diff --git "a/qwen_2_5_7b_math_instruct/limo/checkpoint-1230/trainer_state.json" "b/qwen_2_5_7b_math_instruct/limo/checkpoint-1230/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/qwen_2_5_7b_math_instruct/limo/checkpoint-1230/trainer_state.json"
@@ -0,0 +1,8644 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.0,
+  "eval_steps": 500,
+  "global_step": 1230,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004878048780487805,
+      "grad_norm": 37.90617752075195,
+      "learning_rate": 5e-06,
+      "loss": 5.3731,
+      "step": 1
+    },
+    {
+      "epoch": 0.00975609756097561,
+      "grad_norm": 32.92101287841797,
+      "learning_rate": 4.999997064365715e-06,
+      "loss": 3.7922,
+      "step": 2
+    },
+    {
+      "epoch": 0.014634146341463415,
+      "grad_norm": 28.488866806030273,
+      "learning_rate": 4.999988257469751e-06,
+      "loss": 3.3238,
+      "step": 3
+    },
+    {
+      "epoch": 0.01951219512195122,
+      "grad_norm": 28.274654388427734,
+      "learning_rate": 4.999973579332793e-06,
+      "loss": 3.5405,
+      "step": 4
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 18.761985778808594,
+      "learning_rate": 4.999953029989312e-06,
+      "loss": 2.7492,
+      "step": 5
+    },
+    {
+      "epoch": 0.02926829268292683,
+      "grad_norm": 9.646541595458984,
+      "learning_rate": 4.999926609487568e-06,
+      "loss": 1.7146,
+      "step": 6
+    },
+    {
+      "epoch": 0.03414634146341464,
+      "grad_norm": 15.763612747192383,
+      "learning_rate": 4.9998943178896106e-06,
+      "loss": 2.8485,
+      "step": 7
+    },
+    {
+      "epoch": 0.03902439024390244,
+      "grad_norm": 13.853349685668945,
+      "learning_rate": 4.999856155271276e-06,
+      "loss": 3.0975,
+      "step": 8
+    },
+    {
+      "epoch": 0.04390243902439024,
+      "grad_norm": 16.893617630004883,
+      "learning_rate": 4.999812121722191e-06,
+      "loss": 2.8534,
+      "step": 9
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 15.47554874420166,
+      "learning_rate": 4.999762217345766e-06,
+      "loss": 3.4569,
+      "step": 10
+    },
+    {
+      "epoch": 0.05365853658536585,
+      "grad_norm": 6.277958393096924,
+      "learning_rate": 4.999706442259205e-06,
+      "loss": 1.8088,
+      "step": 11
+    },
+    {
+      "epoch": 0.05853658536585366,
+      "grad_norm": 5.451688289642334,
+      "learning_rate": 4.999644796593492e-06,
+      "loss": 1.8104,
+      "step": 12
+    },
+    {
+      "epoch": 0.06341463414634146,
+      "grad_norm": 4.333822727203369,
+      "learning_rate": 4.999577280493407e-06,
+      "loss": 1.7201,
+      "step": 13
+    },
+    {
+      "epoch": 0.06829268292682927,
+      "grad_norm": 4.488610744476318,
+      "learning_rate": 4.99950389411751e-06,
+      "loss": 1.9804,
+      "step": 14
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 1.8763024806976318,
+      "learning_rate": 4.999424637638148e-06,
+      "loss": 0.9227,
+      "step": 15
+    },
+    {
+      "epoch": 0.07804878048780488,
+      "grad_norm": 2.98995304107666,
+      "learning_rate": 4.999339511241458e-06,
+      "loss": 1.2368,
+      "step": 16
+    },
+    {
+      "epoch": 0.08292682926829269,
+      "grad_norm": 5.778397083282471,
+      "learning_rate": 4.9992485151273584e-06,
+      "loss": 2.5745,
+      "step": 17
+    },
+    {
+      "epoch": 0.08780487804878048,
+      "grad_norm": 5.222808361053467,
+      "learning_rate": 4.999151649509554e-06,
+      "loss": 2.0987,
+      "step": 18
+    },
+    {
+      "epoch": 0.09268292682926829,
+      "grad_norm": 4.075173377990723,
+      "learning_rate": 4.9990489146155356e-06,
+      "loss": 1.7732,
+      "step": 19
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 4.210625171661377,
+      "learning_rate": 4.9989403106865765e-06,
+      "loss": 1.8928,
+      "step": 20
+    },
+    {
+      "epoch": 0.1024390243902439,
+      "grad_norm": 1.7968504428863525,
+      "learning_rate": 4.9988258379777334e-06,
+      "loss": 1.0457,
+      "step": 21
+    },
+    {
+      "epoch": 0.1073170731707317,
+      "grad_norm": 2.494248151779175,
+      "learning_rate": 4.998705496757846e-06,
+      "loss": 1.3236,
+      "step": 22
+    },
+    {
+      "epoch": 0.11219512195121951,
+      "grad_norm": 4.312766075134277,
+      "learning_rate": 4.998579287309538e-06,
+      "loss": 2.4289,
+      "step": 23
+    },
+    {
+      "epoch": 0.11707317073170732,
+      "grad_norm": 5.361437797546387,
+      "learning_rate": 4.998447209929211e-06,
+      "loss": 1.7608,
+      "step": 24
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 1.4642046689987183,
+      "learning_rate": 4.998309264927053e-06,
+      "loss": 0.7571,
+      "step": 25
+    },
+    {
+      "epoch": 0.12682926829268293,
+      "grad_norm": 3.642005205154419,
+      "learning_rate": 4.998165452627025e-06,
+      "loss": 1.0634,
+      "step": 26
+    },
+    {
+      "epoch": 0.13170731707317074,
+      "grad_norm": 3.2727863788604736,
+      "learning_rate": 4.998015773366874e-06,
+      "loss": 1.4016,
+      "step": 27
+    },
+    {
+      "epoch": 0.13658536585365855,
+      "grad_norm": 1.5661338567733765,
+      "learning_rate": 4.997860227498122e-06,
+      "loss": 0.8305,
+      "step": 28
+    },
+    {
+      "epoch": 0.14146341463414633,
+      "grad_norm": 3.3989508152008057,
+      "learning_rate": 4.99769881538607e-06,
+      "loss": 1.8744,
+      "step": 29
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 3.512221574783325,
+      "learning_rate": 4.997531537409794e-06,
+      "loss": 1.7701,
+      "step": 30
+    },
+    {
+      "epoch": 0.15121951219512195,
+      "grad_norm": 2.261887311935425,
+      "learning_rate": 4.99735839396215e-06,
+      "loss": 1.0653,
+      "step": 31
+    },
+    {
+      "epoch": 0.15609756097560976,
+      "grad_norm": 2.2495696544647217,
+      "learning_rate": 4.9971793854497655e-06,
+      "loss": 1.1914,
+      "step": 32
+    },
+    {
+      "epoch": 0.16097560975609757,
+      "grad_norm": 4.262424468994141,
+      "learning_rate": 4.996994512293042e-06,
+      "loss": 1.7311,
+      "step": 33
+    },
+    {
+      "epoch": 0.16585365853658537,
+      "grad_norm": 3.361311674118042,
+      "learning_rate": 4.996803774926157e-06,
+      "loss": 1.4482,
+      "step": 34
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 1.5855786800384521,
+      "learning_rate": 4.996607173797059e-06,
+      "loss": 1.3959,
+      "step": 35
+    },
+    {
+      "epoch": 0.17560975609756097,
+      "grad_norm": 2.5001468658447266,
+      "learning_rate": 4.996404709367466e-06,
+      "loss": 1.3452,
+      "step": 36
+    },
+    {
+      "epoch": 0.18048780487804877,
+      "grad_norm": 2.7903096675872803,
+      "learning_rate": 4.996196382112868e-06,
+      "loss": 1.018,
+      "step": 37
+    },
+    {
+      "epoch": 0.18536585365853658,
+      "grad_norm": 2.7586171627044678,
+      "learning_rate": 4.9959821925225235e-06,
+      "loss": 1.291,
+      "step": 38
+    },
+    {
+      "epoch": 0.1902439024390244,
+      "grad_norm": 2.873840808868408,
+      "learning_rate": 4.995762141099456e-06,
+      "loss": 1.1418,
+      "step": 39
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 1.6896076202392578,
+      "learning_rate": 4.995536228360461e-06,
+      "loss": 1.1662,
+      "step": 40
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4597285985946655,
+      "learning_rate": 4.995304454836095e-06,
+      "loss": 1.0373,
+      "step": 41
+    },
+    {
+      "epoch": 0.2048780487804878,
+      "grad_norm": 2.4222421646118164,
+      "learning_rate": 4.9950668210706795e-06,
+      "loss": 1.1895,
+      "step": 42
+    },
+    {
+      "epoch": 0.2097560975609756,
+      "grad_norm": 3.295085906982422,
+      "learning_rate": 4.994823327622299e-06,
+      "loss": 1.4166,
+      "step": 43
+    },
+    {
+      "epoch": 0.2146341463414634,
+      "grad_norm": 1.8216716051101685,
+      "learning_rate": 4.9945739750628e-06,
+      "loss": 1.0009,
+      "step": 44
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 2.2421364784240723,
+      "learning_rate": 4.994318763977789e-06,
+      "loss": 1.1999,
+      "step": 45
+    },
+    {
+      "epoch": 0.22439024390243903,
+      "grad_norm": 5.6812744140625,
+      "learning_rate": 4.994057694966632e-06,
+      "loss": 1.3831,
+      "step": 46
+    },
+    {
+      "epoch": 0.22926829268292684,
+      "grad_norm": 1.477816104888916,
+      "learning_rate": 4.993790768642449e-06,
+      "loss": 1.1419,
+      "step": 47
+    },
+    {
+      "epoch": 0.23414634146341465,
+      "grad_norm": 4.485177993774414,
+      "learning_rate": 4.99351798563212e-06,
+      "loss": 1.9434,
+      "step": 48
+    },
+    {
+      "epoch": 0.23902439024390243,
+      "grad_norm": 2.770219326019287,
+      "learning_rate": 4.993239346576278e-06,
+      "loss": 1.3214,
+      "step": 49
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 2.721611976623535,
+      "learning_rate": 4.99295485212931e-06,
+      "loss": 1.4329,
+      "step": 50
+    },
+    {
+      "epoch": 0.24878048780487805,
+      "grad_norm": 3.714306592941284,
+      "learning_rate": 4.992664502959351e-06,
+      "loss": 1.372,
+      "step": 51
+    },
+    {
+      "epoch": 0.25365853658536586,
+      "grad_norm": 1.2679803371429443,
+      "learning_rate": 4.99236829974829e-06,
+      "loss": 0.8086,
+      "step": 52
+    },
+    {
+      "epoch": 0.25853658536585367,
+      "grad_norm": 1.6986169815063477,
+      "learning_rate": 4.992066243191762e-06,
+      "loss": 1.2012,
+      "step": 53
+    },
+    {
+      "epoch": 0.2634146341463415,
+      "grad_norm": 1.3443604707717896,
+      "learning_rate": 4.991758333999148e-06,
+      "loss": 0.8986,
+      "step": 54
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 2.0896975994110107,
+      "learning_rate": 4.991444572893575e-06,
+      "loss": 0.8819,
+      "step": 55
+    },
+    {
+      "epoch": 0.2731707317073171,
+      "grad_norm": 1.255516529083252,
+      "learning_rate": 4.991124960611916e-06,
+      "loss": 0.6255,
+      "step": 56
+    },
+    {
+      "epoch": 0.2780487804878049,
+      "grad_norm": 2.125410318374634,
+      "learning_rate": 4.99079949790478e-06,
+      "loss": 0.9983,
+      "step": 57
+    },
+    {
+      "epoch": 0.28292682926829266,
+      "grad_norm": 2.0312907695770264,
+      "learning_rate": 4.99046818553652e-06,
+      "loss": 1.1137,
+      "step": 58
+    },
+    {
+      "epoch": 0.28780487804878047,
+      "grad_norm": 2.904625177383423,
+      "learning_rate": 4.9901310242852246e-06,
+      "loss": 1.2009,
+      "step": 59
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 2.512932777404785,
+      "learning_rate": 4.9897880149427206e-06,
+      "loss": 1.2234,
+      "step": 60
+    },
+    {
+      "epoch": 0.2975609756097561,
+      "grad_norm": 2.5621752738952637,
+      "learning_rate": 4.989439158314566e-06,
+      "loss": 1.2354,
+      "step": 61
+    },
+    {
+      "epoch": 0.3024390243902439,
+      "grad_norm": 1.0051912069320679,
+      "learning_rate": 4.989084455220056e-06,
+      "loss": 0.614,
+      "step": 62
+    },
+    {
+      "epoch": 0.3073170731707317,
+      "grad_norm": 1.5008565187454224,
+      "learning_rate": 4.988723906492212e-06,
+      "loss": 1.002,
+      "step": 63
+    },
+    {
+      "epoch": 0.3121951219512195,
+      "grad_norm": 1.1436376571655273,
+      "learning_rate": 4.988357512977785e-06,
+      "loss": 0.6483,
+      "step": 64
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 1.16792893409729,
+      "learning_rate": 4.987985275537252e-06,
+      "loss": 0.638,
+      "step": 65
+    },
+    {
+      "epoch": 0.32195121951219513,
+      "grad_norm": 1.5670536756515503,
+      "learning_rate": 4.9876071950448185e-06,
+      "loss": 1.0274,
+      "step": 66
+    },
+    {
+      "epoch": 0.32682926829268294,
+      "grad_norm": 1.357515811920166,
+      "learning_rate": 4.987223272388407e-06,
+      "loss": 0.7612,
+      "step": 67
+    },
+    {
+      "epoch": 0.33170731707317075,
+      "grad_norm": 1.8383222818374634,
+      "learning_rate": 4.986833508469663e-06,
+      "loss": 1.1496,
+      "step": 68
+    },
+    {
+      "epoch": 0.33658536585365856,
+      "grad_norm": 2.8976657390594482,
+      "learning_rate": 4.98643790420395e-06,
+      "loss": 0.9924,
+      "step": 69
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 1.7679390907287598,
+      "learning_rate": 4.986036460520348e-06,
+      "loss": 0.9408,
+      "step": 70
+    },
+    {
+      "epoch": 0.3463414634146341,
+      "grad_norm": 2.3815059661865234,
+      "learning_rate": 4.98562917836165e-06,
+      "loss": 1.2923,
+      "step": 71
+    },
+    {
+      "epoch": 0.35121951219512193,
+      "grad_norm": 1.679026484489441,
+      "learning_rate": 4.985216058684362e-06,
+      "loss": 0.8476,
+      "step": 72
+    },
+    {
+      "epoch": 0.35609756097560974,
+      "grad_norm": 1.7254586219787598,
+      "learning_rate": 4.984797102458697e-06,
+      "loss": 1.1775,
+      "step": 73
+    },
+    {
+      "epoch": 0.36097560975609755,
+      "grad_norm": 1.195844292640686,
+      "learning_rate": 4.984372310668579e-06,
+      "loss": 0.6664,
+      "step": 74
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 1.754431128501892,
+      "learning_rate": 4.983941684311633e-06,
+      "loss": 1.4361,
+      "step": 75
+    },
+    {
+      "epoch": 0.37073170731707317,
+      "grad_norm": 1.8255788087844849,
+      "learning_rate": 4.983505224399188e-06,
+      "loss": 0.8977,
+      "step": 76
+    },
+    {
+      "epoch": 0.375609756097561,
+      "grad_norm": 1.340451717376709,
+      "learning_rate": 4.983062931956275e-06,
+      "loss": 0.9432,
+      "step": 77
+    },
+    {
+      "epoch": 0.3804878048780488,
+      "grad_norm": 1.7810503244400024,
+      "learning_rate": 4.9826148080216195e-06,
+      "loss": 1.0463,
+      "step": 78
+    },
+    {
+      "epoch": 0.3853658536585366,
+      "grad_norm": 1.729826807975769,
+      "learning_rate": 4.9821608536476445e-06,
+      "loss": 1.4451,
+      "step": 79
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 2.6638681888580322,
+      "learning_rate": 4.981701069900465e-06,
+      "loss": 1.0409,
+      "step": 80
+    },
+    {
+      "epoch": 0.3951219512195122,
+      "grad_norm": 1.467076301574707,
+      "learning_rate": 4.9812354578598876e-06,
+      "loss": 0.8596,
+      "step": 81
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.453657627105713,
+      "learning_rate": 4.980764018619405e-06,
+      "loss": 0.9033,
+      "step": 82
+    },
+    {
+      "epoch": 0.40487804878048783,
+      "grad_norm": 1.9025623798370361,
+      "learning_rate": 4.980286753286196e-06,
+      "loss": 1.0586,
+      "step": 83
+    },
+    {
+      "epoch": 0.4097560975609756,
+      "grad_norm": 1.4698103666305542,
+      "learning_rate": 4.97980366298112e-06,
+      "loss": 0.8611,
+      "step": 84
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 1.6086381673812866,
+      "learning_rate": 4.97931474883872e-06,
+      "loss": 0.8884,
+      "step": 85
+    },
+    {
+      "epoch": 0.4195121951219512,
+      "grad_norm": 1.6679224967956543,
+      "learning_rate": 4.978820012007213e-06,
+      "loss": 0.969,
+      "step": 86
+    },
+    {
+      "epoch": 0.424390243902439,
+      "grad_norm": 1.5308334827423096,
+      "learning_rate": 4.978319453648495e-06,
+      "loss": 1.0345,
+      "step": 87
+    },
+    {
+      "epoch": 0.4292682926829268,
+      "grad_norm": 1.7788817882537842,
+      "learning_rate": 4.977813074938128e-06,
+      "loss": 0.9649,
+      "step": 88
+    },
+    {
+      "epoch": 0.43414634146341463,
+      "grad_norm": 1.591383934020996,
+      "learning_rate": 4.977300877065347e-06,
+      "loss": 0.9299,
+      "step": 89
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 1.525985836982727,
+      "learning_rate": 4.976782861233053e-06,
+      "loss": 0.7272,
+      "step": 90
+    },
+    {
+      "epoch": 0.44390243902439025,
+      "grad_norm": 2.7333128452301025,
+      "learning_rate": 4.976259028657812e-06,
+      "loss": 0.6694,
+      "step": 91
+    },
+    {
+      "epoch": 0.44878048780487806,
+      "grad_norm": 1.2718108892440796,
+      "learning_rate": 4.975729380569845e-06,
+      "loss": 0.8037,
+      "step": 92
+    },
+    {
+      "epoch": 0.45365853658536587,
+      "grad_norm": 1.157771110534668,
+      "learning_rate": 4.975193918213035e-06,
+      "loss": 0.52,
+      "step": 93
+    },
+    {
+      "epoch": 0.4585365853658537,
+      "grad_norm": 1.0602248907089233,
+      "learning_rate": 4.974652642844921e-06,
+      "loss": 0.7135,
+      "step": 94
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 2.418555974960327,
+      "learning_rate": 4.974105555736693e-06,
+      "loss": 1.0996,
+      "step": 95
+    },
+    {
+      "epoch": 0.4682926829268293,
+      "grad_norm": 1.1728110313415527,
+      "learning_rate": 4.973552658173186e-06,
+      "loss": 0.7268,
+      "step": 96
+    },
+    {
+      "epoch": 0.47317073170731705,
+      "grad_norm": 1.943113088607788,
+      "learning_rate": 4.972993951452887e-06,
+      "loss": 0.9091,
+      "step": 97
+    },
+    {
+      "epoch": 0.47804878048780486,
+      "grad_norm": 5.063096523284912,
+      "learning_rate": 4.9724294368879214e-06,
+      "loss": 0.8242,
+      "step": 98
+    },
+    {
+      "epoch": 0.48292682926829267,
+      "grad_norm": 1.5232555866241455,
+      "learning_rate": 4.971859115804055e-06,
+      "loss": 1.0543,
+      "step": 99
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 1.1453967094421387,
+      "learning_rate": 4.9712829895406935e-06,
+      "loss": 0.8209,
+      "step": 100
+    },
+    {
+      "epoch": 0.4926829268292683,
+      "grad_norm": 2.12345814704895,
+      "learning_rate": 4.970701059450872e-06,
+      "loss": 0.5849,
+      "step": 101
+    },
+    {
+      "epoch": 0.4975609756097561,
+      "grad_norm": 1.2201842069625854,
+      "learning_rate": 4.970113326901258e-06,
+      "loss": 0.9969,
+      "step": 102
+    },
+    {
+      "epoch": 0.5024390243902439,
+      "grad_norm": 1.3047524690628052,
+      "learning_rate": 4.9695197932721455e-06,
+      "loss": 0.9339,
+      "step": 103
+    },
+    {
+      "epoch": 0.5073170731707317,
+      "grad_norm": 1.6083660125732422,
+      "learning_rate": 4.968920459957453e-06,
+      "loss": 0.9702,
+      "step": 104
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 1.5247286558151245,
+      "learning_rate": 4.968315328364719e-06,
+      "loss": 1.0449,
+      "step": 105
+    },
+    {
+      "epoch": 0.5170731707317073,
+      "grad_norm": 1.354805588722229,
+      "learning_rate": 4.9677043999151e-06,
+      "loss": 1.1431,
+      "step": 106
+    },
+    {
+      "epoch": 0.5219512195121951,
+      "grad_norm": 1.088321328163147,
+      "learning_rate": 4.967087676043366e-06,
+      "loss": 0.519,
+      "step": 107
+    },
+    {
+      "epoch": 0.526829268292683,
+      "grad_norm": 1.5086662769317627,
+      "learning_rate": 4.966465158197897e-06,
+      "loss": 0.9357,
+      "step": 108
+    },
+    {
+      "epoch": 0.5317073170731708,
+      "grad_norm": 1.3161298036575317,
+      "learning_rate": 4.965836847840681e-06,
+      "loss": 0.7234,
+      "step": 109
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 1.4465640783309937,
+      "learning_rate": 4.96520274644731e-06,
+      "loss": 0.8911,
+      "step": 110
+    },
+    {
+      "epoch": 0.5414634146341464,
+      "grad_norm": 1.0576995611190796,
+      "learning_rate": 4.964562855506976e-06,
+      "loss": 0.7254,
+      "step": 111
+    },
+    {
+      "epoch": 0.5463414634146342,
+      "grad_norm": 1.1018916368484497,
+      "learning_rate": 4.963917176522466e-06,
+      "loss": 0.6603,
+      "step": 112
+    },
+    {
+      "epoch": 0.551219512195122,
+      "grad_norm": 2.150622606277466,
+      "learning_rate": 4.963265711010164e-06,
+      "loss": 1.0451,
+      "step": 113
+    },
+    {
+      "epoch": 0.5560975609756098,
+      "grad_norm": 1.7743186950683594,
+      "learning_rate": 4.9626084605000395e-06,
+      "loss": 0.8717,
+      "step": 114
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 1.205291509628296,
+      "learning_rate": 4.961945426535652e-06,
+      "loss": 0.5802,
+      "step": 115
+    },
+    {
+      "epoch": 0.5658536585365853,
+      "grad_norm": 1.3969353437423706,
+      "learning_rate": 4.961276610674141e-06,
+      "loss": 0.9158,
+      "step": 116
+    },
+    {
+      "epoch": 0.5707317073170731,
+      "grad_norm": 1.2318240404129028,
+      "learning_rate": 4.960602014486225e-06,
+      "loss": 1.0086,
+      "step": 117
+    },
+    {
+      "epoch": 0.5756097560975609,
+      "grad_norm": 1.2202470302581787,
+      "learning_rate": 4.959921639556199e-06,
+      "loss": 0.7888,
+      "step": 118
+    },
+    {
+      "epoch": 0.5804878048780487,
+      "grad_norm": 1.1564440727233887,
+      "learning_rate": 4.959235487481928e-06,
+      "loss": 1.0053,
+      "step": 119
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 1.2278865575790405,
+      "learning_rate": 4.958543559874846e-06,
+      "loss": 0.5486,
+      "step": 120
+    },
+    {
+      "epoch": 0.5902439024390244,
+      "grad_norm": 1.5465888977050781,
+      "learning_rate": 4.9578458583599495e-06,
+      "loss": 0.8232,
+      "step": 121
+    },
+    {
+      "epoch": 0.5951219512195122,
+      "grad_norm": 1.6284047365188599,
+      "learning_rate": 4.957142384575795e-06,
+      "loss": 0.7773,
+      "step": 122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1794490814208984,
+      "learning_rate": 4.956433140174498e-06,
+      "loss": 0.8236,
+      "step": 123
+    },
+    {
+      "epoch": 0.6048780487804878,
+      "grad_norm": 1.2404091358184814,
+      "learning_rate": 4.9557181268217225e-06,
+      "loss": 0.8611,
+      "step": 124
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 1.2188373804092407,
+      "learning_rate": 4.954997346196683e-06,
+      "loss": 1.2922,
+      "step": 125
+    },
+    {
+      "epoch": 0.6146341463414634,
+      "grad_norm": 1.1514503955841064,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.7073,
+      "step": 126
+    },
+    {
+      "epoch": 0.6195121951219512,
+      "grad_norm": 1.285913109779358,
+      "learning_rate": 4.953538489914387e-06,
+      "loss": 0.729,
+      "step": 127
+    },
+    {
+      "epoch": 0.624390243902439,
+      "grad_norm": 1.8677104711532593,
+      "learning_rate": 4.9528004176832654e-06,
+      "loss": 0.6211,
+      "step": 128
+    },
+    {
+      "epoch": 0.6292682926829268,
+      "grad_norm": 1.2369698286056519,
+      "learning_rate": 4.952056585032142e-06,
+      "loss": 1.0311,
+      "step": 129
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 1.186990737915039,
+      "learning_rate": 4.951306993707913e-06,
+      "loss": 0.8207,
+      "step": 130
+    },
+    {
+      "epoch": 0.6390243902439025,
+      "grad_norm": 1.189424991607666,
+      "learning_rate": 4.950551645470998e-06,
+      "loss": 0.7853,
+      "step": 131
+    },
+    {
+      "epoch": 0.6439024390243903,
+      "grad_norm": 1.3477216958999634,
+      "learning_rate": 4.9497905420953406e-06,
+      "loss": 0.7271,
+      "step": 132
+    },
+    {
+      "epoch": 0.6487804878048781,
+      "grad_norm": 1.368592619895935,
+      "learning_rate": 4.949023685368395e-06,
+      "loss": 0.7394,
+      "step": 133
+    },
+    {
+      "epoch": 0.6536585365853659,
+      "grad_norm": 1.2769527435302734,
+      "learning_rate": 4.948251077091131e-06,
+      "loss": 1.0713,
+      "step": 134
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 1.1477972269058228,
+      "learning_rate": 4.947472719078025e-06,
+      "loss": 0.841,
+      "step": 135
+    },
+    {
+      "epoch": 0.6634146341463415,
+      "grad_norm": 1.0983368158340454,
+      "learning_rate": 4.9466886131570565e-06,
+      "loss": 0.8665,
+      "step": 136
+    },
+    {
+      "epoch": 0.6682926829268293,
+      "grad_norm": 1.2526847124099731,
+      "learning_rate": 4.945898761169704e-06,
+      "loss": 1.0613,
+      "step": 137
+    },
+    {
+      "epoch": 0.6731707317073171,
+      "grad_norm": 1.1426396369934082,
+      "learning_rate": 4.945103164970941e-06,
+      "loss": 0.6248,
+      "step": 138
+    },
+    {
+      "epoch": 0.6780487804878049,
+      "grad_norm": 1.3275880813598633,
+      "learning_rate": 4.9443018264292304e-06,
+      "loss": 0.755,
+      "step": 139
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 1.4942073822021484,
+      "learning_rate": 4.9434947474265225e-06,
+      "loss": 0.9451,
+      "step": 140
+    },
+    {
+      "epoch": 0.6878048780487804,
+      "grad_norm": 1.1777600049972534,
+      "learning_rate": 4.942681929858249e-06,
+      "loss": 1.0505,
+      "step": 141
+    },
+    {
+      "epoch": 0.6926829268292682,
+      "grad_norm": 1.2474116086959839,
+      "learning_rate": 4.941863375633315e-06,
+      "loss": 0.9174,
+      "step": 142
+    },
+    {
+      "epoch": 0.697560975609756,
+      "grad_norm": 1.5067697763442993,
+      "learning_rate": 4.9410390866741056e-06,
+      "loss": 0.7737,
+      "step": 143
+    },
+    {
+      "epoch": 0.7024390243902439,
+      "grad_norm": 1.3016574382781982,
+      "learning_rate": 4.9402090649164655e-06,
+      "loss": 0.7588,
+      "step": 144
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 1.652600884437561,
+      "learning_rate": 4.9393733123097085e-06,
+      "loss": 1.0588,
+      "step": 145
+    },
+    {
+      "epoch": 0.7121951219512195,
+      "grad_norm": 1.267997145652771,
+      "learning_rate": 4.9385318308166065e-06,
+      "loss": 0.8996,
+      "step": 146
+    },
+    {
+      "epoch": 0.7170731707317073,
+      "grad_norm": 1.8560484647750854,
+      "learning_rate": 4.937684622413385e-06,
+      "loss": 0.6286,
+      "step": 147
+    },
+    {
+      "epoch": 0.7219512195121951,
+      "grad_norm": 1.4419782161712646,
+      "learning_rate": 4.9368316890897185e-06,
+      "loss": 0.9818,
+      "step": 148
+    },
+    {
+      "epoch": 0.7268292682926829,
+      "grad_norm": 1.134084701538086,
+      "learning_rate": 4.9359730328487264e-06,
+      "loss": 0.5246,
+      "step": 149
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 1.0102615356445312,
+      "learning_rate": 4.935108655706972e-06,
+      "loss": 0.7948,
+      "step": 150
+    },
+    {
+      "epoch": 0.7365853658536585,
+      "grad_norm": 1.3480703830718994,
+      "learning_rate": 4.934238559694448e-06,
+      "loss": 1.0951,
+      "step": 151
+    },
+    {
+      "epoch": 0.7414634146341463,
+      "grad_norm": 1.4133951663970947,
+      "learning_rate": 4.9333627468545845e-06,
+      "loss": 0.6936,
+      "step": 152
+    },
+    {
+      "epoch": 0.7463414634146341,
+      "grad_norm": 1.3072413206100464,
+      "learning_rate": 4.932481219244231e-06,
+      "loss": 0.799,
+      "step": 153
+    },
+    {
+      "epoch": 0.751219512195122,
+      "grad_norm": 1.3893049955368042,
+      "learning_rate": 4.931593978933666e-06,
+      "loss": 0.7375,
+      "step": 154
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 1.2313531637191772,
+      "learning_rate": 4.930701028006577e-06,
+      "loss": 0.9487,
+      "step": 155
+    },
+    {
+      "epoch": 0.7609756097560976,
+      "grad_norm": 1.3426295518875122,
+      "learning_rate": 4.929802368560066e-06,
+      "loss": 0.7542,
+      "step": 156
+    },
+    {
+      "epoch": 0.7658536585365854,
+      "grad_norm": 1.5280004739761353,
+      "learning_rate": 4.928898002704642e-06,
+      "loss": 0.8784,
+      "step": 157
+    },
+    {
+      "epoch": 0.7707317073170732,
+      "grad_norm": 1.3149527311325073,
+      "learning_rate": 4.927987932564215e-06,
+      "loss": 0.7247,
+      "step": 158
+    },
+    {
+      "epoch": 0.775609756097561,
+      "grad_norm": 1.073188304901123,
+      "learning_rate": 4.927072160276092e-06,
+      "loss": 0.7826,
+      "step": 159
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 1.426188588142395,
+      "learning_rate": 4.926150687990969e-06,
+      "loss": 0.6129,
+      "step": 160
+    },
+    {
+      "epoch": 0.7853658536585366,
+      "grad_norm": 1.2348767518997192,
+      "learning_rate": 4.925223517872934e-06,
+      "loss": 0.9191,
+      "step": 161
+    },
+    {
+      "epoch": 0.7902439024390244,
+      "grad_norm": 1.599665641784668,
+      "learning_rate": 4.9242906520994484e-06,
+      "loss": 1.0177,
+      "step": 162
+    },
+    {
+      "epoch": 0.7951219512195122,
+      "grad_norm": 1.1841332912445068,
+      "learning_rate": 4.923352092861358e-06,
+      "loss": 0.8342,
+      "step": 163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0213048458099365,
+      "learning_rate": 4.922407842362875e-06,
+      "loss": 0.6292,
+      "step": 164
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 1.1878992319107056,
+      "learning_rate": 4.921457902821578e-06,
+      "loss": 0.9357,
+      "step": 165
+    },
+    {
+      "epoch": 0.8097560975609757,
+      "grad_norm": 1.297462821006775,
+      "learning_rate": 4.920502276468408e-06,
+      "loss": 0.8829,
+      "step": 166
+    },
+    {
+      "epoch": 0.8146341463414634,
+      "grad_norm": 1.1800835132598877,
+      "learning_rate": 4.9195409655476605e-06,
+      "loss": 0.712,
+      "step": 167
+    },
+    {
+      "epoch": 0.8195121951219512,
+      "grad_norm": 1.4254546165466309,
+      "learning_rate": 4.918573972316982e-06,
+      "loss": 0.9997,
+      "step": 168
+    },
+    {
+      "epoch": 0.824390243902439,
+      "grad_norm": 1.3243224620819092,
+      "learning_rate": 4.917601299047361e-06,
+      "loss": 0.7944,
+      "step": 169
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 1.3879033327102661,
+      "learning_rate": 4.916622948023129e-06,
+      "loss": 0.7778,
+      "step": 170
+    },
+    {
+      "epoch": 0.8341463414634146,
+      "grad_norm": 1.245430588722229,
+      "learning_rate": 4.915638921541952e-06,
+      "loss": 0.6247,
+      "step": 171
+    },
+    {
+      "epoch": 0.8390243902439024,
+      "grad_norm": 1.3728258609771729,
+      "learning_rate": 4.914649221914822e-06,
+      "loss": 0.8762,
+      "step": 172
+    },
+    {
+      "epoch": 0.8439024390243902,
+      "grad_norm": 1.3080862760543823,
+      "learning_rate": 4.913653851466057e-06,
+      "loss": 0.6381,
+      "step": 173
+    },
+    {
+      "epoch": 0.848780487804878,
+      "grad_norm": 1.5109484195709229,
+      "learning_rate": 4.912652812533291e-06,
+      "loss": 0.8127,
+      "step": 174
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 1.237879753112793,
+      "learning_rate": 4.911646107467472e-06,
+      "loss": 0.8257,
+      "step": 175
+    },
+    {
+      "epoch": 0.8585365853658536,
+      "grad_norm": 1.2725176811218262,
+      "learning_rate": 4.9106337386328524e-06,
+      "loss": 0.9758,
+      "step": 176
+    },
+    {
+      "epoch": 0.8634146341463415,
+      "grad_norm": 1.271756887435913,
+      "learning_rate": 4.909615708406991e-06,
+      "loss": 0.8436,
+      "step": 177
+    },
+    {
+      "epoch": 0.8682926829268293,
+      "grad_norm": 1.0920095443725586,
+      "learning_rate": 4.908592019180738e-06,
+      "loss": 0.6657,
+      "step": 178
+    },
+    {
+      "epoch": 0.8731707317073171,
+      "grad_norm": 0.9080491065979004,
+      "learning_rate": 4.907562673358234e-06,
+      "loss": 0.6322,
+      "step": 179
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 0.9868524074554443,
+      "learning_rate": 4.906527673356907e-06,
+      "loss": 0.5918,
+      "step": 180
+    },
+    {
+      "epoch": 0.8829268292682927,
+      "grad_norm": 1.49118173122406,
+      "learning_rate": 4.905487021607462e-06,
+      "loss": 0.4939,
+      "step": 181
+    },
+    {
+      "epoch": 0.8878048780487805,
+      "grad_norm": 1.3323795795440674,
+      "learning_rate": 4.904440720553876e-06,
+      "loss": 0.8444,
+      "step": 182
+    },
+    {
+      "epoch": 0.8926829268292683,
+      "grad_norm": 1.340374231338501,
+      "learning_rate": 4.903388772653396e-06,
+      "loss": 0.765,
+      "step": 183
+    },
+    {
+      "epoch": 0.8975609756097561,
+      "grad_norm": 1.2728081941604614,
+      "learning_rate": 4.902331180376529e-06,
+      "loss": 0.8126,
+      "step": 184
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 1.154461145401001,
+      "learning_rate": 4.901267946207038e-06,
+      "loss": 0.9256,
+      "step": 185
+    },
+    {
+      "epoch": 0.9073170731707317,
+      "grad_norm": 1.0781828165054321,
+      "learning_rate": 4.900199072641937e-06,
+      "loss": 0.7057,
+      "step": 186
+    },
+    {
+      "epoch": 0.9121951219512195,
+      "grad_norm": 1.0474439859390259,
+      "learning_rate": 4.899124562191484e-06,
+      "loss": 0.6024,
+      "step": 187
+    },
+    {
+      "epoch": 0.9170731707317074,
+      "grad_norm": 1.019089937210083,
+      "learning_rate": 4.8980444173791735e-06,
+      "loss": 0.5001,
+      "step": 188
+    },
+    {
+      "epoch": 0.9219512195121952,
+      "grad_norm": 1.3556936979293823,
+      "learning_rate": 4.896958640741735e-06,
+      "loss": 0.8636,
+      "step": 189
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 1.2974662780761719,
+      "learning_rate": 4.895867234829121e-06,
+      "loss": 0.9257,
+      "step": 190
+    },
+    {
+      "epoch": 0.9317073170731708,
+      "grad_norm": 1.3130805492401123,
+      "learning_rate": 4.894770202204509e-06,
+      "loss": 0.7026,
+      "step": 191
+    },
+    {
+      "epoch": 0.9365853658536586,
+      "grad_norm": 1.5521293878555298,
+      "learning_rate": 4.893667545444285e-06,
+      "loss": 0.8671,
+      "step": 192
+    },
+    {
+      "epoch": 0.9414634146341463,
+      "grad_norm": 1.0167640447616577,
+      "learning_rate": 4.8925592671380495e-06,
+      "loss": 0.6643,
+      "step": 193
+    },
+    {
+      "epoch": 0.9463414634146341,
+      "grad_norm": 0.8992295861244202,
+      "learning_rate": 4.891445369888601e-06,
+      "loss": 0.5294,
+      "step": 194
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 1.053918480873108,
+      "learning_rate": 4.890325856311936e-06,
+      "loss": 0.8734,
+      "step": 195
+    },
+    {
+      "epoch": 0.9560975609756097,
+      "grad_norm": 1.1650863885879517,
+      "learning_rate": 4.889200729037241e-06,
+      "loss": 0.7858,
+      "step": 196
+    },
+    {
+      "epoch": 0.9609756097560975,
+      "grad_norm": 1.0390702486038208,
+      "learning_rate": 4.888069990706884e-06,
+      "loss": 0.6428,
+      "step": 197
+    },
+    {
+      "epoch": 0.9658536585365853,
+      "grad_norm": 1.4749332666397095,
+      "learning_rate": 4.886933643976414e-06,
+      "loss": 0.55,
+      "step": 198
+    },
+    {
+      "epoch": 0.9707317073170731,
+      "grad_norm": 0.9483368992805481,
+      "learning_rate": 4.885791691514548e-06,
+      "loss": 0.4786,
+      "step": 199
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 1.0259203910827637,
+      "learning_rate": 4.884644136003172e-06,
+      "loss": 0.6137,
+      "step": 200
+    },
+    {
+      "epoch": 0.9804878048780488,
+      "grad_norm": 1.1431037187576294,
+      "learning_rate": 4.883490980137327e-06,
+      "loss": 1.2973,
+      "step": 201
+    },
+    {
+      "epoch": 0.9853658536585366,
+      "grad_norm": 1.1985281705856323,
+      "learning_rate": 4.882332226625208e-06,
+      "loss": 0.7734,
+      "step": 202
+    },
+    {
+      "epoch": 0.9902439024390244,
+      "grad_norm": 0.9396414160728455,
+      "learning_rate": 4.881167878188158e-06,
+      "loss": 0.8172,
+      "step": 203
+    },
+    {
+      "epoch": 0.9951219512195122,
+      "grad_norm": 1.3293070793151855,
+      "learning_rate": 4.8799979375606565e-06,
+      "loss": 0.727,
+      "step": 204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0133116245269775,
+      "learning_rate": 4.878822407490319e-06,
+      "loss": 0.7099,
+      "step": 205
+    },
+    {
+      "epoch": 1.0048780487804878,
+      "grad_norm": 1.0823330879211426,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.6611,
+      "step": 206
+    },
+    {
+      "epoch": 1.0097560975609756,
+      "grad_norm": 0.9297066926956177,
+      "learning_rate": 4.876454590077216e-06,
+      "loss": 0.5825,
+      "step": 207
+    },
+    {
+      "epoch": 1.0146341463414634,
+      "grad_norm": 1.0025253295898438,
+      "learning_rate": 4.875262308295289e-06,
+      "loss": 0.8154,
+      "step": 208
+    },
+    {
+      "epoch": 1.0195121951219512,
+      "grad_norm": 1.1356369256973267,
+      "learning_rate": 4.874064448192185e-06,
+      "loss": 0.7126,
+      "step": 209
+    },
+    {
+      "epoch": 1.024390243902439,
+      "grad_norm": 1.4361388683319092,
+      "learning_rate": 4.872861012581088e-06,
+      "loss": 0.6323,
+      "step": 210
+    },
+    {
+      "epoch": 1.0292682926829269,
+      "grad_norm": 1.2314097881317139,
+      "learning_rate": 4.871652004288275e-06,
+      "loss": 0.6683,
+      "step": 211
+    },
+    {
+      "epoch": 1.0341463414634147,
+      "grad_norm": 1.1820945739746094,
+      "learning_rate": 4.870437426153113e-06,
+      "loss": 0.6659,
+      "step": 212
+    },
+    {
+      "epoch": 1.0390243902439025,
+      "grad_norm": 1.1955009698867798,
+      "learning_rate": 4.869217281028045e-06,
+      "loss": 0.8964,
+      "step": 213
+    },
+    {
+      "epoch": 1.0439024390243903,
+      "grad_norm": 1.2589020729064941,
+      "learning_rate": 4.867991571778592e-06,
+      "loss": 0.894,
+      "step": 214
+    },
+    {
+      "epoch": 1.048780487804878,
+      "grad_norm": 0.8933411836624146,
+      "learning_rate": 4.866760301283342e-06,
+      "loss": 0.5199,
+      "step": 215
+    },
+    {
+      "epoch": 1.053658536585366,
+      "grad_norm": 1.126219630241394,
+      "learning_rate": 4.865523472433942e-06,
+      "loss": 0.7014,
+      "step": 216
+    },
+    {
+      "epoch": 1.0585365853658537,
+      "grad_norm": 1.1173880100250244,
+      "learning_rate": 4.8642810881350935e-06,
+      "loss": 0.6085,
+      "step": 217
+    },
+    {
+      "epoch": 1.0634146341463415,
+      "grad_norm": 0.9508454203605652,
+      "learning_rate": 4.863033151304546e-06,
+      "loss": 0.6333,
+      "step": 218
+    },
+    {
+      "epoch": 1.0682926829268293,
+      "grad_norm": 1.8204004764556885,
+      "learning_rate": 4.861779664873088e-06,
+      "loss": 0.8899,
+      "step": 219
+    },
+    {
+      "epoch": 1.0731707317073171,
+      "grad_norm": 1.5721564292907715,
+      "learning_rate": 4.8605206317845425e-06,
+      "loss": 0.8105,
+      "step": 220
+    },
+    {
+      "epoch": 1.078048780487805,
+      "grad_norm": 1.3572758436203003,
+      "learning_rate": 4.859256054995758e-06,
+      "loss": 0.8634,
+      "step": 221
+    },
+    {
+      "epoch": 1.0829268292682928,
+      "grad_norm": 0.8531430959701538,
+      "learning_rate": 4.8579859374766e-06,
+      "loss": 0.4826,
+      "step": 222
+    },
+    {
+      "epoch": 1.0878048780487806,
+      "grad_norm": 0.771088719367981,
+      "learning_rate": 4.856710282209952e-06,
+      "loss": 0.4085,
+      "step": 223
+    },
+    {
+      "epoch": 1.0926829268292684,
+      "grad_norm": 1.0056614875793457,
+      "learning_rate": 4.855429092191698e-06,
+      "loss": 0.7168,
+      "step": 224
+    },
+    {
+      "epoch": 1.0975609756097562,
+      "grad_norm": 0.9825501441955566,
+      "learning_rate": 4.854142370430725e-06,
+      "loss": 0.6276,
+      "step": 225
+    },
+    {
+      "epoch": 1.102439024390244,
+      "grad_norm": 1.0209708213806152,
+      "learning_rate": 4.8528501199489045e-06,
+      "loss": 0.6521,
+      "step": 226
+    },
+    {
+      "epoch": 1.1073170731707318,
+      "grad_norm": 1.3162622451782227,
+      "learning_rate": 4.851552343781099e-06,
+      "loss": 0.8244,
+      "step": 227
+    },
+    {
+      "epoch": 1.1121951219512196,
+      "grad_norm": 1.150299072265625,
+      "learning_rate": 4.850249044975145e-06,
+      "loss": 0.7804,
+      "step": 228
+    },
+    {
+      "epoch": 1.1170731707317074,
+      "grad_norm": 1.142996907234192,
+      "learning_rate": 4.848940226591849e-06,
+      "loss": 0.9935,
+      "step": 229
+    },
+    {
+      "epoch": 1.1219512195121952,
+      "grad_norm": 0.9757675528526306,
+      "learning_rate": 4.847625891704982e-06,
+      "loss": 0.6041,
+      "step": 230
+    },
+    {
+      "epoch": 1.126829268292683,
+      "grad_norm": 1.1686437129974365,
+      "learning_rate": 4.846306043401268e-06,
+      "loss": 0.6873,
+      "step": 231
+    },
+    {
+      "epoch": 1.1317073170731708,
+      "grad_norm": 1.4826396703720093,
+      "learning_rate": 4.844980684780381e-06,
+      "loss": 0.6412,
+      "step": 232
+    },
+    {
+      "epoch": 1.1365853658536587,
+      "grad_norm": 1.204198956489563,
+      "learning_rate": 4.8436498189549345e-06,
+      "loss": 0.5779,
+      "step": 233
+    },
+    {
+      "epoch": 1.1414634146341462,
+      "grad_norm": 0.8927454948425293,
+      "learning_rate": 4.842313449050477e-06,
+      "loss": 0.5894,
+      "step": 234
+    },
+    {
+      "epoch": 1.146341463414634,
+      "grad_norm": 0.9504485726356506,
+      "learning_rate": 4.840971578205486e-06,
+      "loss": 0.5195,
+      "step": 235
+    },
+    {
+      "epoch": 1.1512195121951219,
+      "grad_norm": 0.8682957291603088,
+      "learning_rate": 4.839624209571352e-06,
+      "loss": 0.3927,
+      "step": 236
+    },
+    {
+      "epoch": 1.1560975609756097,
+      "grad_norm": 1.0243151187896729,
+      "learning_rate": 4.838271346312381e-06,
+      "loss": 0.8549,
+      "step": 237
+    },
+    {
+      "epoch": 1.1609756097560975,
+      "grad_norm": 0.9931624531745911,
+      "learning_rate": 4.836912991605782e-06,
+      "loss": 0.8616,
+      "step": 238
+    },
+    {
+      "epoch": 1.1658536585365853,
+      "grad_norm": 1.0424671173095703,
+      "learning_rate": 4.835549148641663e-06,
+      "loss": 0.5675,
+      "step": 239
+    },
+    {
+      "epoch": 1.170731707317073,
+      "grad_norm": 0.9706572890281677,
+      "learning_rate": 4.834179820623018e-06,
+      "loss": 0.6444,
+      "step": 240
+    },
+    {
+      "epoch": 1.175609756097561,
+      "grad_norm": 0.9557904601097107,
+      "learning_rate": 4.832805010765724e-06,
+      "loss": 0.6006,
+      "step": 241
+    },
+    {
+      "epoch": 1.1804878048780487,
+      "grad_norm": 1.2297061681747437,
+      "learning_rate": 4.831424722298531e-06,
+      "loss": 0.6897,
+      "step": 242
+    },
+    {
+      "epoch": 1.1853658536585365,
+      "grad_norm": 1.3953443765640259,
+      "learning_rate": 4.830038958463061e-06,
+      "loss": 0.772,
+      "step": 243
+    },
+    {
+      "epoch": 1.1902439024390243,
+      "grad_norm": 2.120232105255127,
+      "learning_rate": 4.828647722513785e-06,
+      "loss": 0.9361,
+      "step": 244
+    },
+    {
+      "epoch": 1.1951219512195121,
+      "grad_norm": 1.4457300901412964,
+      "learning_rate": 4.827251017718034e-06,
+      "loss": 0.8036,
+      "step": 245
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0927690267562866,
+      "learning_rate": 4.8258488473559794e-06,
+      "loss": 0.8354,
+      "step": 246
+    },
+    {
+      "epoch": 1.2048780487804878,
+      "grad_norm": 1.4953900575637817,
+      "learning_rate": 4.824441214720629e-06,
+      "loss": 0.9083,
+      "step": 247
+    },
+    {
+      "epoch": 1.2097560975609756,
+      "grad_norm": 0.9229385852813721,
+      "learning_rate": 4.823028123117818e-06,
+      "loss": 0.441,
+      "step": 248
+    },
+    {
+      "epoch": 1.2146341463414634,
+      "grad_norm": 1.8001827001571655,
+      "learning_rate": 4.8216095758662015e-06,
+      "loss": 0.7975,
+      "step": 249
+    },
+    {
+      "epoch": 1.2195121951219512,
+      "grad_norm": 1.0904715061187744,
+      "learning_rate": 4.82018557629725e-06,
+      "loss": 0.8526,
+      "step": 250
+    },
+    {
+      "epoch": 1.224390243902439,
+      "grad_norm": 1.3482294082641602,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.6198,
+      "step": 251
+    },
+    {
+      "epoch": 1.2292682926829268,
+      "grad_norm": 0.9257897138595581,
+      "learning_rate": 4.817321233597232e-06,
+      "loss": 0.8115,
+      "step": 252
+    },
+    {
+      "epoch": 1.2341463414634146,
+      "grad_norm": 1.1533488035202026,
+      "learning_rate": 4.815880897193095e-06,
+      "loss": 0.5524,
+      "step": 253
+    },
+    {
+      "epoch": 1.2390243902439024,
+      "grad_norm": 1.2778701782226562,
+      "learning_rate": 4.814435121925466e-06,
+      "loss": 0.802,
+      "step": 254
+    },
+    {
+      "epoch": 1.2439024390243902,
+      "grad_norm": 1.037063717842102,
+      "learning_rate": 4.812983911189761e-06,
+      "loss": 0.696,
+      "step": 255
+    },
+    {
+      "epoch": 1.248780487804878,
+      "grad_norm": 1.0112067461013794,
+      "learning_rate": 4.811527268394157e-06,
+      "loss": 0.5189,
+      "step": 256
+    },
+    {
+      "epoch": 1.2536585365853659,
+      "grad_norm": 1.1271675825119019,
+      "learning_rate": 4.810065196959591e-06,
+      "loss": 0.6292,
+      "step": 257
+    },
+    {
+      "epoch": 1.2585365853658537,
+      "grad_norm": 1.0987378358840942,
+      "learning_rate": 4.8085977003197496e-06,
+      "loss": 0.8452,
+      "step": 258
+    },
+    {
+      "epoch": 1.2634146341463415,
+      "grad_norm": 1.5386906862258911,
+      "learning_rate": 4.807124781921059e-06,
+      "loss": 1.1094,
+      "step": 259
+    },
+    {
+      "epoch": 1.2682926829268293,
+      "grad_norm": 0.9946249127388,
+      "learning_rate": 4.805646445222679e-06,
+      "loss": 0.6367,
+      "step": 260
+    },
+    {
+      "epoch": 1.273170731707317,
+      "grad_norm": 0.9774323105812073,
+      "learning_rate": 4.804162693696494e-06,
+      "loss": 0.6157,
+      "step": 261
+    },
+    {
+      "epoch": 1.278048780487805,
+      "grad_norm": 1.9447377920150757,
+      "learning_rate": 4.802673530827105e-06,
+      "loss": 0.6076,
+      "step": 262
+    },
+    {
+      "epoch": 1.2829268292682927,
+      "grad_norm": 0.7815698981285095,
+      "learning_rate": 4.801178960111823e-06,
+      "loss": 0.5985,
+      "step": 263
+    },
+    {
+      "epoch": 1.2878048780487805,
+      "grad_norm": 1.014426589012146,
+      "learning_rate": 4.799678985060658e-06,
+      "loss": 0.851,
+      "step": 264
+    },
+    {
+      "epoch": 1.2926829268292683,
+      "grad_norm": 0.9904977083206177,
+      "learning_rate": 4.798173609196314e-06,
+      "loss": 0.8511,
+      "step": 265
+    },
+    {
+      "epoch": 1.2975609756097561,
+      "grad_norm": 1.0354292392730713,
+      "learning_rate": 4.796662836054176e-06,
+      "loss": 0.4885,
+      "step": 266
+    },
+    {
+      "epoch": 1.302439024390244,
+      "grad_norm": 0.9485008120536804,
+      "learning_rate": 4.795146669182304e-06,
+      "loss": 0.6486,
+      "step": 267
+    },
+    {
+      "epoch": 1.3073170731707318,
+      "grad_norm": 0.9609008431434631,
+      "learning_rate": 4.793625112141431e-06,
+      "loss": 0.5446,
+      "step": 268
+    },
+    {
+      "epoch": 1.3121951219512196,
+      "grad_norm": 0.8880930542945862,
+      "learning_rate": 4.792098168504943e-06,
+      "loss": 0.5741,
+      "step": 269
+    },
+    {
+      "epoch": 1.3170731707317074,
+      "grad_norm": 0.8475173711776733,
+      "learning_rate": 4.790565841858879e-06,
+      "loss": 0.488,
+      "step": 270
+    },
+    {
+      "epoch": 1.3219512195121952,
+      "grad_norm": 1.04447603225708,
+      "learning_rate": 4.789028135801919e-06,
+      "loss": 0.7792,
+      "step": 271
+    },
+    {
+      "epoch": 1.326829268292683,
+      "grad_norm": 1.2021688222885132,
+      "learning_rate": 4.787485053945377e-06,
+      "loss": 0.8513,
+      "step": 272
+    },
+    {
+      "epoch": 1.3317073170731708,
+      "grad_norm": 0.9609706401824951,
+      "learning_rate": 4.785936599913193e-06,
+      "loss": 0.7186,
+      "step": 273
+    },
+    {
+      "epoch": 1.3365853658536586,
+      "grad_norm": 0.9922477602958679,
+      "learning_rate": 4.784382777341922e-06,
+      "loss": 0.7355,
+      "step": 274
+    },
+    {
+      "epoch": 1.3414634146341464,
+      "grad_norm": 1.2870302200317383,
+      "learning_rate": 4.782823589880729e-06,
+      "loss": 1.0985,
+      "step": 275
+    },
+    {
+      "epoch": 1.346341463414634,
+      "grad_norm": 0.9369707107543945,
+      "learning_rate": 4.7812590411913755e-06,
+      "loss": 0.7729,
+      "step": 276
+    },
+    {
+      "epoch": 1.3512195121951218,
+      "grad_norm": 1.1406941413879395,
+      "learning_rate": 4.779689134948217e-06,
+      "loss": 1.1142,
+      "step": 277
+    },
+    {
+      "epoch": 1.3560975609756096,
+      "grad_norm": 1.1078243255615234,
+      "learning_rate": 4.77811387483819e-06,
+      "loss": 0.5348,
+      "step": 278
+    },
+    {
+      "epoch": 1.3609756097560974,
+      "grad_norm": 1.1744791269302368,
+      "learning_rate": 4.776533264560804e-06,
+      "loss": 0.7202,
+      "step": 279
+    },
+    {
+      "epoch": 1.3658536585365852,
+      "grad_norm": 1.2643998861312866,
+      "learning_rate": 4.774947307828134e-06,
+      "loss": 0.9659,
+      "step": 280
+    },
+    {
+      "epoch": 1.370731707317073,
+      "grad_norm": 1.069071888923645,
+      "learning_rate": 4.773356008364812e-06,
+      "loss": 0.6257,
+      "step": 281
+    },
+    {
+      "epoch": 1.3756097560975609,
+      "grad_norm": 0.8661286234855652,
+      "learning_rate": 4.771759369908017e-06,
+      "loss": 0.4138,
+      "step": 282
+    },
+    {
+      "epoch": 1.3804878048780487,
+      "grad_norm": 0.9560813903808594,
+      "learning_rate": 4.7701573962074635e-06,
+      "loss": 0.6435,
+      "step": 283
+    },
+    {
+      "epoch": 1.3853658536585365,
+      "grad_norm": 1.0229038000106812,
+      "learning_rate": 4.7685500910254015e-06,
+      "loss": 0.5709,
+      "step": 284
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 1.7641900777816772,
+      "learning_rate": 4.766937458136598e-06,
+      "loss": 0.7815,
+      "step": 285
+    },
+    {
+      "epoch": 1.395121951219512,
+      "grad_norm": 1.011093258857727,
+      "learning_rate": 4.765319501328332e-06,
+      "loss": 0.7428,
+      "step": 286
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0194127559661865,
+      "learning_rate": 4.763696224400391e-06,
+      "loss": 0.542,
+      "step": 287
+    },
+    {
+      "epoch": 1.4048780487804877,
+      "grad_norm": 0.9469794034957886,
+      "learning_rate": 4.762067631165049e-06,
+      "loss": 0.5758,
+      "step": 288
+    },
+    {
+      "epoch": 1.4097560975609755,
+      "grad_norm": 1.1177825927734375,
+      "learning_rate": 4.760433725447071e-06,
+      "loss": 0.8141,
+      "step": 289
+    },
+    {
+      "epoch": 1.4146341463414633,
+      "grad_norm": 1.388083577156067,
+      "learning_rate": 4.758794511083697e-06,
+      "loss": 0.8286,
+      "step": 290
+    },
+    {
+      "epoch": 1.4195121951219511,
+      "grad_norm": 1.1623952388763428,
+      "learning_rate": 4.757149991924633e-06,
+      "loss": 0.6733,
+      "step": 291
+    },
+    {
+      "epoch": 1.424390243902439,
+      "grad_norm": 1.328235149383545,
+      "learning_rate": 4.755500171832045e-06,
+      "loss": 0.5397,
+      "step": 292
+    },
+    {
+      "epoch": 1.4292682926829268,
+      "grad_norm": 1.0121268033981323,
+      "learning_rate": 4.753845054680548e-06,
+      "loss": 0.6813,
+      "step": 293
+    },
+    {
+      "epoch": 1.4341463414634146,
+      "grad_norm": 1.0799837112426758,
+      "learning_rate": 4.752184644357197e-06,
+      "loss": 0.5136,
+      "step": 294
+    },
+    {
+      "epoch": 1.4390243902439024,
+      "grad_norm": 1.010602593421936,
+      "learning_rate": 4.750518944761477e-06,
+      "loss": 0.5768,
+      "step": 295
+    },
+    {
+      "epoch": 1.4439024390243902,
+      "grad_norm": 1.1127134561538696,
+      "learning_rate": 4.748847959805297e-06,
+      "loss": 0.5663,
+      "step": 296
+    },
+    {
+      "epoch": 1.448780487804878,
+      "grad_norm": 0.9182597398757935,
+      "learning_rate": 4.7471716934129774e-06,
+      "loss": 0.5599,
+      "step": 297
+    },
+    {
+      "epoch": 1.4536585365853658,
+      "grad_norm": 1.0173683166503906,
+      "learning_rate": 4.745490149521242e-06,
+      "loss": 0.5155,
+      "step": 298
+    },
+    {
+      "epoch": 1.4585365853658536,
+      "grad_norm": 0.9683080911636353,
+      "learning_rate": 4.743803332079209e-06,
+      "loss": 0.5744,
+      "step": 299
+    },
+    {
+      "epoch": 1.4634146341463414,
+      "grad_norm": 0.9615181684494019,
+      "learning_rate": 4.742111245048382e-06,
+      "loss": 0.5961,
+      "step": 300
+    },
+    {
+      "epoch": 1.4682926829268292,
+      "grad_norm": 1.1113585233688354,
+      "learning_rate": 4.740413892402639e-06,
+      "loss": 0.5751,
+      "step": 301
+    },
+    {
+      "epoch": 1.473170731707317,
+      "grad_norm": 1.1533280611038208,
+      "learning_rate": 4.738711278128228e-06,
+      "loss": 0.6668,
+      "step": 302
+    },
+    {
+      "epoch": 1.4780487804878049,
+      "grad_norm": 1.086147665977478,
+      "learning_rate": 4.7370034062237476e-06,
+      "loss": 0.4249,
+      "step": 303
+    },
+    {
+      "epoch": 1.4829268292682927,
+      "grad_norm": 3.0501999855041504,
+      "learning_rate": 4.73529028070015e-06,
+      "loss": 0.6284,
+      "step": 304
+    },
+    {
+      "epoch": 1.4878048780487805,
+      "grad_norm": 1.2545751333236694,
+      "learning_rate": 4.733571905580723e-06,
+      "loss": 0.8968,
+      "step": 305
+    },
+    {
+      "epoch": 1.4926829268292683,
+      "grad_norm": 0.9740838408470154,
+      "learning_rate": 4.731848284901082e-06,
+      "loss": 0.7402,
+      "step": 306
+    },
+    {
+      "epoch": 1.497560975609756,
+      "grad_norm": 1.0430322885513306,
+      "learning_rate": 4.730119422709165e-06,
+      "loss": 0.4697,
+      "step": 307
+    },
+    {
+      "epoch": 1.502439024390244,
+      "grad_norm": 1.2707469463348389,
+      "learning_rate": 4.728385323065215e-06,
+      "loss": 0.7548,
+      "step": 308
+    },
+    {
+      "epoch": 1.5073170731707317,
+      "grad_norm": 0.9956101775169373,
+      "learning_rate": 4.7266459900417815e-06,
+      "loss": 0.5444,
+      "step": 309
+    },
+    {
+      "epoch": 1.5121951219512195,
+      "grad_norm": 1.0116164684295654,
+      "learning_rate": 4.724901427723698e-06,
+      "loss": 0.7939,
+      "step": 310
+    },
+    {
+      "epoch": 1.5170731707317073,
+      "grad_norm": 1.0668343305587769,
+      "learning_rate": 4.723151640208084e-06,
+      "loss": 0.3966,
+      "step": 311
+    },
+    {
+      "epoch": 1.5219512195121951,
+      "grad_norm": 0.9897716045379639,
+      "learning_rate": 4.721396631604327e-06,
+      "loss": 0.4675,
+      "step": 312
+    },
+    {
+      "epoch": 1.526829268292683,
+      "grad_norm": 0.9885281920433044,
+      "learning_rate": 4.7196364060340785e-06,
+      "loss": 0.5411,
+      "step": 313
+    },
+    {
+      "epoch": 1.5317073170731708,
+      "grad_norm": 1.1385118961334229,
+      "learning_rate": 4.7178709676312416e-06,
+      "loss": 0.8041,
+      "step": 314
+    },
+    {
+      "epoch": 1.5365853658536586,
+      "grad_norm": 1.2253623008728027,
+      "learning_rate": 4.716100320541961e-06,
+      "loss": 1.0583,
+      "step": 315
+    },
+    {
+      "epoch": 1.5414634146341464,
+      "grad_norm": 1.1313822269439697,
+      "learning_rate": 4.714324468924614e-06,
+      "loss": 0.7701,
+      "step": 316
+    },
+    {
+      "epoch": 1.5463414634146342,
+      "grad_norm": 1.120343804359436,
+      "learning_rate": 4.712543416949803e-06,
+      "loss": 0.7407,
+      "step": 317
+    },
+    {
+      "epoch": 1.551219512195122,
+      "grad_norm": 1.5084882974624634,
+      "learning_rate": 4.71075716880034e-06,
+      "loss": 0.753,
+      "step": 318
+    },
+    {
+      "epoch": 1.5560975609756098,
+      "grad_norm": 1.3213189840316772,
+      "learning_rate": 4.708965728671243e-06,
+      "loss": 0.8935,
+      "step": 319
+    },
+    {
+      "epoch": 1.5609756097560976,
+      "grad_norm": 1.170746922492981,
+      "learning_rate": 4.7071691007697214e-06,
+      "loss": 0.6782,
+      "step": 320
+    },
+    {
+      "epoch": 1.5658536585365854,
+      "grad_norm": 1.2023199796676636,
+      "learning_rate": 4.705367289315172e-06,
+      "loss": 0.8011,
+      "step": 321
+    },
+    {
+      "epoch": 1.5707317073170732,
+      "grad_norm": 1.0213698148727417,
+      "learning_rate": 4.703560298539158e-06,
+      "loss": 0.5011,
+      "step": 322
+    },
+    {
+      "epoch": 1.575609756097561,
+      "grad_norm": 1.045581579208374,
+      "learning_rate": 4.701748132685415e-06,
+      "loss": 0.5501,
+      "step": 323
+    },
+    {
+      "epoch": 1.5804878048780489,
+      "grad_norm": 0.9141654372215271,
+      "learning_rate": 4.699930796009825e-06,
+      "loss": 0.5551,
+      "step": 324
+    },
+    {
+      "epoch": 1.5853658536585367,
+      "grad_norm": 1.3005549907684326,
+      "learning_rate": 4.698108292780418e-06,
+      "loss": 0.7293,
+      "step": 325
+    },
+    {
+      "epoch": 1.5902439024390245,
+      "grad_norm": 0.985907793045044,
+      "learning_rate": 4.696280627277356e-06,
+      "loss": 0.5366,
+      "step": 326
+    },
+    {
+      "epoch": 1.5951219512195123,
+      "grad_norm": 0.9095384478569031,
+      "learning_rate": 4.6944478037929255e-06,
+      "loss": 0.5508,
+      "step": 327
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.346676230430603,
+      "learning_rate": 4.692609826631525e-06,
+      "loss": 0.6719,
+      "step": 328
+    },
+    {
+      "epoch": 1.604878048780488,
+      "grad_norm": 1.088921308517456,
+      "learning_rate": 4.690766700109659e-06,
+      "loss": 0.4088,
+      "step": 329
+    },
+    {
+      "epoch": 1.6097560975609757,
+      "grad_norm": 0.8905205726623535,
+      "learning_rate": 4.6889184285559234e-06,
+      "loss": 0.4671,
+      "step": 330
+    },
+    {
+      "epoch": 1.6146341463414635,
+      "grad_norm": 1.2066144943237305,
+      "learning_rate": 4.687065016310996e-06,
+      "loss": 0.7891,
+      "step": 331
+    },
+    {
+      "epoch": 1.6195121951219513,
+      "grad_norm": 1.0449296236038208,
+      "learning_rate": 4.685206467727631e-06,
+      "loss": 0.6103,
+      "step": 332
+    },
+    {
+      "epoch": 1.6243902439024391,
+      "grad_norm": 1.15915048122406,
+      "learning_rate": 4.683342787170644e-06,
+      "loss": 0.605,
+      "step": 333
+    },
+    {
+      "epoch": 1.629268292682927,
+      "grad_norm": 1.0918726921081543,
+      "learning_rate": 4.6814739790169006e-06,
+      "loss": 0.5444,
+      "step": 334
+    },
+    {
+      "epoch": 1.6341463414634148,
+      "grad_norm": 1.0298805236816406,
+      "learning_rate": 4.679600047655313e-06,
+      "loss": 0.7902,
+      "step": 335
+    },
+    {
+      "epoch": 1.6390243902439026,
+      "grad_norm": 1.3017504215240479,
+      "learning_rate": 4.6777209974868194e-06,
+      "loss": 1.1195,
+      "step": 336
+    },
+    {
+      "epoch": 1.6439024390243904,
+      "grad_norm": 1.45652174949646,
+      "learning_rate": 4.675836832924387e-06,
+      "loss": 0.6358,
+      "step": 337
+    },
+    {
+      "epoch": 1.6487804878048782,
+      "grad_norm": 0.8610002398490906,
+      "learning_rate": 4.673947558392989e-06,
+      "loss": 0.4196,
+      "step": 338
+    },
+    {
+      "epoch": 1.653658536585366,
+      "grad_norm": 0.8891443014144897,
+      "learning_rate": 4.6720531783296e-06,
+      "loss": 0.5593,
+      "step": 339
+    },
+    {
+      "epoch": 1.6585365853658538,
+      "grad_norm": 0.9679135680198669,
+      "learning_rate": 4.670153697183185e-06,
+      "loss": 0.6149,
+      "step": 340
+    },
+    {
+      "epoch": 1.6634146341463416,
+      "grad_norm": 1.019687294960022,
+      "learning_rate": 4.668249119414692e-06,
+      "loss": 0.5855,
+      "step": 341
+    },
+    {
+      "epoch": 1.6682926829268294,
+      "grad_norm": 0.9645085334777832,
+      "learning_rate": 4.666339449497033e-06,
+      "loss": 0.6321,
+      "step": 342
+    },
+    {
+      "epoch": 1.6731707317073172,
+      "grad_norm": 1.1760913133621216,
+      "learning_rate": 4.664424691915084e-06,
+      "loss": 0.5839,
+      "step": 343
+    },
+    {
+      "epoch": 1.678048780487805,
+      "grad_norm": 1.1706181764602661,
+      "learning_rate": 4.6625048511656675e-06,
+      "loss": 0.5806,
+      "step": 344
+    },
+    {
+      "epoch": 1.6829268292682928,
+      "grad_norm": 1.1575871706008911,
+      "learning_rate": 4.660579931757543e-06,
+      "loss": 0.4945,
+      "step": 345
+    },
+    {
+      "epoch": 1.6878048780487804,
+      "grad_norm": 0.9929284453392029,
+      "learning_rate": 4.6586499382113985e-06,
+      "loss": 0.5662,
+      "step": 346
+    },
+    {
+      "epoch": 1.6926829268292682,
+      "grad_norm": 0.9940921068191528,
+      "learning_rate": 4.6567148750598375e-06,
+      "loss": 0.8166,
+      "step": 347
+    },
+    {
+      "epoch": 1.697560975609756,
+      "grad_norm": 1.1893478631973267,
+      "learning_rate": 4.6547747468473705e-06,
+      "loss": 0.9291,
+      "step": 348
+    },
+    {
+      "epoch": 1.7024390243902439,
+      "grad_norm": 0.8776846528053284,
+      "learning_rate": 4.652829558130404e-06,
+      "loss": 0.4457,
+      "step": 349
+    },
+    {
+      "epoch": 1.7073170731707317,
+      "grad_norm": 1.2133142948150635,
+      "learning_rate": 4.6508793134772265e-06,
+      "loss": 0.6364,
+      "step": 350
+    },
+    {
+      "epoch": 1.7121951219512195,
+      "grad_norm": 0.8865175247192383,
+      "learning_rate": 4.648924017468003e-06,
+      "loss": 0.5514,
+      "step": 351
+    },
+    {
+      "epoch": 1.7170731707317073,
+      "grad_norm": 1.3637226819992065,
+      "learning_rate": 4.646963674694761e-06,
+      "loss": 0.8656,
+      "step": 352
+    },
+    {
+      "epoch": 1.721951219512195,
+      "grad_norm": 0.875629723072052,
+      "learning_rate": 4.64499828976138e-06,
+      "loss": 0.3992,
+      "step": 353
+    },
+    {
+      "epoch": 1.726829268292683,
+      "grad_norm": 1.0361976623535156,
+      "learning_rate": 4.64302786728358e-06,
+      "loss": 0.5056,
+      "step": 354
+    },
+    {
+      "epoch": 1.7317073170731707,
+      "grad_norm": 0.9224256873130798,
+      "learning_rate": 4.641052411888913e-06,
+      "loss": 0.5352,
+      "step": 355
+    },
+    {
+      "epoch": 1.7365853658536585,
+      "grad_norm": 1.2034342288970947,
+      "learning_rate": 4.6390719282167515e-06,
+      "loss": 0.4753,
+      "step": 356
+    },
+    {
+      "epoch": 1.7414634146341463,
+      "grad_norm": 1.056547999382019,
+      "learning_rate": 4.637086420918276e-06,
+      "loss": 0.7975,
+      "step": 357
+    },
+    {
+      "epoch": 1.7463414634146341,
+      "grad_norm": 0.9398707151412964,
+      "learning_rate": 4.635095894656465e-06,
+      "loss": 0.6944,
+      "step": 358
+    },
+    {
+      "epoch": 1.751219512195122,
+      "grad_norm": 1.3796380758285522,
+      "learning_rate": 4.633100354106085e-06,
+      "loss": 0.377,
+      "step": 359
+    },
+    {
+      "epoch": 1.7560975609756098,
+      "grad_norm": 1.1999366283416748,
+      "learning_rate": 4.631099803953677e-06,
+      "loss": 0.8488,
+      "step": 360
+    },
+    {
+      "epoch": 1.7609756097560976,
+      "grad_norm": 1.1613068580627441,
+      "learning_rate": 4.629094248897546e-06,
+      "loss": 0.4856,
+      "step": 361
+    },
+    {
+      "epoch": 1.7658536585365854,
+      "grad_norm": 1.407758116722107,
+      "learning_rate": 4.627083693647757e-06,
+      "loss": 0.5932,
+      "step": 362
+    },
+    {
+      "epoch": 1.7707317073170732,
+      "grad_norm": 1.1262705326080322,
+      "learning_rate": 4.625068142926111e-06,
+      "loss": 0.862,
+      "step": 363
+    },
+    {
+      "epoch": 1.775609756097561,
+      "grad_norm": 1.0207730531692505,
+      "learning_rate": 4.623047601466144e-06,
+      "loss": 0.7695,
+      "step": 364
+    },
+    {
+      "epoch": 1.7804878048780488,
+      "grad_norm": 1.0084431171417236,
+      "learning_rate": 4.621022074013114e-06,
+      "loss": 0.6608,
+      "step": 365
+    },
+    {
+      "epoch": 1.7853658536585366,
+      "grad_norm": 1.044545292854309,
+      "learning_rate": 4.618991565323987e-06,
+      "loss": 0.5231,
+      "step": 366
+    },
+    {
+      "epoch": 1.7902439024390244,
+      "grad_norm": 0.9962389469146729,
+      "learning_rate": 4.616956080167426e-06,
+      "loss": 0.5736,
+      "step": 367
+    },
+    {
+      "epoch": 1.7951219512195122,
+      "grad_norm": 1.3212288618087769,
+      "learning_rate": 4.614915623323786e-06,
+      "loss": 0.9586,
+      "step": 368
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9554499983787537,
+      "learning_rate": 4.612870199585092e-06,
+      "loss": 0.4346,
+      "step": 369
+    },
+    {
+      "epoch": 1.8048780487804879,
+      "grad_norm": 1.082366943359375,
+      "learning_rate": 4.610819813755038e-06,
+      "loss": 0.505,
+      "step": 370
+    },
+    {
+      "epoch": 1.8097560975609757,
+      "grad_norm": 1.225993037223816,
+      "learning_rate": 4.608764470648971e-06,
+      "loss": 0.4801,
+      "step": 371
+    },
+    {
+      "epoch": 1.8146341463414632,
+      "grad_norm": 1.0881706476211548,
+      "learning_rate": 4.606704175093879e-06,
+      "loss": 0.4478,
+      "step": 372
+    },
+    {
+      "epoch": 1.819512195121951,
+      "grad_norm": 1.114046335220337,
+      "learning_rate": 4.604638931928383e-06,
+      "loss": 0.8015,
+      "step": 373
+    },
+    {
+      "epoch": 1.8243902439024389,
+      "grad_norm": 0.9838706851005554,
+      "learning_rate": 4.602568746002718e-06,
+      "loss": 0.5204,
+      "step": 374
+    },
+    {
+      "epoch": 1.8292682926829267,
+      "grad_norm": 1.038713812828064,
+      "learning_rate": 4.600493622178734e-06,
+      "loss": 0.8388,
+      "step": 375
+    },
+    {
+      "epoch": 1.8341463414634145,
+      "grad_norm": 1.0684245824813843,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.5744,
+      "step": 376
+    },
+    {
+      "epoch": 1.8390243902439023,
+      "grad_norm": 0.8907456994056702,
+      "learning_rate": 4.596328580341169e-06,
+      "loss": 0.5621,
+      "step": 377
+    },
+    {
+      "epoch": 1.84390243902439,
+      "grad_norm": 0.9921515583992004,
+      "learning_rate": 4.5942386721092195e-06,
+      "loss": 0.6967,
+      "step": 378
+    },
+    {
+      "epoch": 1.848780487804878,
+      "grad_norm": 1.1683647632598877,
+      "learning_rate": 4.592143845542189e-06,
+      "loss": 0.6489,
+      "step": 379
+    },
+    {
+      "epoch": 1.8536585365853657,
+      "grad_norm": 1.0409291982650757,
+      "learning_rate": 4.590044105559797e-06,
+      "loss": 0.8945,
+      "step": 380
+    },
+    {
+      "epoch": 1.8585365853658535,
+      "grad_norm": 1.0684564113616943,
+      "learning_rate": 4.587939457093296e-06,
+      "loss": 0.5476,
+      "step": 381
+    },
+    {
+      "epoch": 1.8634146341463413,
+      "grad_norm": 1.3661733865737915,
+      "learning_rate": 4.585829905085468e-06,
+      "loss": 0.6763,
+      "step": 382
+    },
+    {
+      "epoch": 1.8682926829268292,
+      "grad_norm": 1.1465227603912354,
+      "learning_rate": 4.5837154544906135e-06,
+      "loss": 0.7817,
+      "step": 383
+    },
+    {
+      "epoch": 1.873170731707317,
+      "grad_norm": 1.0232677459716797,
+      "learning_rate": 4.581596110274535e-06,
+      "loss": 0.7276,
+      "step": 384
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 1.0359809398651123,
+      "learning_rate": 4.579471877414527e-06,
+      "loss": 0.9635,
+      "step": 385
+    },
+    {
+      "epoch": 1.8829268292682926,
+      "grad_norm": 1.2472409009933472,
+      "learning_rate": 4.577342760899368e-06,
+      "loss": 0.6782,
+      "step": 386
+    },
+    {
+      "epoch": 1.8878048780487804,
+      "grad_norm": 0.9241912961006165,
+      "learning_rate": 4.575208765729302e-06,
+      "loss": 0.5327,
+      "step": 387
+    },
+    {
+      "epoch": 1.8926829268292682,
+      "grad_norm": 1.3745805025100708,
+      "learning_rate": 4.573069896916035e-06,
+      "loss": 0.8436,
+      "step": 388
+    },
+    {
+      "epoch": 1.897560975609756,
+      "grad_norm": 1.0620322227478027,
+      "learning_rate": 4.5709261594827125e-06,
+      "loss": 0.659,
+      "step": 389
+    },
+    {
+      "epoch": 1.9024390243902438,
+      "grad_norm": 1.3068687915802002,
+      "learning_rate": 4.568777558463922e-06,
+      "loss": 0.5219,
+      "step": 390
+    },
+    {
+      "epoch": 1.9073170731707316,
+      "grad_norm": 1.0368698835372925,
+      "learning_rate": 4.566624098905665e-06,
+      "loss": 0.7099,
+      "step": 391
+    },
+    {
+      "epoch": 1.9121951219512194,
+      "grad_norm": 1.0876407623291016,
+      "learning_rate": 4.564465785865359e-06,
+      "loss": 0.5276,
+      "step": 392
+    },
+    {
+      "epoch": 1.9170731707317072,
+      "grad_norm": 0.9230280518531799,
+      "learning_rate": 4.56230262441182e-06,
+      "loss": 0.4292,
+      "step": 393
+    },
+    {
+      "epoch": 1.921951219512195,
+      "grad_norm": 0.9985240697860718,
+      "learning_rate": 4.560134619625247e-06,
+      "loss": 0.6602,
+      "step": 394
+    },
+    {
+      "epoch": 1.9268292682926829,
+      "grad_norm": 0.8872730135917664,
+      "learning_rate": 4.5579617765972155e-06,
+      "loss": 0.5711,
+      "step": 395
+    },
+    {
+      "epoch": 1.9317073170731707,
+      "grad_norm": 0.9474479556083679,
+      "learning_rate": 4.555784100430662e-06,
+      "loss": 0.5299,
+      "step": 396
+    },
+    {
+      "epoch": 1.9365853658536585,
+      "grad_norm": 1.0377358198165894,
+      "learning_rate": 4.553601596239877e-06,
+      "loss": 0.4649,
+      "step": 397
+    },
+    {
+      "epoch": 1.9414634146341463,
+      "grad_norm": 1.0765342712402344,
+      "learning_rate": 4.551414269150489e-06,
+      "loss": 0.5578,
+      "step": 398
+    },
+    {
+      "epoch": 1.946341463414634,
+      "grad_norm": 1.0844486951828003,
+      "learning_rate": 4.54922212429945e-06,
+      "loss": 0.5486,
+      "step": 399
+    },
+    {
+      "epoch": 1.951219512195122,
+      "grad_norm": 1.1640657186508179,
+      "learning_rate": 4.547025166835027e-06,
+      "loss": 0.7462,
+      "step": 400
+    },
+    {
+      "epoch": 1.9560975609756097,
+      "grad_norm": 1.2733076810836792,
+      "learning_rate": 4.544823401916794e-06,
+      "loss": 0.881,
+      "step": 401
+    },
+    {
+      "epoch": 1.9609756097560975,
+      "grad_norm": 1.155869960784912,
+      "learning_rate": 4.542616834715612e-06,
+      "loss": 0.6374,
+      "step": 402
+    },
+    {
+      "epoch": 1.9658536585365853,
+      "grad_norm": 0.8623374700546265,
+      "learning_rate": 4.540405470413618e-06,
+      "loss": 0.4188,
+      "step": 403
+    },
+    {
+      "epoch": 1.9707317073170731,
+      "grad_norm": 1.0502699613571167,
+      "learning_rate": 4.53818931420422e-06,
+      "loss": 0.7233,
+      "step": 404
+    },
+    {
+      "epoch": 1.975609756097561,
+      "grad_norm": 0.8630202412605286,
+      "learning_rate": 4.535968371292076e-06,
+      "loss": 0.5896,
+      "step": 405
+    },
+    {
+      "epoch": 1.9804878048780488,
+      "grad_norm": 1.1017824411392212,
+      "learning_rate": 4.533742646893086e-06,
+      "loss": 0.6971,
+      "step": 406
+    },
+    {
+      "epoch": 1.9853658536585366,
+      "grad_norm": 0.9387734532356262,
+      "learning_rate": 4.531512146234383e-06,
+      "loss": 0.6718,
+      "step": 407
+    },
+    {
+      "epoch": 1.9902439024390244,
+      "grad_norm": 1.0347439050674438,
+      "learning_rate": 4.529276874554312e-06,
+      "loss": 0.8829,
+      "step": 408
+    },
+    {
+      "epoch": 1.9951219512195122,
+      "grad_norm": 1.0173542499542236,
+      "learning_rate": 4.527036837102426e-06,
+      "loss": 0.5154,
+      "step": 409
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1297523975372314,
+      "learning_rate": 4.524792039139471e-06,
+      "loss": 0.7721,
+      "step": 410
+    },
+    {
+      "epoch": 2.004878048780488,
+      "grad_norm": 0.9763960242271423,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.3978,
+      "step": 411
+    },
+    {
+      "epoch": 2.0097560975609756,
+      "grad_norm": 0.9650730490684509,
+      "learning_rate": 4.520288182779214e-06,
+      "loss": 0.6826,
+      "step": 412
+    },
+    {
+      "epoch": 2.0146341463414634,
+      "grad_norm": 0.7682514190673828,
+      "learning_rate": 4.518029134959253e-06,
+      "loss": 0.351,
+      "step": 413
+    },
+    {
+      "epoch": 2.0195121951219512,
+      "grad_norm": 1.0681227445602417,
+      "learning_rate": 4.515765347782878e-06,
+      "loss": 0.6467,
+      "step": 414
+    },
+    {
+      "epoch": 2.024390243902439,
+      "grad_norm": 1.2426350116729736,
+      "learning_rate": 4.5134968265666085e-06,
+      "loss": 0.8831,
+      "step": 415
+    },
+    {
+      "epoch": 2.029268292682927,
+      "grad_norm": 0.9794759154319763,
+      "learning_rate": 4.511223576638084e-06,
+      "loss": 0.6419,
+      "step": 416
+    },
+    {
+      "epoch": 2.0341463414634147,
+      "grad_norm": 1.119649887084961,
+      "learning_rate": 4.508945603336049e-06,
+      "loss": 0.8023,
+      "step": 417
+    },
+    {
+      "epoch": 2.0390243902439025,
+      "grad_norm": 0.858971118927002,
+      "learning_rate": 4.50666291201034e-06,
+      "loss": 0.4974,
+      "step": 418
+    },
+    {
+      "epoch": 2.0439024390243903,
+      "grad_norm": 1.0508594512939453,
+      "learning_rate": 4.504375508021876e-06,
+      "loss": 0.3603,
+      "step": 419
+    },
+    {
+      "epoch": 2.048780487804878,
+      "grad_norm": 0.9566763043403625,
+      "learning_rate": 4.50208339674264e-06,
+      "loss": 0.7813,
+      "step": 420
+    },
+    {
+      "epoch": 2.053658536585366,
+      "grad_norm": 1.074040412902832,
+      "learning_rate": 4.499786583555675e-06,
+      "loss": 0.8065,
+      "step": 421
+    },
+    {
+      "epoch": 2.0585365853658537,
+      "grad_norm": 0.8816580772399902,
+      "learning_rate": 4.497485073855061e-06,
+      "loss": 0.4439,
+      "step": 422
+    },
+    {
+      "epoch": 2.0634146341463415,
+      "grad_norm": 1.0733896493911743,
+      "learning_rate": 4.495178873045913e-06,
+      "loss": 0.4481,
+      "step": 423
+    },
+    {
+      "epoch": 2.0682926829268293,
+      "grad_norm": 0.9010451436042786,
+      "learning_rate": 4.4928679865443605e-06,
+      "loss": 0.4407,
+      "step": 424
+    },
+    {
+      "epoch": 2.073170731707317,
+      "grad_norm": 0.9359092712402344,
+      "learning_rate": 4.4905524197775366e-06,
+      "loss": 0.5847,
+      "step": 425
+    },
+    {
+      "epoch": 2.078048780487805,
+      "grad_norm": 0.9389141798019409,
+      "learning_rate": 4.4882321781835666e-06,
+      "loss": 0.5377,
+      "step": 426
+    },
+    {
+      "epoch": 2.0829268292682928,
+      "grad_norm": 0.89751797914505,
+      "learning_rate": 4.4859072672115565e-06,
+      "loss": 0.3879,
+      "step": 427
+    },
+    {
+      "epoch": 2.0878048780487806,
+      "grad_norm": 1.0244700908660889,
+      "learning_rate": 4.483577692321577e-06,
+      "loss": 0.8237,
+      "step": 428
+    },
+    {
+      "epoch": 2.0926829268292684,
+      "grad_norm": 1.0703409910202026,
+      "learning_rate": 4.481243458984651e-06,
+      "loss": 0.5826,
+      "step": 429
+    },
+    {
+      "epoch": 2.097560975609756,
+      "grad_norm": 0.9995833039283752,
+      "learning_rate": 4.478904572682743e-06,
+      "loss": 0.6091,
+      "step": 430
+    },
+    {
+      "epoch": 2.102439024390244,
+      "grad_norm": 0.8649471998214722,
+      "learning_rate": 4.476561038908745e-06,
+      "loss": 0.4863,
+      "step": 431
+    },
+    {
+      "epoch": 2.107317073170732,
+      "grad_norm": 1.1670926809310913,
+      "learning_rate": 4.474212863166464e-06,
+      "loss": 0.6584,
+      "step": 432
+    },
+    {
+      "epoch": 2.1121951219512196,
+      "grad_norm": 1.2743312120437622,
+      "learning_rate": 4.471860050970608e-06,
+      "loss": 0.6777,
+      "step": 433
+    },
+    {
+      "epoch": 2.1170731707317074,
+      "grad_norm": 1.2678401470184326,
+      "learning_rate": 4.469502607846774e-06,
+      "loss": 0.9609,
+      "step": 434
+    },
+    {
+      "epoch": 2.1219512195121952,
+      "grad_norm": 0.9796558618545532,
+      "learning_rate": 4.467140539331434e-06,
+      "loss": 0.4574,
+      "step": 435
+    },
+    {
+      "epoch": 2.126829268292683,
+      "grad_norm": 1.0830684900283813,
+      "learning_rate": 4.464773850971924e-06,
+      "loss": 0.3067,
+      "step": 436
+    },
+    {
+      "epoch": 2.131707317073171,
+      "grad_norm": 1.002589464187622,
+      "learning_rate": 4.46240254832643e-06,
+      "loss": 0.5383,
+      "step": 437
+    },
+    {
+      "epoch": 2.1365853658536587,
+      "grad_norm": 1.1145734786987305,
+      "learning_rate": 4.460026636963971e-06,
+      "loss": 0.6173,
+      "step": 438
+    },
+    {
+      "epoch": 2.1414634146341465,
+      "grad_norm": 0.94740891456604,
+      "learning_rate": 4.4576461224643965e-06,
+      "loss": 0.4991,
+      "step": 439
+    },
+    {
+      "epoch": 2.1463414634146343,
+      "grad_norm": 0.8613864183425903,
+      "learning_rate": 4.455261010418359e-06,
+      "loss": 0.3956,
+      "step": 440
+    },
+    {
+      "epoch": 2.151219512195122,
+      "grad_norm": 0.9509091377258301,
+      "learning_rate": 4.452871306427314e-06,
+      "loss": 0.7165,
+      "step": 441
+    },
+    {
+      "epoch": 2.15609756097561,
+      "grad_norm": 1.259364128112793,
+      "learning_rate": 4.450477016103498e-06,
+      "loss": 0.5682,
+      "step": 442
+    },
+    {
+      "epoch": 2.1609756097560977,
+      "grad_norm": 1.1716279983520508,
+      "learning_rate": 4.4480781450699205e-06,
+      "loss": 0.4917,
+      "step": 443
+    },
+    {
+      "epoch": 2.1658536585365855,
+      "grad_norm": 1.0395866632461548,
+      "learning_rate": 4.4456746989603464e-06,
+      "loss": 0.4338,
+      "step": 444
+    },
+    {
+      "epoch": 2.1707317073170733,
+      "grad_norm": 1.232602834701538,
+      "learning_rate": 4.443266683419289e-06,
+      "loss": 0.6356,
+      "step": 445
+    },
+    {
+      "epoch": 2.175609756097561,
+      "grad_norm": 1.254172921180725,
+      "learning_rate": 4.440854104101988e-06,
+      "loss": 0.472,
+      "step": 446
+    },
+    {
+      "epoch": 2.180487804878049,
+      "grad_norm": 1.2319004535675049,
+      "learning_rate": 4.438436966674406e-06,
+      "loss": 0.6408,
+      "step": 447
+    },
+    {
+      "epoch": 2.1853658536585368,
+      "grad_norm": 1.0100780725479126,
+      "learning_rate": 4.436015276813208e-06,
+      "loss": 0.4524,
+      "step": 448
+    },
+    {
+      "epoch": 2.1902439024390246,
+      "grad_norm": 0.9656887650489807,
+      "learning_rate": 4.4335890402057505e-06,
+      "loss": 0.5999,
+      "step": 449
+    },
+    {
+      "epoch": 2.1951219512195124,
+      "grad_norm": 1.1730879545211792,
+      "learning_rate": 4.431158262550067e-06,
+      "loss": 0.5953,
+      "step": 450
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9255422949790955,
+      "learning_rate": 4.428722949554858e-06,
+      "loss": 0.3487,
+      "step": 451
+    },
+    {
+      "epoch": 2.204878048780488,
+      "grad_norm": 0.9878072142601013,
+      "learning_rate": 4.426283106939474e-06,
+      "loss": 0.4937,
+      "step": 452
+    },
+    {
+      "epoch": 2.209756097560976,
+      "grad_norm": 0.982023298740387,
+      "learning_rate": 4.423838740433903e-06,
+      "loss": 0.6299,
+      "step": 453
+    },
+    {
+      "epoch": 2.2146341463414636,
+      "grad_norm": 0.7727266550064087,
+      "learning_rate": 4.4213898557787586e-06,
+      "loss": 0.2789,
+      "step": 454
+    },
+    {
+      "epoch": 2.2195121951219514,
+      "grad_norm": 1.5341951847076416,
+      "learning_rate": 4.4189364587252636e-06,
+      "loss": 0.8498,
+      "step": 455
+    },
+    {
+      "epoch": 2.2243902439024392,
+      "grad_norm": 1.1611250638961792,
+      "learning_rate": 4.416478555035241e-06,
+      "loss": 0.4075,
+      "step": 456
+    },
+    {
+      "epoch": 2.229268292682927,
+      "grad_norm": 1.0459867715835571,
+      "learning_rate": 4.4140161504810935e-06,
+      "loss": 0.4946,
+      "step": 457
+    },
+    {
+      "epoch": 2.234146341463415,
+      "grad_norm": 0.9366090297698975,
+      "learning_rate": 4.4115492508457986e-06,
+      "loss": 0.3479,
+      "step": 458
+    },
+    {
+      "epoch": 2.2390243902439027,
+      "grad_norm": 1.0325448513031006,
+      "learning_rate": 4.409077861922887e-06,
+      "loss": 0.5437,
+      "step": 459
+    },
+    {
+      "epoch": 2.2439024390243905,
+      "grad_norm": 0.9326527118682861,
+      "learning_rate": 4.406601989516435e-06,
+      "loss": 0.4594,
+      "step": 460
+    },
+    {
+      "epoch": 2.2487804878048783,
+      "grad_norm": 0.7127180099487305,
+      "learning_rate": 4.404121639441047e-06,
+      "loss": 0.3067,
+      "step": 461
+    },
+    {
+      "epoch": 2.253658536585366,
+      "grad_norm": 1.0416815280914307,
+      "learning_rate": 4.401636817521843e-06,
+      "loss": 0.5402,
+      "step": 462
+    },
+    {
+      "epoch": 2.258536585365854,
+      "grad_norm": 1.8258185386657715,
+      "learning_rate": 4.399147529594447e-06,
+      "loss": 0.3964,
+      "step": 463
+    },
+    {
+      "epoch": 2.2634146341463417,
+      "grad_norm": 0.9795071482658386,
+      "learning_rate": 4.3966537815049686e-06,
+      "loss": 0.5118,
+      "step": 464
+    },
+    {
+      "epoch": 2.2682926829268295,
+      "grad_norm": 1.1920483112335205,
+      "learning_rate": 4.394155579109994e-06,
+      "loss": 0.6511,
+      "step": 465
+    },
+    {
+      "epoch": 2.2731707317073173,
+      "grad_norm": 1.361159324645996,
+      "learning_rate": 4.391652928276572e-06,
+      "loss": 0.6874,
+      "step": 466
+    },
+    {
+      "epoch": 2.278048780487805,
+      "grad_norm": 0.9973228573799133,
+      "learning_rate": 4.389145834882195e-06,
+      "loss": 0.6057,
+      "step": 467
+    },
+    {
+      "epoch": 2.2829268292682925,
+      "grad_norm": 1.3514574766159058,
+      "learning_rate": 4.386634304814789e-06,
+      "loss": 0.4762,
+      "step": 468
+    },
+    {
+      "epoch": 2.2878048780487803,
+      "grad_norm": 1.2089687585830688,
+      "learning_rate": 4.384118343972704e-06,
+      "loss": 0.689,
+      "step": 469
+    },
+    {
+      "epoch": 2.292682926829268,
+      "grad_norm": 0.9414058327674866,
+      "learning_rate": 4.381597958264692e-06,
+      "loss": 0.7257,
+      "step": 470
+    },
+    {
+      "epoch": 2.297560975609756,
+      "grad_norm": 1.0120850801467896,
+      "learning_rate": 4.379073153609896e-06,
+      "loss": 0.7515,
+      "step": 471
+    },
+    {
+      "epoch": 2.3024390243902437,
+      "grad_norm": 1.586024522781372,
+      "learning_rate": 4.37654393593784e-06,
+      "loss": 0.64,
+      "step": 472
+    },
+    {
+      "epoch": 2.3073170731707315,
+      "grad_norm": 0.9766375422477722,
+      "learning_rate": 4.3740103111884096e-06,
+      "loss": 0.5162,
+      "step": 473
+    },
+    {
+      "epoch": 2.3121951219512193,
+      "grad_norm": 1.001685380935669,
+      "learning_rate": 4.371472285311842e-06,
+      "loss": 0.4464,
+      "step": 474
+    },
+    {
+      "epoch": 2.317073170731707,
+      "grad_norm": 0.7923868894577026,
+      "learning_rate": 4.368929864268709e-06,
+      "loss": 0.365,
+      "step": 475
+    },
+    {
+      "epoch": 2.321951219512195,
+      "grad_norm": 1.0744857788085938,
+      "learning_rate": 4.366383054029907e-06,
+      "loss": 0.6253,
+      "step": 476
+    },
+    {
+      "epoch": 2.3268292682926828,
+      "grad_norm": 0.9156374931335449,
+      "learning_rate": 4.363831860576638e-06,
+      "loss": 0.6512,
+      "step": 477
+    },
+    {
+      "epoch": 2.3317073170731706,
+      "grad_norm": 1.0944534540176392,
+      "learning_rate": 4.361276289900396e-06,
+      "loss": 0.5627,
+      "step": 478
+    },
+    {
+      "epoch": 2.3365853658536584,
+      "grad_norm": 1.1438108682632446,
+      "learning_rate": 4.358716348002962e-06,
+      "loss": 0.7402,
+      "step": 479
+    },
+    {
+      "epoch": 2.341463414634146,
+      "grad_norm": 1.1678388118743896,
+      "learning_rate": 4.356152040896376e-06,
+      "loss": 0.512,
+      "step": 480
+    },
+    {
+      "epoch": 2.346341463414634,
+      "grad_norm": 0.8130245208740234,
+      "learning_rate": 4.3535833746029335e-06,
+      "loss": 0.3934,
+      "step": 481
+    },
+    {
+      "epoch": 2.351219512195122,
+      "grad_norm": 1.229127287864685,
+      "learning_rate": 4.351010355155165e-06,
+      "loss": 0.4782,
+      "step": 482
+    },
+    {
+      "epoch": 2.3560975609756096,
+      "grad_norm": 0.9830904006958008,
+      "learning_rate": 4.348432988595828e-06,
+      "loss": 0.3879,
+      "step": 483
+    },
+    {
+      "epoch": 2.3609756097560974,
+      "grad_norm": 1.3584911823272705,
+      "learning_rate": 4.345851280977885e-06,
+      "loss": 0.7305,
+      "step": 484
+    },
+    {
+      "epoch": 2.3658536585365852,
+      "grad_norm": 1.0106158256530762,
+      "learning_rate": 4.343265238364496e-06,
+      "loss": 0.4247,
+      "step": 485
+    },
+    {
+      "epoch": 2.370731707317073,
+      "grad_norm": 1.1921676397323608,
+      "learning_rate": 4.340674866829001e-06,
+      "loss": 0.5446,
+      "step": 486
+    },
+    {
+      "epoch": 2.375609756097561,
+      "grad_norm": 1.1733497381210327,
+      "learning_rate": 4.338080172454908e-06,
+      "loss": 0.8386,
+      "step": 487
+    },
+    {
+      "epoch": 2.3804878048780487,
+      "grad_norm": 0.8914453387260437,
+      "learning_rate": 4.335481161335875e-06,
+      "loss": 0.5393,
+      "step": 488
+    },
+    {
+      "epoch": 2.3853658536585365,
+      "grad_norm": 1.122434377670288,
+      "learning_rate": 4.332877839575699e-06,
+      "loss": 0.4645,
+      "step": 489
+    },
+    {
+      "epoch": 2.3902439024390243,
+      "grad_norm": 1.078754186630249,
+      "learning_rate": 4.330270213288301e-06,
+      "loss": 0.6284,
+      "step": 490
+    },
+    {
+      "epoch": 2.395121951219512,
+      "grad_norm": 1.0400331020355225,
+      "learning_rate": 4.32765828859771e-06,
+      "loss": 0.8431,
+      "step": 491
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9895585775375366,
+      "learning_rate": 4.325042071638051e-06,
+      "loss": 0.6601,
+      "step": 492
+    },
+    {
+      "epoch": 2.4048780487804877,
+      "grad_norm": 1.1766878366470337,
+      "learning_rate": 4.322421568553529e-06,
+      "loss": 0.4766,
+      "step": 493
+    },
+    {
+      "epoch": 2.4097560975609755,
+      "grad_norm": 0.9032670855522156,
+      "learning_rate": 4.319796785498416e-06,
+      "loss": 0.4363,
+      "step": 494
+    },
+    {
+      "epoch": 2.4146341463414633,
+      "grad_norm": 0.9736960530281067,
+      "learning_rate": 4.317167728637032e-06,
+      "loss": 0.6109,
+      "step": 495
+    },
+    {
+      "epoch": 2.419512195121951,
+      "grad_norm": 0.7923660278320312,
+      "learning_rate": 4.314534404143738e-06,
+      "loss": 0.4987,
+      "step": 496
+    },
+    {
+      "epoch": 2.424390243902439,
+      "grad_norm": 0.9087777137756348,
+      "learning_rate": 4.3118968182029155e-06,
+      "loss": 0.6411,
+      "step": 497
+    },
+    {
+      "epoch": 2.4292682926829268,
+      "grad_norm": 0.851117730140686,
+      "learning_rate": 4.3092549770089566e-06,
+      "loss": 0.3541,
+      "step": 498
+    },
+    {
+      "epoch": 2.4341463414634146,
+      "grad_norm": 0.9581378102302551,
+      "learning_rate": 4.306608886766243e-06,
+      "loss": 0.5448,
+      "step": 499
+    },
+    {
+      "epoch": 2.4390243902439024,
+      "grad_norm": 1.081851601600647,
+      "learning_rate": 4.303958553689137e-06,
+      "loss": 0.5593,
+      "step": 500
+    },
+    {
+      "epoch": 2.44390243902439,
+      "grad_norm": 1.1111576557159424,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.7566,
+      "step": 501
+    },
+    {
+      "epoch": 2.448780487804878,
+      "grad_norm": 1.0168198347091675,
+      "learning_rate": 4.2986451839390105e-06,
+      "loss": 0.3996,
+      "step": 502
+    },
+    {
+      "epoch": 2.453658536585366,
+      "grad_norm": 0.9412428140640259,
+      "learning_rate": 4.295982159744476e-06,
+      "loss": 0.5602,
+      "step": 503
+    },
+    {
+      "epoch": 2.4585365853658536,
+      "grad_norm": 1.0618679523468018,
+      "learning_rate": 4.293314917672498e-06,
+      "loss": 0.6466,
+      "step": 504
+    },
+    {
+      "epoch": 2.4634146341463414,
+      "grad_norm": 0.8414422273635864,
+      "learning_rate": 4.290643463987114e-06,
+      "loss": 0.357,
+      "step": 505
+    },
+    {
+      "epoch": 2.4682926829268292,
+      "grad_norm": 0.93071448802948,
+      "learning_rate": 4.287967804962252e-06,
+      "loss": 0.4179,
+      "step": 506
+    },
+    {
+      "epoch": 2.473170731707317,
+      "grad_norm": 0.9793124794960022,
+      "learning_rate": 4.285287946881718e-06,
+      "loss": 0.2698,
+      "step": 507
+    },
+    {
+      "epoch": 2.478048780487805,
+      "grad_norm": 0.9545714259147644,
+      "learning_rate": 4.282603896039178e-06,
+      "loss": 0.6855,
+      "step": 508
+    },
+    {
+      "epoch": 2.4829268292682927,
+      "grad_norm": 1.1140731573104858,
+      "learning_rate": 4.279915658738145e-06,
+      "loss": 0.5114,
+      "step": 509
+    },
+    {
+      "epoch": 2.4878048780487805,
+      "grad_norm": 1.0547738075256348,
+      "learning_rate": 4.277223241291966e-06,
+      "loss": 0.7367,
+      "step": 510
+    },
+    {
+      "epoch": 2.4926829268292683,
+      "grad_norm": 0.8616530895233154,
+      "learning_rate": 4.274526650023801e-06,
+      "loss": 0.5337,
+      "step": 511
+    },
+    {
+      "epoch": 2.497560975609756,
+      "grad_norm": 1.3530237674713135,
+      "learning_rate": 4.271825891266617e-06,
+      "loss": 0.5597,
+      "step": 512
+    },
+    {
+      "epoch": 2.502439024390244,
+      "grad_norm": 1.4124853610992432,
+      "learning_rate": 4.269120971363164e-06,
+      "loss": 0.8461,
+      "step": 513
+    },
+    {
+      "epoch": 2.5073170731707317,
+      "grad_norm": 1.0308994054794312,
+      "learning_rate": 4.266411896665967e-06,
+      "loss": 0.4304,
+      "step": 514
+    },
+    {
+      "epoch": 2.5121951219512195,
+      "grad_norm": 1.0831527709960938,
+      "learning_rate": 4.263698673537309e-06,
+      "loss": 0.5428,
+      "step": 515
+    },
+    {
+      "epoch": 2.5170731707317073,
+      "grad_norm": 1.2008475065231323,
+      "learning_rate": 4.260981308349214e-06,
+      "loss": 0.6922,
+      "step": 516
+    },
+    {
+      "epoch": 2.521951219512195,
+      "grad_norm": 1.1929224729537964,
+      "learning_rate": 4.258259807483434e-06,
+      "loss": 0.5716,
+      "step": 517
+    },
+    {
+      "epoch": 2.526829268292683,
+      "grad_norm": 1.016539216041565,
+      "learning_rate": 4.255534177331435e-06,
+      "loss": 0.5806,
+      "step": 518
+    },
+    {
+      "epoch": 2.5317073170731708,
+      "grad_norm": 1.018875002861023,
+      "learning_rate": 4.252804424294378e-06,
+      "loss": 0.5581,
+      "step": 519
+    },
+    {
+      "epoch": 2.5365853658536586,
+      "grad_norm": 0.9992810487747192,
+      "learning_rate": 4.25007055478311e-06,
+      "loss": 0.6786,
+      "step": 520
+    },
+    {
+      "epoch": 2.5414634146341464,
+      "grad_norm": 1.1207003593444824,
+      "learning_rate": 4.247332575218144e-06,
+      "loss": 0.4548,
+      "step": 521
+    },
+    {
+      "epoch": 2.546341463414634,
+      "grad_norm": 1.1575409173965454,
+      "learning_rate": 4.244590492029643e-06,
+      "loss": 0.6846,
+      "step": 522
+    },
+    {
+      "epoch": 2.551219512195122,
+      "grad_norm": 0.9805243015289307,
+      "learning_rate": 4.241844311657411e-06,
+      "loss": 0.4301,
+      "step": 523
+    },
+    {
+      "epoch": 2.55609756097561,
+      "grad_norm": 0.9760981202125549,
+      "learning_rate": 4.239094040550875e-06,
+      "loss": 0.3545,
+      "step": 524
+    },
+    {
+      "epoch": 2.5609756097560976,
+      "grad_norm": 0.8702017664909363,
+      "learning_rate": 4.236339685169065e-06,
+      "loss": 0.5429,
+      "step": 525
+    },
+    {
+      "epoch": 2.5658536585365854,
+      "grad_norm": 1.0681567192077637,
+      "learning_rate": 4.233581251980604e-06,
+      "loss": 0.3289,
+      "step": 526
+    },
+    {
+      "epoch": 2.5707317073170732,
+      "grad_norm": 1.1807548999786377,
+      "learning_rate": 4.230818747463696e-06,
+      "loss": 0.4876,
+      "step": 527
+    },
+    {
+      "epoch": 2.575609756097561,
+      "grad_norm": 0.9812930226325989,
+      "learning_rate": 4.228052178106101e-06,
+      "loss": 0.5025,
+      "step": 528
+    },
+    {
+      "epoch": 2.580487804878049,
+      "grad_norm": 0.8600794672966003,
+      "learning_rate": 4.2252815504051285e-06,
+      "loss": 0.3133,
+      "step": 529
+    },
+    {
+      "epoch": 2.5853658536585367,
+      "grad_norm": 1.89119553565979,
+      "learning_rate": 4.222506870867618e-06,
+      "loss": 0.6036,
+      "step": 530
+    },
+    {
+      "epoch": 2.5902439024390245,
+      "grad_norm": 1.0424220561981201,
+      "learning_rate": 4.2197281460099245e-06,
+      "loss": 0.6877,
+      "step": 531
+    },
+    {
+      "epoch": 2.5951219512195123,
+      "grad_norm": 1.027593731880188,
+      "learning_rate": 4.216945382357905e-06,
+      "loss": 0.6352,
+      "step": 532
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.1954094171524048,
+      "learning_rate": 4.214158586446901e-06,
+      "loss": 0.966,
+      "step": 533
+    },
+    {
+      "epoch": 2.604878048780488,
+      "grad_norm": 1.2490975856781006,
+      "learning_rate": 4.211367764821722e-06,
+      "loss": 0.9059,
+      "step": 534
+    },
+    {
+      "epoch": 2.6097560975609757,
+      "grad_norm": 1.1297111511230469,
+      "learning_rate": 4.208572924036634e-06,
+      "loss": 0.5006,
+      "step": 535
+    },
+    {
+      "epoch": 2.6146341463414635,
+      "grad_norm": 0.9446660876274109,
+      "learning_rate": 4.2057740706553415e-06,
+      "loss": 0.4905,
+      "step": 536
+    },
+    {
+      "epoch": 2.6195121951219513,
+      "grad_norm": 0.916692316532135,
+      "learning_rate": 4.202971211250971e-06,
+      "loss": 0.6694,
+      "step": 537
+    },
+    {
+      "epoch": 2.624390243902439,
+      "grad_norm": 0.840929388999939,
+      "learning_rate": 4.200164352406061e-06,
+      "loss": 0.3739,
+      "step": 538
+    },
+    {
+      "epoch": 2.629268292682927,
+      "grad_norm": 1.069427728652954,
+      "learning_rate": 4.197353500712539e-06,
+      "loss": 0.6359,
+      "step": 539
+    },
+    {
+      "epoch": 2.6341463414634148,
+      "grad_norm": 1.0660371780395508,
+      "learning_rate": 4.1945386627717115e-06,
+      "loss": 0.5006,
+      "step": 540
+    },
+    {
+      "epoch": 2.6390243902439026,
+      "grad_norm": 1.9662373065948486,
+      "learning_rate": 4.191719845194246e-06,
+      "loss": 0.649,
+      "step": 541
+    },
+    {
+      "epoch": 2.6439024390243904,
+      "grad_norm": 0.9865717887878418,
+      "learning_rate": 4.188897054600156e-06,
+      "loss": 0.6179,
+      "step": 542
+    },
+    {
+      "epoch": 2.648780487804878,
+      "grad_norm": 1.0393004417419434,
+      "learning_rate": 4.186070297618787e-06,
+      "loss": 0.7156,
+      "step": 543
+    },
+    {
+      "epoch": 2.653658536585366,
+      "grad_norm": 0.9797636270523071,
+      "learning_rate": 4.183239580888799e-06,
+      "loss": 0.7249,
+      "step": 544
+    },
+    {
+      "epoch": 2.658536585365854,
+      "grad_norm": 1.180819034576416,
+      "learning_rate": 4.18040491105815e-06,
+      "loss": 0.4961,
+      "step": 545
+    },
+    {
+      "epoch": 2.6634146341463416,
+      "grad_norm": 0.9986240863800049,
+      "learning_rate": 4.177566294784085e-06,
+      "loss": 0.4397,
+      "step": 546
+    },
+    {
+      "epoch": 2.6682926829268294,
+      "grad_norm": 1.2416610717773438,
+      "learning_rate": 4.174723738733114e-06,
+      "loss": 0.7625,
+      "step": 547
+    },
+    {
+      "epoch": 2.6731707317073172,
+      "grad_norm": 0.9271990656852722,
+      "learning_rate": 4.171877249581001e-06,
+      "loss": 0.6626,
+      "step": 548
+    },
+    {
+      "epoch": 2.678048780487805,
+      "grad_norm": 0.9085447788238525,
+      "learning_rate": 4.169026834012748e-06,
+      "loss": 0.4209,
+      "step": 549
+    },
+    {
+      "epoch": 2.682926829268293,
+      "grad_norm": 0.9767999649047852,
+      "learning_rate": 4.166172498722577e-06,
+      "loss": 0.4466,
+      "step": 550
+    },
+    {
+      "epoch": 2.68780487804878,
+      "grad_norm": 1.405003309249878,
+      "learning_rate": 4.163314250413913e-06,
+      "loss": 0.8207,
+      "step": 551
+    },
+    {
+      "epoch": 2.692682926829268,
+      "grad_norm": 1.001278042793274,
+      "learning_rate": 4.160452095799378e-06,
+      "loss": 0.4949,
+      "step": 552
+    },
+    {
+      "epoch": 2.697560975609756,
+      "grad_norm": 0.8813621401786804,
+      "learning_rate": 4.157586041600759e-06,
+      "loss": 0.2726,
+      "step": 553
+    },
+    {
+      "epoch": 2.7024390243902436,
+      "grad_norm": 1.087396502494812,
+      "learning_rate": 4.154716094549008e-06,
+      "loss": 0.6358,
+      "step": 554
+    },
+    {
+      "epoch": 2.7073170731707314,
+      "grad_norm": 0.864211916923523,
+      "learning_rate": 4.151842261384217e-06,
+      "loss": 0.4146,
+      "step": 555
+    },
+    {
+      "epoch": 2.7121951219512193,
+      "grad_norm": 1.580068588256836,
+      "learning_rate": 4.148964548855603e-06,
+      "loss": 0.9695,
+      "step": 556
+    },
+    {
+      "epoch": 2.717073170731707,
+      "grad_norm": 0.7623794674873352,
+      "learning_rate": 4.146082963721496e-06,
+      "loss": 0.3429,
+      "step": 557
+    },
+    {
+      "epoch": 2.721951219512195,
+      "grad_norm": 1.2673579454421997,
+      "learning_rate": 4.143197512749322e-06,
+      "loss": 1.1916,
+      "step": 558
+    },
+    {
+      "epoch": 2.7268292682926827,
+      "grad_norm": 1.0848994255065918,
+      "learning_rate": 4.140308202715581e-06,
+      "loss": 0.8112,
+      "step": 559
+    },
+    {
+      "epoch": 2.7317073170731705,
+      "grad_norm": 0.9205752015113831,
+      "learning_rate": 4.13741504040584e-06,
+      "loss": 0.4364,
+      "step": 560
+    },
+    {
+      "epoch": 2.7365853658536583,
+      "grad_norm": 1.0304152965545654,
+      "learning_rate": 4.134518032614713e-06,
+      "loss": 0.4841,
+      "step": 561
+    },
+    {
+      "epoch": 2.741463414634146,
+      "grad_norm": 0.9854786992073059,
+      "learning_rate": 4.1316171861458445e-06,
+      "loss": 0.418,
+      "step": 562
+    },
+    {
+      "epoch": 2.746341463414634,
+      "grad_norm": 1.0625019073486328,
+      "learning_rate": 4.128712507811893e-06,
+      "loss": 0.6479,
+      "step": 563
+    },
+    {
+      "epoch": 2.7512195121951217,
+      "grad_norm": 1.0722914934158325,
+      "learning_rate": 4.125804004434517e-06,
+      "loss": 0.6462,
+      "step": 564
+    },
+    {
+      "epoch": 2.7560975609756095,
+      "grad_norm": 0.890087366104126,
+      "learning_rate": 4.12289168284436e-06,
+      "loss": 0.4462,
+      "step": 565
+    },
+    {
+      "epoch": 2.7609756097560973,
+      "grad_norm": 1.0615348815917969,
+      "learning_rate": 4.119975549881029e-06,
+      "loss": 0.6229,
+      "step": 566
+    },
+    {
+      "epoch": 2.765853658536585,
+      "grad_norm": 0.8919638395309448,
+      "learning_rate": 4.1170556123930846e-06,
+      "loss": 0.3847,
+      "step": 567
+    },
+    {
+      "epoch": 2.770731707317073,
+      "grad_norm": 0.9881047606468201,
+      "learning_rate": 4.114131877238021e-06,
+      "loss": 0.5972,
+      "step": 568
+    },
+    {
+      "epoch": 2.7756097560975608,
+      "grad_norm": 0.8665289878845215,
+      "learning_rate": 4.111204351282254e-06,
+      "loss": 0.4755,
+      "step": 569
+    },
+    {
+      "epoch": 2.7804878048780486,
+      "grad_norm": 0.7870835661888123,
+      "learning_rate": 4.108273041401098e-06,
+      "loss": 0.4341,
+      "step": 570
+    },
+    {
+      "epoch": 2.7853658536585364,
+      "grad_norm": 1.3610732555389404,
+      "learning_rate": 4.105337954478756e-06,
+      "loss": 0.8646,
+      "step": 571
+    },
+    {
+      "epoch": 2.790243902439024,
+      "grad_norm": 1.0715464353561401,
+      "learning_rate": 4.102399097408304e-06,
+      "loss": 0.7017,
+      "step": 572
+    },
+    {
+      "epoch": 2.795121951219512,
+      "grad_norm": 0.952342689037323,
+      "learning_rate": 4.099456477091667e-06,
+      "loss": 0.3381,
+      "step": 573
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.151577115058899,
+      "learning_rate": 4.096510100439611e-06,
+      "loss": 0.8217,
+      "step": 574
+    },
+    {
+      "epoch": 2.8048780487804876,
+      "grad_norm": 0.8553835153579712,
+      "learning_rate": 4.093559974371725e-06,
+      "loss": 0.3334,
+      "step": 575
+    },
+    {
+      "epoch": 2.8097560975609754,
+      "grad_norm": 1.004453420639038,
+      "learning_rate": 4.0906061058164e-06,
+      "loss": 0.8324,
+      "step": 576
+    },
+    {
+      "epoch": 2.8146341463414632,
+      "grad_norm": 0.9378971457481384,
+      "learning_rate": 4.087648501710819e-06,
+      "loss": 0.3753,
+      "step": 577
+    },
+    {
+      "epoch": 2.819512195121951,
+      "grad_norm": 0.9435027241706848,
+      "learning_rate": 4.084687169000938e-06,
+      "loss": 0.5675,
+      "step": 578
+    },
+    {
+      "epoch": 2.824390243902439,
+      "grad_norm": 0.856401801109314,
+      "learning_rate": 4.081722114641469e-06,
+      "loss": 0.5309,
+      "step": 579
+    },
+    {
+      "epoch": 2.8292682926829267,
+      "grad_norm": 1.1497118473052979,
+      "learning_rate": 4.0787533455958626e-06,
+      "loss": 0.3339,
+      "step": 580
+    },
+    {
+      "epoch": 2.8341463414634145,
+      "grad_norm": 1.0275132656097412,
+      "learning_rate": 4.075780868836296e-06,
+      "loss": 0.4303,
+      "step": 581
+    },
+    {
+      "epoch": 2.8390243902439023,
+      "grad_norm": 0.903195321559906,
+      "learning_rate": 4.072804691343653e-06,
+      "loss": 0.495,
+      "step": 582
+    },
+    {
+      "epoch": 2.84390243902439,
+      "grad_norm": 1.1491634845733643,
+      "learning_rate": 4.069824820107507e-06,
+      "loss": 0.9994,
+      "step": 583
+    },
+    {
+      "epoch": 2.848780487804878,
+      "grad_norm": 0.7706964015960693,
+      "learning_rate": 4.06684126212611e-06,
+      "loss": 0.3226,
+      "step": 584
+    },
+    {
+      "epoch": 2.8536585365853657,
+      "grad_norm": 0.9241564273834229,
+      "learning_rate": 4.063854024406369e-06,
+      "loss": 0.5793,
+      "step": 585
+    },
+    {
+      "epoch": 2.8585365853658535,
+      "grad_norm": 0.9884312152862549,
+      "learning_rate": 4.060863113963835e-06,
+      "loss": 0.4683,
+      "step": 586
+    },
+    {
+      "epoch": 2.8634146341463413,
+      "grad_norm": 0.9997614026069641,
+      "learning_rate": 4.057868537822683e-06,
+      "loss": 0.519,
+      "step": 587
+    },
+    {
+      "epoch": 2.868292682926829,
+      "grad_norm": 0.9449941515922546,
+      "learning_rate": 4.054870303015695e-06,
+      "loss": 0.381,
+      "step": 588
+    },
+    {
+      "epoch": 2.873170731707317,
+      "grad_norm": 1.0271875858306885,
+      "learning_rate": 4.05186841658425e-06,
+      "loss": 0.4554,
+      "step": 589
+    },
+    {
+      "epoch": 2.8780487804878048,
+      "grad_norm": 0.9256722331047058,
+      "learning_rate": 4.048862885578301e-06,
+      "loss": 0.5297,
+      "step": 590
+    },
+    {
+      "epoch": 2.8829268292682926,
+      "grad_norm": 1.0047836303710938,
+      "learning_rate": 4.045853717056358e-06,
+      "loss": 0.5968,
+      "step": 591
+    },
+    {
+      "epoch": 2.8878048780487804,
+      "grad_norm": 0.9485352635383606,
+      "learning_rate": 4.0428409180854775e-06,
+      "loss": 0.5042,
+      "step": 592
+    },
+    {
+      "epoch": 2.892682926829268,
+      "grad_norm": 0.9246886372566223,
+      "learning_rate": 4.039824495741238e-06,
+      "loss": 0.4622,
+      "step": 593
+    },
+    {
+      "epoch": 2.897560975609756,
+      "grad_norm": 0.8539214134216309,
+      "learning_rate": 4.036804457107733e-06,
+      "loss": 0.53,
+      "step": 594
+    },
+    {
+      "epoch": 2.902439024390244,
+      "grad_norm": 1.2358900308609009,
+      "learning_rate": 4.0337808092775435e-06,
+      "loss": 0.701,
+      "step": 595
+    },
+    {
+      "epoch": 2.9073170731707316,
+      "grad_norm": 0.8977146744728088,
+      "learning_rate": 4.030753559351728e-06,
+      "loss": 0.3942,
+      "step": 596
+    },
+    {
+      "epoch": 2.9121951219512194,
+      "grad_norm": 0.8575205206871033,
+      "learning_rate": 4.027722714439808e-06,
+      "loss": 0.3713,
+      "step": 597
+    },
+    {
+      "epoch": 2.9170731707317072,
+      "grad_norm": 1.2680315971374512,
+      "learning_rate": 4.024688281659743e-06,
+      "loss": 0.9398,
+      "step": 598
+    },
+    {
+      "epoch": 2.921951219512195,
+      "grad_norm": 1.7810138463974,
+      "learning_rate": 4.021650268137924e-06,
+      "loss": 0.6056,
+      "step": 599
+    },
+    {
+      "epoch": 2.926829268292683,
+      "grad_norm": 0.8538106083869934,
+      "learning_rate": 4.018608681009143e-06,
+      "loss": 0.4939,
+      "step": 600
+    },
+    {
+      "epoch": 2.9317073170731707,
+      "grad_norm": 1.0676621198654175,
+      "learning_rate": 4.015563527416596e-06,
+      "loss": 0.5676,
+      "step": 601
+    },
+    {
+      "epoch": 2.9365853658536585,
+      "grad_norm": 0.987902820110321,
+      "learning_rate": 4.012514814511844e-06,
+      "loss": 0.5004,
+      "step": 602
+    },
+    {
+      "epoch": 2.9414634146341463,
+      "grad_norm": 0.9196493625640869,
+      "learning_rate": 4.009462549454816e-06,
+      "loss": 0.5476,
+      "step": 603
+    },
+    {
+      "epoch": 2.946341463414634,
+      "grad_norm": 1.1288981437683105,
+      "learning_rate": 4.006406739413775e-06,
+      "loss": 0.5804,
+      "step": 604
+    },
+    {
+      "epoch": 2.951219512195122,
+      "grad_norm": 0.8905384540557861,
+      "learning_rate": 4.003347391565317e-06,
+      "loss": 0.4979,
+      "step": 605
+    },
+    {
+      "epoch": 2.9560975609756097,
+      "grad_norm": 1.000335693359375,
+      "learning_rate": 4.000284513094342e-06,
+      "loss": 0.5727,
+      "step": 606
+    },
+    {
+      "epoch": 2.9609756097560975,
+      "grad_norm": 0.9347658753395081,
+      "learning_rate": 3.997218111194042e-06,
+      "loss": 0.5235,
+      "step": 607
+    },
+    {
+      "epoch": 2.9658536585365853,
+      "grad_norm": 0.8476413488388062,
+      "learning_rate": 3.994148193065886e-06,
+      "loss": 0.3922,
+      "step": 608
+    },
+    {
+      "epoch": 2.970731707317073,
+      "grad_norm": 0.8186416029930115,
+      "learning_rate": 3.991074765919598e-06,
+      "loss": 0.3482,
+      "step": 609
+    },
+    {
+      "epoch": 2.975609756097561,
+      "grad_norm": 1.2038166522979736,
+      "learning_rate": 3.987997836973147e-06,
+      "loss": 0.4684,
+      "step": 610
+    },
+    {
+      "epoch": 2.9804878048780488,
+      "grad_norm": 1.037007212638855,
+      "learning_rate": 3.984917413452721e-06,
+      "loss": 0.4811,
+      "step": 611
+    },
+    {
+      "epoch": 2.9853658536585366,
+      "grad_norm": 1.2915143966674805,
+      "learning_rate": 3.981833502592717e-06,
+      "loss": 0.7361,
+      "step": 612
+    },
+    {
+      "epoch": 2.9902439024390244,
+      "grad_norm": 1.5253301858901978,
+      "learning_rate": 3.978746111635725e-06,
+      "loss": 0.312,
+      "step": 613
+    },
+    {
+      "epoch": 2.995121951219512,
+      "grad_norm": 0.8432179093360901,
+      "learning_rate": 3.9756552478325045e-06,
+      "loss": 0.4931,
+      "step": 614
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8148512840270996,
+      "learning_rate": 3.972560918441972e-06,
+      "loss": 0.3185,
+      "step": 615
+    },
+    {
+      "epoch": 3.004878048780488,
+      "grad_norm": 0.7907199859619141,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.3731,
+      "step": 616
+    },
+    {
+      "epoch": 3.0097560975609756,
+      "grad_norm": 0.8440088033676147,
+      "learning_rate": 3.966361891975316e-06,
+      "loss": 0.4135,
+      "step": 617
+    },
+    {
+      "epoch": 3.0146341463414634,
+      "grad_norm": 0.9237042665481567,
+      "learning_rate": 3.963257209457652e-06,
+      "loss": 0.4563,
+      "step": 618
+    },
+    {
+      "epoch": 3.0195121951219512,
+      "grad_norm": 0.7932549118995667,
+      "learning_rate": 3.960149090469561e-06,
+      "loss": 0.2539,
+      "step": 619
+    },
+    {
+      "epoch": 3.024390243902439,
+      "grad_norm": 0.9151872396469116,
+      "learning_rate": 3.957037542310484e-06,
+      "loss": 0.2685,
+      "step": 620
+    },
+    {
+      "epoch": 3.029268292682927,
+      "grad_norm": 0.8039369583129883,
+      "learning_rate": 3.953922572287915e-06,
+      "loss": 0.4401,
+      "step": 621
+    },
+    {
+      "epoch": 3.0341463414634147,
+      "grad_norm": 0.9228289723396301,
+      "learning_rate": 3.950804187717384e-06,
+      "loss": 0.5399,
+      "step": 622
+    },
+    {
+      "epoch": 3.0390243902439025,
+      "grad_norm": 1.0876184701919556,
+      "learning_rate": 3.947682395922439e-06,
+      "loss": 0.6546,
+      "step": 623
+    },
+    {
+      "epoch": 3.0439024390243903,
+      "grad_norm": 0.8611107468605042,
+      "learning_rate": 3.9445572042346346e-06,
+      "loss": 0.2007,
+      "step": 624
+    },
+    {
+      "epoch": 3.048780487804878,
+      "grad_norm": 0.8184982538223267,
+      "learning_rate": 3.941428619993505e-06,
+      "loss": 0.3513,
+      "step": 625
+    },
+    {
+      "epoch": 3.053658536585366,
+      "grad_norm": 1.0696086883544922,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.2816,
+      "step": 626
+    },
+    {
+      "epoch": 3.0585365853658537,
+      "grad_norm": 1.4805349111557007,
+      "learning_rate": 3.935161303249231e-06,
+      "loss": 0.4662,
+      "step": 627
+    },
+    {
+      "epoch": 3.0634146341463415,
+      "grad_norm": 1.0260307788848877,
+      "learning_rate": 3.932022585464928e-06,
+      "loss": 0.4517,
+      "step": 628
+    },
+    {
+      "epoch": 3.0682926829268293,
+      "grad_norm": 0.972687304019928,
+      "learning_rate": 3.928880504564943e-06,
+      "loss": 0.3406,
+      "step": 629
+    },
+    {
+      "epoch": 3.073170731707317,
+      "grad_norm": 0.9252133369445801,
+      "learning_rate": 3.92573506792848e-06,
+      "loss": 0.3444,
+      "step": 630
+    },
+    {
+      "epoch": 3.078048780487805,
+      "grad_norm": 1.0907679796218872,
+      "learning_rate": 3.9225862829426184e-06,
+      "loss": 0.3427,
+      "step": 631
+    },
+    {
+      "epoch": 3.0829268292682928,
+      "grad_norm": 1.0673291683197021,
+      "learning_rate": 3.919434157002303e-06,
+      "loss": 0.4476,
+      "step": 632
+    },
+    {
+      "epoch": 3.0878048780487806,
+      "grad_norm": 0.758525550365448,
+      "learning_rate": 3.916278697510325e-06,
+      "loss": 0.274,
+      "step": 633
+    },
+    {
+      "epoch": 3.0926829268292684,
+      "grad_norm": 0.9257010221481323,
+      "learning_rate": 3.913119911877305e-06,
+      "loss": 0.5037,
+      "step": 634
+    },
+    {
+      "epoch": 3.097560975609756,
+      "grad_norm": 1.0138721466064453,
+      "learning_rate": 3.909957807521674e-06,
+      "loss": 0.253,
+      "step": 635
+    },
+    {
+      "epoch": 3.102439024390244,
+      "grad_norm": 0.9045038819313049,
+      "learning_rate": 3.906792391869657e-06,
+      "loss": 0.381,
+      "step": 636
+    },
+    {
+      "epoch": 3.107317073170732,
+      "grad_norm": 0.9882903099060059,
+      "learning_rate": 3.903623672355258e-06,
+      "loss": 0.3979,
+      "step": 637
+    },
+    {
+      "epoch": 3.1121951219512196,
+      "grad_norm": 0.9428558945655823,
+      "learning_rate": 3.900451656420237e-06,
+      "loss": 0.3531,
+      "step": 638
+    },
+    {
+      "epoch": 3.1170731707317074,
+      "grad_norm": 0.9153510332107544,
+      "learning_rate": 3.897276351514097e-06,
+      "loss": 0.2724,
+      "step": 639
+    },
+    {
+      "epoch": 3.1219512195121952,
+      "grad_norm": 1.1453444957733154,
+      "learning_rate": 3.894097765094065e-06,
+      "loss": 0.5488,
+      "step": 640
+    },
+    {
+      "epoch": 3.126829268292683,
+      "grad_norm": 0.7873973250389099,
+      "learning_rate": 3.890915904625075e-06,
+      "loss": 0.2855,
+      "step": 641
+    },
+    {
+      "epoch": 3.131707317073171,
+      "grad_norm": 1.2348054647445679,
+      "learning_rate": 3.887730777579751e-06,
+      "loss": 0.5758,
+      "step": 642
+    },
+    {
+      "epoch": 3.1365853658536587,
+      "grad_norm": 1.1336091756820679,
+      "learning_rate": 3.884542391438387e-06,
+      "loss": 0.681,
+      "step": 643
+    },
+    {
+      "epoch": 3.1414634146341465,
+      "grad_norm": 1.1599171161651611,
+      "learning_rate": 3.88135075368893e-06,
+      "loss": 0.8581,
+      "step": 644
+    },
+    {
+      "epoch": 3.1463414634146343,
+      "grad_norm": 0.9442818760871887,
+      "learning_rate": 3.878155871826968e-06,
+      "loss": 0.4053,
+      "step": 645
+    },
+    {
+      "epoch": 3.151219512195122,
+      "grad_norm": 0.9527677297592163,
+      "learning_rate": 3.874957753355701e-06,
+      "loss": 0.3673,
+      "step": 646
+    },
+    {
+      "epoch": 3.15609756097561,
+      "grad_norm": 0.9200324416160583,
+      "learning_rate": 3.8717564057859365e-06,
+      "loss": 0.6203,
+      "step": 647
+    },
+    {
+      "epoch": 3.1609756097560977,
+      "grad_norm": 1.0606499910354614,
+      "learning_rate": 3.868551836636063e-06,
+      "loss": 0.564,
+      "step": 648
+    },
+    {
+      "epoch": 3.1658536585365855,
+      "grad_norm": 1.064496397972107,
+      "learning_rate": 3.865344053432035e-06,
+      "loss": 0.316,
+      "step": 649
+    },
+    {
+      "epoch": 3.1707317073170733,
+      "grad_norm": 1.259859323501587,
+      "learning_rate": 3.862133063707353e-06,
+      "loss": 0.4876,
+      "step": 650
+    },
+    {
+      "epoch": 3.175609756097561,
+      "grad_norm": 0.9357886910438538,
+      "learning_rate": 3.858918875003053e-06,
+      "loss": 0.3449,
+      "step": 651
+    },
+    {
+      "epoch": 3.180487804878049,
+      "grad_norm": 0.8545693755149841,
+      "learning_rate": 3.855701494867679e-06,
+      "loss": 0.3919,
+      "step": 652
+    },
+    {
+      "epoch": 3.1853658536585368,
+      "grad_norm": 0.9744400382041931,
+      "learning_rate": 3.852480930857275e-06,
+      "loss": 0.5602,
+      "step": 653
+    },
+    {
+      "epoch": 3.1902439024390246,
+      "grad_norm": 0.9290996193885803,
+      "learning_rate": 3.849257190535356e-06,
+      "loss": 0.3391,
+      "step": 654
+    },
+    {
+      "epoch": 3.1951219512195124,
+      "grad_norm": 1.150904655456543,
+      "learning_rate": 3.846030281472902e-06,
+      "loss": 0.6754,
+      "step": 655
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.9049080610275269,
+      "learning_rate": 3.842800211248333e-06,
+      "loss": 0.342,
+      "step": 656
+    },
+    {
+      "epoch": 3.204878048780488,
+      "grad_norm": 1.6837809085845947,
+      "learning_rate": 3.839566987447492e-06,
+      "loss": 0.5241,
+      "step": 657
+    },
+    {
+      "epoch": 3.209756097560976,
+      "grad_norm": 0.9485814571380615,
+      "learning_rate": 3.8363306176636296e-06,
+      "loss": 0.6228,
+      "step": 658
+    },
+    {
+      "epoch": 3.2146341463414636,
+      "grad_norm": 1.029710054397583,
+      "learning_rate": 3.833091109497384e-06,
+      "loss": 0.6792,
+      "step": 659
+    },
+    {
+      "epoch": 3.2195121951219514,
+      "grad_norm": 0.881193220615387,
+      "learning_rate": 3.829848470556765e-06,
+      "loss": 0.25,
+      "step": 660
+    },
+    {
+      "epoch": 3.2243902439024392,
+      "grad_norm": 0.8923963904380798,
+      "learning_rate": 3.8266027084571335e-06,
+      "loss": 0.4637,
+      "step": 661
+    },
+    {
+      "epoch": 3.229268292682927,
+      "grad_norm": 1.039754867553711,
+      "learning_rate": 3.823353830821187e-06,
+      "loss": 0.2432,
+      "step": 662
+    },
+    {
+      "epoch": 3.234146341463415,
+      "grad_norm": 0.8456618189811707,
+      "learning_rate": 3.820101845278937e-06,
+      "loss": 0.4113,
+      "step": 663
+    },
+    {
+      "epoch": 3.2390243902439027,
+      "grad_norm": 1.1256377696990967,
+      "learning_rate": 3.816846759467696e-06,
+      "loss": 0.4089,
+      "step": 664
+    },
+    {
+      "epoch": 3.2439024390243905,
+      "grad_norm": 1.1781972646713257,
+      "learning_rate": 3.8135885810320587e-06,
+      "loss": 0.5524,
+      "step": 665
+    },
+    {
+      "epoch": 3.2487804878048783,
+      "grad_norm": 0.8150308132171631,
+      "learning_rate": 3.810327317623881e-06,
+      "loss": 0.3115,
+      "step": 666
+    },
+    {
+      "epoch": 3.253658536585366,
+      "grad_norm": 1.0984582901000977,
+      "learning_rate": 3.8070629769022628e-06,
+      "loss": 0.4814,
+      "step": 667
+    },
+    {
+      "epoch": 3.258536585365854,
+      "grad_norm": 0.9676326513290405,
+      "learning_rate": 3.8037955665335335e-06,
+      "loss": 0.3787,
+      "step": 668
+    },
+    {
+      "epoch": 3.2634146341463417,
+      "grad_norm": 1.1023277044296265,
+      "learning_rate": 3.800525094191231e-06,
+      "loss": 0.4983,
+      "step": 669
+    },
+    {
+      "epoch": 3.2682926829268295,
+      "grad_norm": 0.8557276725769043,
+      "learning_rate": 3.797251567556083e-06,
+      "loss": 0.3505,
+      "step": 670
+    },
+    {
+      "epoch": 3.2731707317073173,
+      "grad_norm": 0.8645808100700378,
+      "learning_rate": 3.793974994315991e-06,
+      "loss": 0.2222,
+      "step": 671
+    },
+    {
+      "epoch": 3.278048780487805,
+      "grad_norm": 1.0875083208084106,
+      "learning_rate": 3.790695382166013e-06,
+      "loss": 0.454,
+      "step": 672
+    },
+    {
+      "epoch": 3.2829268292682925,
+      "grad_norm": 0.9899719953536987,
+      "learning_rate": 3.7874127388083415e-06,
+      "loss": 0.3548,
+      "step": 673
+    },
+    {
+      "epoch": 3.2878048780487803,
+      "grad_norm": 1.0319510698318481,
+      "learning_rate": 3.7841270719522895e-06,
+      "loss": 0.4269,
+      "step": 674
+    },
+    {
+      "epoch": 3.292682926829268,
+      "grad_norm": 1.1482504606246948,
+      "learning_rate": 3.7808383893142692e-06,
+      "loss": 0.2515,
+      "step": 675
+    },
+    {
+      "epoch": 3.297560975609756,
+      "grad_norm": 0.977626383304596,
+      "learning_rate": 3.7775466986177763e-06,
+      "loss": 0.3434,
+      "step": 676
+    },
+    {
+      "epoch": 3.3024390243902437,
+      "grad_norm": 0.8783503770828247,
+      "learning_rate": 3.774252007593371e-06,
+      "loss": 0.2183,
+      "step": 677
+    },
+    {
+      "epoch": 3.3073170731707315,
+      "grad_norm": 1.0656304359436035,
+      "learning_rate": 3.7709543239786593e-06,
+      "loss": 0.5429,
+      "step": 678
+    },
+    {
+      "epoch": 3.3121951219512193,
+      "grad_norm": 0.9340170621871948,
+      "learning_rate": 3.767653655518277e-06,
+      "loss": 0.3605,
+      "step": 679
+    },
+    {
+      "epoch": 3.317073170731707,
+      "grad_norm": 1.228542685508728,
+      "learning_rate": 3.7643500099638673e-06,
+      "loss": 0.3435,
+      "step": 680
+    },
+    {
+      "epoch": 3.321951219512195,
+      "grad_norm": 0.9645997881889343,
+      "learning_rate": 3.7610433950740667e-06,
+      "loss": 0.6613,
+      "step": 681
+    },
+    {
+      "epoch": 3.3268292682926828,
+      "grad_norm": 1.077845811843872,
+      "learning_rate": 3.757733818614485e-06,
+      "loss": 0.4901,
+      "step": 682
+    },
+    {
+      "epoch": 3.3317073170731706,
+      "grad_norm": 0.8504129648208618,
+      "learning_rate": 3.7544212883576856e-06,
+      "loss": 0.4069,
+      "step": 683
+    },
+    {
+      "epoch": 3.3365853658536584,
+      "grad_norm": 1.0994789600372314,
+      "learning_rate": 3.751105812083172e-06,
+      "loss": 0.3476,
+      "step": 684
+    },
+    {
+      "epoch": 3.341463414634146,
+      "grad_norm": 1.2599692344665527,
+      "learning_rate": 3.7477873975773655e-06,
+      "loss": 0.5327,
+      "step": 685
+    },
+    {
+      "epoch": 3.346341463414634,
+      "grad_norm": 1.0701467990875244,
+      "learning_rate": 3.7444660526335853e-06,
+      "loss": 0.473,
+      "step": 686
+    },
+    {
+      "epoch": 3.351219512195122,
+      "grad_norm": 1.7928602695465088,
+      "learning_rate": 3.741141785052036e-06,
+      "loss": 0.8412,
+      "step": 687
+    },
+    {
+      "epoch": 3.3560975609756096,
+      "grad_norm": 1.075492262840271,
+      "learning_rate": 3.737814602639784e-06,
+      "loss": 0.5616,
+      "step": 688
+    },
+    {
+      "epoch": 3.3609756097560974,
+      "grad_norm": 0.9147948622703552,
+      "learning_rate": 3.7344845132107427e-06,
+      "loss": 0.4454,
+      "step": 689
+    },
+    {
+      "epoch": 3.3658536585365852,
+      "grad_norm": 1.0845904350280762,
+      "learning_rate": 3.731151524585651e-06,
+      "loss": 0.4776,
+      "step": 690
+    },
+    {
+      "epoch": 3.370731707317073,
+      "grad_norm": 1.2276109457015991,
+      "learning_rate": 3.7278156445920584e-06,
+      "loss": 0.8544,
+      "step": 691
+    },
+    {
+      "epoch": 3.375609756097561,
+      "grad_norm": 0.866237998008728,
+      "learning_rate": 3.724476881064303e-06,
+      "loss": 0.297,
+      "step": 692
+    },
+    {
+      "epoch": 3.3804878048780487,
+      "grad_norm": 0.8218240737915039,
+      "learning_rate": 3.721135241843496e-06,
+      "loss": 0.3716,
+      "step": 693
+    },
+    {
+      "epoch": 3.3853658536585365,
+      "grad_norm": 1.1857250928878784,
+      "learning_rate": 3.7177907347775016e-06,
+      "loss": 0.4965,
+      "step": 694
+    },
+    {
+      "epoch": 3.3902439024390243,
+      "grad_norm": 1.726030945777893,
+      "learning_rate": 3.71444336772092e-06,
+      "loss": 0.6302,
+      "step": 695
+    },
+    {
+      "epoch": 3.395121951219512,
+      "grad_norm": 1.3070801496505737,
+      "learning_rate": 3.711093148535068e-06,
+      "loss": 0.7139,
+      "step": 696
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.210959553718567,
+      "learning_rate": 3.707740085087959e-06,
+      "loss": 0.3059,
+      "step": 697
+    },
+    {
+      "epoch": 3.4048780487804877,
+      "grad_norm": 0.9498068690299988,
+      "learning_rate": 3.7043841852542884e-06,
+      "loss": 0.4441,
+      "step": 698
+    },
+    {
+      "epoch": 3.4097560975609755,
+      "grad_norm": 0.8833593130111694,
+      "learning_rate": 3.701025456915411e-06,
+      "loss": 0.3382,
+      "step": 699
+    },
+    {
+      "epoch": 3.4146341463414633,
+      "grad_norm": 0.9266107678413391,
+      "learning_rate": 3.697663907959327e-06,
+      "loss": 0.4165,
+      "step": 700
+    },
+    {
+      "epoch": 3.419512195121951,
+      "grad_norm": 1.4498791694641113,
+      "learning_rate": 3.6942995462806574e-06,
+      "loss": 0.7236,
+      "step": 701
+    },
+    {
+      "epoch": 3.424390243902439,
+      "grad_norm": 0.8553938269615173,
+      "learning_rate": 3.6909323797806314e-06,
+      "loss": 0.3514,
+      "step": 702
+    },
+    {
+      "epoch": 3.4292682926829268,
+      "grad_norm": 1.291072130203247,
+      "learning_rate": 3.6875624163670635e-06,
+      "loss": 0.6465,
+      "step": 703
+    },
+    {
+      "epoch": 3.4341463414634146,
+      "grad_norm": 1.2313672304153442,
+      "learning_rate": 3.6841896639543394e-06,
+      "loss": 0.3876,
+      "step": 704
+    },
+    {
+      "epoch": 3.4390243902439024,
+      "grad_norm": 0.8775885105133057,
+      "learning_rate": 3.6808141304633924e-06,
+      "loss": 0.3682,
+      "step": 705
+    },
+    {
+      "epoch": 3.44390243902439,
+      "grad_norm": 0.9601542949676514,
+      "learning_rate": 3.6774358238216878e-06,
+      "loss": 0.3753,
+      "step": 706
+    },
+    {
+      "epoch": 3.448780487804878,
+      "grad_norm": 1.371579885482788,
+      "learning_rate": 3.6740547519632048e-06,
+      "loss": 0.3362,
+      "step": 707
+    },
+    {
+      "epoch": 3.453658536585366,
+      "grad_norm": 0.7805762887001038,
+      "learning_rate": 3.670670922828414e-06,
+      "loss": 0.3909,
+      "step": 708
+    },
+    {
+      "epoch": 3.4585365853658536,
+      "grad_norm": 1.2143738269805908,
+      "learning_rate": 3.667284344364264e-06,
+      "loss": 0.3859,
+      "step": 709
+    },
+    {
+      "epoch": 3.4634146341463414,
+      "grad_norm": 0.9730637073516846,
+      "learning_rate": 3.6638950245241604e-06,
+      "loss": 0.5365,
+      "step": 710
+    },
+    {
+      "epoch": 3.4682926829268292,
+      "grad_norm": 1.291444182395935,
+      "learning_rate": 3.660502971267945e-06,
+      "loss": 0.3862,
+      "step": 711
+    },
+    {
+      "epoch": 3.473170731707317,
+      "grad_norm": 0.9339273571968079,
+      "learning_rate": 3.65710819256188e-06,
+      "loss": 0.1731,
+      "step": 712
+    },
+    {
+      "epoch": 3.478048780487805,
+      "grad_norm": 0.9810922741889954,
+      "learning_rate": 3.65371069637863e-06,
+      "loss": 0.3927,
+      "step": 713
+    },
+    {
+      "epoch": 3.4829268292682927,
+      "grad_norm": 1.027161955833435,
+      "learning_rate": 3.650310490697238e-06,
+      "loss": 0.5772,
+      "step": 714
+    },
+    {
+      "epoch": 3.4878048780487805,
+      "grad_norm": 1.0883551836013794,
+      "learning_rate": 3.646907583503114e-06,
+      "loss": 0.5687,
+      "step": 715
+    },
+    {
+      "epoch": 3.4926829268292683,
+      "grad_norm": 1.0536428689956665,
+      "learning_rate": 3.6435019827880093e-06,
+      "loss": 0.3598,
+      "step": 716
+    },
+    {
+      "epoch": 3.497560975609756,
+      "grad_norm": 0.8690207004547119,
+      "learning_rate": 3.640093696550003e-06,
+      "loss": 0.4317,
+      "step": 717
+    },
+    {
+      "epoch": 3.502439024390244,
+      "grad_norm": 0.9786996245384216,
+      "learning_rate": 3.6366827327934817e-06,
+      "loss": 0.428,
+      "step": 718
+    },
+    {
+      "epoch": 3.5073170731707317,
+      "grad_norm": 1.1201248168945312,
+      "learning_rate": 3.6332690995291176e-06,
+      "loss": 0.5273,
+      "step": 719
+    },
+    {
+      "epoch": 3.5121951219512195,
+      "grad_norm": 1.447026252746582,
+      "learning_rate": 3.6298528047738545e-06,
+      "loss": 1.3048,
+      "step": 720
+    },
+    {
+      "epoch": 3.5170731707317073,
+      "grad_norm": 0.988138735294342,
+      "learning_rate": 3.626433856550886e-06,
+      "loss": 0.6072,
+      "step": 721
+    },
+    {
+      "epoch": 3.521951219512195,
+      "grad_norm": 1.019419550895691,
+      "learning_rate": 3.623012262889637e-06,
+      "loss": 0.5362,
+      "step": 722
+    },
+    {
+      "epoch": 3.526829268292683,
+      "grad_norm": 1.2538626194000244,
+      "learning_rate": 3.6195880318257465e-06,
+      "loss": 0.5864,
+      "step": 723
+    },
+    {
+      "epoch": 3.5317073170731708,
+      "grad_norm": 1.2680046558380127,
+      "learning_rate": 3.616161171401046e-06,
+      "loss": 0.7208,
+      "step": 724
+    },
+    {
+      "epoch": 3.5365853658536586,
+      "grad_norm": 1.0290186405181885,
+      "learning_rate": 3.612731689663542e-06,
+      "loss": 0.3631,
+      "step": 725
+    },
+    {
+      "epoch": 3.5414634146341464,
+      "grad_norm": 0.941155731678009,
+      "learning_rate": 3.6092995946673996e-06,
+      "loss": 0.592,
+      "step": 726
+    },
+    {
+      "epoch": 3.546341463414634,
+      "grad_norm": 1.253089427947998,
+      "learning_rate": 3.605864894472918e-06,
+      "loss": 0.388,
+      "step": 727
+    },
+    {
+      "epoch": 3.551219512195122,
+      "grad_norm": 1.1095308065414429,
+      "learning_rate": 3.602427597146516e-06,
+      "loss": 0.5986,
+      "step": 728
+    },
+    {
+      "epoch": 3.55609756097561,
+      "grad_norm": 1.3642429113388062,
+      "learning_rate": 3.5989877107607134e-06,
+      "loss": 0.5178,
+      "step": 729
+    },
+    {
+      "epoch": 3.5609756097560976,
+      "grad_norm": 1.2215455770492554,
+      "learning_rate": 3.5955452433941075e-06,
+      "loss": 0.562,
+      "step": 730
+    },
+    {
+      "epoch": 3.5658536585365854,
+      "grad_norm": 0.955337405204773,
+      "learning_rate": 3.5921002031313586e-06,
+      "loss": 0.337,
+      "step": 731
+    },
+    {
+      "epoch": 3.5707317073170732,
+      "grad_norm": 0.9776975512504578,
+      "learning_rate": 3.58865259806317e-06,
+      "loss": 0.3314,
+      "step": 732
+    },
+    {
+      "epoch": 3.575609756097561,
+      "grad_norm": 1.1856746673583984,
+      "learning_rate": 3.585202436286267e-06,
+      "loss": 0.5582,
+      "step": 733
+    },
+    {
+      "epoch": 3.580487804878049,
+      "grad_norm": 1.1398861408233643,
+      "learning_rate": 3.581749725903381e-06,
+      "loss": 0.5148,
+      "step": 734
+    },
+    {
+      "epoch": 3.5853658536585367,
+      "grad_norm": 0.9702528119087219,
+      "learning_rate": 3.5782944750232274e-06,
+      "loss": 0.4666,
+      "step": 735
+    },
+    {
+      "epoch": 3.5902439024390245,
+      "grad_norm": 0.951115608215332,
+      "learning_rate": 3.574836691760489e-06,
+      "loss": 0.1839,
+      "step": 736
+    },
+    {
+      "epoch": 3.5951219512195123,
+      "grad_norm": 1.1828500032424927,
+      "learning_rate": 3.571376384235795e-06,
+      "loss": 0.4325,
+      "step": 737
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.0837311744689941,
+      "learning_rate": 3.5679135605757035e-06,
+      "loss": 0.3254,
+      "step": 738
+    },
+    {
+      "epoch": 3.604878048780488,
+      "grad_norm": 0.9672379493713379,
+      "learning_rate": 3.564448228912682e-06,
+      "loss": 0.3019,
+      "step": 739
+    },
+    {
+      "epoch": 3.6097560975609757,
+      "grad_norm": 1.1051793098449707,
+      "learning_rate": 3.5609803973850877e-06,
+      "loss": 0.4066,
+      "step": 740
+    },
+    {
+      "epoch": 3.6146341463414635,
+      "grad_norm": 1.0380092859268188,
+      "learning_rate": 3.557510074137147e-06,
+      "loss": 0.4469,
+      "step": 741
+    },
+    {
+      "epoch": 3.6195121951219513,
+      "grad_norm": 0.8451574444770813,
+      "learning_rate": 3.554037267318942e-06,
+      "loss": 0.4371,
+      "step": 742
+    },
+    {
+      "epoch": 3.624390243902439,
+      "grad_norm": 0.9186508655548096,
+      "learning_rate": 3.5505619850863847e-06,
+      "loss": 0.3235,
+      "step": 743
+    },
+    {
+      "epoch": 3.629268292682927,
+      "grad_norm": 0.9366088509559631,
+      "learning_rate": 3.5470842356012007e-06,
+      "loss": 0.2157,
+      "step": 744
+    },
+    {
+      "epoch": 3.6341463414634148,
+      "grad_norm": 1.0610483884811401,
+      "learning_rate": 3.5436040270309113e-06,
+      "loss": 0.5039,
+      "step": 745
+    },
+    {
+      "epoch": 3.6390243902439026,
+      "grad_norm": 0.9160799980163574,
+      "learning_rate": 3.540121367548811e-06,
+      "loss": 0.2946,
+      "step": 746
+    },
+    {
+      "epoch": 3.6439024390243904,
+      "grad_norm": 1.1034797430038452,
+      "learning_rate": 3.5366362653339524e-06,
+      "loss": 0.6456,
+      "step": 747
+    },
+    {
+      "epoch": 3.648780487804878,
+      "grad_norm": 1.0821247100830078,
+      "learning_rate": 3.533148728571124e-06,
+      "loss": 0.2399,
+      "step": 748
+    },
+    {
+      "epoch": 3.653658536585366,
+      "grad_norm": 1.058131456375122,
+      "learning_rate": 3.5296587654508317e-06,
+      "loss": 0.4446,
+      "step": 749
+    },
+    {
+      "epoch": 3.658536585365854,
+      "grad_norm": 2.5837414264678955,
+      "learning_rate": 3.526166384169279e-06,
+      "loss": 0.6744,
+      "step": 750
+    },
+    {
+      "epoch": 3.6634146341463416,
+      "grad_norm": 1.1982769966125488,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.3842,
+      "step": 751
+    },
+    {
+      "epoch": 3.6682926829268294,
+      "grad_norm": 0.849543035030365,
+      "learning_rate": 3.519174399935588e-06,
+      "loss": 0.2888,
+      "step": 752
+    },
+    {
+      "epoch": 3.6731707317073172,
+      "grad_norm": 0.9241414666175842,
+      "learning_rate": 3.5156748134041767e-06,
+      "loss": 0.2104,
+      "step": 753
+    },
+    {
+      "epoch": 3.678048780487805,
+      "grad_norm": 1.1095383167266846,
+      "learning_rate": 3.5121728415529203e-06,
+      "loss": 0.7921,
+      "step": 754
+    },
+    {
+      "epoch": 3.682926829268293,
+      "grad_norm": 1.0041614770889282,
+      "learning_rate": 3.5086684926062266e-06,
+      "loss": 0.3725,
+      "step": 755
+    },
+    {
+      "epoch": 3.68780487804878,
+      "grad_norm": 1.1060123443603516,
+      "learning_rate": 3.505161774794085e-06,
+      "loss": 0.4381,
+      "step": 756
+    },
+    {
+      "epoch": 3.692682926829268,
+      "grad_norm": 0.917390763759613,
+      "learning_rate": 3.5016526963520474e-06,
+      "loss": 0.3079,
+      "step": 757
+    },
+    {
+      "epoch": 3.697560975609756,
+      "grad_norm": 1.10245943069458,
+      "learning_rate": 3.498141265521212e-06,
+      "loss": 0.8131,
+      "step": 758
+    },
+    {
+      "epoch": 3.7024390243902436,
+      "grad_norm": 0.9380331039428711,
+      "learning_rate": 3.4946274905481997e-06,
+      "loss": 0.3149,
+      "step": 759
+    },
+    {
+      "epoch": 3.7073170731707314,
+      "grad_norm": 1.1269862651824951,
+      "learning_rate": 3.4911113796851364e-06,
+      "loss": 0.4485,
+      "step": 760
+    },
+    {
+      "epoch": 3.7121951219512193,
+      "grad_norm": 0.9700295925140381,
+      "learning_rate": 3.487592941189636e-06,
+      "loss": 0.2876,
+      "step": 761
+    },
+    {
+      "epoch": 3.717073170731707,
+      "grad_norm": 1.2680857181549072,
+      "learning_rate": 3.484072183324776e-06,
+      "loss": 0.7641,
+      "step": 762
+    },
+    {
+      "epoch": 3.721951219512195,
+      "grad_norm": 1.0977661609649658,
+      "learning_rate": 3.4805491143590823e-06,
+      "loss": 0.5787,
+      "step": 763
+    },
+    {
+      "epoch": 3.7268292682926827,
+      "grad_norm": 1.1288927793502808,
+      "learning_rate": 3.4770237425665103e-06,
+      "loss": 0.5136,
+      "step": 764
+    },
+    {
+      "epoch": 3.7317073170731705,
+      "grad_norm": 1.0365726947784424,
+      "learning_rate": 3.4734960762264204e-06,
+      "loss": 0.6113,
+      "step": 765
+    },
+    {
+      "epoch": 3.7365853658536583,
+      "grad_norm": 1.0862468481063843,
+      "learning_rate": 3.469966123623563e-06,
+      "loss": 0.6132,
+      "step": 766
+    },
+    {
+      "epoch": 3.741463414634146,
+      "grad_norm": 1.078573226928711,
+      "learning_rate": 3.46643389304806e-06,
+      "loss": 0.465,
+      "step": 767
+    },
+    {
+      "epoch": 3.746341463414634,
+      "grad_norm": 1.3158698081970215,
+      "learning_rate": 3.4628993927953786e-06,
+      "loss": 0.9712,
+      "step": 768
+    },
+    {
+      "epoch": 3.7512195121951217,
+      "grad_norm": 0.9415583610534668,
+      "learning_rate": 3.45936263116632e-06,
+      "loss": 0.2983,
+      "step": 769
+    },
+    {
+      "epoch": 3.7560975609756095,
+      "grad_norm": 0.8315958380699158,
+      "learning_rate": 3.4558236164669957e-06,
+      "loss": 0.3424,
+      "step": 770
+    },
+    {
+      "epoch": 3.7609756097560973,
+      "grad_norm": 1.0617963075637817,
+      "learning_rate": 3.4522823570088073e-06,
+      "loss": 0.2894,
+      "step": 771
+    },
+    {
+      "epoch": 3.765853658536585,
+      "grad_norm": 1.1169861555099487,
+      "learning_rate": 3.4487388611084295e-06,
+      "loss": 0.4487,
+      "step": 772
+    },
+    {
+      "epoch": 3.770731707317073,
+      "grad_norm": 1.055111050605774,
+      "learning_rate": 3.445193137087788e-06,
+      "loss": 0.2635,
+      "step": 773
+    },
+    {
+      "epoch": 3.7756097560975608,
+      "grad_norm": 0.9245924949645996,
+      "learning_rate": 3.4416451932740424e-06,
+      "loss": 0.4559,
+      "step": 774
+    },
+    {
+      "epoch": 3.7804878048780486,
+      "grad_norm": 1.205375075340271,
+      "learning_rate": 3.4380950379995652e-06,
+      "loss": 0.5927,
+      "step": 775
+    },
+    {
+      "epoch": 3.7853658536585364,
+      "grad_norm": 0.9713656902313232,
+      "learning_rate": 3.434542679601922e-06,
+      "loss": 0.469,
+      "step": 776
+    },
+    {
+      "epoch": 3.790243902439024,
+      "grad_norm": 1.037654161453247,
+      "learning_rate": 3.4309881264238538e-06,
+      "loss": 0.2425,
+      "step": 777
+    },
+    {
+      "epoch": 3.795121951219512,
+      "grad_norm": 1.1205179691314697,
+      "learning_rate": 3.4274313868132547e-06,
+      "loss": 0.3137,
+      "step": 778
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.0268551111221313,
+      "learning_rate": 3.4238724691231534e-06,
+      "loss": 0.3677,
+      "step": 779
+    },
+    {
+      "epoch": 3.8048780487804876,
+      "grad_norm": 1.1247752904891968,
+      "learning_rate": 3.4203113817116955e-06,
+      "loss": 0.5795,
+      "step": 780
+    },
+    {
+      "epoch": 3.8097560975609754,
+      "grad_norm": 1.093456745147705,
+      "learning_rate": 3.4167481329421204e-06,
+      "loss": 0.3284,
+      "step": 781
+    },
+    {
+      "epoch": 3.8146341463414632,
+      "grad_norm": 1.1700347661972046,
+      "learning_rate": 3.4131827311827447e-06,
+      "loss": 0.5421,
+      "step": 782
+    },
+    {
+      "epoch": 3.819512195121951,
+      "grad_norm": 0.8392572999000549,
+      "learning_rate": 3.4096151848069416e-06,
+      "loss": 0.2708,
+      "step": 783
+    },
+    {
+      "epoch": 3.824390243902439,
+      "grad_norm": 0.8408678770065308,
+      "learning_rate": 3.4060455021931195e-06,
+      "loss": 0.3812,
+      "step": 784
+    },
+    {
+      "epoch": 3.8292682926829267,
+      "grad_norm": 1.0810067653656006,
+      "learning_rate": 3.402473691724704e-06,
+      "loss": 0.3683,
+      "step": 785
+    },
+    {
+      "epoch": 3.8341463414634145,
+      "grad_norm": 0.7941387295722961,
+      "learning_rate": 3.39889976179012e-06,
+      "loss": 0.3159,
+      "step": 786
+    },
+    {
+      "epoch": 3.8390243902439023,
+      "grad_norm": 0.9092013835906982,
+      "learning_rate": 3.3953237207827673e-06,
+      "loss": 0.488,
+      "step": 787
+    },
+    {
+      "epoch": 3.84390243902439,
+      "grad_norm": 1.1730678081512451,
+      "learning_rate": 3.391745577101005e-06,
+      "loss": 0.729,
+      "step": 788
+    },
+    {
+      "epoch": 3.848780487804878,
+      "grad_norm": 1.2216498851776123,
+      "learning_rate": 3.3881653391481306e-06,
+      "loss": 0.4317,
+      "step": 789
+    },
+    {
+      "epoch": 3.8536585365853657,
+      "grad_norm": 1.2773486375808716,
+      "learning_rate": 3.384583015332359e-06,
+      "loss": 0.4652,
+      "step": 790
+    },
+    {
+      "epoch": 3.8585365853658535,
+      "grad_norm": 1.0143038034439087,
+      "learning_rate": 3.380998614066805e-06,
+      "loss": 0.3218,
+      "step": 791
+    },
+    {
+      "epoch": 3.8634146341463413,
+      "grad_norm": 1.1280455589294434,
+      "learning_rate": 3.3774121437694606e-06,
+      "loss": 0.4218,
+      "step": 792
+    },
+    {
+      "epoch": 3.868292682926829,
+      "grad_norm": 1.0607877969741821,
+      "learning_rate": 3.3738236128631786e-06,
+      "loss": 0.2838,
+      "step": 793
+    },
+    {
+      "epoch": 3.873170731707317,
+      "grad_norm": 1.145294189453125,
+      "learning_rate": 3.3702330297756503e-06,
+      "loss": 0.5267,
+      "step": 794
+    },
+    {
+      "epoch": 3.8780487804878048,
+      "grad_norm": 1.0045257806777954,
+      "learning_rate": 3.366640402939387e-06,
+      "loss": 0.1956,
+      "step": 795
+    },
+    {
+      "epoch": 3.8829268292682926,
+      "grad_norm": 1.9531723260879517,
+      "learning_rate": 3.363045740791698e-06,
+      "loss": 0.465,
+      "step": 796
+    },
+    {
+      "epoch": 3.8878048780487804,
+      "grad_norm": 0.8978796005249023,
+      "learning_rate": 3.3594490517746774e-06,
+      "loss": 0.3693,
+      "step": 797
+    },
+    {
+      "epoch": 3.892682926829268,
+      "grad_norm": 0.9718804955482483,
+      "learning_rate": 3.3558503443351733e-06,
+      "loss": 0.4562,
+      "step": 798
+    },
+    {
+      "epoch": 3.897560975609756,
+      "grad_norm": 0.888634204864502,
+      "learning_rate": 3.352249626924777e-06,
+      "loss": 0.3944,
+      "step": 799
+    },
+    {
+      "epoch": 3.902439024390244,
+      "grad_norm": 1.355827808380127,
+      "learning_rate": 3.348646907999801e-06,
+      "loss": 0.932,
+      "step": 800
+    },
+    {
+      "epoch": 3.9073170731707316,
+      "grad_norm": 1.005395770072937,
+      "learning_rate": 3.345042196021257e-06,
+      "loss": 0.4999,
+      "step": 801
+    },
+    {
+      "epoch": 3.9121951219512194,
+      "grad_norm": 1.253622055053711,
+      "learning_rate": 3.3414354994548385e-06,
+      "loss": 0.7352,
+      "step": 802
+    },
+    {
+      "epoch": 3.9170731707317072,
+      "grad_norm": 1.058511734008789,
+      "learning_rate": 3.337826826770898e-06,
+      "loss": 0.4555,
+      "step": 803
+    },
+    {
+      "epoch": 3.921951219512195,
+      "grad_norm": 1.0374521017074585,
+      "learning_rate": 3.3342161864444312e-06,
+      "loss": 0.3344,
+      "step": 804
+    },
+    {
+      "epoch": 3.926829268292683,
+      "grad_norm": 0.8886129856109619,
+      "learning_rate": 3.3306035869550534e-06,
+      "loss": 0.2988,
+      "step": 805
+    },
+    {
+      "epoch": 3.9317073170731707,
+      "grad_norm": 1.144066572189331,
+      "learning_rate": 3.326989036786981e-06,
+      "loss": 0.2734,
+      "step": 806
+    },
+    {
+      "epoch": 3.9365853658536585,
+      "grad_norm": 1.2550209760665894,
+      "learning_rate": 3.3233725444290126e-06,
+      "loss": 0.4412,
+      "step": 807
+    },
+    {
+      "epoch": 3.9414634146341463,
+      "grad_norm": 0.9990758299827576,
+      "learning_rate": 3.3197541183745065e-06,
+      "loss": 0.6003,
+      "step": 808
+    },
+    {
+      "epoch": 3.946341463414634,
+      "grad_norm": 0.9522280097007751,
+      "learning_rate": 3.3161337671213634e-06,
+      "loss": 0.3911,
+      "step": 809
+    },
+    {
+      "epoch": 3.951219512195122,
+      "grad_norm": 0.8392520546913147,
+      "learning_rate": 3.312511499172006e-06,
+      "loss": 0.3038,
+      "step": 810
+    },
+    {
+      "epoch": 3.9560975609756097,
+      "grad_norm": 1.0443915128707886,
+      "learning_rate": 3.3088873230333562e-06,
+      "loss": 0.4902,
+      "step": 811
+    },
+    {
+      "epoch": 3.9609756097560975,
+      "grad_norm": 1.5163060426712036,
+      "learning_rate": 3.3052612472168193e-06,
+      "loss": 0.3193,
+      "step": 812
+    },
+    {
+      "epoch": 3.9658536585365853,
+      "grad_norm": 1.0339828729629517,
+      "learning_rate": 3.3016332802382618e-06,
+      "loss": 0.4381,
+      "step": 813
+    },
+    {
+      "epoch": 3.970731707317073,
+      "grad_norm": 1.1660975217819214,
+      "learning_rate": 3.2980034306179897e-06,
+      "loss": 0.3802,
+      "step": 814
+    },
+    {
+      "epoch": 3.975609756097561,
+      "grad_norm": 0.9736506938934326,
+      "learning_rate": 3.294371706880733e-06,
+      "loss": 0.4898,
+      "step": 815
+    },
+    {
+      "epoch": 3.9804878048780488,
+      "grad_norm": 0.9874656796455383,
+      "learning_rate": 3.290738117555622e-06,
+      "loss": 0.3163,
+      "step": 816
+    },
+    {
+      "epoch": 3.9853658536585366,
+      "grad_norm": 1.0968129634857178,
+      "learning_rate": 3.2871026711761666e-06,
+      "loss": 0.5979,
+      "step": 817
+    },
+    {
+      "epoch": 3.9902439024390244,
+      "grad_norm": 1.0595239400863647,
+      "learning_rate": 3.2834653762802414e-06,
+      "loss": 0.3729,
+      "step": 818
+    },
+    {
+      "epoch": 3.995121951219512,
+      "grad_norm": 1.129920482635498,
+      "learning_rate": 3.2798262414100594e-06,
+      "loss": 0.3179,
+      "step": 819
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9918786883354187,
+      "learning_rate": 3.2761852751121566e-06,
+      "loss": 0.3184,
+      "step": 820
+    },
+    {
+      "epoch": 4.004878048780488,
+      "grad_norm": 1.0081284046173096,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.4687,
+      "step": 821
+    },
+    {
+      "epoch": 4.009756097560976,
+      "grad_norm": 1.025944471359253,
+      "learning_rate": 3.2688978824408136e-06,
+      "loss": 0.2941,
+      "step": 822
+    },
+    {
+      "epoch": 4.014634146341463,
+      "grad_norm": 0.9274433255195618,
+      "learning_rate": 3.2652514731818698e-06,
+      "loss": 0.2709,
+      "step": 823
+    },
+    {
+      "epoch": 4.019512195121951,
+      "grad_norm": 0.9776120185852051,
+      "learning_rate": 3.2616032667241564e-06,
+      "loss": 0.1915,
+      "step": 824
+    },
+    {
+      "epoch": 4.024390243902439,
+      "grad_norm": 1.0483906269073486,
+      "learning_rate": 3.257953271635513e-06,
+      "loss": 0.5291,
+      "step": 825
+    },
+    {
+      "epoch": 4.029268292682927,
+      "grad_norm": 0.9861155152320862,
+      "learning_rate": 3.2543014964879814e-06,
+      "loss": 0.2661,
+      "step": 826
+    },
+    {
+      "epoch": 4.034146341463415,
+      "grad_norm": 1.320236086845398,
+      "learning_rate": 3.250647949857781e-06,
+      "loss": 0.2301,
+      "step": 827
+    },
+    {
+      "epoch": 4.0390243902439025,
+      "grad_norm": 1.469008207321167,
+      "learning_rate": 3.2469926403252932e-06,
+      "loss": 0.6312,
+      "step": 828
+    },
+    {
+      "epoch": 4.04390243902439,
+      "grad_norm": 1.1796139478683472,
+      "learning_rate": 3.2433355764750417e-06,
+      "loss": 0.1662,
+      "step": 829
+    },
+    {
+      "epoch": 4.048780487804878,
+      "grad_norm": 1.3211941719055176,
+      "learning_rate": 3.2396767668956656e-06,
+      "loss": 0.514,
+      "step": 830
+    },
+    {
+      "epoch": 4.053658536585366,
+      "grad_norm": 1.2660949230194092,
+      "learning_rate": 3.2360162201799085e-06,
+      "loss": 0.4094,
+      "step": 831
+    },
+    {
+      "epoch": 4.058536585365854,
+      "grad_norm": 1.0212953090667725,
+      "learning_rate": 3.2323539449245906e-06,
+      "loss": 0.2655,
+      "step": 832
+    },
+    {
+      "epoch": 4.0634146341463415,
+      "grad_norm": 1.0902808904647827,
+      "learning_rate": 3.2286899497305917e-06,
+      "loss": 0.2253,
+      "step": 833
+    },
+    {
+      "epoch": 4.068292682926829,
+      "grad_norm": 0.9108849167823792,
+      "learning_rate": 3.2250242432028335e-06,
+      "loss": 0.3198,
+      "step": 834
+    },
+    {
+      "epoch": 4.073170731707317,
+      "grad_norm": 1.349802851676941,
+      "learning_rate": 3.221356833950254e-06,
+      "loss": 0.7985,
+      "step": 835
+    },
+    {
+      "epoch": 4.078048780487805,
+      "grad_norm": 0.9956075549125671,
+      "learning_rate": 3.21768773058579e-06,
+      "loss": 0.2803,
+      "step": 836
+    },
+    {
+      "epoch": 4.082926829268293,
+      "grad_norm": 1.123693823814392,
+      "learning_rate": 3.21401694172636e-06,
+      "loss": 0.2803,
+      "step": 837
+    },
+    {
+      "epoch": 4.087804878048781,
+      "grad_norm": 0.7673684358596802,
+      "learning_rate": 3.2103444759928383e-06,
+      "loss": 0.1641,
+      "step": 838
+    },
+    {
+      "epoch": 4.092682926829268,
+      "grad_norm": 0.8822228312492371,
+      "learning_rate": 3.2066703420100377e-06,
+      "loss": 0.1986,
+      "step": 839
+    },
+    {
+      "epoch": 4.097560975609756,
+      "grad_norm": 1.0781583786010742,
+      "learning_rate": 3.2029945484066883e-06,
+      "loss": 0.4772,
+      "step": 840
+    },
+    {
+      "epoch": 4.102439024390244,
+      "grad_norm": 1.1041297912597656,
+      "learning_rate": 3.1993171038154203e-06,
+      "loss": 0.2845,
+      "step": 841
+    },
+    {
+      "epoch": 4.107317073170732,
+      "grad_norm": 0.8251579403877258,
+      "learning_rate": 3.1956380168727385e-06,
+      "loss": 0.267,
+      "step": 842
+    },
+    {
+      "epoch": 4.11219512195122,
+      "grad_norm": 1.3584140539169312,
+      "learning_rate": 3.191957296219007e-06,
+      "loss": 0.4082,
+      "step": 843
+    },
+    {
+      "epoch": 4.117073170731707,
+      "grad_norm": 1.0828088521957397,
+      "learning_rate": 3.1882749504984247e-06,
+      "loss": 0.235,
+      "step": 844
+    },
+    {
+      "epoch": 4.121951219512195,
+      "grad_norm": 0.7371436357498169,
+      "learning_rate": 3.1845909883590076e-06,
+      "loss": 0.2477,
+      "step": 845
+    },
+    {
+      "epoch": 4.126829268292683,
+      "grad_norm": 1.1333115100860596,
+      "learning_rate": 3.180905418452569e-06,
+      "loss": 0.5014,
+      "step": 846
+    },
+    {
+      "epoch": 4.131707317073171,
+      "grad_norm": 0.8755896687507629,
+      "learning_rate": 3.1772182494346963e-06,
+      "loss": 0.3227,
+      "step": 847
+    },
+    {
+      "epoch": 4.136585365853659,
+      "grad_norm": 1.198217749595642,
+      "learning_rate": 3.1735294899647344e-06,
+      "loss": 0.415,
+      "step": 848
+    },
+    {
+      "epoch": 4.1414634146341465,
+      "grad_norm": 1.0420305728912354,
+      "learning_rate": 3.169839148705762e-06,
+      "loss": 0.2631,
+      "step": 849
+    },
+    {
+      "epoch": 4.146341463414634,
+      "grad_norm": 1.231887698173523,
+      "learning_rate": 3.1661472343245725e-06,
+      "loss": 0.7334,
+      "step": 850
+    },
+    {
+      "epoch": 4.151219512195122,
+      "grad_norm": 1.086815357208252,
+      "learning_rate": 3.162453755491655e-06,
+      "loss": 0.4588,
+      "step": 851
+    },
+    {
+      "epoch": 4.15609756097561,
+      "grad_norm": 1.2308528423309326,
+      "learning_rate": 3.158758720881171e-06,
+      "loss": 0.431,
+      "step": 852
+    },
+    {
+      "epoch": 4.160975609756098,
+      "grad_norm": 1.3028631210327148,
+      "learning_rate": 3.155062139170937e-06,
+      "loss": 0.351,
+      "step": 853
+    },
+    {
+      "epoch": 4.1658536585365855,
+      "grad_norm": 1.2701951265335083,
+      "learning_rate": 3.1513640190424034e-06,
+      "loss": 0.2138,
+      "step": 854
+    },
+    {
+      "epoch": 4.170731707317073,
+      "grad_norm": 1.1779377460479736,
+      "learning_rate": 3.147664369180632e-06,
+      "loss": 0.2532,
+      "step": 855
+    },
+    {
+      "epoch": 4.175609756097561,
+      "grad_norm": 1.287402868270874,
+      "learning_rate": 3.143963198274278e-06,
+      "loss": 0.4129,
+      "step": 856
+    },
+    {
+      "epoch": 4.180487804878049,
+      "grad_norm": 0.9432743787765503,
+      "learning_rate": 3.140260515015569e-06,
+      "loss": 0.2941,
+      "step": 857
+    },
+    {
+      "epoch": 4.185365853658537,
+      "grad_norm": 1.0425447225570679,
+      "learning_rate": 3.136556328100284e-06,
+      "loss": 0.3346,
+      "step": 858
+    },
+    {
+      "epoch": 4.190243902439025,
+      "grad_norm": 1.240814447402954,
+      "learning_rate": 3.132850646227734e-06,
+      "loss": 0.6134,
+      "step": 859
+    },
+    {
+      "epoch": 4.195121951219512,
+      "grad_norm": 1.1277090311050415,
+      "learning_rate": 3.12914347810074e-06,
+      "loss": 0.3543,
+      "step": 860
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.275528073310852,
+      "learning_rate": 3.125434832425613e-06,
+      "loss": 0.2518,
+      "step": 861
+    },
+    {
+      "epoch": 4.204878048780488,
+      "grad_norm": 0.7928908467292786,
+      "learning_rate": 3.121724717912138e-06,
+      "loss": 0.1532,
+      "step": 862
+    },
+    {
+      "epoch": 4.209756097560976,
+      "grad_norm": 1.1028555631637573,
+      "learning_rate": 3.118013143273542e-06,
+      "loss": 0.3368,
+      "step": 863
+    },
+    {
+      "epoch": 4.214634146341464,
+      "grad_norm": 1.8106095790863037,
+      "learning_rate": 3.1143001172264893e-06,
+      "loss": 0.2334,
+      "step": 864
+    },
+    {
+      "epoch": 4.219512195121951,
+      "grad_norm": 1.2573802471160889,
+      "learning_rate": 3.1105856484910474e-06,
+      "loss": 0.3072,
+      "step": 865
+    },
+    {
+      "epoch": 4.224390243902439,
+      "grad_norm": 0.870959460735321,
+      "learning_rate": 3.1068697457906736e-06,
+      "loss": 0.2215,
+      "step": 866
+    },
+    {
+      "epoch": 4.229268292682927,
+      "grad_norm": 1.066684603691101,
+      "learning_rate": 3.1031524178521938e-06,
+      "loss": 0.4311,
+      "step": 867
+    },
+    {
+      "epoch": 4.234146341463415,
+      "grad_norm": 1.2743711471557617,
+      "learning_rate": 3.0994336734057804e-06,
+      "loss": 0.1342,
+      "step": 868
+    },
+    {
+      "epoch": 4.239024390243903,
+      "grad_norm": 0.8251144289970398,
+      "learning_rate": 3.0957135211849315e-06,
+      "loss": 0.2224,
+      "step": 869
+    },
+    {
+      "epoch": 4.2439024390243905,
+      "grad_norm": 1.1386126279830933,
+      "learning_rate": 3.0919919699264535e-06,
+      "loss": 0.2674,
+      "step": 870
+    },
+    {
+      "epoch": 4.248780487804878,
+      "grad_norm": 1.1335779428482056,
+      "learning_rate": 3.0882690283704355e-06,
+      "loss": 0.8506,
+      "step": 871
+    },
+    {
+      "epoch": 4.253658536585366,
+      "grad_norm": 0.9600904583930969,
+      "learning_rate": 3.084544705260234e-06,
+      "loss": 0.2926,
+      "step": 872
+    },
+    {
+      "epoch": 4.258536585365854,
+      "grad_norm": 0.896686851978302,
+      "learning_rate": 3.080819009342451e-06,
+      "loss": 0.1941,
+      "step": 873
+    },
+    {
+      "epoch": 4.263414634146342,
+      "grad_norm": 0.8546966314315796,
+      "learning_rate": 3.077091949366908e-06,
+      "loss": 0.2209,
+      "step": 874
+    },
+    {
+      "epoch": 4.2682926829268295,
+      "grad_norm": 0.9326870441436768,
+      "learning_rate": 3.073363534086636e-06,
+      "loss": 0.2608,
+      "step": 875
+    },
+    {
+      "epoch": 4.273170731707317,
+      "grad_norm": 0.8635675311088562,
+      "learning_rate": 3.0696337722578444e-06,
+      "loss": 0.1911,
+      "step": 876
+    },
+    {
+      "epoch": 4.278048780487805,
+      "grad_norm": 1.2039563655853271,
+      "learning_rate": 3.0659026726399072e-06,
+      "loss": 0.3195,
+      "step": 877
+    },
+    {
+      "epoch": 4.282926829268293,
+      "grad_norm": 1.041063666343689,
+      "learning_rate": 3.0621702439953393e-06,
+      "loss": 0.4438,
+      "step": 878
+    },
+    {
+      "epoch": 4.287804878048781,
+      "grad_norm": 0.9710636734962463,
+      "learning_rate": 3.0584364950897768e-06,
+      "loss": 0.1192,
+      "step": 879
+    },
+    {
+      "epoch": 4.2926829268292686,
+      "grad_norm": 1.001789927482605,
+      "learning_rate": 3.0547014346919574e-06,
+      "loss": 0.2768,
+      "step": 880
+    },
+    {
+      "epoch": 4.297560975609756,
+      "grad_norm": 1.0699266195297241,
+      "learning_rate": 3.0509650715736977e-06,
+      "loss": 0.2787,
+      "step": 881
+    },
+    {
+      "epoch": 4.302439024390244,
+      "grad_norm": 0.9404529929161072,
+      "learning_rate": 3.0472274145098744e-06,
+      "loss": 0.4201,
+      "step": 882
+    },
+    {
+      "epoch": 4.307317073170732,
+      "grad_norm": 1.296057105064392,
+      "learning_rate": 3.0434884722784026e-06,
+      "loss": 0.3718,
+      "step": 883
+    },
+    {
+      "epoch": 4.31219512195122,
+      "grad_norm": 0.9642940163612366,
+      "learning_rate": 3.0397482536602168e-06,
+      "loss": 0.2024,
+      "step": 884
+    },
+    {
+      "epoch": 4.317073170731708,
+      "grad_norm": 1.0013500452041626,
+      "learning_rate": 3.0360067674392475e-06,
+      "loss": 0.2857,
+      "step": 885
+    },
+    {
+      "epoch": 4.321951219512195,
+      "grad_norm": 1.1276105642318726,
+      "learning_rate": 3.0322640224024024e-06,
+      "loss": 0.2326,
+      "step": 886
+    },
+    {
+      "epoch": 4.326829268292683,
+      "grad_norm": 1.1642167568206787,
+      "learning_rate": 3.0285200273395478e-06,
+      "loss": 0.3897,
+      "step": 887
+    },
+    {
+      "epoch": 4.331707317073171,
+      "grad_norm": 1.1692358255386353,
+      "learning_rate": 3.024774791043481e-06,
+      "loss": 0.4172,
+      "step": 888
+    },
+    {
+      "epoch": 4.336585365853659,
+      "grad_norm": 1.2507195472717285,
+      "learning_rate": 3.021028322309921e-06,
+      "loss": 0.4637,
+      "step": 889
+    },
+    {
+      "epoch": 4.341463414634147,
+      "grad_norm": 1.2589614391326904,
+      "learning_rate": 3.0172806299374734e-06,
+      "loss": 0.4284,
+      "step": 890
+    },
+    {
+      "epoch": 4.3463414634146345,
+      "grad_norm": 1.159298300743103,
+      "learning_rate": 3.0135317227276247e-06,
+      "loss": 0.3921,
+      "step": 891
+    },
+    {
+      "epoch": 4.351219512195122,
+      "grad_norm": 1.0309828519821167,
+      "learning_rate": 3.0097816094847104e-06,
+      "loss": 0.3799,
+      "step": 892
+    },
+    {
+      "epoch": 4.35609756097561,
+      "grad_norm": 0.9921269416809082,
+      "learning_rate": 3.0060302990158984e-06,
+      "loss": 0.2218,
+      "step": 893
+    },
+    {
+      "epoch": 4.360975609756098,
+      "grad_norm": 0.9881783127784729,
+      "learning_rate": 3.002277800131171e-06,
+      "loss": 0.1882,
+      "step": 894
+    },
+    {
+      "epoch": 4.365853658536586,
+      "grad_norm": 1.0279638767242432,
+      "learning_rate": 2.998524121643298e-06,
+      "loss": 0.2855,
+      "step": 895
+    },
+    {
+      "epoch": 4.3707317073170735,
+      "grad_norm": 1.2783422470092773,
+      "learning_rate": 2.994769272367822e-06,
+      "loss": 0.3899,
+      "step": 896
+    },
+    {
+      "epoch": 4.375609756097561,
+      "grad_norm": 0.9480025172233582,
+      "learning_rate": 2.991013261123035e-06,
+      "loss": 0.2136,
+      "step": 897
+    },
+    {
+      "epoch": 4.380487804878049,
+      "grad_norm": 1.1166763305664062,
+      "learning_rate": 2.9872560967299554e-06,
+      "loss": 0.2992,
+      "step": 898
+    },
+    {
+      "epoch": 4.385365853658537,
+      "grad_norm": 1.1016902923583984,
+      "learning_rate": 2.9834977880123132e-06,
+      "loss": 0.1948,
+      "step": 899
+    },
+    {
+      "epoch": 4.390243902439025,
+      "grad_norm": 1.0177924633026123,
+      "learning_rate": 2.9797383437965243e-06,
+      "loss": 0.2447,
+      "step": 900
+    },
+    {
+      "epoch": 4.3951219512195125,
+      "grad_norm": 0.910971999168396,
+      "learning_rate": 2.975977772911671e-06,
+      "loss": 0.2198,
+      "step": 901
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.0623188018798828,
+      "learning_rate": 2.972216084189482e-06,
+      "loss": 0.2353,
+      "step": 902
+    },
+    {
+      "epoch": 4.404878048780488,
+      "grad_norm": 0.8617135286331177,
+      "learning_rate": 2.9684532864643123e-06,
+      "loss": 0.1998,
+      "step": 903
+    },
+    {
+      "epoch": 4.409756097560976,
+      "grad_norm": 1.0422427654266357,
+      "learning_rate": 2.964689388573118e-06,
+      "loss": 0.1688,
+      "step": 904
+    },
+    {
+      "epoch": 4.414634146341464,
+      "grad_norm": 1.0696886777877808,
+      "learning_rate": 2.9609243993554434e-06,
+      "loss": 0.4587,
+      "step": 905
+    },
+    {
+      "epoch": 4.419512195121952,
+      "grad_norm": 1.1068525314331055,
+      "learning_rate": 2.9571583276533923e-06,
+      "loss": 0.2282,
+      "step": 906
+    },
+    {
+      "epoch": 4.424390243902439,
+      "grad_norm": 1.363254189491272,
+      "learning_rate": 2.9533911823116124e-06,
+      "loss": 0.7017,
+      "step": 907
+    },
+    {
+      "epoch": 4.429268292682927,
+      "grad_norm": 1.504699945449829,
+      "learning_rate": 2.9496229721772734e-06,
+      "loss": 0.4154,
+      "step": 908
+    },
+    {
+      "epoch": 4.434146341463415,
+      "grad_norm": 1.0644068717956543,
+      "learning_rate": 2.9458537061000435e-06,
+      "loss": 0.3262,
+      "step": 909
+    },
+    {
+      "epoch": 4.439024390243903,
+      "grad_norm": 1.1175642013549805,
+      "learning_rate": 2.9420833929320726e-06,
+      "loss": 0.3613,
+      "step": 910
+    },
+    {
+      "epoch": 4.443902439024391,
+      "grad_norm": 1.418470859527588,
+      "learning_rate": 2.93831204152797e-06,
+      "loss": 0.5274,
+      "step": 911
+    },
+    {
+      "epoch": 4.4487804878048784,
+      "grad_norm": 0.8108303546905518,
+      "learning_rate": 2.9345396607447807e-06,
+      "loss": 0.1267,
+      "step": 912
+    },
+    {
+      "epoch": 4.453658536585366,
+      "grad_norm": 1.049116611480713,
+      "learning_rate": 2.9307662594419704e-06,
+      "loss": 0.2678,
+      "step": 913
+    },
+    {
+      "epoch": 4.458536585365854,
+      "grad_norm": 1.158342719078064,
+      "learning_rate": 2.9269918464814e-06,
+      "loss": 0.3162,
+      "step": 914
+    },
+    {
+      "epoch": 4.463414634146342,
+      "grad_norm": 1.4344679117202759,
+      "learning_rate": 2.923216430727306e-06,
+      "loss": 0.4786,
+      "step": 915
+    },
+    {
+      "epoch": 4.46829268292683,
+      "grad_norm": 1.156111478805542,
+      "learning_rate": 2.9194400210462808e-06,
+      "loss": 0.4763,
+      "step": 916
+    },
+    {
+      "epoch": 4.473170731707317,
+      "grad_norm": 1.1826188564300537,
+      "learning_rate": 2.91566262630725e-06,
+      "loss": 0.5285,
+      "step": 917
+    },
+    {
+      "epoch": 4.478048780487805,
+      "grad_norm": 1.2508891820907593,
+      "learning_rate": 2.9118842553814526e-06,
+      "loss": 0.2565,
+      "step": 918
+    },
+    {
+      "epoch": 4.482926829268292,
+      "grad_norm": 0.9462976455688477,
+      "learning_rate": 2.9081049171424223e-06,
+      "loss": 0.2208,
+      "step": 919
+    },
+    {
+      "epoch": 4.487804878048781,
+      "grad_norm": 0.9479988813400269,
+      "learning_rate": 2.9043246204659624e-06,
+      "loss": 0.2,
+      "step": 920
+    },
+    {
+      "epoch": 4.492682926829268,
+      "grad_norm": 1.0272135734558105,
+      "learning_rate": 2.9005433742301274e-06,
+      "loss": 0.4585,
+      "step": 921
+    },
+    {
+      "epoch": 4.4975609756097565,
+      "grad_norm": 1.7547694444656372,
+      "learning_rate": 2.8967611873152037e-06,
+      "loss": 0.3949,
+      "step": 922
+    },
+    {
+      "epoch": 4.5024390243902435,
+      "grad_norm": 1.0447901487350464,
+      "learning_rate": 2.892978068603683e-06,
+      "loss": 0.1928,
+      "step": 923
+    },
+    {
+      "epoch": 4.507317073170732,
+      "grad_norm": 1.0131272077560425,
+      "learning_rate": 2.889194026980249e-06,
+      "loss": 0.3458,
+      "step": 924
+    },
+    {
+      "epoch": 4.512195121951219,
+      "grad_norm": 1.1343997716903687,
+      "learning_rate": 2.8854090713317514e-06,
+      "loss": 0.1519,
+      "step": 925
+    },
+    {
+      "epoch": 4.517073170731708,
+      "grad_norm": 1.2432359457015991,
+      "learning_rate": 2.8816232105471864e-06,
+      "loss": 0.1789,
+      "step": 926
+    },
+    {
+      "epoch": 4.521951219512195,
+      "grad_norm": 1.2051594257354736,
+      "learning_rate": 2.877836453517677e-06,
+      "loss": 0.518,
+      "step": 927
+    },
+    {
+      "epoch": 4.526829268292683,
+      "grad_norm": 1.1518090963363647,
+      "learning_rate": 2.8740488091364492e-06,
+      "loss": 0.4737,
+      "step": 928
+    },
+    {
+      "epoch": 4.53170731707317,
+      "grad_norm": 1.456160068511963,
+      "learning_rate": 2.870260286298814e-06,
+      "loss": 0.6425,
+      "step": 929
+    },
+    {
+      "epoch": 4.536585365853659,
+      "grad_norm": 1.2971833944320679,
+      "learning_rate": 2.866470893902147e-06,
+      "loss": 0.3322,
+      "step": 930
+    },
+    {
+      "epoch": 4.541463414634146,
+      "grad_norm": 1.0898277759552002,
+      "learning_rate": 2.8626806408458626e-06,
+      "loss": 0.258,
+      "step": 931
+    },
+    {
+      "epoch": 4.546341463414635,
+      "grad_norm": 0.9179545044898987,
+      "learning_rate": 2.8588895360313983e-06,
+      "loss": 0.2085,
+      "step": 932
+    },
+    {
+      "epoch": 4.5512195121951216,
+      "grad_norm": 1.0029069185256958,
+      "learning_rate": 2.8550975883621935e-06,
+      "loss": 0.3558,
+      "step": 933
+    },
+    {
+      "epoch": 4.55609756097561,
+      "grad_norm": 1.0553942918777466,
+      "learning_rate": 2.8513048067436644e-06,
+      "loss": 0.3892,
+      "step": 934
+    },
+    {
+      "epoch": 4.560975609756097,
+      "grad_norm": 1.0313464403152466,
+      "learning_rate": 2.847511200083187e-06,
+      "loss": 0.2626,
+      "step": 935
+    },
+    {
+      "epoch": 4.565853658536585,
+      "grad_norm": 0.9714272618293762,
+      "learning_rate": 2.843716777290074e-06,
+      "loss": 0.2896,
+      "step": 936
+    },
+    {
+      "epoch": 4.570731707317073,
+      "grad_norm": 1.239687204360962,
+      "learning_rate": 2.839921547275556e-06,
+      "loss": 0.2688,
+      "step": 937
+    },
+    {
+      "epoch": 4.575609756097561,
+      "grad_norm": 1.451963186264038,
+      "learning_rate": 2.836125518952759e-06,
+      "loss": 0.2953,
+      "step": 938
+    },
+    {
+      "epoch": 4.580487804878048,
+      "grad_norm": 1.0113798379898071,
+      "learning_rate": 2.8323287012366845e-06,
+      "loss": 0.4247,
+      "step": 939
+    },
+    {
+      "epoch": 4.585365853658536,
+      "grad_norm": 1.1036227941513062,
+      "learning_rate": 2.828531103044186e-06,
+      "loss": 0.2388,
+      "step": 940
+    },
+    {
+      "epoch": 4.590243902439024,
+      "grad_norm": 1.0381038188934326,
+      "learning_rate": 2.8247327332939512e-06,
+      "loss": 0.3913,
+      "step": 941
+    },
+    {
+      "epoch": 4.595121951219512,
+      "grad_norm": 1.0053890943527222,
+      "learning_rate": 2.82093360090648e-06,
+      "loss": 0.4356,
+      "step": 942
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.2142903804779053,
+      "learning_rate": 2.8171337148040636e-06,
+      "loss": 0.4161,
+      "step": 943
+    },
+    {
+      "epoch": 4.6048780487804875,
+      "grad_norm": 1.8541486263275146,
+      "learning_rate": 2.813333083910761e-06,
+      "loss": 0.2803,
+      "step": 944
+    },
+    {
+      "epoch": 4.609756097560975,
+      "grad_norm": 0.9036626815795898,
+      "learning_rate": 2.8095317171523835e-06,
+      "loss": 0.1578,
+      "step": 945
+    },
+    {
+      "epoch": 4.614634146341463,
+      "grad_norm": 1.1875560283660889,
+      "learning_rate": 2.805729623456469e-06,
+      "loss": 0.4107,
+      "step": 946
+    },
+    {
+      "epoch": 4.619512195121951,
+      "grad_norm": 1.259789228439331,
+      "learning_rate": 2.8019268117522624e-06,
+      "loss": 0.5262,
+      "step": 947
+    },
+    {
+      "epoch": 4.624390243902439,
+      "grad_norm": 1.116606593132019,
+      "learning_rate": 2.798123290970695e-06,
+      "loss": 0.3334,
+      "step": 948
+    },
+    {
+      "epoch": 4.6292682926829265,
+      "grad_norm": 1.0191316604614258,
+      "learning_rate": 2.794319070044365e-06,
+      "loss": 0.395,
+      "step": 949
+    },
+    {
+      "epoch": 4.634146341463414,
+      "grad_norm": 1.1099485158920288,
+      "learning_rate": 2.790514157907512e-06,
+      "loss": 0.2684,
+      "step": 950
+    },
+    {
+      "epoch": 4.639024390243902,
+      "grad_norm": 0.9431113004684448,
+      "learning_rate": 2.786708563496002e-06,
+      "loss": 0.2696,
+      "step": 951
+    },
+    {
+      "epoch": 4.64390243902439,
+      "grad_norm": 1.19779372215271,
+      "learning_rate": 2.782902295747299e-06,
+      "loss": 0.5113,
+      "step": 952
+    },
+    {
+      "epoch": 4.648780487804878,
+      "grad_norm": 1.1742883920669556,
+      "learning_rate": 2.7790953636004536e-06,
+      "loss": 0.6055,
+      "step": 953
+    },
+    {
+      "epoch": 4.6536585365853655,
+      "grad_norm": 1.2542123794555664,
+      "learning_rate": 2.775287775996074e-06,
+      "loss": 0.5556,
+      "step": 954
+    },
+    {
+      "epoch": 4.658536585365853,
+      "grad_norm": 1.0845754146575928,
+      "learning_rate": 2.7714795418763067e-06,
+      "loss": 0.2445,
+      "step": 955
+    },
+    {
+      "epoch": 4.663414634146341,
+      "grad_norm": 1.3217185735702515,
+      "learning_rate": 2.7676706701848187e-06,
+      "loss": 0.4524,
+      "step": 956
+    },
+    {
+      "epoch": 4.668292682926829,
+      "grad_norm": 0.9640854597091675,
+      "learning_rate": 2.763861169866774e-06,
+      "loss": 0.4953,
+      "step": 957
+    },
+    {
+      "epoch": 4.673170731707317,
+      "grad_norm": 0.864771842956543,
+      "learning_rate": 2.7600510498688104e-06,
+      "loss": 0.1544,
+      "step": 958
+    },
+    {
+      "epoch": 4.678048780487805,
+      "grad_norm": 2.0469486713409424,
+      "learning_rate": 2.7562403191390246e-06,
+      "loss": 0.2998,
+      "step": 959
+    },
+    {
+      "epoch": 4.682926829268292,
+      "grad_norm": 1.0981053113937378,
+      "learning_rate": 2.7524289866269467e-06,
+      "loss": 0.3188,
+      "step": 960
+    },
+    {
+      "epoch": 4.68780487804878,
+      "grad_norm": 0.959457278251648,
+      "learning_rate": 2.748617061283518e-06,
+      "loss": 0.2297,
+      "step": 961
+    },
+    {
+      "epoch": 4.692682926829268,
+      "grad_norm": 1.2737903594970703,
+      "learning_rate": 2.744804552061074e-06,
+      "loss": 0.2563,
+      "step": 962
+    },
+    {
+      "epoch": 4.697560975609756,
+      "grad_norm": 1.0287935733795166,
+      "learning_rate": 2.740991467913321e-06,
+      "loss": 0.4164,
+      "step": 963
+    },
+    {
+      "epoch": 4.702439024390244,
+      "grad_norm": 1.0602225065231323,
+      "learning_rate": 2.737177817795315e-06,
+      "loss": 0.2201,
+      "step": 964
+    },
+    {
+      "epoch": 4.7073170731707314,
+      "grad_norm": 1.0084565877914429,
+      "learning_rate": 2.7333636106634414e-06,
+      "loss": 0.283,
+      "step": 965
+    },
+    {
+      "epoch": 4.712195121951219,
+      "grad_norm": 1.2561463117599487,
+      "learning_rate": 2.7295488554753957e-06,
+      "loss": 0.4592,
+      "step": 966
+    },
+    {
+      "epoch": 4.717073170731707,
+      "grad_norm": 1.0347182750701904,
+      "learning_rate": 2.725733561190157e-06,
+      "loss": 0.2642,
+      "step": 967
+    },
+    {
+      "epoch": 4.721951219512195,
+      "grad_norm": 1.312893033027649,
+      "learning_rate": 2.721917736767973e-06,
+      "loss": 0.335,
+      "step": 968
+    },
+    {
+      "epoch": 4.726829268292683,
+      "grad_norm": 1.0608656406402588,
+      "learning_rate": 2.7181013911703357e-06,
+      "loss": 0.2288,
+      "step": 969
+    },
+    {
+      "epoch": 4.7317073170731705,
+      "grad_norm": 1.0253140926361084,
+      "learning_rate": 2.714284533359961e-06,
+      "loss": 0.2827,
+      "step": 970
+    },
+    {
+      "epoch": 4.736585365853658,
+      "grad_norm": 0.9973790049552917,
+      "learning_rate": 2.710467172300768e-06,
+      "loss": 0.4102,
+      "step": 971
+    },
+    {
+      "epoch": 4.741463414634146,
+      "grad_norm": 1.2393323183059692,
+      "learning_rate": 2.706649316957857e-06,
+      "loss": 0.4489,
+      "step": 972
+    },
+    {
+      "epoch": 4.746341463414634,
+      "grad_norm": 1.1221544742584229,
+      "learning_rate": 2.7028309762974897e-06,
+      "loss": 0.1774,
+      "step": 973
+    },
+    {
+      "epoch": 4.751219512195122,
+      "grad_norm": 1.07940673828125,
+      "learning_rate": 2.699012159287069e-06,
+      "loss": 0.3329,
+      "step": 974
+    },
+    {
+      "epoch": 4.7560975609756095,
+      "grad_norm": 1.027277946472168,
+      "learning_rate": 2.6951928748951125e-06,
+      "loss": 0.2044,
+      "step": 975
+    },
+    {
+      "epoch": 4.760975609756097,
+      "grad_norm": 1.0281989574432373,
+      "learning_rate": 2.69137313209124e-06,
+      "loss": 0.3622,
+      "step": 976
+    },
+    {
+      "epoch": 4.765853658536585,
+      "grad_norm": 1.1298357248306274,
+      "learning_rate": 2.687552939846145e-06,
+      "loss": 0.4628,
+      "step": 977
+    },
+    {
+      "epoch": 4.770731707317073,
+      "grad_norm": 1.0114378929138184,
+      "learning_rate": 2.6837323071315766e-06,
+      "loss": 0.1181,
+      "step": 978
+    },
+    {
+      "epoch": 4.775609756097561,
+      "grad_norm": 1.000329613685608,
+      "learning_rate": 2.679911242920321e-06,
+      "loss": 0.2879,
+      "step": 979
+    },
+    {
+      "epoch": 4.780487804878049,
+      "grad_norm": 1.3905553817749023,
+      "learning_rate": 2.6760897561861742e-06,
+      "loss": 0.6606,
+      "step": 980
+    },
+    {
+      "epoch": 4.785365853658536,
+      "grad_norm": 0.9317846894264221,
+      "learning_rate": 2.672267855903927e-06,
+      "loss": 0.1451,
+      "step": 981
+    },
+    {
+      "epoch": 4.790243902439024,
+      "grad_norm": 0.9949944615364075,
+      "learning_rate": 2.6684455510493413e-06,
+      "loss": 0.3478,
+      "step": 982
+    },
+    {
+      "epoch": 4.795121951219512,
+      "grad_norm": 0.9357514381408691,
+      "learning_rate": 2.6646228505991267e-06,
+      "loss": 0.3766,
+      "step": 983
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.4344565868377686,
+      "learning_rate": 2.6607997635309246e-06,
+      "loss": 0.3122,
+      "step": 984
+    },
+    {
+      "epoch": 4.804878048780488,
+      "grad_norm": 1.2585278749465942,
+      "learning_rate": 2.6569762988232838e-06,
+      "loss": 0.356,
+      "step": 985
+    },
+    {
+      "epoch": 4.809756097560975,
+      "grad_norm": 1.0323916673660278,
+      "learning_rate": 2.653152465455639e-06,
+      "loss": 0.5199,
+      "step": 986
+    },
+    {
+      "epoch": 4.814634146341463,
+      "grad_norm": 1.1831841468811035,
+      "learning_rate": 2.6493282724082913e-06,
+      "loss": 0.5155,
+      "step": 987
+    },
+    {
+      "epoch": 4.819512195121951,
+      "grad_norm": 1.177829384803772,
+      "learning_rate": 2.6455037286623864e-06,
+      "loss": 0.263,
+      "step": 988
+    },
+    {
+      "epoch": 4.824390243902439,
+      "grad_norm": 0.7737200260162354,
+      "learning_rate": 2.6416788431998935e-06,
+      "loss": 0.2256,
+      "step": 989
+    },
+    {
+      "epoch": 4.829268292682927,
+      "grad_norm": 1.0419354438781738,
+      "learning_rate": 2.637853625003585e-06,
+      "loss": 0.4914,
+      "step": 990
+    },
+    {
+      "epoch": 4.8341463414634145,
+      "grad_norm": 1.000624418258667,
+      "learning_rate": 2.6340280830570142e-06,
+      "loss": 0.3594,
+      "step": 991
+    },
+    {
+      "epoch": 4.839024390243902,
+      "grad_norm": 1.0777456760406494,
+      "learning_rate": 2.6302022263444947e-06,
+      "loss": 0.263,
+      "step": 992
+    },
+    {
+      "epoch": 4.84390243902439,
+      "grad_norm": 1.333800196647644,
+      "learning_rate": 2.6263760638510793e-06,
+      "loss": 0.5697,
+      "step": 993
+    },
+    {
+      "epoch": 4.848780487804878,
+      "grad_norm": 1.0035558938980103,
+      "learning_rate": 2.6225496045625394e-06,
+      "loss": 0.2984,
+      "step": 994
+    },
+    {
+      "epoch": 4.853658536585366,
+      "grad_norm": 1.214937448501587,
+      "learning_rate": 2.6187228574653428e-06,
+      "loss": 0.3698,
+      "step": 995
+    },
+    {
+      "epoch": 4.8585365853658535,
+      "grad_norm": 0.9136359095573425,
+      "learning_rate": 2.614895831546633e-06,
+      "loss": 0.2602,
+      "step": 996
+    },
+    {
+      "epoch": 4.863414634146341,
+      "grad_norm": 1.2129359245300293,
+      "learning_rate": 2.6110685357942096e-06,
+      "loss": 0.4928,
+      "step": 997
+    },
+    {
+      "epoch": 4.868292682926829,
+      "grad_norm": 1.2663754224777222,
+      "learning_rate": 2.6072409791965048e-06,
+      "loss": 0.5422,
+      "step": 998
+    },
+    {
+      "epoch": 4.873170731707317,
+      "grad_norm": 1.4935377836227417,
+      "learning_rate": 2.6034131707425638e-06,
+      "loss": 0.7989,
+      "step": 999
+    },
+    {
+      "epoch": 4.878048780487805,
+      "grad_norm": 0.999937891960144,
+      "learning_rate": 2.5995851194220223e-06,
+      "loss": 0.3162,
+      "step": 1000
+    },
+    {
+      "epoch": 4.882926829268293,
+      "grad_norm": 1.0912432670593262,
+      "learning_rate": 2.595756834225089e-06,
+      "loss": 0.329,
+      "step": 1001
+    },
+    {
+      "epoch": 4.88780487804878,
+      "grad_norm": 1.195001244544983,
+      "learning_rate": 2.5919283241425188e-06,
+      "loss": 0.2323,
+      "step": 1002
+    },
+    {
+      "epoch": 4.892682926829268,
+      "grad_norm": 0.7830819487571716,
+      "learning_rate": 2.5880995981655965e-06,
+      "loss": 0.2372,
+      "step": 1003
+    },
+    {
+      "epoch": 4.897560975609756,
+      "grad_norm": 1.099342703819275,
+      "learning_rate": 2.584270665286113e-06,
+      "loss": 0.1749,
+      "step": 1004
+    },
+    {
+      "epoch": 4.902439024390244,
+      "grad_norm": 1.270918846130371,
+      "learning_rate": 2.580441534496346e-06,
+      "loss": 0.3421,
+      "step": 1005
+    },
+    {
+      "epoch": 4.907317073170732,
+      "grad_norm": 0.9069335460662842,
+      "learning_rate": 2.576612214789039e-06,
+      "loss": 0.2538,
+      "step": 1006
+    },
+    {
+      "epoch": 4.912195121951219,
+      "grad_norm": 1.377794623374939,
+      "learning_rate": 2.5727827151573747e-06,
+      "loss": 0.4034,
+      "step": 1007
+    },
+    {
+      "epoch": 4.917073170731707,
+      "grad_norm": 1.1331335306167603,
+      "learning_rate": 2.568953044594964e-06,
+      "loss": 0.2865,
+      "step": 1008
+    },
+    {
+      "epoch": 4.921951219512195,
+      "grad_norm": 1.0799779891967773,
+      "learning_rate": 2.5651232120958157e-06,
+      "loss": 0.3365,
+      "step": 1009
+    },
+    {
+      "epoch": 4.926829268292683,
+      "grad_norm": 1.1052212715148926,
+      "learning_rate": 2.56129322665432e-06,
+      "loss": 0.2303,
+      "step": 1010
+    },
+    {
+      "epoch": 4.931707317073171,
+      "grad_norm": 0.913105845451355,
+      "learning_rate": 2.5574630972652263e-06,
+      "loss": 0.1745,
+      "step": 1011
+    },
+    {
+      "epoch": 4.9365853658536585,
+      "grad_norm": 1.0110231637954712,
+      "learning_rate": 2.553632832923622e-06,
+      "loss": 0.3202,
+      "step": 1012
+    },
+    {
+      "epoch": 4.941463414634146,
+      "grad_norm": 1.1608706712722778,
+      "learning_rate": 2.5498024426249107e-06,
+      "loss": 0.5757,
+      "step": 1013
+    },
+    {
+      "epoch": 4.946341463414634,
+      "grad_norm": 0.9768334627151489,
+      "learning_rate": 2.545971935364794e-06,
+      "loss": 0.1663,
+      "step": 1014
+    },
+    {
+      "epoch": 4.951219512195122,
+      "grad_norm": 1.130802035331726,
+      "learning_rate": 2.5421413201392443e-06,
+      "loss": 0.3053,
+      "step": 1015
+    },
+    {
+      "epoch": 4.95609756097561,
+      "grad_norm": 1.5004934072494507,
+      "learning_rate": 2.538310605944491e-06,
+      "loss": 0.2943,
+      "step": 1016
+    },
+    {
+      "epoch": 4.9609756097560975,
+      "grad_norm": 0.9939762353897095,
+      "learning_rate": 2.534479801776996e-06,
+      "loss": 0.2933,
+      "step": 1017
+    },
+    {
+      "epoch": 4.965853658536585,
+      "grad_norm": 0.9881106019020081,
+      "learning_rate": 2.53064891663343e-06,
+      "loss": 0.3813,
+      "step": 1018
+    },
+    {
+      "epoch": 4.970731707317073,
+      "grad_norm": 0.9731859564781189,
+      "learning_rate": 2.526817959510655e-06,
+      "loss": 0.3543,
+      "step": 1019
+    },
+    {
+      "epoch": 4.975609756097561,
+      "grad_norm": 1.3951330184936523,
+      "learning_rate": 2.5229869394057038e-06,
+      "loss": 0.4254,
+      "step": 1020
+    },
+    {
+      "epoch": 4.980487804878049,
+      "grad_norm": 0.9658120274543762,
+      "learning_rate": 2.5191558653157542e-06,
+      "loss": 0.2484,
+      "step": 1021
+    },
+    {
+      "epoch": 4.985365853658537,
+      "grad_norm": 0.8997960090637207,
+      "learning_rate": 2.515324746238113e-06,
+      "loss": 0.1693,
+      "step": 1022
+    },
+    {
+      "epoch": 4.990243902439024,
+      "grad_norm": 0.9819786548614502,
+      "learning_rate": 2.511493591170191e-06,
+      "loss": 0.3336,
+      "step": 1023
+    },
+    {
+      "epoch": 4.995121951219512,
+      "grad_norm": 1.3949546813964844,
+      "learning_rate": 2.5076624091094846e-06,
+      "loss": 0.2995,
+      "step": 1024
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.9916773438453674,
+      "learning_rate": 2.503831209053554e-06,
+      "loss": 0.3243,
+      "step": 1025
+    },
+    {
+      "epoch": 5.004878048780488,
+      "grad_norm": 0.9542043805122375,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1981,
+      "step": 1026
+    },
+    {
+      "epoch": 5.009756097560976,
+      "grad_norm": 1.2487229108810425,
+      "learning_rate": 2.4961687909464462e-06,
+      "loss": 0.3443,
+      "step": 1027
+    },
+    {
+      "epoch": 5.014634146341463,
+      "grad_norm": 1.5230785608291626,
+      "learning_rate": 2.492337590890516e-06,
+      "loss": 0.5778,
+      "step": 1028
+    },
+    {
+      "epoch": 5.019512195121951,
+      "grad_norm": 1.013128638267517,
+      "learning_rate": 2.4885064088298097e-06,
+      "loss": 0.4135,
+      "step": 1029
+    },
+    {
+      "epoch": 5.024390243902439,
+      "grad_norm": 0.944968044757843,
+      "learning_rate": 2.4846752537618875e-06,
+      "loss": 0.1586,
+      "step": 1030
+    },
+    {
+      "epoch": 5.029268292682927,
+      "grad_norm": 1.3642313480377197,
+      "learning_rate": 2.480844134684246e-06,
+      "loss": 0.407,
+      "step": 1031
+    },
+    {
+      "epoch": 5.034146341463415,
+      "grad_norm": 1.2477118968963623,
+      "learning_rate": 2.4770130605942966e-06,
+      "loss": 0.2376,
+      "step": 1032
+    },
+    {
+      "epoch": 5.0390243902439025,
+      "grad_norm": 1.0908373594284058,
+      "learning_rate": 2.4731820404893457e-06,
+      "loss": 0.2063,
+      "step": 1033
+    },
+    {
+      "epoch": 5.04390243902439,
+      "grad_norm": 1.423384666442871,
+      "learning_rate": 2.469351083366571e-06,
+      "loss": 0.255,
+      "step": 1034
+    },
+    {
+      "epoch": 5.048780487804878,
+      "grad_norm": 1.0463767051696777,
+      "learning_rate": 2.4655201982230044e-06,
+      "loss": 0.1164,
+      "step": 1035
+    },
+    {
+      "epoch": 5.053658536585366,
+      "grad_norm": 1.1025980710983276,
+      "learning_rate": 2.4616893940555094e-06,
+      "loss": 0.1895,
+      "step": 1036
+    },
+    {
+      "epoch": 5.058536585365854,
+      "grad_norm": 1.50823175907135,
+      "learning_rate": 2.457858679860757e-06,
+      "loss": 0.6323,
+      "step": 1037
+    },
+    {
+      "epoch": 5.0634146341463415,
+      "grad_norm": 1.2428679466247559,
+      "learning_rate": 2.4540280646352072e-06,
+      "loss": 0.303,
+      "step": 1038
+    },
+    {
+      "epoch": 5.068292682926829,
+      "grad_norm": 1.2156243324279785,
+      "learning_rate": 2.45019755737509e-06,
+      "loss": 0.2281,
+      "step": 1039
+    },
+    {
+      "epoch": 5.073170731707317,
+      "grad_norm": 1.9944231510162354,
+      "learning_rate": 2.4463671670763787e-06,
+      "loss": 0.3072,
+      "step": 1040
+    },
+    {
+      "epoch": 5.078048780487805,
+      "grad_norm": 1.210920810699463,
+      "learning_rate": 2.4425369027347746e-06,
+      "loss": 0.3228,
+      "step": 1041
+    },
+    {
+      "epoch": 5.082926829268293,
+      "grad_norm": 0.9305552244186401,
+      "learning_rate": 2.4387067733456804e-06,
+      "loss": 0.2808,
+      "step": 1042
+    },
+    {
+      "epoch": 5.087804878048781,
+      "grad_norm": 1.200379490852356,
+      "learning_rate": 2.4348767879041847e-06,
+      "loss": 0.2998,
+      "step": 1043
+    },
+    {
+      "epoch": 5.092682926829268,
+      "grad_norm": 1.2235733270645142,
+      "learning_rate": 2.4310469554050366e-06,
+      "loss": 0.2274,
+      "step": 1044
+    },
+    {
+      "epoch": 5.097560975609756,
+      "grad_norm": 0.9284917116165161,
+      "learning_rate": 2.4272172848426257e-06,
+      "loss": 0.3156,
+      "step": 1045
+    },
+    {
+      "epoch": 5.102439024390244,
+      "grad_norm": 1.1812609434127808,
+      "learning_rate": 2.423387785210962e-06,
+      "loss": 0.2529,
+      "step": 1046
+    },
+    {
+      "epoch": 5.107317073170732,
+      "grad_norm": 1.2920304536819458,
+      "learning_rate": 2.4195584655036544e-06,
+      "loss": 0.4466,
+      "step": 1047
+    },
+    {
+      "epoch": 5.11219512195122,
+      "grad_norm": 1.0265544652938843,
+      "learning_rate": 2.4157293347138877e-06,
+      "loss": 0.1946,
+      "step": 1048
+    },
+    {
+      "epoch": 5.117073170731707,
+      "grad_norm": 2.3565609455108643,
+      "learning_rate": 2.4119004018344043e-06,
+      "loss": 0.3177,
+      "step": 1049
+    },
+    {
+      "epoch": 5.121951219512195,
+      "grad_norm": 1.1925452947616577,
+      "learning_rate": 2.408071675857482e-06,
+      "loss": 0.2563,
+      "step": 1050
+    },
+    {
+      "epoch": 5.126829268292683,
+      "grad_norm": 1.3194458484649658,
+      "learning_rate": 2.404243165774912e-06,
+      "loss": 0.4448,
+      "step": 1051
+    },
+    {
+      "epoch": 5.131707317073171,
+      "grad_norm": 0.8365583419799805,
+      "learning_rate": 2.4004148805779785e-06,
+      "loss": 0.1032,
+      "step": 1052
+    },
+    {
+      "epoch": 5.136585365853659,
+      "grad_norm": 1.257830262184143,
+      "learning_rate": 2.3965868292574375e-06,
+      "loss": 0.2855,
+      "step": 1053
+    },
+    {
+      "epoch": 5.1414634146341465,
+      "grad_norm": 1.3698722124099731,
+      "learning_rate": 2.392759020803496e-06,
+      "loss": 0.2996,
+      "step": 1054
+    },
+    {
+      "epoch": 5.146341463414634,
+      "grad_norm": 1.1302275657653809,
+      "learning_rate": 2.3889314642057916e-06,
+      "loss": 0.2131,
+      "step": 1055
+    },
+    {
+      "epoch": 5.151219512195122,
+      "grad_norm": 1.1752660274505615,
+      "learning_rate": 2.3851041684533677e-06,
+      "loss": 0.4146,
+      "step": 1056
+    },
+    {
+      "epoch": 5.15609756097561,
+      "grad_norm": 0.8562495112419128,
+      "learning_rate": 2.381277142534658e-06,
+      "loss": 0.1059,
+      "step": 1057
+    },
+    {
+      "epoch": 5.160975609756098,
+      "grad_norm": 1.279471516609192,
+      "learning_rate": 2.3774503954374614e-06,
+      "loss": 0.1535,
+      "step": 1058
+    },
+    {
+      "epoch": 5.1658536585365855,
+      "grad_norm": 1.7923734188079834,
+      "learning_rate": 2.373623936148921e-06,
+      "loss": 0.3611,
+      "step": 1059
+    },
+    {
+      "epoch": 5.170731707317073,
+      "grad_norm": 1.5222898721694946,
+      "learning_rate": 2.369797773655506e-06,
+      "loss": 0.3535,
+      "step": 1060
+    },
+    {
+      "epoch": 5.175609756097561,
+      "grad_norm": 1.0636627674102783,
+      "learning_rate": 2.3659719169429866e-06,
+      "loss": 0.2679,
+      "step": 1061
+    },
+    {
+      "epoch": 5.180487804878049,
+      "grad_norm": 1.0380195379257202,
+      "learning_rate": 2.3621463749964153e-06,
+      "loss": 0.1407,
+      "step": 1062
+    },
+    {
+      "epoch": 5.185365853658537,
+      "grad_norm": 0.9535474181175232,
+      "learning_rate": 2.3583211568001073e-06,
+      "loss": 0.1263,
+      "step": 1063
+    },
+    {
+      "epoch": 5.190243902439025,
+      "grad_norm": 1.0084075927734375,
+      "learning_rate": 2.3544962713376144e-06,
+      "loss": 0.1556,
+      "step": 1064
+    },
+    {
+      "epoch": 5.195121951219512,
+      "grad_norm": 0.9850101470947266,
+      "learning_rate": 2.3506717275917095e-06,
+      "loss": 0.1819,
+      "step": 1065
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 0.9447112083435059,
+      "learning_rate": 2.346847534544362e-06,
+      "loss": 0.1648,
+      "step": 1066
+    },
+    {
+      "epoch": 5.204878048780488,
+      "grad_norm": 1.0958189964294434,
+      "learning_rate": 2.3430237011767166e-06,
+      "loss": 0.2921,
+      "step": 1067
+    },
+    {
+      "epoch": 5.209756097560976,
+      "grad_norm": 0.9113654494285583,
+      "learning_rate": 2.3392002364690762e-06,
+      "loss": 0.1123,
+      "step": 1068
+    },
+    {
+      "epoch": 5.214634146341464,
+      "grad_norm": 1.5014578104019165,
+      "learning_rate": 2.335377149400874e-06,
+      "loss": 0.2756,
+      "step": 1069
+    },
+    {
+      "epoch": 5.219512195121951,
+      "grad_norm": 1.297300100326538,
+      "learning_rate": 2.3315544489506596e-06,
+      "loss": 0.3623,
+      "step": 1070
+    },
+    {
+      "epoch": 5.224390243902439,
+      "grad_norm": 1.3639144897460938,
+      "learning_rate": 2.3277321440960733e-06,
+      "loss": 0.2713,
+      "step": 1071
+    },
+    {
+      "epoch": 5.229268292682927,
+      "grad_norm": 0.932643473148346,
+      "learning_rate": 2.323910243813826e-06,
+      "loss": 0.1127,
+      "step": 1072
+    },
+    {
+      "epoch": 5.234146341463415,
+      "grad_norm": 1.1686640977859497,
+      "learning_rate": 2.3200887570796798e-06,
+      "loss": 0.4201,
+      "step": 1073
+    },
+    {
+      "epoch": 5.239024390243903,
+      "grad_norm": 1.193697214126587,
+      "learning_rate": 2.316267692868424e-06,
+      "loss": 0.2283,
+      "step": 1074
+    },
+    {
+      "epoch": 5.2439024390243905,
+      "grad_norm": 1.6448438167572021,
+      "learning_rate": 2.312447060153856e-06,
+      "loss": 0.2204,
+      "step": 1075
+    },
+    {
+      "epoch": 5.248780487804878,
+      "grad_norm": 1.203925371170044,
+      "learning_rate": 2.308626867908761e-06,
+      "loss": 0.2119,
+      "step": 1076
+    },
+    {
+      "epoch": 5.253658536585366,
+      "grad_norm": 1.2128314971923828,
+      "learning_rate": 2.3048071251048884e-06,
+      "loss": 0.3189,
+      "step": 1077
+    },
+    {
+      "epoch": 5.258536585365854,
+      "grad_norm": 1.2930793762207031,
+      "learning_rate": 2.300987840712932e-06,
+      "loss": 0.3482,
+      "step": 1078
+    },
+    {
+      "epoch": 5.263414634146342,
+      "grad_norm": 1.1673916578292847,
+      "learning_rate": 2.297169023702511e-06,
+      "loss": 0.4532,
+      "step": 1079
+    },
+    {
+      "epoch": 5.2682926829268295,
+      "grad_norm": 1.0989543199539185,
+      "learning_rate": 2.2933506830421436e-06,
+      "loss": 0.3365,
+      "step": 1080
+    },
+    {
+      "epoch": 5.273170731707317,
+      "grad_norm": 1.0421515703201294,
+      "learning_rate": 2.2895328276992325e-06,
+      "loss": 0.1207,
+      "step": 1081
+    },
+    {
+      "epoch": 5.278048780487805,
+      "grad_norm": 1.3659073114395142,
+      "learning_rate": 2.28571546664004e-06,
+      "loss": 0.3638,
+      "step": 1082
+    },
+    {
+      "epoch": 5.282926829268293,
+      "grad_norm": 0.975866436958313,
+      "learning_rate": 2.281898608829665e-06,
+      "loss": 0.1151,
+      "step": 1083
+    },
+    {
+      "epoch": 5.287804878048781,
+      "grad_norm": 0.9325199127197266,
+      "learning_rate": 2.2780822632320273e-06,
+      "loss": 0.1909,
+      "step": 1084
+    },
+    {
+      "epoch": 5.2926829268292686,
+      "grad_norm": 1.0251420736312866,
+      "learning_rate": 2.2742664388098435e-06,
+      "loss": 0.156,
+      "step": 1085
+    },
+    {
+      "epoch": 5.297560975609756,
+      "grad_norm": 0.8853452205657959,
+      "learning_rate": 2.270451144524605e-06,
+      "loss": 0.2,
+      "step": 1086
+    },
+    {
+      "epoch": 5.302439024390244,
+      "grad_norm": 1.1478848457336426,
+      "learning_rate": 2.266636389336559e-06,
+      "loss": 0.2425,
+      "step": 1087
+    },
+    {
+      "epoch": 5.307317073170732,
+      "grad_norm": 0.9325762987136841,
+      "learning_rate": 2.262822182204686e-06,
+      "loss": 0.1064,
+      "step": 1088
+    },
+    {
+      "epoch": 5.31219512195122,
+      "grad_norm": 1.0397100448608398,
+      "learning_rate": 2.2590085320866798e-06,
+      "loss": 0.1143,
+      "step": 1089
+    },
+    {
+      "epoch": 5.317073170731708,
+      "grad_norm": 0.9676294922828674,
+      "learning_rate": 2.255195447938927e-06,
+      "loss": 0.1028,
+      "step": 1090
+    },
+    {
+      "epoch": 5.321951219512195,
+      "grad_norm": 1.4252179861068726,
+      "learning_rate": 2.251382938716482e-06,
+      "loss": 0.2209,
+      "step": 1091
+    },
+    {
+      "epoch": 5.326829268292683,
+      "grad_norm": 1.0721783638000488,
+      "learning_rate": 2.2475710133730533e-06,
+      "loss": 0.1538,
+      "step": 1092
+    },
+    {
+      "epoch": 5.331707317073171,
+      "grad_norm": 0.9011938571929932,
+      "learning_rate": 2.243759680860975e-06,
+      "loss": 0.1627,
+      "step": 1093
+    },
+    {
+      "epoch": 5.336585365853659,
+      "grad_norm": 1.1225794553756714,
+      "learning_rate": 2.2399489501311896e-06,
+      "loss": 0.2376,
+      "step": 1094
+    },
+    {
+      "epoch": 5.341463414634147,
+      "grad_norm": 1.106941819190979,
+      "learning_rate": 2.2361388301332265e-06,
+      "loss": 0.4092,
+      "step": 1095
+    },
+    {
+      "epoch": 5.3463414634146345,
+      "grad_norm": 0.7896466255187988,
+      "learning_rate": 2.2323293298151817e-06,
+      "loss": 0.1019,
+      "step": 1096
+    },
+    {
+      "epoch": 5.351219512195122,
+      "grad_norm": 1.0374608039855957,
+      "learning_rate": 2.2285204581236937e-06,
+      "loss": 0.1808,
+      "step": 1097
+    },
+    {
+      "epoch": 5.35609756097561,
+      "grad_norm": 1.07212495803833,
+      "learning_rate": 2.2247122240039268e-06,
+      "loss": 0.1773,
+      "step": 1098
+    },
+    {
+      "epoch": 5.360975609756098,
+      "grad_norm": 0.9184301495552063,
+      "learning_rate": 2.2209046363995464e-06,
+      "loss": 0.0608,
+      "step": 1099
+    },
+    {
+      "epoch": 5.365853658536586,
+      "grad_norm": 1.1575993299484253,
+      "learning_rate": 2.217097704252701e-06,
+      "loss": 0.289,
+      "step": 1100
+    },
+    {
+      "epoch": 5.3707317073170735,
+      "grad_norm": 0.869531512260437,
+      "learning_rate": 2.2132914365039993e-06,
+      "loss": 0.1568,
+      "step": 1101
+    },
+    {
+      "epoch": 5.375609756097561,
+      "grad_norm": 1.1020842790603638,
+      "learning_rate": 2.2094858420924882e-06,
+      "loss": 0.0779,
+      "step": 1102
+    },
+    {
+      "epoch": 5.380487804878049,
+      "grad_norm": 1.4796274900436401,
+      "learning_rate": 2.205680929955635e-06,
+      "loss": 0.3573,
+      "step": 1103
+    },
+    {
+      "epoch": 5.385365853658537,
+      "grad_norm": 1.0098387002944946,
+      "learning_rate": 2.201876709029305e-06,
+      "loss": 0.2391,
+      "step": 1104
+    },
+    {
+      "epoch": 5.390243902439025,
+      "grad_norm": 1.0524678230285645,
+      "learning_rate": 2.198073188247738e-06,
+      "loss": 0.1946,
+      "step": 1105
+    },
+    {
+      "epoch": 5.3951219512195125,
+      "grad_norm": 0.8759175539016724,
+      "learning_rate": 2.1942703765435317e-06,
+      "loss": 0.0508,
+      "step": 1106
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.378225564956665,
+      "learning_rate": 2.190468282847617e-06,
+      "loss": 0.4062,
+      "step": 1107
+    },
+    {
+      "epoch": 5.404878048780488,
+      "grad_norm": 1.21510910987854,
+      "learning_rate": 2.186666916089239e-06,
+      "loss": 0.539,
+      "step": 1108
+    },
+    {
+      "epoch": 5.409756097560976,
+      "grad_norm": 1.0114099979400635,
+      "learning_rate": 2.1828662851959377e-06,
+      "loss": 0.1577,
+      "step": 1109
+    },
+    {
+      "epoch": 5.414634146341464,
+      "grad_norm": 1.0858491659164429,
+      "learning_rate": 2.1790663990935203e-06,
+      "loss": 0.2497,
+      "step": 1110
+    },
+    {
+      "epoch": 5.419512195121952,
+      "grad_norm": 1.2288187742233276,
+      "learning_rate": 2.1752672667060488e-06,
+      "loss": 0.1893,
+      "step": 1111
+    },
+    {
+      "epoch": 5.424390243902439,
+      "grad_norm": 1.244839072227478,
+      "learning_rate": 2.1714688969558146e-06,
+      "loss": 0.1314,
+      "step": 1112
+    },
+    {
+      "epoch": 5.429268292682927,
+      "grad_norm": 1.169716715812683,
+      "learning_rate": 2.167671298763316e-06,
+      "loss": 0.4354,
+      "step": 1113
+    },
+    {
+      "epoch": 5.434146341463415,
+      "grad_norm": 1.0661805868148804,
+      "learning_rate": 2.1638744810472414e-06,
+      "loss": 0.3079,
+      "step": 1114
+    },
+    {
+      "epoch": 5.439024390243903,
+      "grad_norm": 0.9970805644989014,
+      "learning_rate": 2.1600784527244445e-06,
+      "loss": 0.1573,
+      "step": 1115
+    },
+    {
+      "epoch": 5.443902439024391,
+      "grad_norm": 1.458986759185791,
+      "learning_rate": 2.1562832227099266e-06,
+      "loss": 0.5079,
+      "step": 1116
+    },
+    {
+      "epoch": 5.4487804878048784,
+      "grad_norm": 1.2743233442306519,
+      "learning_rate": 2.152488799916814e-06,
+      "loss": 0.4,
+      "step": 1117
+    },
+    {
+      "epoch": 5.453658536585366,
+      "grad_norm": 1.822080135345459,
+      "learning_rate": 2.148695193256336e-06,
+      "loss": 0.5168,
+      "step": 1118
+    },
+    {
+      "epoch": 5.458536585365854,
+      "grad_norm": 0.9681462645530701,
+      "learning_rate": 2.1449024116378064e-06,
+      "loss": 0.2588,
+      "step": 1119
+    },
+    {
+      "epoch": 5.463414634146342,
+      "grad_norm": 1.0483061075210571,
+      "learning_rate": 2.1411104639686013e-06,
+      "loss": 0.1395,
+      "step": 1120
+    },
+    {
+      "epoch": 5.46829268292683,
+      "grad_norm": 1.3658936023712158,
+      "learning_rate": 2.137319359154138e-06,
+      "loss": 0.2823,
+      "step": 1121
+    },
+    {
+      "epoch": 5.473170731707317,
+      "grad_norm": 0.9887218475341797,
+      "learning_rate": 2.133529106097853e-06,
+      "loss": 0.1195,
+      "step": 1122
+    },
+    {
+      "epoch": 5.478048780487805,
+      "grad_norm": 1.1662182807922363,
+      "learning_rate": 2.1297397137011862e-06,
+      "loss": 0.2621,
+      "step": 1123
+    },
+    {
+      "epoch": 5.482926829268292,
+      "grad_norm": 1.2363375425338745,
+      "learning_rate": 2.125951190863551e-06,
+      "loss": 0.2179,
+      "step": 1124
+    },
+    {
+      "epoch": 5.487804878048781,
+      "grad_norm": 0.7923662662506104,
+      "learning_rate": 2.1221635464823237e-06,
+      "loss": 0.1064,
+      "step": 1125
+    },
+    {
+      "epoch": 5.492682926829268,
+      "grad_norm": 1.751186728477478,
+      "learning_rate": 2.1183767894528135e-06,
+      "loss": 0.5021,
+      "step": 1126
+    },
+    {
+      "epoch": 5.4975609756097565,
+      "grad_norm": 0.9320492148399353,
+      "learning_rate": 2.114590928668249e-06,
+      "loss": 0.1303,
+      "step": 1127
+    },
+    {
+      "epoch": 5.5024390243902435,
+      "grad_norm": 0.9334157109260559,
+      "learning_rate": 2.1108059730197517e-06,
+      "loss": 0.1716,
+      "step": 1128
+    },
+    {
+      "epoch": 5.507317073170732,
+      "grad_norm": 1.0579321384429932,
+      "learning_rate": 2.1070219313963173e-06,
+      "loss": 0.1431,
+      "step": 1129
+    },
+    {
+      "epoch": 5.512195121951219,
+      "grad_norm": 1.2380542755126953,
+      "learning_rate": 2.1032388126847967e-06,
+      "loss": 0.1667,
+      "step": 1130
+    },
+    {
+      "epoch": 5.517073170731708,
+      "grad_norm": 0.8622035384178162,
+      "learning_rate": 2.099456625769872e-06,
+      "loss": 0.0722,
+      "step": 1131
+    },
+    {
+      "epoch": 5.521951219512195,
+      "grad_norm": 1.125844955444336,
+      "learning_rate": 2.0956753795340376e-06,
+      "loss": 0.1991,
+      "step": 1132
+    },
+    {
+      "epoch": 5.526829268292683,
+      "grad_norm": 1.1052132844924927,
+      "learning_rate": 2.091895082857578e-06,
+      "loss": 0.357,
+      "step": 1133
+    },
+    {
+      "epoch": 5.53170731707317,
+      "grad_norm": 1.0329958200454712,
+      "learning_rate": 2.0881157446185474e-06,
+      "loss": 0.172,
+      "step": 1134
+    },
+    {
+      "epoch": 5.536585365853659,
+      "grad_norm": 1.0203710794448853,
+      "learning_rate": 2.0843373736927506e-06,
+      "loss": 0.1389,
+      "step": 1135
+    },
+    {
+      "epoch": 5.541463414634146,
+      "grad_norm": 0.7079288959503174,
+      "learning_rate": 2.08055997895372e-06,
+      "loss": 0.0853,
+      "step": 1136
+    },
+    {
+      "epoch": 5.546341463414635,
+      "grad_norm": 1.1098510026931763,
+      "learning_rate": 2.0767835692726944e-06,
+      "loss": 0.1492,
+      "step": 1137
+    },
+    {
+      "epoch": 5.5512195121951216,
+      "grad_norm": 1.1874253749847412,
+      "learning_rate": 2.0730081535186e-06,
+      "loss": 0.457,
+      "step": 1138
+    },
+    {
+      "epoch": 5.55609756097561,
+      "grad_norm": 1.0629135370254517,
+      "learning_rate": 2.06923374055803e-06,
+      "loss": 0.2759,
+      "step": 1139
+    },
+    {
+      "epoch": 5.560975609756097,
+      "grad_norm": 0.9080162048339844,
+      "learning_rate": 2.0654603392552193e-06,
+      "loss": 0.0708,
+      "step": 1140
+    },
+    {
+      "epoch": 5.565853658536585,
+      "grad_norm": 1.2526748180389404,
+      "learning_rate": 2.0616879584720305e-06,
+      "loss": 0.2727,
+      "step": 1141
+    },
+    {
+      "epoch": 5.570731707317073,
+      "grad_norm": 0.9930171966552734,
+      "learning_rate": 2.057916607067928e-06,
+      "loss": 0.1473,
+      "step": 1142
+    },
+    {
+      "epoch": 5.575609756097561,
+      "grad_norm": 0.9729718565940857,
+      "learning_rate": 2.054146293899957e-06,
+      "loss": 0.1325,
+      "step": 1143
+    },
+    {
+      "epoch": 5.580487804878048,
+      "grad_norm": 0.9204704761505127,
+      "learning_rate": 2.0503770278227274e-06,
+      "loss": 0.1246,
+      "step": 1144
+    },
+    {
+      "epoch": 5.585365853658536,
+      "grad_norm": 1.0536152124404907,
+      "learning_rate": 2.0466088176883876e-06,
+      "loss": 0.1,
+      "step": 1145
+    },
+    {
+      "epoch": 5.590243902439024,
+      "grad_norm": 1.211815595626831,
+      "learning_rate": 2.042841672346608e-06,
+      "loss": 0.2188,
+      "step": 1146
+    },
+    {
+      "epoch": 5.595121951219512,
+      "grad_norm": 1.4135674238204956,
+      "learning_rate": 2.039075600644557e-06,
+      "loss": 0.1599,
+      "step": 1147
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.0430986881256104,
+      "learning_rate": 2.0353106114268824e-06,
+      "loss": 0.1691,
+      "step": 1148
+    },
+    {
+      "epoch": 5.6048780487804875,
+      "grad_norm": 0.8738663196563721,
+      "learning_rate": 2.031546713535688e-06,
+      "loss": 0.1175,
+      "step": 1149
+    },
+    {
+      "epoch": 5.609756097560975,
+      "grad_norm": 1.015977144241333,
+      "learning_rate": 2.027783915810518e-06,
+      "loss": 0.1638,
+      "step": 1150
+    },
+    {
+      "epoch": 5.614634146341463,
+      "grad_norm": 1.184048056602478,
+      "learning_rate": 2.024022227088329e-06,
+      "loss": 0.4066,
+      "step": 1151
+    },
+    {
+      "epoch": 5.619512195121951,
+      "grad_norm": 1.1206215620040894,
+      "learning_rate": 2.020261656203476e-06,
+      "loss": 0.4061,
+      "step": 1152
+    },
+    {
+      "epoch": 5.624390243902439,
+      "grad_norm": 1.150066614151001,
+      "learning_rate": 2.016502211987687e-06,
+      "loss": 0.2951,
+      "step": 1153
+    },
+    {
+      "epoch": 5.6292682926829265,
+      "grad_norm": 1.465074896812439,
+      "learning_rate": 2.0127439032700446e-06,
+      "loss": 0.2386,
+      "step": 1154
+    },
+    {
+      "epoch": 5.634146341463414,
+      "grad_norm": 1.1083053350448608,
+      "learning_rate": 2.0089867388769664e-06,
+      "loss": 0.2346,
+      "step": 1155
+    },
+    {
+      "epoch": 5.639024390243902,
+      "grad_norm": 0.8865154981613159,
+      "learning_rate": 2.0052307276321793e-06,
+      "loss": 0.1226,
+      "step": 1156
+    },
+    {
+      "epoch": 5.64390243902439,
+      "grad_norm": 1.0869574546813965,
+      "learning_rate": 2.001475878356703e-06,
+      "loss": 0.2506,
+      "step": 1157
+    },
+    {
+      "epoch": 5.648780487804878,
+      "grad_norm": 1.066562533378601,
+      "learning_rate": 1.99772219986883e-06,
+      "loss": 0.3113,
+      "step": 1158
+    },
+    {
+      "epoch": 5.6536585365853655,
+      "grad_norm": 1.489364504814148,
+      "learning_rate": 1.9939697009841024e-06,
+      "loss": 0.1878,
+      "step": 1159
+    },
+    {
+      "epoch": 5.658536585365853,
+      "grad_norm": 1.1426478624343872,
+      "learning_rate": 1.990218390515291e-06,
+      "loss": 0.2903,
+      "step": 1160
+    },
+    {
+      "epoch": 5.663414634146341,
+      "grad_norm": 1.1296617984771729,
+      "learning_rate": 1.9864682772723757e-06,
+      "loss": 0.2505,
+      "step": 1161
+    },
+    {
+      "epoch": 5.668292682926829,
+      "grad_norm": 0.7900092005729675,
+      "learning_rate": 1.9827193700625274e-06,
+      "loss": 0.1204,
+      "step": 1162
+    },
+    {
+      "epoch": 5.673170731707317,
+      "grad_norm": 1.2168926000595093,
+      "learning_rate": 1.978971677690081e-06,
+      "loss": 0.4264,
+      "step": 1163
+    },
+    {
+      "epoch": 5.678048780487805,
+      "grad_norm": 1.0507045984268188,
+      "learning_rate": 1.97522520895652e-06,
+      "loss": 0.0916,
+      "step": 1164
+    },
+    {
+      "epoch": 5.682926829268292,
+      "grad_norm": 1.5506707429885864,
+      "learning_rate": 1.971479972660454e-06,
+      "loss": 0.2858,
+      "step": 1165
+    },
+    {
+      "epoch": 5.68780487804878,
+      "grad_norm": 0.9136821031570435,
+      "learning_rate": 1.967735977597598e-06,
+      "loss": 0.0917,
+      "step": 1166
+    },
+    {
+      "epoch": 5.692682926829268,
+      "grad_norm": 0.7896243929862976,
+      "learning_rate": 1.9639932325607538e-06,
+      "loss": 0.1089,
+      "step": 1167
+    },
+    {
+      "epoch": 5.697560975609756,
+      "grad_norm": 1.396902084350586,
+      "learning_rate": 1.9602517463397845e-06,
+      "loss": 0.1683,
+      "step": 1168
+    },
+    {
+      "epoch": 5.702439024390244,
+      "grad_norm": 1.012985348701477,
+      "learning_rate": 1.9565115277215978e-06,
+      "loss": 0.2019,
+      "step": 1169
+    },
+    {
+      "epoch": 5.7073170731707314,
+      "grad_norm": 0.9816607236862183,
+      "learning_rate": 1.952772585490127e-06,
+      "loss": 0.163,
+      "step": 1170
+    },
+    {
+      "epoch": 5.712195121951219,
+      "grad_norm": 0.9821074604988098,
+      "learning_rate": 1.9490349284263036e-06,
+      "loss": 0.0972,
+      "step": 1171
+    },
+    {
+      "epoch": 5.717073170731707,
+      "grad_norm": 0.9475919604301453,
+      "learning_rate": 1.9452985653080443e-06,
+      "loss": 0.1485,
+      "step": 1172
+    },
+    {
+      "epoch": 5.721951219512195,
+      "grad_norm": 0.921571671962738,
+      "learning_rate": 1.9415635049102245e-06,
+      "loss": 0.1658,
+      "step": 1173
+    },
+    {
+      "epoch": 5.726829268292683,
+      "grad_norm": 1.297415852546692,
+      "learning_rate": 1.937829756004662e-06,
+      "loss": 0.4056,
+      "step": 1174
+    },
+    {
+      "epoch": 5.7317073170731705,
+      "grad_norm": 1.2068060636520386,
+      "learning_rate": 1.9340973273600944e-06,
+      "loss": 0.228,
+      "step": 1175
+    },
+    {
+      "epoch": 5.736585365853658,
+      "grad_norm": 1.0924711227416992,
+      "learning_rate": 1.930366227742157e-06,
+      "loss": 0.3379,
+      "step": 1176
+    },
+    {
+      "epoch": 5.741463414634146,
+      "grad_norm": 1.2374088764190674,
+      "learning_rate": 1.9266364659133653e-06,
+      "loss": 0.2372,
+      "step": 1177
+    },
+    {
+      "epoch": 5.746341463414634,
+      "grad_norm": 0.913871705532074,
+      "learning_rate": 1.922908050633093e-06,
+      "loss": 0.113,
+      "step": 1178
+    },
+    {
+      "epoch": 5.751219512195122,
+      "grad_norm": 1.136725902557373,
+      "learning_rate": 1.919180990657551e-06,
+      "loss": 0.2227,
+      "step": 1179
+    },
+    {
+      "epoch": 5.7560975609756095,
+      "grad_norm": 0.9916288256645203,
+      "learning_rate": 1.9154552947397668e-06,
+      "loss": 0.1983,
+      "step": 1180
+    },
+    {
+      "epoch": 5.760975609756097,
+      "grad_norm": 1.3981820344924927,
+      "learning_rate": 1.9117309716295658e-06,
+      "loss": 0.3343,
+      "step": 1181
+    },
+    {
+      "epoch": 5.765853658536585,
+      "grad_norm": 1.008737325668335,
+      "learning_rate": 1.9080080300735478e-06,
+      "loss": 0.1742,
+      "step": 1182
+    },
+    {
+      "epoch": 5.770731707317073,
+      "grad_norm": 1.3809062242507935,
+      "learning_rate": 1.9042864788150695e-06,
+      "loss": 0.3167,
+      "step": 1183
+    },
+    {
+      "epoch": 5.775609756097561,
+      "grad_norm": 0.7168273329734802,
+      "learning_rate": 1.9005663265942206e-06,
+      "loss": 0.0745,
+      "step": 1184
+    },
+    {
+      "epoch": 5.780487804878049,
+      "grad_norm": 0.9125983715057373,
+      "learning_rate": 1.8968475821478066e-06,
+      "loss": 0.1352,
+      "step": 1185
+    },
+    {
+      "epoch": 5.785365853658536,
+      "grad_norm": 1.2017624378204346,
+      "learning_rate": 1.8931302542093274e-06,
+      "loss": 0.2144,
+      "step": 1186
+    },
+    {
+      "epoch": 5.790243902439024,
+      "grad_norm": 0.902391254901886,
+      "learning_rate": 1.8894143515089539e-06,
+      "loss": 0.1071,
+      "step": 1187
+    },
+    {
+      "epoch": 5.795121951219512,
+      "grad_norm": 0.9119297862052917,
+      "learning_rate": 1.8856998827735118e-06,
+      "loss": 0.141,
+      "step": 1188
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.308137059211731,
+      "learning_rate": 1.8819868567264588e-06,
+      "loss": 0.4523,
+      "step": 1189
+    },
+    {
+      "epoch": 5.804878048780488,
+      "grad_norm": 1.2294268608093262,
+      "learning_rate": 1.8782752820878636e-06,
+      "loss": 0.1783,
+      "step": 1190
+    },
+    {
+      "epoch": 5.809756097560975,
+      "grad_norm": 1.2038543224334717,
+      "learning_rate": 1.8745651675743876e-06,
+      "loss": 0.3132,
+      "step": 1191
+    },
+    {
+      "epoch": 5.814634146341463,
+      "grad_norm": 1.4764758348464966,
+      "learning_rate": 1.870856521899261e-06,
+      "loss": 0.2718,
+      "step": 1192
+    },
+    {
+      "epoch": 5.819512195121951,
+      "grad_norm": 0.9900771975517273,
+      "learning_rate": 1.867149353772267e-06,
+      "loss": 0.0695,
+      "step": 1193
+    },
+    {
+      "epoch": 5.824390243902439,
+      "grad_norm": 1.202021837234497,
+      "learning_rate": 1.863443671899717e-06,
+      "loss": 0.1107,
+      "step": 1194
+    },
+    {
+      "epoch": 5.829268292682927,
+      "grad_norm": 1.2088758945465088,
+      "learning_rate": 1.8597394849844319e-06,
+      "loss": 0.2713,
+      "step": 1195
+    },
+    {
+      "epoch": 5.8341463414634145,
+      "grad_norm": 0.9177408814430237,
+      "learning_rate": 1.8560368017257229e-06,
+      "loss": 0.1473,
+      "step": 1196
+    },
+    {
+      "epoch": 5.839024390243902,
+      "grad_norm": 1.2501375675201416,
+      "learning_rate": 1.8523356308193696e-06,
+      "loss": 0.4487,
+      "step": 1197
+    },
+    {
+      "epoch": 5.84390243902439,
+      "grad_norm": 1.118666410446167,
+      "learning_rate": 1.8486359809575977e-06,
+      "loss": 0.2547,
+      "step": 1198
+    },
+    {
+      "epoch": 5.848780487804878,
+      "grad_norm": 1.0372276306152344,
+      "learning_rate": 1.8449378608290638e-06,
+      "loss": 0.176,
+      "step": 1199
+    },
+    {
+      "epoch": 5.853658536585366,
+      "grad_norm": 1.4009885787963867,
+      "learning_rate": 1.8412412791188306e-06,
+      "loss": 0.2859,
+      "step": 1200
+    },
+    {
+      "epoch": 5.8585365853658535,
+      "grad_norm": 1.327064037322998,
+      "learning_rate": 1.8375462445083464e-06,
+      "loss": 0.282,
+      "step": 1201
+    },
+    {
+      "epoch": 5.863414634146341,
+      "grad_norm": 1.2236543893814087,
+      "learning_rate": 1.8338527656754285e-06,
+      "loss": 0.138,
+      "step": 1202
+    },
+    {
+      "epoch": 5.868292682926829,
+      "grad_norm": 1.3589495420455933,
+      "learning_rate": 1.830160851294239e-06,
+      "loss": 0.2383,
+      "step": 1203
+    },
+    {
+      "epoch": 5.873170731707317,
+      "grad_norm": 1.2593518495559692,
+      "learning_rate": 1.8264705100352662e-06,
+      "loss": 0.3166,
+      "step": 1204
+    },
+    {
+      "epoch": 5.878048780487805,
+      "grad_norm": 1.3545321226119995,
+      "learning_rate": 1.8227817505653045e-06,
+      "loss": 0.2123,
+      "step": 1205
+    },
+    {
+      "epoch": 5.882926829268293,
+      "grad_norm": 1.2606163024902344,
+      "learning_rate": 1.8190945815474323e-06,
+      "loss": 0.3234,
+      "step": 1206
+    },
+    {
+      "epoch": 5.88780487804878,
+      "grad_norm": 1.1344720125198364,
+      "learning_rate": 1.8154090116409934e-06,
+      "loss": 0.2473,
+      "step": 1207
+    },
+    {
+      "epoch": 5.892682926829268,
+      "grad_norm": 1.3704792261123657,
+      "learning_rate": 1.811725049501577e-06,
+      "loss": 0.2684,
+      "step": 1208
+    },
+    {
+      "epoch": 5.897560975609756,
+      "grad_norm": 1.3613022565841675,
+      "learning_rate": 1.8080427037809941e-06,
+      "loss": 0.4087,
+      "step": 1209
+    },
+    {
+      "epoch": 5.902439024390244,
+      "grad_norm": 1.125900387763977,
+      "learning_rate": 1.8043619831272623e-06,
+      "loss": 0.2208,
+      "step": 1210
+    },
+    {
+      "epoch": 5.907317073170732,
+      "grad_norm": 1.4554516077041626,
+      "learning_rate": 1.8006828961845807e-06,
+      "loss": 0.4104,
+      "step": 1211
+    },
+    {
+      "epoch": 5.912195121951219,
+      "grad_norm": 1.507084846496582,
+      "learning_rate": 1.7970054515933124e-06,
+      "loss": 0.4952,
+      "step": 1212
+    },
+    {
+      "epoch": 5.917073170731707,
+      "grad_norm": 1.6275372505187988,
+      "learning_rate": 1.793329657989964e-06,
+      "loss": 0.5594,
+      "step": 1213
+    },
+    {
+      "epoch": 5.921951219512195,
+      "grad_norm": 1.0565378665924072,
+      "learning_rate": 1.7896555240071627e-06,
+      "loss": 0.1108,
+      "step": 1214
+    },
+    {
+      "epoch": 5.926829268292683,
+      "grad_norm": 0.8863050937652588,
+      "learning_rate": 1.7859830582736406e-06,
+      "loss": 0.1669,
+      "step": 1215
+    },
+    {
+      "epoch": 5.931707317073171,
+      "grad_norm": 0.9882891774177551,
+      "learning_rate": 1.782312269414211e-06,
+      "loss": 0.2115,
+      "step": 1216
+    },
+    {
+      "epoch": 5.9365853658536585,
+      "grad_norm": 1.5111620426177979,
+      "learning_rate": 1.7786431660497474e-06,
+      "loss": 0.5776,
+      "step": 1217
+    },
+    {
+      "epoch": 5.941463414634146,
+      "grad_norm": 1.2369654178619385,
+      "learning_rate": 1.7749757567971678e-06,
+      "loss": 0.2491,
+      "step": 1218
+    },
+    {
+      "epoch": 5.946341463414634,
+      "grad_norm": 1.2180466651916504,
+      "learning_rate": 1.7713100502694091e-06,
+      "loss": 0.2229,
+      "step": 1219
+    },
+    {
+      "epoch": 5.951219512195122,
+      "grad_norm": 0.8823730945587158,
+      "learning_rate": 1.7676460550754104e-06,
+      "loss": 0.0664,
+      "step": 1220
+    },
+    {
+      "epoch": 5.95609756097561,
+      "grad_norm": 1.2817288637161255,
+      "learning_rate": 1.7639837798200923e-06,
+      "loss": 0.293,
+      "step": 1221
+    },
+    {
+      "epoch": 5.9609756097560975,
+      "grad_norm": 0.871067464351654,
+      "learning_rate": 1.7603232331043346e-06,
+      "loss": 0.1633,
+      "step": 1222
+    },
+    {
+      "epoch": 5.965853658536585,
+      "grad_norm": 1.4113450050354004,
+      "learning_rate": 1.7566644235249591e-06,
+      "loss": 0.6671,
+      "step": 1223
+    },
+    {
+      "epoch": 5.970731707317073,
+      "grad_norm": 1.4699230194091797,
+      "learning_rate": 1.7530073596747072e-06,
+      "loss": 0.1407,
+      "step": 1224
+    },
+    {
+      "epoch": 5.975609756097561,
+      "grad_norm": 0.8759057521820068,
+      "learning_rate": 1.74935205014222e-06,
+      "loss": 0.0762,
+      "step": 1225
+    },
+    {
+      "epoch": 5.980487804878049,
+      "grad_norm": 0.9844141006469727,
+      "learning_rate": 1.7456985035120194e-06,
+      "loss": 0.1108,
+      "step": 1226
+    },
+    {
+      "epoch": 5.985365853658537,
+      "grad_norm": 0.9976269006729126,
+      "learning_rate": 1.7420467283644877e-06,
+      "loss": 0.2162,
+      "step": 1227
+    },
+    {
+      "epoch": 5.990243902439024,
+      "grad_norm": 1.125545620918274,
+      "learning_rate": 1.738396733275844e-06,
+      "loss": 0.1998,
+      "step": 1228
+    },
+    {
+      "epoch": 5.995121951219512,
+      "grad_norm": 2.0705325603485107,
+      "learning_rate": 1.7347485268181309e-06,
+      "loss": 0.4108,
+      "step": 1229
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.8654595017433167,
+      "learning_rate": 1.7311021175591868e-06,
+      "loss": 0.1534,
+      "step": 1230
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 208,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.6315135228248064e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}