034315731089018,
+      "learning_rate": 1.4474535021531099e-06,
+      "loss": 0.2774,
+      "step": 557
+    },
+    {
+      "epoch": 1.6889897843359818,
+      "grad_norm": 0.7090907931747564,
+      "learning_rate": 1.4201658676502294e-06,
+      "loss": 0.2856,
+      "step": 558
+    },
+    {
+      "epoch": 1.6920166477487704,
+      "grad_norm": 0.8107688135477413,
+      "learning_rate": 1.3931182295516965e-06,
+      "loss": 0.2944,
+      "step": 559
+    },
+    {
+      "epoch": 1.695043511161559,
+      "grad_norm": 0.7299819831699966,
+      "learning_rate": 1.3663113444380905e-06,
+      "loss": 0.3055,
+      "step": 560
+    },
+    {
+      "epoch": 1.6980703745743473,
+      "grad_norm": 0.7904948671328375,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.2907,
+      "step": 561
+    },
+    {
+      "epoch": 1.7010972379871359,
+      "grad_norm": 0.7634403547908745,
+      "learning_rate": 1.3134228257951142e-06,
+      "loss": 0.2807,
+      "step": 562
+    },
+    {
+      "epoch": 1.7041241013999242,
+      "grad_norm": 0.755616668842537,
+      "learning_rate": 1.2873426716713012e-06,
+      "loss": 0.29,
+      "step": 563
+    },
+    {
+      "epoch": 1.7071509648127128,
+      "grad_norm": 0.7096807194845756,
+      "learning_rate": 1.2615062293021508e-06,
+      "loss": 0.2999,
+      "step": 564
+    },
+    {
+      "epoch": 1.7101778282255014,
+      "grad_norm": 0.6725688197294233,
+      "learning_rate": 1.2359142213884933e-06,
+      "loss": 0.2894,
+      "step": 565
+    },
+    {
+      "epoch": 1.71320469163829,
+      "grad_norm": 0.7999411093376383,
+      "learning_rate": 1.2105673637938054e-06,
+      "loss": 0.2921,
+      "step": 566
+    },
+    {
+      "epoch": 1.7162315550510783,
+      "grad_norm": 0.8025753081590201,
+      "learning_rate": 1.1854663655241804e-06,
+      "loss": 0.2927,
+      "step": 567
+    },
+    {
+      "epoch": 1.7192584184638668,
+      "grad_norm": 0.7345654511665216,
+      "learning_rate": 1.1606119287084982e-06,
+      "loss": 0.3067,
+      "step": 568
+    },
+    {
+      "epoch": 1.7222852818766552,
+      "grad_norm": 0.7264021021624221,
+      "learning_rate": 1.136004748578785e-06,
+      "loss": 0.3034,
+      "step": 569
+    },
+    {
+      "epoch": 1.7253121452894438,
+      "grad_norm": 0.6769106712997419,
+      "learning_rate": 1.1116455134507665e-06,
+      "loss": 0.2902,
+      "step": 570
+    },
+    {
+      "epoch": 1.7283390087022323,
+      "grad_norm": 0.7460040881168952,
+      "learning_rate": 1.0875349047046113e-06,
+      "loss": 0.2908,
+      "step": 571
+    },
+    {
+      "epoch": 1.731365872115021,
+      "grad_norm": 0.7622578965355392,
+      "learning_rate": 1.0636735967658785e-06,
+      "loss": 0.2999,
+      "step": 572
+    },
+    {
+      "epoch": 1.7343927355278093,
+      "grad_norm": 0.7091907311876559,
+      "learning_rate": 1.0400622570866426e-06,
+      "loss": 0.2928,
+      "step": 573
+    },
+    {
+      "epoch": 1.7374195989405978,
+      "grad_norm": 0.7233251298182816,
+      "learning_rate": 1.0167015461268303e-06,
+      "loss": 0.3,
+      "step": 574
+    },
+    {
+      "epoch": 1.7404464623533862,
+      "grad_norm": 0.7390840905700544,
+      "learning_rate": 9.935921173357444e-07,
+      "loss": 0.2844,
+      "step": 575
+    },
+    {
+      "epoch": 1.7434733257661748,
+      "grad_norm": 0.7157760630639298,
+      "learning_rate": 9.707346171337895e-07,
+      "loss": 0.293,
+      "step": 576
+    },
+    {
+      "epoch": 1.7465001891789633,
+      "grad_norm": 0.7104519753290416,
+      "learning_rate": 9.481296848943744e-07,
+      "loss": 0.2834,
+      "step": 577
+    },
+    {
+      "epoch": 1.749527052591752,
+      "grad_norm": 0.7192815082713216,
+      "learning_rate": 9.257779529260558e-07,
+      "loss": 0.2945,
+      "step": 578
+    },
+    {
+      "epoch": 1.7525539160045402,
+      "grad_norm": 0.6943516725355893,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.2777,
+      "step": 579
+    },
+    {
+      "epoch": 1.7555807794173288,
+      "grad_norm": 0.7000854485395912,
+      "learning_rate": 8.818365836066101e-07,
+      "loss": 0.2837,
+      "step": 580
+    },
+    {
+      "epoch": 1.7586076428301172,
+      "grad_norm": 0.7020299005867426,
+      "learning_rate": 8.602481753900427e-07,
+      "loss": 0.2896,
+      "step": 581
+    },
+    {
+      "epoch": 1.7616345062429057,
+      "grad_norm": 0.6890802839214029,
+      "learning_rate": 8.389154256793042e-07,
+      "loss": 0.2843,
+      "step": 582
+    },
+    {
+      "epoch": 1.7646613696556943,
+      "grad_norm": 0.7313949301325045,
+      "learning_rate": 8.178389311972612e-07,
+      "loss": 0.3132,
+      "step": 583
+    },
+    {
+      "epoch": 1.7676882330684829,
+      "grad_norm": 0.6940746713971585,
+      "learning_rate": 7.970192814987676e-07,
+      "loss": 0.2877,
+      "step": 584
+    },
+    {
+      "epoch": 1.7707150964812712,
+      "grad_norm": 0.6625527790218431,
+      "learning_rate": 7.764570589541876e-07,
+      "loss": 0.282,
+      "step": 585
+    },
+    {
+      "epoch": 1.7737419598940598,
+      "grad_norm": 0.7113927498585739,
+      "learning_rate": 7.561528387330797e-07,
+      "loss": 0.2842,
+      "step": 586
+    },
+    {
+      "epoch": 1.7767688233068482,
+      "grad_norm": 0.7540098620036764,
+      "learning_rate": 7.361071887881376e-07,
+      "loss": 0.2871,
+      "step": 587
+    },
+    {
+      "epoch": 1.7797956867196367,
+      "grad_norm": 0.7060225608387275,
+      "learning_rate": 7.163206698392744e-07,
+      "loss": 0.2803,
+      "step": 588
+    },
+    {
+      "epoch": 1.7828225501324253,
+      "grad_norm": 0.7367595735687928,
+      "learning_rate": 6.96793835357964e-07,
+      "loss": 0.2962,
+      "step": 589
+    },
+    {
+      "epoch": 1.7858494135452139,
+      "grad_norm": 0.7300560324264508,
+      "learning_rate": 6.775272315517423e-07,
+      "loss": 0.2868,
+      "step": 590
+    },
+    {
+      "epoch": 1.7888762769580022,
+      "grad_norm": 0.769974841977717,
+      "learning_rate": 6.585213973489335e-07,
+      "loss": 0.3074,
+      "step": 591
+    },
+    {
+      "epoch": 1.7919031403707908,
+      "grad_norm": 0.7075338812192769,
+      "learning_rate": 6.397768643835755e-07,
+      "loss": 0.2847,
+      "step": 592
+    },
+    {
+      "epoch": 1.7949300037835791,
+      "grad_norm": 0.7860472120832016,
+      "learning_rate": 6.212941569805508e-07,
+      "loss": 0.3128,
+      "step": 593
+    },
+    {
+      "epoch": 1.7979568671963677,
+      "grad_norm": 0.7424808020712871,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.2939,
+      "step": 594
+    },
+    {
+      "epoch": 1.8009837306091563,
+      "grad_norm": 0.7147902042809061,
+      "learning_rate": 5.851162795274445e-07,
+      "loss": 0.2988,
+      "step": 595
+    },
+    {
+      "epoch": 1.8040105940219449,
+      "grad_norm": 0.7349984474714761,
+      "learning_rate": 5.674221214503639e-07,
+      "loss": 0.291,
+      "step": 596
+    },
+    {
+      "epoch": 1.8070374574347332,
+      "grad_norm": 0.712818638558487,
+      "learning_rate": 5.499918128533155e-07,
+      "loss": 0.2865,
+      "step": 597
+    },
+    {
+      "epoch": 1.8100643208475218,
+      "grad_norm": 0.7281994584534587,
+      "learning_rate": 5.328258412994958e-07,
+      "loss": 0.2933,
+      "step": 598
+    },
+    {
+      "epoch": 1.8130911842603101,
+      "grad_norm": 0.7171320565467246,
+      "learning_rate": 5.159246869580348e-07,
+      "loss": 0.3033,
+      "step": 599
+    },
+    {
+      "epoch": 1.8161180476730987,
+      "grad_norm": 0.7350552399068235,
+      "learning_rate": 4.992888225905467e-07,
+      "loss": 0.2821,
+      "step": 600
+    },
+    {
+      "epoch": 1.8191449110858873,
+      "grad_norm": 0.7290514589779898,
+      "learning_rate": 4.829187135379221e-07,
+      "loss": 0.2993,
+      "step": 601
+    },
+    {
+      "epoch": 1.8221717744986758,
+      "grad_norm": 0.6876702072513811,
+      "learning_rate": 4.6681481770729844e-07,
+      "loss": 0.2817,
+      "step": 602
+    },
+    {
+      "epoch": 1.8251986379114642,
+      "grad_norm": 0.7063796041305779,
+      "learning_rate": 4.509775855592613e-07,
+      "loss": 0.2935,
+      "step": 603
+    },
+    {
+      "epoch": 1.8282255013242528,
+      "grad_norm": 0.7140753206256399,
+      "learning_rate": 4.354074600952407e-07,
+      "loss": 0.2935,
+      "step": 604
+    },
+    {
+      "epoch": 1.8312523647370411,
+      "grad_norm": 0.705819114584338,
+      "learning_rate": 4.2010487684511105e-07,
+      "loss": 0.2875,
+      "step": 605
+    },
+    {
+      "epoch": 1.8342792281498297,
+      "grad_norm": 0.6922950151564867,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2789,
+      "step": 606
+    },
+    {
+      "epoch": 1.8373060915626183,
+      "grad_norm": 0.7108856218425824,
+      "learning_rate": 3.9030404167542777e-07,
+      "loss": 0.2954,
+      "step": 607
+    },
+    {
+      "epoch": 1.8403329549754068,
+      "grad_norm": 0.7131065344324313,
+      "learning_rate": 3.7580662334929517e-07,
+      "loss": 0.3045,
+      "step": 608
+    },
+    {
+      "epoch": 1.8433598183881952,
+      "grad_norm": 0.694490583764877,
+      "learning_rate": 3.615784144005796e-07,
+      "loss": 0.2953,
+      "step": 609
+    },
+    {
+      "epoch": 1.8463866818009838,
+      "grad_norm": 0.6983769385292246,
+      "learning_rate": 3.476198128228736e-07,
+      "loss": 0.2907,
+      "step": 610
+    },
+    {
+      "epoch": 1.849413545213772,
+      "grad_norm": 0.7255269411362608,
+      "learning_rate": 3.339312090682689e-07,
+      "loss": 0.296,
+      "step": 611
+    },
+    {
+      "epoch": 1.8524404086265607,
+      "grad_norm": 0.724686450762986,
+      "learning_rate": 3.2051298603643754e-07,
+      "loss": 0.284,
+      "step": 612
+    },
+    {
+      "epoch": 1.8554672720393492,
+      "grad_norm": 0.7337537218252735,
+      "learning_rate": 3.0736551906392354e-07,
+      "loss": 0.2913,
+      "step": 613
+    },
+    {
+      "epoch": 1.8584941354521378,
+      "grad_norm": 0.6756588196176819,
+      "learning_rate": 2.9448917591363923e-07,
+      "loss": 0.2797,
+      "step": 614
+    },
+    {
+      "epoch": 1.8615209988649262,
+      "grad_norm": 0.710607675130199,
+      "learning_rate": 2.818843167645835e-07,
+      "loss": 0.2986,
+      "step": 615
+    },
+    {
+      "epoch": 1.8645478622777147,
+      "grad_norm": 0.7420231707074685,
+      "learning_rate": 2.6955129420176193e-07,
+      "loss": 0.2979,
+      "step": 616
+    },
+    {
+      "epoch": 1.867574725690503,
+      "grad_norm": 0.7018411742603732,
+      "learning_rate": 2.5749045320632824e-07,
+      "loss": 0.2809,
+      "step": 617
+    },
+    {
+      "epoch": 1.8706015891032917,
+      "grad_norm": 0.6592805843635714,
+      "learning_rate": 2.4570213114592957e-07,
+      "loss": 0.2753,
+      "step": 618
+    },
+    {
+      "epoch": 1.8736284525160802,
+      "grad_norm": 0.7004749847628583,
+      "learning_rate": 2.3418665776527738e-07,
+      "loss": 0.294,
+      "step": 619
+    },
+    {
+      "epoch": 1.8766553159288688,
+      "grad_norm": 0.7192419469712777,
+      "learning_rate": 2.2294435517691504e-07,
+      "loss": 0.2808,
+      "step": 620
+    },
+    {
+      "epoch": 1.8796821793416572,
+      "grad_norm": 0.6853355870792595,
+      "learning_rate": 2.119755378522137e-07,
+      "loss": 0.2825,
+      "step": 621
+    },
+    {
+      "epoch": 1.8827090427544457,
+      "grad_norm": 0.6893891241524042,
+      "learning_rate": 2.0128051261257165e-07,
+      "loss": 0.2846,
+      "step": 622
+    },
+    {
+      "epoch": 1.885735906167234,
+      "grad_norm": 0.7153684181001242,
+      "learning_rate": 1.908595786208367e-07,
+      "loss": 0.2861,
+      "step": 623
+    },
+    {
+      "epoch": 1.8887627695800226,
+      "grad_norm": 0.698568295870308,
+      "learning_rate": 1.8071302737293294e-07,
+      "loss": 0.298,
+      "step": 624
+    },
+    {
+      "epoch": 1.8917896329928112,
+      "grad_norm": 0.7349928847810268,
+      "learning_rate": 1.7084114268971275e-07,
+      "loss": 0.297,
+      "step": 625
+    },
+    {
+      "epoch": 1.8948164964055998,
+      "grad_norm": 0.6690847929482501,
+      "learning_rate": 1.612442007090076e-07,
+      "loss": 0.2822,
+      "step": 626
+    },
+    {
+      "epoch": 1.8978433598183881,
+      "grad_norm": 0.7065372661864638,
+      "learning_rate": 1.519224698779198e-07,
+      "loss": 0.3005,
+      "step": 627
+    },
+    {
+      "epoch": 1.9008702232311767,
+      "grad_norm": 0.6877797646675535,
+      "learning_rate": 1.4287621094529524e-07,
+      "loss": 0.2894,
+      "step": 628
+    },
+    {
+      "epoch": 1.903897086643965,
+      "grad_norm": 0.7582375581056698,
+      "learning_rate": 1.3410567695444576e-07,
+      "loss": 0.3236,
+      "step": 629
+    },
+    {
+      "epoch": 1.9069239500567536,
+      "grad_norm": 0.6867580892115934,
+      "learning_rate": 1.2561111323605714e-07,
+      "loss": 0.2965,
+      "step": 630
+    },
+    {
+      "epoch": 1.9099508134695422,
+      "grad_norm": 0.7082460824460324,
+      "learning_rate": 1.1739275740134004e-07,
+      "loss": 0.3034,
+      "step": 631
+    },
+    {
+      "epoch": 1.9129776768823308,
+      "grad_norm": 0.7101455131229634,
+      "learning_rate": 1.0945083933537104e-07,
+      "loss": 0.2992,
+      "step": 632
+    },
+    {
+      "epoch": 1.9160045402951191,
+      "grad_norm": 0.6807635270867676,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2752,
+      "step": 633
+    },
+    {
+      "epoch": 1.9190314037079077,
+      "grad_norm": 0.7334421079682671,
+      "learning_rate": 9.439719738099318e-08,
+      "loss": 0.2899,
+      "step": 634
+    },
+    {
+      "epoch": 1.922058267120696,
+      "grad_norm": 0.7223888672456802,
+      "learning_rate": 8.728589457530857e-08,
+      "loss": 0.2885,
+      "step": 635
+    },
+    {
+      "epoch": 1.9250851305334846,
+      "grad_norm": 0.6806843503386794,
+      "learning_rate": 8.04518716920466e-08,
+      "loss": 0.2827,
+      "step": 636
+    },
+    {
+      "epoch": 1.9281119939462732,
+      "grad_norm": 0.6938827482888975,
+      "learning_rate": 7.389531989351773e-08,
+      "loss": 0.2745,
+      "step": 637
+    },
+    {
+      "epoch": 1.9311388573590618,
+      "grad_norm": 0.7068605149308287,
+      "learning_rate": 6.761642258056977e-08,
+      "loss": 0.2889,
+      "step": 638
+    },
+    {
+      "epoch": 1.9341657207718501,
+      "grad_norm": 0.6731692509261796,
+      "learning_rate": 6.161535538745877e-08,
+      "loss": 0.3021,
+      "step": 639
+    },
+    {
+      "epoch": 1.9371925841846387,
+      "grad_norm": 0.7260919692175745,
+      "learning_rate": 5.5892286176932875e-08,
+      "loss": 0.2834,
+      "step": 640
+    },
+    {
+      "epoch": 1.940219447597427,
+      "grad_norm": 0.738444646759649,
+      "learning_rate": 5.044737503554165e-08,
+      "loss": 0.3006,
+      "step": 641
+    },
+    {
+      "epoch": 1.9432463110102156,
+      "grad_norm": 0.7319099546064092,
+      "learning_rate": 4.528077426915412e-08,
+      "loss": 0.3005,
+      "step": 642
+    },
+    {
+      "epoch": 1.9462731744230042,
+      "grad_norm": 0.7214957439661821,
+      "learning_rate": 4.0392628398699954e-08,
+      "loss": 0.2947,
+      "step": 643
+    },
+    {
+      "epoch": 1.9493000378357928,
+      "grad_norm": 0.689313696487217,
+      "learning_rate": 3.578307415612714e-08,
+      "loss": 0.2846,
+      "step": 644
+    },
+    {
+      "epoch": 1.952326901248581,
+      "grad_norm": 0.7063603792863032,
+      "learning_rate": 3.1452240480577265e-08,
+      "loss": 0.2809,
+      "step": 645
+    },
+    {
+      "epoch": 1.9553537646613697,
+      "grad_norm": 0.7280860301661419,
+      "learning_rate": 2.7400248514776184e-08,
+      "loss": 0.3012,
+      "step": 646
+    },
+    {
+      "epoch": 1.958380628074158,
+      "grad_norm": 0.7429925114255341,
+      "learning_rate": 2.3627211601651157e-08,
+      "loss": 0.2928,
+      "step": 647
+    },
+    {
+      "epoch": 1.9614074914869466,
+      "grad_norm": 0.7228465789561147,
+      "learning_rate": 2.013323528115674e-08,
+      "loss": 0.3039,
+      "step": 648
+    },
+    {
+      "epoch": 1.9644343548997352,
+      "grad_norm": 0.7217595831575006,
+      "learning_rate": 1.6918417287318245e-08,
+      "loss": 0.2883,
+      "step": 649
+    },
+    {
+      "epoch": 1.9674612183125237,
+      "grad_norm": 0.7427981017802576,
+      "learning_rate": 1.3982847545507271e-08,
+      "loss": 0.2767,
+      "step": 650
+    },
+    {
+      "epoch": 1.970488081725312,
+      "grad_norm": 0.7389104295907093,
+      "learning_rate": 1.1326608169920373e-08,
+      "loss": 0.2958,
+      "step": 651
+    },
+    {
+      "epoch": 1.9735149451381007,
+      "grad_norm": 0.7184615710477206,
+      "learning_rate": 8.949773461282008e-09,
+      "loss": 0.2836,
+      "step": 652
+    },
+    {
+      "epoch": 1.976541808550889,
+      "grad_norm": 0.7031585221499792,
+      "learning_rate": 6.8524099047695415e-09,
+      "loss": 0.2977,
+      "step": 653
+    },
+    {
+      "epoch": 1.9795686719636776,
+      "grad_norm": 0.7185394320500036,
+      "learning_rate": 5.034576168149175e-09,
+      "loss": 0.2863,
+      "step": 654
+    },
+    {
+      "epoch": 1.9825955353764662,
+      "grad_norm": 0.680821347329333,
+      "learning_rate": 3.4963231001383657e-09,
+      "loss": 0.2847,
+      "step": 655
+    },
+    {
+      "epoch": 1.9856223987892547,
+      "grad_norm": 0.7003998717407318,
+      "learning_rate": 2.237693728981416e-09,
+      "loss": 0.286,
+      "step": 656
+    },
+    {
+      "epoch": 1.988649262202043,
+      "grad_norm": 0.7081021315379157,
+      "learning_rate": 1.2587232612493172e-09,
+      "loss": 0.2976,
+      "step": 657
+    },
+    {
+      "epoch": 1.9916761256148316,
+      "grad_norm": 0.7096586253975485,
+      "learning_rate": 5.594390808494332e-10,
+      "loss": 0.2999,
+      "step": 658
+    },
+    {
+      "epoch": 1.99470298902762,
+      "grad_norm": 0.7058911771954404,
+      "learning_rate": 1.3986074826388697e-10,
+      "loss": 0.2773,
+      "step": 659
+    },
+    {
+      "epoch": 1.9977298524404086,
+      "grad_norm": 0.7219760069293065,
+      "learning_rate": 0.0,
+      "loss": 0.3018,
+      "step": 660
+    },
+    {
+      "epoch": 1.9977298524404086,
+      "step": 660,
+      "total_flos": 71086025687040.0,
+      "train_loss": 0.3834661602973938,
+      "train_runtime": 2405.362,
+      "train_samples_per_second": 35.15,
+      "train_steps_per_second": 0.274
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 660,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 71086025687040.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}