{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 305,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01639344262295082,
      "grad_norm": 5.58112545754218,
      "learning_rate": 2.580645161290323e-06,
      "loss": 0.9597,
      "step": 1
    },
    {
      "epoch": 0.03278688524590164,
      "grad_norm": 5.5621177659627685,
      "learning_rate": 5.161290322580646e-06,
      "loss": 0.9618,
      "step": 2
    },
    {
      "epoch": 0.04918032786885246,
      "grad_norm": 5.2002830302431375,
      "learning_rate": 7.741935483870968e-06,
      "loss": 0.9528,
      "step": 3
    },
    {
      "epoch": 0.06557377049180328,
      "grad_norm": 2.429288619179718,
      "learning_rate": 1.0322580645161291e-05,
      "loss": 0.8743,
      "step": 4
    },
    {
      "epoch": 0.08196721311475409,
      "grad_norm": 2.303682448858842,
      "learning_rate": 1.2903225806451613e-05,
      "loss": 0.8371,
      "step": 5
    },
    {
      "epoch": 0.09836065573770492,
      "grad_norm": 3.7903444701868207,
      "learning_rate": 1.5483870967741936e-05,
      "loss": 0.8396,
      "step": 6
    },
    {
      "epoch": 0.11475409836065574,
      "grad_norm": 4.276574856653076,
      "learning_rate": 1.806451612903226e-05,
      "loss": 0.8132,
      "step": 7
    },
    {
      "epoch": 0.13114754098360656,
      "grad_norm": 3.2100580662432443,
      "learning_rate": 2.0645161290322582e-05,
      "loss": 0.8073,
      "step": 8
    },
    {
      "epoch": 0.14754098360655737,
      "grad_norm": 2.400368523717908,
      "learning_rate": 2.3225806451612906e-05,
      "loss": 0.7532,
      "step": 9
    },
    {
      "epoch": 0.16393442622950818,
      "grad_norm": 2.107929978755255,
      "learning_rate": 2.5806451612903226e-05,
      "loss": 0.7392,
      "step": 10
    },
    {
      "epoch": 0.18032786885245902,
      "grad_norm": 1.706605478358681,
      "learning_rate": 2.8387096774193552e-05,
      "loss": 0.7263,
      "step": 11
    },
    {
      "epoch": 0.19672131147540983,
      "grad_norm": 1.4882561058595487,
      "learning_rate": 3.096774193548387e-05,
      "loss": 0.709,
      "step": 12
    },
    {
      "epoch": 0.21311475409836064,
      "grad_norm": 1.2623324140261931,
      "learning_rate": 3.3548387096774195e-05,
      "loss": 0.6926,
      "step": 13
    },
    {
      "epoch": 0.22950819672131148,
      "grad_norm": 1.0042413251607623,
      "learning_rate": 3.612903225806452e-05,
      "loss": 0.6882,
      "step": 14
    },
    {
      "epoch": 0.2459016393442623,
      "grad_norm": 1.2054555229493642,
      "learning_rate": 3.870967741935484e-05,
      "loss": 0.6705,
      "step": 15
    },
    {
      "epoch": 0.26229508196721313,
      "grad_norm": 1.404103110949475,
      "learning_rate": 4.1290322580645165e-05,
      "loss": 0.6702,
      "step": 16
    },
    {
      "epoch": 0.2786885245901639,
      "grad_norm": 1.4484944472411994,
      "learning_rate": 4.387096774193548e-05,
      "loss": 0.6655,
      "step": 17
    },
    {
      "epoch": 0.29508196721311475,
      "grad_norm": 0.9845835703253549,
      "learning_rate": 4.645161290322581e-05,
      "loss": 0.6602,
      "step": 18
    },
    {
      "epoch": 0.3114754098360656,
      "grad_norm": 1.8228708849031587,
      "learning_rate": 4.9032258064516135e-05,
      "loss": 0.6586,
      "step": 19
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 1.0073875744998761,
      "learning_rate": 5.161290322580645e-05,
      "loss": 0.6283,
      "step": 20
    },
    {
      "epoch": 0.3442622950819672,
      "grad_norm": 2.1477016621670657,
      "learning_rate": 5.4193548387096774e-05,
      "loss": 0.645,
      "step": 21
    },
    {
      "epoch": 0.36065573770491804,
      "grad_norm": 1.4144920525799276,
      "learning_rate": 5.6774193548387104e-05,
      "loss": 0.6402,
      "step": 22
    },
    {
      "epoch": 0.3770491803278688,
      "grad_norm": 1.7423289786096712,
      "learning_rate": 5.935483870967743e-05,
      "loss": 0.6422,
      "step": 23
    },
    {
      "epoch": 0.39344262295081966,
      "grad_norm": 1.5173523064275696,
      "learning_rate": 6.193548387096774e-05,
      "loss": 0.6368,
      "step": 24
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 1.400015090802773,
      "learning_rate": 6.451612903225807e-05,
      "loss": 0.6331,
      "step": 25
    },
    {
      "epoch": 0.4262295081967213,
      "grad_norm": 1.4981731922098314,
      "learning_rate": 6.709677419354839e-05,
      "loss": 0.6154,
      "step": 26
    },
    {
      "epoch": 0.4426229508196721,
      "grad_norm": 1.3910445077945457,
      "learning_rate": 6.967741935483871e-05,
      "loss": 0.6248,
      "step": 27
    },
    {
      "epoch": 0.45901639344262296,
      "grad_norm": 2.287234175350141,
      "learning_rate": 7.225806451612904e-05,
      "loss": 0.6323,
      "step": 28
    },
    {
      "epoch": 0.47540983606557374,
      "grad_norm": 1.8671369685906136,
      "learning_rate": 7.483870967741936e-05,
      "loss": 0.6233,
      "step": 29
    },
    {
      "epoch": 0.4918032786885246,
      "grad_norm": 1.465283599887322,
      "learning_rate": 7.741935483870968e-05,
      "loss": 0.6156,
      "step": 30
    },
    {
      "epoch": 0.5081967213114754,
      "grad_norm": 2.013369833154031,
      "learning_rate": 8e-05,
      "loss": 0.6138,
      "step": 31
    },
    {
      "epoch": 0.5245901639344263,
      "grad_norm": 1.4445848293844874,
      "learning_rate": 7.99973707985831e-05,
      "loss": 0.6199,
      "step": 32
    },
    {
      "epoch": 0.5409836065573771,
      "grad_norm": 2.008002806928902,
      "learning_rate": 7.998948353996737e-05,
      "loss": 0.6296,
      "step": 33
    },
    {
      "epoch": 0.5573770491803278,
      "grad_norm": 1.6006692484547813,
      "learning_rate": 7.99763392610124e-05,
      "loss": 0.6103,
      "step": 34
    },
    {
      "epoch": 0.5737704918032787,
      "grad_norm": 1.7952820693638532,
      "learning_rate": 7.995793968966601e-05,
      "loss": 0.6189,
      "step": 35
    },
    {
      "epoch": 0.5901639344262295,
      "grad_norm": 1.260255034241788,
      "learning_rate": 7.993428724473719e-05,
      "loss": 0.604,
      "step": 36
    },
    {
      "epoch": 0.6065573770491803,
      "grad_norm": 2.918515643648626,
      "learning_rate": 7.990538503557801e-05,
      "loss": 0.6133,
      "step": 37
    },
    {
      "epoch": 0.6229508196721312,
      "grad_norm": 1.5936203819891444,
      "learning_rate": 7.987123686167492e-05,
      "loss": 0.6041,
      "step": 38
    },
    {
      "epoch": 0.639344262295082,
      "grad_norm": 3.246485277477806,
      "learning_rate": 7.98318472121493e-05,
      "loss": 0.6172,
      "step": 39
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 2.818701756393196,
      "learning_rate": 7.978722126516727e-05,
      "loss": 0.6167,
      "step": 40
    },
    {
      "epoch": 0.6721311475409836,
      "grad_norm": 1.7221453757855867,
      "learning_rate": 7.973736488725896e-05,
      "loss": 0.5973,
      "step": 41
    },
    {
      "epoch": 0.6885245901639344,
      "grad_norm": 1.802719546933628,
      "learning_rate": 7.968228463254735e-05,
      "loss": 0.6016,
      "step": 42
    },
    {
      "epoch": 0.7049180327868853,
      "grad_norm": 1.471860307081128,
      "learning_rate": 7.962198774188664e-05,
      "loss": 0.5971,
      "step": 43
    },
    {
      "epoch": 0.7213114754098361,
      "grad_norm": 1.2373741203327389,
      "learning_rate": 7.955648214191034e-05,
      "loss": 0.5889,
      "step": 44
    },
    {
      "epoch": 0.7377049180327869,
      "grad_norm": 1.261142820035152,
      "learning_rate": 7.948577644398924e-05,
      "loss": 0.5849,
      "step": 45
    },
    {
      "epoch": 0.7540983606557377,
      "grad_norm": 2.0142144188297166,
      "learning_rate": 7.940987994309944e-05,
      "loss": 0.5871,
      "step": 46
    },
    {
      "epoch": 0.7704918032786885,
      "grad_norm": 1.1871765895293158,
      "learning_rate": 7.932880261660031e-05,
      "loss": 0.577,
      "step": 47
    },
    {
      "epoch": 0.7868852459016393,
      "grad_norm": 1.2630851815058795,
      "learning_rate": 7.924255512292291e-05,
      "loss": 0.5856,
      "step": 48
    },
    {
      "epoch": 0.8032786885245902,
      "grad_norm": 1.6849343569275448,
      "learning_rate": 7.91511488001689e-05,
      "loss": 0.5931,
      "step": 49
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 0.9859431187761973,
      "learning_rate": 7.905459566461993e-05,
      "loss": 0.571,
      "step": 50
    },
    {
      "epoch": 0.8360655737704918,
      "grad_norm": 1.7623450712427522,
      "learning_rate": 7.895290840915803e-05,
      "loss": 0.5915,
      "step": 51
    },
    {
      "epoch": 0.8524590163934426,
      "grad_norm": 1.2069052100241535,
      "learning_rate": 7.884610040159704e-05,
      "loss": 0.5879,
      "step": 52
    },
    {
      "epoch": 0.8688524590163934,
      "grad_norm": 1.2236642927441974,
      "learning_rate": 7.873418568292517e-05,
      "loss": 0.5779,
      "step": 53
    },
    {
      "epoch": 0.8852459016393442,
      "grad_norm": 5.906328363997742,
      "learning_rate": 7.861717896545928e-05,
      "loss": 0.6193,
      "step": 54
    },
    {
      "epoch": 0.9016393442622951,
      "grad_norm": 2.8517859472775013,
      "learning_rate": 7.849509563091073e-05,
      "loss": 0.6092,
      "step": 55
    },
    {
      "epoch": 0.9180327868852459,
      "grad_norm": 1.6402739036234257,
      "learning_rate": 7.836795172836335e-05,
      "loss": 0.5895,
      "step": 56
    },
    {
      "epoch": 0.9344262295081968,
      "grad_norm": 1.071478048704771,
      "learning_rate": 7.823576397216354e-05,
      "loss": 0.5821,
      "step": 57
    },
    {
      "epoch": 0.9508196721311475,
      "grad_norm": 2.6529573722112563,
      "learning_rate": 7.809854973972311e-05,
      "loss": 0.5988,
      "step": 58
    },
    {
      "epoch": 0.9672131147540983,
      "grad_norm": 2.1179614996092426,
      "learning_rate": 7.795632706923478e-05,
      "loss": 0.6166,
      "step": 59
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 1.9008009613755237,
      "learning_rate": 7.78091146573009e-05,
      "loss": 0.5992,
      "step": 60
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.894221127540697,
      "learning_rate": 7.765693185647557e-05,
      "loss": 0.5826,
      "step": 61
    },
    {
      "epoch": 1.0163934426229508,
      "grad_norm": 1.708613207423494,
      "learning_rate": 7.749979867272055e-05,
      "loss": 0.5692,
      "step": 62
    },
    {
      "epoch": 1.0327868852459017,
      "grad_norm": 1.8649041648138827,
      "learning_rate": 7.733773576277533e-05,
      "loss": 0.5729,
      "step": 63
    },
    {
      "epoch": 1.0491803278688525,
      "grad_norm": 1.41612070780247,
      "learning_rate": 7.71707644314415e-05,
      "loss": 0.5579,
      "step": 64
    },
    {
      "epoch": 1.0655737704918034,
      "grad_norm": 1.9035698103601548,
      "learning_rate": 7.699890662878214e-05,
      "loss": 0.5545,
      "step": 65
    },
    {
      "epoch": 1.0819672131147542,
      "grad_norm": 4.94355544144476,
      "learning_rate": 7.682218494723615e-05,
      "loss": 0.5625,
      "step": 66
    },
    {
      "epoch": 1.098360655737705,
      "grad_norm": 1.4759904398074293,
      "learning_rate": 7.66406226186483e-05,
      "loss": 0.5791,
      "step": 67
    },
    {
      "epoch": 1.1147540983606556,
      "grad_norm": 1.5331088373110235,
      "learning_rate": 7.645424351121519e-05,
      "loss": 0.5604,
      "step": 68
    },
    {
      "epoch": 1.1311475409836065,
      "grad_norm": 1.289579106246813,
      "learning_rate": 7.626307212634745e-05,
      "loss": 0.5585,
      "step": 69
    },
    {
      "epoch": 1.1475409836065573,
      "grad_norm": 1.3594377937672748,
      "learning_rate": 7.606713359544891e-05,
      "loss": 0.5563,
      "step": 70
    },
    {
      "epoch": 1.1639344262295082,
      "grad_norm": 1.4718812627084297,
      "learning_rate": 7.586645367661272e-05,
      "loss": 0.5581,
      "step": 71
    },
    {
      "epoch": 1.180327868852459,
      "grad_norm": 0.9356040160597947,
      "learning_rate": 7.566105875123523e-05,
      "loss": 0.5361,
      "step": 72
    },
    {
      "epoch": 1.1967213114754098,
      "grad_norm": 0.801036937648957,
      "learning_rate": 7.545097582054786e-05,
      "loss": 0.555,
      "step": 73
    },
    {
      "epoch": 1.2131147540983607,
      "grad_norm": 0.9230784392329119,
      "learning_rate": 7.523623250206758e-05,
      "loss": 0.5514,
      "step": 74
    },
    {
      "epoch": 1.2295081967213115,
      "grad_norm": 1.27446883289555,
      "learning_rate": 7.501685702596625e-05,
      "loss": 0.5514,
      "step": 75
    },
    {
      "epoch": 1.2459016393442623,
      "grad_norm": 1.089059775020075,
      "learning_rate": 7.47928782313595e-05,
      "loss": 0.5494,
      "step": 76
    },
    {
      "epoch": 1.2622950819672132,
      "grad_norm": 0.9532311404835904,
      "learning_rate": 7.456432556251552e-05,
      "loss": 0.5369,
      "step": 77
    },
    {
      "epoch": 1.278688524590164,
      "grad_norm": 1.1037696208490315,
      "learning_rate": 7.433122906498436e-05,
      "loss": 0.5454,
      "step": 78
    },
    {
      "epoch": 1.2950819672131146,
      "grad_norm": 1.0044629833456924,
      "learning_rate": 7.409361938164811e-05,
      "loss": 0.5536,
      "step": 79
    },
    {
      "epoch": 1.3114754098360657,
      "grad_norm": 1.139817816380103,
      "learning_rate": 7.385152774869257e-05,
      "loss": 0.5451,
      "step": 80
    },
    {
      "epoch": 1.3278688524590163,
      "grad_norm": 0.9956186270275268,
      "learning_rate": 7.360498599150095e-05,
      "loss": 0.5444,
      "step": 81
    },
    {
      "epoch": 1.3442622950819672,
      "grad_norm": 1.4777189178807848,
      "learning_rate": 7.335402652047011e-05,
      "loss": 0.5482,
      "step": 82
    },
    {
      "epoch": 1.360655737704918,
      "grad_norm": 1.0684258214405637,
      "learning_rate": 7.309868232674992e-05,
      "loss": 0.5425,
      "step": 83
    },
    {
      "epoch": 1.3770491803278688,
      "grad_norm": 1.045469949816082,
      "learning_rate": 7.283898697790615e-05,
      "loss": 0.5326,
      "step": 84
    },
    {
      "epoch": 1.3934426229508197,
      "grad_norm": 1.2884574146396586,
      "learning_rate": 7.257497461350778e-05,
      "loss": 0.541,
      "step": 85
    },
    {
      "epoch": 1.4098360655737705,
      "grad_norm": 1.2291748562831601,
      "learning_rate": 7.23066799406389e-05,
      "loss": 0.5463,
      "step": 86
    },
    {
      "epoch": 1.4262295081967213,
      "grad_norm": 0.8801073224983168,
      "learning_rate": 7.203413822933626e-05,
      "loss": 0.5362,
      "step": 87
    },
    {
      "epoch": 1.4426229508196722,
      "grad_norm": 0.9677124160991732,
      "learning_rate": 7.17573853079525e-05,
      "loss": 0.5405,
      "step": 88
    },
    {
      "epoch": 1.459016393442623,
      "grad_norm": 1.3289631307779222,
      "learning_rate": 7.14764575584463e-05,
      "loss": 0.5393,
      "step": 89
    },
    {
      "epoch": 1.4754098360655736,
      "grad_norm": 0.6274914847275181,
      "learning_rate": 7.119139191159949e-05,
      "loss": 0.5329,
      "step": 90
    },
    {
      "epoch": 1.4918032786885247,
      "grad_norm": 0.5931193088659492,
      "learning_rate": 7.090222584216221e-05,
      "loss": 0.5369,
      "step": 91
    },
    {
      "epoch": 1.5081967213114753,
      "grad_norm": 1.0213786001116576,
      "learning_rate": 7.060899736392643e-05,
      "loss": 0.5404,
      "step": 92
    },
    {
      "epoch": 1.5245901639344264,
      "grad_norm": 1.2526087387777403,
      "learning_rate": 7.031174502472868e-05,
      "loss": 0.5445,
      "step": 93
    },
    {
      "epoch": 1.540983606557377,
      "grad_norm": 0.9943578918521142,
      "learning_rate": 7.001050790138252e-05,
      "loss": 0.5315,
      "step": 94
    },
    {
      "epoch": 1.5573770491803278,
      "grad_norm": 1.2583383738970864,
      "learning_rate": 6.970532559454152e-05,
      "loss": 0.5369,
      "step": 95
    },
    {
      "epoch": 1.5737704918032787,
      "grad_norm": 0.8624167954684502,
      "learning_rate": 6.939623822349339e-05,
      "loss": 0.5212,
      "step": 96
    },
    {
      "epoch": 1.5901639344262295,
      "grad_norm": 0.8714228266723913,
      "learning_rate": 6.908328642088579e-05,
      "loss": 0.5325,
      "step": 97
    },
    {
      "epoch": 1.6065573770491803,
      "grad_norm": 0.9559074422174284,
      "learning_rate": 6.87665113273849e-05,
      "loss": 0.5354,
      "step": 98
    },
    {
      "epoch": 1.6229508196721312,
      "grad_norm": 1.020705004388792,
      "learning_rate": 6.844595458626691e-05,
      "loss": 0.5305,
      "step": 99
    },
    {
      "epoch": 1.639344262295082,
      "grad_norm": 1.2718848911597436,
      "learning_rate": 6.812165833794375e-05,
      "loss": 0.534,
      "step": 100
    },
    {
      "epoch": 1.6557377049180326,
      "grad_norm": 0.8646369335110455,
      "learning_rate": 6.779366521442318e-05,
      "loss": 0.5298,
      "step": 101
    },
    {
      "epoch": 1.6721311475409837,
      "grad_norm": 1.0431466383974872,
      "learning_rate": 6.746201833370447e-05,
      "loss": 0.5258,
      "step": 102
    },
    {
      "epoch": 1.6885245901639343,
      "grad_norm": 1.2838018594279825,
      "learning_rate": 6.712676129411005e-05,
      "loss": 0.5363,
      "step": 103
    },
    {
      "epoch": 1.7049180327868854,
      "grad_norm": 0.7735313109171527,
      "learning_rate": 6.678793816855408e-05,
      "loss": 0.5256,
      "step": 104
    },
    {
      "epoch": 1.721311475409836,
      "grad_norm": 1.1292517886407496,
      "learning_rate": 6.644559349874869e-05,
      "loss": 0.5366,
      "step": 105
    },
    {
      "epoch": 1.737704918032787,
      "grad_norm": 1.1330330672560118,
      "learning_rate": 6.609977228934838e-05,
      "loss": 0.522,
      "step": 106
    },
    {
      "epoch": 1.7540983606557377,
      "grad_norm": 0.8960440091301521,
      "learning_rate": 6.575052000203386e-05,
      "loss": 0.5321,
      "step": 107
    },
    {
      "epoch": 1.7704918032786885,
      "grad_norm": 0.8001970911153957,
      "learning_rate": 6.539788254953557e-05,
      "loss": 0.5292,
      "step": 108
    },
    {
      "epoch": 1.7868852459016393,
      "grad_norm": 0.826768977068413,
      "learning_rate": 6.504190628959798e-05,
      "loss": 0.5266,
      "step": 109
    },
    {
      "epoch": 1.8032786885245902,
      "grad_norm": 1.047081785970805,
      "learning_rate": 6.468263801888546e-05,
      "loss": 0.5249,
      "step": 110
    },
    {
      "epoch": 1.819672131147541,
      "grad_norm": 1.2237949792369915,
      "learning_rate": 6.432012496683032e-05,
      "loss": 0.5252,
      "step": 111
    },
    {
      "epoch": 1.8360655737704918,
      "grad_norm": 0.5326723853060618,
      "learning_rate": 6.395441478942404e-05,
      "loss": 0.5286,
      "step": 112
    },
    {
      "epoch": 1.8524590163934427,
      "grad_norm": 0.8184475093441959,
      "learning_rate": 6.358555556295252e-05,
      "loss": 0.5271,
      "step": 113
    },
    {
      "epoch": 1.8688524590163933,
      "grad_norm": 1.0030170729677295,
      "learning_rate": 6.321359577767574e-05,
      "loss": 0.5355,
      "step": 114
    },
    {
      "epoch": 1.8852459016393444,
      "grad_norm": 0.8438447158790536,
      "learning_rate": 6.283858433145345e-05,
      "loss": 0.5256,
      "step": 115
    },
    {
      "epoch": 1.901639344262295,
      "grad_norm": 0.6775475091945453,
      "learning_rate": 6.246057052331693e-05,
      "loss": 0.524,
      "step": 116
    },
    {
      "epoch": 1.918032786885246,
      "grad_norm": 0.5277047106995992,
      "learning_rate": 6.207960404698818e-05,
      "loss": 0.5243,
      "step": 117
    },
    {
      "epoch": 1.9344262295081966,
      "grad_norm": 0.5241956608413472,
      "learning_rate": 6.169573498434717e-05,
      "loss": 0.5257,
      "step": 118
    },
    {
      "epoch": 1.9508196721311475,
      "grad_norm": 0.6035995248402194,
      "learning_rate": 6.130901379884808e-05,
      "loss": 0.5202,
      "step": 119
    },
    {
      "epoch": 1.9672131147540983,
      "grad_norm": 0.779564589475285,
      "learning_rate": 6.091949132888534e-05,
      "loss": 0.527,
      "step": 120
    },
    {
      "epoch": 1.9836065573770492,
      "grad_norm": 1.022288473196731,
      "learning_rate": 6.052721878111045e-05,
      "loss": 0.5281,
      "step": 121
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2764091263174406,
      "learning_rate": 6.0132247723700324e-05,
      "loss": 0.5128,
      "step": 122
    },
    {
      "epoch": 2.0163934426229506,
      "grad_norm": 0.7708344672192273,
      "learning_rate": 5.973463007957817e-05,
      "loss": 0.4906,
      "step": 123
    },
    {
      "epoch": 2.0327868852459017,
      "grad_norm": 0.7977537199838536,
      "learning_rate": 5.933441811958764e-05,
      "loss": 0.4939,
      "step": 124
    },
    {
      "epoch": 2.0491803278688523,
      "grad_norm": 1.0793702753193162,
      "learning_rate": 5.8931664455621345e-05,
      "loss": 0.4952,
      "step": 125
    },
    {
      "epoch": 2.0655737704918034,
      "grad_norm": 0.8986778009783392,
      "learning_rate": 5.8526422033704494e-05,
      "loss": 0.4892,
      "step": 126
    },
    {
      "epoch": 2.081967213114754,
      "grad_norm": 1.030605766210909,
      "learning_rate": 5.811874412703457e-05,
      "loss": 0.4903,
      "step": 127
    },
    {
      "epoch": 2.098360655737705,
      "grad_norm": 1.0660924489319643,
      "learning_rate": 5.770868432897809e-05,
      "loss": 0.4874,
      "step": 128
    },
    {
      "epoch": 2.1147540983606556,
      "grad_norm": 0.940262946228767,
      "learning_rate": 5.729629654602513e-05,
      "loss": 0.4829,
      "step": 129
    },
    {
      "epoch": 2.1311475409836067,
      "grad_norm": 0.9888397367489762,
      "learning_rate": 5.6881634990702855e-05,
      "loss": 0.4918,
      "step": 130
    },
    {
      "epoch": 2.1475409836065573,
      "grad_norm": 0.8484560909641439,
      "learning_rate": 5.646475417444873e-05,
      "loss": 0.4901,
      "step": 131
    },
    {
      "epoch": 2.1639344262295084,
      "grad_norm": 0.521192253144461,
      "learning_rate": 5.6045708900444364e-05,
      "loss": 0.4949,
      "step": 132
    },
    {
      "epoch": 2.180327868852459,
      "grad_norm": 0.771777999740359,
      "learning_rate": 5.5624554256411175e-05,
      "loss": 0.4811,
      "step": 133
    },
    {
      "epoch": 2.19672131147541,
      "grad_norm": 0.8227491310955285,
      "learning_rate": 5.520134560736852e-05,
      "loss": 0.4885,
      "step": 134
    },
    {
      "epoch": 2.2131147540983607,
      "grad_norm": 0.5992373916210076,
      "learning_rate": 5.477613858835536e-05,
      "loss": 0.4872,
      "step": 135
    },
    {
      "epoch": 2.2295081967213113,
      "grad_norm": 0.5052047355472387,
      "learning_rate": 5.434898909711656e-05,
      "loss": 0.4805,
      "step": 136
    },
    {
      "epoch": 2.2459016393442623,
      "grad_norm": 0.7157168717143663,
      "learning_rate": 5.3919953286754477e-05,
      "loss": 0.4825,
      "step": 137
    },
    {
      "epoch": 2.262295081967213,
      "grad_norm": 0.5410573992811272,
      "learning_rate": 5.348908755834716e-05,
      "loss": 0.4834,
      "step": 138
    },
    {
      "epoch": 2.278688524590164,
      "grad_norm": 0.4166121333458553,
      "learning_rate": 5.305644855353378e-05,
      "loss": 0.4833,
      "step": 139
    },
    {
      "epoch": 2.2950819672131146,
      "grad_norm": 0.5233638295240889,
      "learning_rate": 5.262209314706856e-05,
      "loss": 0.4844,
      "step": 140
    },
    {
      "epoch": 2.3114754098360657,
      "grad_norm": 0.4090985726833338,
      "learning_rate": 5.218607843934401e-05,
      "loss": 0.487,
      "step": 141
    },
    {
      "epoch": 2.3278688524590163,
      "grad_norm": 0.363420954496579,
      "learning_rate": 5.1748461748884476e-05,
      "loss": 0.4774,
      "step": 142
    },
    {
      "epoch": 2.3442622950819674,
      "grad_norm": 0.4690278843251207,
      "learning_rate": 5.130930060481113e-05,
      "loss": 0.4811,
      "step": 143
    },
    {
      "epoch": 2.360655737704918,
      "grad_norm": 0.46496866291592687,
      "learning_rate": 5.0868652739279045e-05,
      "loss": 0.4799,
      "step": 144
    },
    {
      "epoch": 2.3770491803278686,
      "grad_norm": 0.4067588783945542,
      "learning_rate": 5.042657607988787e-05,
      "loss": 0.4884,
      "step": 145
    },
    {
      "epoch": 2.3934426229508197,
      "grad_norm": 0.3249832363624369,
      "learning_rate": 4.9983128742066553e-05,
      "loss": 0.4819,
      "step": 146
    },
    {
      "epoch": 2.4098360655737707,
      "grad_norm": 0.3155267675853147,
      "learning_rate": 4.953836902143354e-05,
      "loss": 0.4817,
      "step": 147
    },
    {
      "epoch": 2.4262295081967213,
      "grad_norm": 0.30276002082546505,
      "learning_rate": 4.909235538613321e-05,
      "loss": 0.4817,
      "step": 148
    },
    {
      "epoch": 2.442622950819672,
      "grad_norm": 0.3578858890853383,
      "learning_rate": 4.864514646914968e-05,
      "loss": 0.4855,
      "step": 149
    },
    {
      "epoch": 2.459016393442623,
      "grad_norm": 0.28701615085250826,
      "learning_rate": 4.819680106059884e-05,
      "loss": 0.4829,
      "step": 150
    },
    {
      "epoch": 2.4754098360655736,
      "grad_norm": 0.37224620973373557,
      "learning_rate": 4.7747378099999866e-05,
      "loss": 0.491,
      "step": 151
    },
    {
      "epoch": 2.4918032786885247,
      "grad_norm": 0.34012579144317995,
      "learning_rate": 4.729693666852699e-05,
      "loss": 0.4884,
      "step": 152
    },
    {
      "epoch": 2.5081967213114753,
      "grad_norm": 0.3774404172171865,
      "learning_rate": 4.684553598124272e-05,
      "loss": 0.4851,
      "step": 153
    },
    {
      "epoch": 2.5245901639344264,
      "grad_norm": 0.3549426068916513,
      "learning_rate": 4.639323537931339e-05,
      "loss": 0.4819,
      "step": 154
    },
    {
      "epoch": 2.540983606557377,
      "grad_norm": 0.39031877291102196,
      "learning_rate": 4.5940094322208156e-05,
      "loss": 0.4905,
      "step": 155
    },
    {
      "epoch": 2.557377049180328,
      "grad_norm": 0.295463066655204,
      "learning_rate": 4.5486172379882495e-05,
      "loss": 0.4765,
      "step": 156
    },
    {
      "epoch": 2.5737704918032787,
      "grad_norm": 0.29781720638326936,
      "learning_rate": 4.5031529224947106e-05,
      "loss": 0.4852,
      "step": 157
    },
    {
      "epoch": 2.5901639344262293,
      "grad_norm": 0.2852567559187638,
      "learning_rate": 4.4576224624823334e-05,
      "loss": 0.4836,
      "step": 158
    },
    {
      "epoch": 2.6065573770491803,
      "grad_norm": 0.239161233813967,
      "learning_rate": 4.412031843388618e-05,
      "loss": 0.4889,
      "step": 159
    },
    {
      "epoch": 2.6229508196721314,
      "grad_norm": 0.37163233149883945,
      "learning_rate": 4.36638705855958e-05,
      "loss": 0.481,
      "step": 160
    },
    {
      "epoch": 2.639344262295082,
      "grad_norm": 0.35373235092972494,
      "learning_rate": 4.320694108461868e-05,
      "loss": 0.4738,
      "step": 161
    },
    {
      "epoch": 2.6557377049180326,
      "grad_norm": 0.3602393611325291,
      "learning_rate": 4.2749589998939366e-05,
      "loss": 0.4783,
      "step": 162
    },
    {
      "epoch": 2.6721311475409837,
      "grad_norm": 0.33028586314976544,
      "learning_rate": 4.2291877451964e-05,
      "loss": 0.4807,
      "step": 163
    },
    {
      "epoch": 2.6885245901639343,
      "grad_norm": 0.27454726727065487,
      "learning_rate": 4.183386361461643e-05,
      "loss": 0.4754,
      "step": 164
    },
    {
      "epoch": 2.7049180327868854,
      "grad_norm": 0.2201982880819997,
      "learning_rate": 4.137560869742816e-05,
      "loss": 0.4816,
      "step": 165
    },
    {
      "epoch": 2.721311475409836,
      "grad_norm": 0.24024736204090635,
      "learning_rate": 4.0917172942623074e-05,
      "loss": 0.477,
      "step": 166
    },
    {
      "epoch": 2.737704918032787,
      "grad_norm": 0.2774042453950074,
      "learning_rate": 4.045861661619796e-05,
      "loss": 0.4863,
      "step": 167
    },
    {
      "epoch": 2.7540983606557377,
      "grad_norm": 0.25031478188990913,
      "learning_rate": 4e-05,
      "loss": 0.4806,
      "step": 168
    },
    {
      "epoch": 2.7704918032786887,
      "grad_norm": 0.3146343902104566,
      "learning_rate": 3.954138338380204e-05,
      "loss": 0.4819,
      "step": 169
    },
    {
      "epoch": 2.7868852459016393,
      "grad_norm": 0.29476137692022925,
      "learning_rate": 3.908282705737694e-05,
      "loss": 0.4829,
      "step": 170
    },
    {
      "epoch": 2.80327868852459,
      "grad_norm": 0.2781979864605831,
      "learning_rate": 3.862439130257184e-05,
      "loss": 0.4842,
      "step": 171
    },
    {
      "epoch": 2.819672131147541,
      "grad_norm": 0.2812486506160709,
      "learning_rate": 3.8166136385383585e-05,
      "loss": 0.4848,
      "step": 172
    },
    {
      "epoch": 2.836065573770492,
      "grad_norm": 0.2443137176068297,
      "learning_rate": 3.770812254803601e-05,
      "loss": 0.4814,
      "step": 173
    },
    {
      "epoch": 2.8524590163934427,
      "grad_norm": 0.2139764225453824,
      "learning_rate": 3.725041000106065e-05,
      "loss": 0.4793,
      "step": 174
    },
    {
      "epoch": 2.8688524590163933,
      "grad_norm": 0.2771342287766635,
      "learning_rate": 3.679305891538133e-05,
      "loss": 0.466,
      "step": 175
    },
    {
      "epoch": 2.8852459016393444,
      "grad_norm": 0.26719263531966153,
      "learning_rate": 3.6336129414404214e-05,
      "loss": 0.4818,
      "step": 176
    },
    {
      "epoch": 2.901639344262295,
      "grad_norm": 0.21705964839463288,
      "learning_rate": 3.5879681566113836e-05,
      "loss": 0.4841,
      "step": 177
    },
    {
      "epoch": 2.918032786885246,
      "grad_norm": 0.20964241906287742,
      "learning_rate": 3.5423775375176686e-05,
      "loss": 0.4857,
      "step": 178
    },
    {
      "epoch": 2.9344262295081966,
      "grad_norm": 0.19494840468268942,
      "learning_rate": 3.4968470775052914e-05,
      "loss": 0.475,
      "step": 179
    },
    {
      "epoch": 2.9508196721311473,
      "grad_norm": 0.19411352485607267,
      "learning_rate": 3.451382762011753e-05,
      "loss": 0.4765,
      "step": 180
    },
    {
      "epoch": 2.9672131147540983,
      "grad_norm": 0.2389081408757777,
      "learning_rate": 3.405990567779185e-05,
      "loss": 0.4843,
      "step": 181
    },
    {
      "epoch": 2.9836065573770494,
      "grad_norm": 0.19939844528891387,
      "learning_rate": 3.3606764620686614e-05,
      "loss": 0.4807,
      "step": 182
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.23959467510788,
      "learning_rate": 3.3154464018757284e-05,
      "loss": 0.483,
      "step": 183
    },
    {
      "epoch": 3.0163934426229506,
      "grad_norm": 0.20927947022485016,
      "learning_rate": 3.270306333147301e-05,
      "loss": 0.4458,
      "step": 184
    },
    {
      "epoch": 3.0327868852459017,
      "grad_norm": 0.2064164028547752,
      "learning_rate": 3.225262190000015e-05,
      "loss": 0.46,
      "step": 185
    },
    {
      "epoch": 3.0491803278688523,
      "grad_norm": 0.22391411463107186,
      "learning_rate": 3.180319893940117e-05,
      "loss": 0.4488,
      "step": 186
    },
    {
      "epoch": 3.0655737704918034,
      "grad_norm": 0.21718594350542614,
      "learning_rate": 3.135485353085033e-05,
      "loss": 0.4487,
      "step": 187
    },
    {
      "epoch": 3.081967213114754,
      "grad_norm": 0.2467699050215421,
      "learning_rate": 3.0907644613866795e-05,
      "loss": 0.4466,
      "step": 188
    },
    {
      "epoch": 3.098360655737705,
      "grad_norm": 0.23078659838474802,
      "learning_rate": 3.0461630978566473e-05,
      "loss": 0.4423,
      "step": 189
    },
    {
      "epoch": 3.1147540983606556,
      "grad_norm": 0.2370592376095119,
      "learning_rate": 3.0016871257933457e-05,
      "loss": 0.4438,
      "step": 190
    },
    {
      "epoch": 3.1311475409836067,
      "grad_norm": 0.21570100983782378,
      "learning_rate": 2.9573423920112155e-05,
      "loss": 0.4455,
      "step": 191
    },
    {
      "epoch": 3.1475409836065573,
      "grad_norm": 0.21407395891268235,
      "learning_rate": 2.9131347260720965e-05,
      "loss": 0.4463,
      "step": 192
    },
    {
      "epoch": 3.1639344262295084,
      "grad_norm": 0.20617008048319857,
      "learning_rate": 2.8690699395188873e-05,
      "loss": 0.4465,
      "step": 193
    },
    {
      "epoch": 3.180327868852459,
      "grad_norm": 0.2125529545129492,
      "learning_rate": 2.825153825111553e-05,
      "loss": 0.445,
      "step": 194
    },
    {
      "epoch": 3.19672131147541,
      "grad_norm": 0.2002948390041686,
      "learning_rate": 2.7813921560656003e-05,
      "loss": 0.4472,
      "step": 195
    },
    {
      "epoch": 3.2131147540983607,
      "grad_norm": 0.19399060354592243,
      "learning_rate": 2.737790685293145e-05,
      "loss": 0.4384,
      "step": 196
    },
    {
      "epoch": 3.2295081967213113,
      "grad_norm": 0.20413317551543775,
      "learning_rate": 2.694355144646622e-05,
      "loss": 0.4474,
      "step": 197
    },
    {
      "epoch": 3.2459016393442623,
      "grad_norm": 0.178524571287303,
      "learning_rate": 2.6510912441652855e-05,
      "loss": 0.4476,
      "step": 198
    },
    {
      "epoch": 3.262295081967213,
      "grad_norm": 0.21495762020528217,
      "learning_rate": 2.6080046713245527e-05,
      "loss": 0.4481,
      "step": 199
    },
    {
      "epoch": 3.278688524590164,
      "grad_norm": 0.17779692741231456,
      "learning_rate": 2.5651010902883457e-05,
      "loss": 0.4447,
      "step": 200
    },
    {
      "epoch": 3.2950819672131146,
      "grad_norm": 0.19536488579919992,
      "learning_rate": 2.5223861411644647e-05,
      "loss": 0.4406,
      "step": 201
    },
    {
      "epoch": 3.3114754098360657,
      "grad_norm": 0.21534013434506138,
      "learning_rate": 2.4798654392631502e-05,
      "loss": 0.443,
      "step": 202
    },
    {
      "epoch": 3.3278688524590163,
      "grad_norm": 0.17678888902216489,
      "learning_rate": 2.4375445743588832e-05,
      "loss": 0.4481,
      "step": 203
    },
    {
      "epoch": 3.3442622950819674,
      "grad_norm": 0.21881378852102576,
      "learning_rate": 2.3954291099555652e-05,
      "loss": 0.4484,
      "step": 204
    },
    {
      "epoch": 3.360655737704918,
      "grad_norm": 0.1911371536452111,
      "learning_rate": 2.3535245825551285e-05,
      "loss": 0.4407,
      "step": 205
    },
    {
      "epoch": 3.3770491803278686,
      "grad_norm": 0.1822153835411314,
      "learning_rate": 2.3118365009297145e-05,
      "loss": 0.4474,
      "step": 206
    },
    {
      "epoch": 3.3934426229508197,
      "grad_norm": 0.1967826759425408,
      "learning_rate": 2.2703703453974883e-05,
      "loss": 0.4466,
      "step": 207
    },
    {
      "epoch": 3.4098360655737707,
      "grad_norm": 0.1787783354685364,
      "learning_rate": 2.2291315671021922e-05,
      "loss": 0.4417,
      "step": 208
    },
    {
      "epoch": 3.4262295081967213,
      "grad_norm": 0.18368967889880147,
      "learning_rate": 2.1881255872965434e-05,
      "loss": 0.4438,
      "step": 209
    },
    {
      "epoch": 3.442622950819672,
      "grad_norm": 0.16138571200714233,
      "learning_rate": 2.1473577966295506e-05,
      "loss": 0.4479,
      "step": 210
    },
    {
      "epoch": 3.459016393442623,
      "grad_norm": 0.15851838029612553,
      "learning_rate": 2.1068335544378672e-05,
      "loss": 0.4511,
      "step": 211
    },
    {
      "epoch": 3.4754098360655736,
      "grad_norm": 0.15179774543882815,
      "learning_rate": 2.0665581880412366e-05,
      "loss": 0.4447,
      "step": 212
    },
    {
      "epoch": 3.4918032786885247,
      "grad_norm": 0.15563700688196921,
      "learning_rate": 2.026536992042184e-05,
      "loss": 0.4456,
      "step": 213
    },
    {
      "epoch": 3.5081967213114753,
      "grad_norm": 0.14825148109958275,
      "learning_rate": 1.9867752276299686e-05,
      "loss": 0.4472,
      "step": 214
    },
    {
      "epoch": 3.5245901639344264,
      "grad_norm": 0.12432896908408256,
      "learning_rate": 1.9472781218889565e-05,
      "loss": 0.4491,
      "step": 215
    },
    {
      "epoch": 3.540983606557377,
      "grad_norm": 0.14986131261136762,
      "learning_rate": 1.908050867111468e-05,
      "loss": 0.447,
      "step": 216
    },
    {
      "epoch": 3.557377049180328,
      "grad_norm": 0.1280996425182516,
      "learning_rate": 1.869098620115194e-05,
      "loss": 0.4459,
      "step": 217
    },
    {
      "epoch": 3.5737704918032787,
      "grad_norm": 0.15181510097669454,
      "learning_rate": 1.8304265015652835e-05,
      "loss": 0.444,
      "step": 218
    },
    {
      "epoch": 3.5901639344262293,
      "grad_norm": 0.1402130102169863,
      "learning_rate": 1.792039595301183e-05,
      "loss": 0.4521,
      "step": 219
    },
    {
      "epoch": 3.6065573770491803,
      "grad_norm": 0.16291621368769718,
      "learning_rate": 1.7539429476683088e-05,
      "loss": 0.447,
      "step": 220
    },
    {
      "epoch": 3.6229508196721314,
      "grad_norm": 0.13943515675692342,
      "learning_rate": 1.7161415668546557e-05,
      "loss": 0.4411,
      "step": 221
    },
    {
      "epoch": 3.639344262295082,
      "grad_norm": 0.1419363610523589,
      "learning_rate": 1.6786404222324278e-05,
      "loss": 0.442,
      "step": 222
    },
    {
      "epoch": 3.6557377049180326,
      "grad_norm": 0.13518707011783945,
      "learning_rate": 1.641444443704749e-05,
      "loss": 0.444,
      "step": 223
    },
    {
      "epoch": 3.6721311475409837,
      "grad_norm": 0.12940286831893916,
      "learning_rate": 1.6045585210575955e-05,
      "loss": 0.448,
      "step": 224
    },
    {
      "epoch": 3.6885245901639343,
      "grad_norm": 0.15612865445371793,
      "learning_rate": 1.56798750331697e-05,
      "loss": 0.4439,
      "step": 225
    },
    {
      "epoch": 3.7049180327868854,
      "grad_norm": 0.13228999124818513,
      "learning_rate": 1.531736198111456e-05,
      "loss": 0.4402,
      "step": 226
    },
    {
      "epoch": 3.721311475409836,
      "grad_norm": 0.14034136382445142,
      "learning_rate": 1.4958093710402036e-05,
      "loss": 0.4394,
      "step": 227
    },
    {
      "epoch": 3.737704918032787,
      "grad_norm": 0.14905661892566827,
      "learning_rate": 1.4602117450464453e-05,
      "loss": 0.4481,
      "step": 228
    },
    {
      "epoch": 3.7540983606557377,
      "grad_norm": 0.16177536977164272,
      "learning_rate": 1.4249479997966145e-05,
      "loss": 0.4457,
      "step": 229
    },
    {
      "epoch": 3.7704918032786887,
      "grad_norm": 0.12677113397589643,
      "learning_rate": 1.3900227710651647e-05,
      "loss": 0.4445,
      "step": 230
    },
    {
      "epoch": 3.7868852459016393,
      "grad_norm": 0.17192511217282086,
      "learning_rate": 1.3554406501251332e-05,
      "loss": 0.4459,
      "step": 231
    },
    {
      "epoch": 3.80327868852459,
      "grad_norm": 0.11920529678713093,
      "learning_rate": 1.3212061831445913e-05,
      "loss": 0.4382,
      "step": 232
    },
    {
      "epoch": 3.819672131147541,
      "grad_norm": 0.12712529888330631,
      "learning_rate": 1.2873238705889968e-05,
      "loss": 0.441,
      "step": 233
    },
    {
      "epoch": 3.836065573770492,
      "grad_norm": 0.12199439925977827,
      "learning_rate": 1.2537981666295536e-05,
      "loss": 0.442,
      "step": 234
    },
    {
      "epoch": 3.8524590163934427,
      "grad_norm": 0.13371713843483143,
      "learning_rate": 1.2206334785576822e-05,
      "loss": 0.4381,
      "step": 235
    },
    {
      "epoch": 3.8688524590163933,
      "grad_norm": 0.11420461356422432,
      "learning_rate": 1.1878341662056259e-05,
      "loss": 0.4442,
      "step": 236
    },
    {
      "epoch": 3.8852459016393444,
      "grad_norm": 0.11704524464444101,
      "learning_rate": 1.1554045413733097e-05,
      "loss": 0.4435,
      "step": 237
    },
    {
      "epoch": 3.901639344262295,
      "grad_norm": 0.11799803709109379,
      "learning_rate": 1.123348867261512e-05,
      "loss": 0.4444,
      "step": 238
    },
    {
      "epoch": 3.918032786885246,
      "grad_norm": 0.12095672068157937,
      "learning_rate": 1.0916713579114218e-05,
      "loss": 0.4425,
      "step": 239
    },
    {
      "epoch": 3.9344262295081966,
      "grad_norm": 0.10579439481053957,
      "learning_rate": 1.0603761776506616e-05,
      "loss": 0.4448,
      "step": 240
    },
    {
      "epoch": 3.9508196721311473,
      "grad_norm": 0.12927283153010521,
      "learning_rate": 1.029467440545849e-05,
      "loss": 0.444,
      "step": 241
    },
    {
      "epoch": 3.9672131147540983,
      "grad_norm": 0.11979578396485492,
      "learning_rate": 9.989492098617494e-06,
      "loss": 0.4462,
      "step": 242
    },
    {
      "epoch": 3.9836065573770494,
      "grad_norm": 0.11270512581429702,
      "learning_rate": 9.688254975271327e-06,
      "loss": 0.4439,
      "step": 243
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.12212819186751334,
      "learning_rate": 9.391002636073589e-06,
      "loss": 0.427,
      "step": 244
    },
    {
      "epoch": 4.016393442622951,
      "grad_norm": 0.15550079584648527,
      "learning_rate": 9.0977741578378e-06,
      "loss": 0.4269,
      "step": 245
    },
    {
      "epoch": 4.032786885245901,
      "grad_norm": 0.13688670983262652,
      "learning_rate": 8.808608088400522e-06,
      "loss": 0.4269,
      "step": 246
    },
    {
      "epoch": 4.049180327868853,
      "grad_norm": 0.11888941041824816,
      "learning_rate": 8.523542441553712e-06,
      "loss": 0.4212,
      "step": 247
    },
    {
      "epoch": 4.065573770491803,
      "grad_norm": 0.13424666760515866,
      "learning_rate": 8.242614692047507e-06,
      "loss": 0.4288,
      "step": 248
    },
    {
      "epoch": 4.081967213114754,
      "grad_norm": 0.15658242982421974,
      "learning_rate": 7.96586177066375e-06,
      "loss": 0.4234,
      "step": 249
    },
    {
      "epoch": 4.098360655737705,
      "grad_norm": 0.14858315425443683,
      "learning_rate": 7.693320059361107e-06,
      "loss": 0.4236,
      "step": 250
    },
    {
      "epoch": 4.114754098360656,
      "grad_norm": 0.1432177280213629,
      "learning_rate": 7.425025386492231e-06,
      "loss": 0.4324,
      "step": 251
    },
    {
      "epoch": 4.131147540983607,
      "grad_norm": 0.1363328212537908,
      "learning_rate": 7.161013022093861e-06,
      "loss": 0.4305,
      "step": 252
    },
    {
      "epoch": 4.147540983606557,
      "grad_norm": 0.13831059143305544,
      "learning_rate": 6.901317673250089e-06,
      "loss": 0.4198,
      "step": 253
    },
    {
      "epoch": 4.163934426229508,
      "grad_norm": 0.1479605646882516,
      "learning_rate": 6.645973479529892e-06,
      "loss": 0.4275,
      "step": 254
    },
    {
      "epoch": 4.180327868852459,
      "grad_norm": 0.12133684415312865,
      "learning_rate": 6.395014008499067e-06,
      "loss": 0.4286,
      "step": 255
    },
    {
      "epoch": 4.19672131147541,
      "grad_norm": 0.11552099846590175,
      "learning_rate": 6.148472251307436e-06,
      "loss": 0.4249,
      "step": 256
    },
    {
      "epoch": 4.213114754098361,
      "grad_norm": 0.1286344544278559,
      "learning_rate": 5.906380618351893e-06,
      "loss": 0.4266,
      "step": 257
    },
    {
      "epoch": 4.229508196721311,
      "grad_norm": 0.12137839353942781,
      "learning_rate": 5.668770935015641e-06,
      "loss": 0.4193,
      "step": 258
    },
    {
      "epoch": 4.245901639344262,
      "grad_norm": 0.10879456442085193,
      "learning_rate": 5.435674437484491e-06,
      "loss": 0.4213,
      "step": 259
    },
    {
      "epoch": 4.262295081967213,
      "grad_norm": 0.10654250511384006,
      "learning_rate": 5.207121768640519e-06,
      "loss": 0.415,
      "step": 260
    },
    {
      "epoch": 4.278688524590164,
      "grad_norm": 0.11634254635354951,
      "learning_rate": 4.983142974033764e-06,
      "loss": 0.4278,
      "step": 261
    },
    {
      "epoch": 4.295081967213115,
      "grad_norm": 0.1188048854614297,
      "learning_rate": 4.763767497932419e-06,
      "loss": 0.4186,
      "step": 262
    },
    {
      "epoch": 4.311475409836065,
      "grad_norm": 0.10499559306706431,
      "learning_rate": 4.549024179452151e-06,
      "loss": 0.4192,
      "step": 263
    },
    {
      "epoch": 4.327868852459017,
      "grad_norm": 0.10258020778548946,
      "learning_rate": 4.338941248764781e-06,
      "loss": 0.4221,
      "step": 264
    },
    {
      "epoch": 4.344262295081967,
      "grad_norm": 0.10174308704204628,
      "learning_rate": 4.133546323387281e-06,
      "loss": 0.4291,
      "step": 265
    },
    {
      "epoch": 4.360655737704918,
      "grad_norm": 0.09609276817159343,
      "learning_rate": 3.932866404551092e-06,
      "loss": 0.4183,
      "step": 266
    },
    {
      "epoch": 4.377049180327869,
      "grad_norm": 0.11045174230896027,
      "learning_rate": 3.7369278736525584e-06,
      "loss": 0.4237,
      "step": 267
    },
    {
      "epoch": 4.39344262295082,
      "grad_norm": 0.09919053798520305,
      "learning_rate": 3.545756488784826e-06,
      "loss": 0.4247,
      "step": 268
    },
    {
      "epoch": 4.409836065573771,
      "grad_norm": 0.10170621462280902,
      "learning_rate": 3.359377381351707e-06,
      "loss": 0.4162,
      "step": 269
    },
    {
      "epoch": 4.426229508196721,
      "grad_norm": 0.09599940271102718,
      "learning_rate": 3.1778150527638664e-06,
      "loss": 0.425,
      "step": 270
    },
    {
      "epoch": 4.442622950819672,
      "grad_norm": 0.09230359083631827,
      "learning_rate": 3.0010933712178735e-06,
      "loss": 0.425,
      "step": 271
    },
    {
      "epoch": 4.459016393442623,
      "grad_norm": 0.09161534788091766,
      "learning_rate": 2.8292355685585147e-06,
      "loss": 0.431,
      "step": 272
    },
    {
      "epoch": 4.475409836065574,
      "grad_norm": 0.09181661135879722,
      "learning_rate": 2.6622642372246874e-06,
      "loss": 0.4202,
      "step": 273
    },
    {
      "epoch": 4.491803278688525,
      "grad_norm": 0.09201200891820971,
      "learning_rate": 2.5002013272794654e-06,
      "loss": 0.4199,
      "step": 274
    },
    {
      "epoch": 4.508196721311475,
      "grad_norm": 0.09965795343039312,
      "learning_rate": 2.3430681435244386e-06,
      "loss": 0.4268,
      "step": 275
    },
    {
      "epoch": 4.524590163934426,
      "grad_norm": 0.09118520040005619,
      "learning_rate": 2.1908853426990983e-06,
      "loss": 0.4251,
      "step": 276
    },
    {
      "epoch": 4.540983606557377,
      "grad_norm": 0.0928501159245484,
      "learning_rate": 2.0436729307652213e-06,
      "loss": 0.4213,
      "step": 277
    },
    {
      "epoch": 4.557377049180328,
      "grad_norm": 0.09183472059851619,
      "learning_rate": 1.9014502602769003e-06,
      "loss": 0.4245,
      "step": 278
    },
    {
      "epoch": 4.573770491803279,
      "grad_norm": 0.08782980102005061,
      "learning_rate": 1.7642360278364767e-06,
      "loss": 0.4233,
      "step": 279
    },
    {
      "epoch": 4.590163934426229,
      "grad_norm": 0.09375666869538167,
      "learning_rate": 1.6320482716366726e-06,
      "loss": 0.425,
      "step": 280
    },
    {
      "epoch": 4.60655737704918,
      "grad_norm": 0.08785996758165288,
      "learning_rate": 1.5049043690892729e-06,
      "loss": 0.4145,
      "step": 281
    },
    {
      "epoch": 4.622950819672131,
      "grad_norm": 0.09413477632311608,
      "learning_rate": 1.3828210345407268e-06,
      "loss": 0.4216,
      "step": 282
    },
    {
      "epoch": 4.639344262295082,
      "grad_norm": 0.09447484104691518,
      "learning_rate": 1.2658143170748383e-06,
      "loss": 0.4201,
      "step": 283
    },
    {
      "epoch": 4.655737704918033,
      "grad_norm": 0.09229151843804058,
      "learning_rate": 1.1538995984029654e-06,
      "loss": 0.4163,
      "step": 284
    },
    {
      "epoch": 4.672131147540983,
      "grad_norm": 0.08568358284035806,
      "learning_rate": 1.047091590841971e-06,
      "loss": 0.4263,
      "step": 285
    },
    {
      "epoch": 4.688524590163935,
      "grad_norm": 0.09023433045860603,
      "learning_rate": 9.454043353800801e-07,
      "loss": 0.4221,
      "step": 286
    },
    {
      "epoch": 4.704918032786885,
      "grad_norm": 0.0844734735476981,
      "learning_rate": 8.488511998311133e-07,
      "loss": 0.422,
      "step": 287
    },
    {
      "epoch": 4.721311475409836,
      "grad_norm": 0.08564054109423068,
      "learning_rate": 7.574448770771003e-07,
      "loss": 0.4226,
      "step": 288
    },
    {
      "epoch": 4.737704918032787,
      "grad_norm": 0.08653160576715448,
      "learning_rate": 6.711973833997088e-07,
      "loss": 0.4179,
      "step": 289
    },
    {
      "epoch": 4.754098360655737,
      "grad_norm": 0.08383004695421666,
      "learning_rate": 5.90120056900565e-07,
      "loss": 0.415,
      "step": 290
    },
    {
      "epoch": 4.770491803278689,
      "grad_norm": 0.08834395449932624,
      "learning_rate": 5.142235560107623e-07,
      "loss": 0.4263,
      "step": 291
    },
    {
      "epoch": 4.786885245901639,
      "grad_norm": 0.08320061631305875,
      "learning_rate": 4.4351785808967485e-07,
      "loss": 0.4272,
      "step": 292
    },
    {
      "epoch": 4.80327868852459,
      "grad_norm": 0.08304223152666458,
      "learning_rate": 3.780122581133672e-07,
      "loss": 0.4223,
      "step": 293
    },
    {
      "epoch": 4.8196721311475414,
      "grad_norm": 0.08927719387363141,
      "learning_rate": 3.177153674526556e-07,
      "loss": 0.4262,
      "step": 294
    },
    {
      "epoch": 4.836065573770492,
      "grad_norm": 0.08297799431829674,
      "learning_rate": 2.626351127410498e-07,
      "loss": 0.4223,
      "step": 295
    },
    {
      "epoch": 4.852459016393443,
      "grad_norm": 0.08318865729803561,
      "learning_rate": 2.1277873483274148e-07,
      "loss": 0.4259,
      "step": 296
    },
    {
      "epoch": 4.868852459016393,
      "grad_norm": 0.0828464753884451,
      "learning_rate": 1.6815278785069944e-07,
      "loss": 0.4235,
      "step": 297
    },
    {
      "epoch": 4.885245901639344,
      "grad_norm": 0.08134276640655788,
      "learning_rate": 1.2876313832508314e-07,
      "loss": 0.4193,
      "step": 298
    },
    {
      "epoch": 4.901639344262295,
      "grad_norm": 0.08124195226061579,
      "learning_rate": 9.461496442199735e-08,
      "loss": 0.417,
      "step": 299
    },
    {
      "epoch": 4.918032786885246,
      "grad_norm": 0.08525708385313023,
      "learning_rate": 6.571275526281451e-08,
      "loss": 0.4301,
      "step": 300
    },
    {
      "epoch": 4.934426229508197,
      "grad_norm": 0.08475437381287206,
      "learning_rate": 4.206031033399338e-08,
      "loss": 0.4309,
      "step": 301
    },
    {
      "epoch": 4.950819672131147,
      "grad_norm": 0.08111110985986,
      "learning_rate": 2.366073898761645e-08,
      "loss": 0.4143,
      "step": 302
    },
    {
      "epoch": 4.967213114754099,
      "grad_norm": 0.08973114402396704,
      "learning_rate": 1.0516460032641284e-08,
      "loss": 0.4301,
      "step": 303
    },
    {
      "epoch": 4.983606557377049,
      "grad_norm": 0.07995200669593748,
      "learning_rate": 2.6292014169104764e-09,
      "loss": 0.418,
      "step": 304
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0879869455035565,
      "learning_rate": 0.0,
      "loss": 0.4146,
      "step": 305
    },
    {
      "epoch": 5.0,
      "step": 305,
      "total_flos": 5116723388743680.0,
      "train_loss": 0.5099891699728418,
      "train_runtime": 4463.8352,
      "train_samples_per_second": 34.825,
      "train_steps_per_second": 0.068
    }
  ],
  "logging_steps": 1,
  "max_steps": 305,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5116723388743680.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}