{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 194,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005154639175257732,
      "grad_norm": 16.203728734181727,
      "learning_rate": 0.0,
      "loss": 0.782,
      "step": 1
    },
    {
      "epoch": 0.010309278350515464,
      "grad_norm": 17.966080542459263,
      "learning_rate": 1.6949152542372883e-07,
      "loss": 0.8806,
      "step": 2
    },
    {
      "epoch": 0.015463917525773196,
      "grad_norm": 19.39755361913842,
      "learning_rate": 3.3898305084745766e-07,
      "loss": 0.9713,
      "step": 3
    },
    {
      "epoch": 0.020618556701030927,
      "grad_norm": 18.521681485723096,
      "learning_rate": 5.084745762711865e-07,
      "loss": 0.7594,
      "step": 4
    },
    {
      "epoch": 0.02577319587628866,
      "grad_norm": 16.960097281692487,
      "learning_rate": 6.779661016949153e-07,
      "loss": 1.0095,
      "step": 5
    },
    {
      "epoch": 0.030927835051546393,
      "grad_norm": 17.9303091015193,
      "learning_rate": 8.474576271186441e-07,
      "loss": 1.0024,
      "step": 6
    },
    {
      "epoch": 0.03608247422680412,
      "grad_norm": 16.393493317929863,
      "learning_rate": 1.016949152542373e-06,
      "loss": 0.7936,
      "step": 7
    },
    {
      "epoch": 0.041237113402061855,
      "grad_norm": 16.350575208232332,
      "learning_rate": 1.186440677966102e-06,
      "loss": 0.6727,
      "step": 8
    },
    {
      "epoch": 0.04639175257731959,
      "grad_norm": 13.392733244736176,
      "learning_rate": 1.3559322033898307e-06,
      "loss": 0.6366,
      "step": 9
    },
    {
      "epoch": 0.05154639175257732,
      "grad_norm": 13.99345418325563,
      "learning_rate": 1.5254237288135596e-06,
      "loss": 0.8352,
      "step": 10
    },
    {
      "epoch": 0.05670103092783505,
      "grad_norm": 16.574632940102198,
      "learning_rate": 1.6949152542372882e-06,
      "loss": 0.8211,
      "step": 11
    },
    {
      "epoch": 0.061855670103092786,
      "grad_norm": 7.432819145467484,
      "learning_rate": 1.8644067796610171e-06,
      "loss": 0.4671,
      "step": 12
    },
    {
      "epoch": 0.06701030927835051,
      "grad_norm": 7.368792116734323,
      "learning_rate": 2.033898305084746e-06,
      "loss": 0.4617,
      "step": 13
    },
    {
      "epoch": 0.07216494845360824,
      "grad_norm": 5.189955956808789,
      "learning_rate": 2.203389830508475e-06,
      "loss": 0.4118,
      "step": 14
    },
    {
      "epoch": 0.07731958762886598,
      "grad_norm": 6.110418564319445,
      "learning_rate": 2.372881355932204e-06,
      "loss": 0.4772,
      "step": 15
    },
    {
      "epoch": 0.08247422680412371,
      "grad_norm": 2.524741574511695,
      "learning_rate": 2.5423728813559323e-06,
      "loss": 0.1693,
      "step": 16
    },
    {
      "epoch": 0.08762886597938144,
      "grad_norm": 4.093704704249702,
      "learning_rate": 2.7118644067796613e-06,
      "loss": 0.5946,
      "step": 17
    },
    {
      "epoch": 0.09278350515463918,
      "grad_norm": 2.792475441128467,
      "learning_rate": 2.8813559322033903e-06,
      "loss": 0.3907,
      "step": 18
    },
    {
      "epoch": 0.0979381443298969,
      "grad_norm": 2.8142697054484422,
      "learning_rate": 3.0508474576271192e-06,
      "loss": 0.4128,
      "step": 19
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 3.0557753995059835,
      "learning_rate": 3.2203389830508473e-06,
      "loss": 0.4128,
      "step": 20
    },
    {
      "epoch": 0.10824742268041238,
      "grad_norm": 2.407462983991879,
      "learning_rate": 3.3898305084745763e-06,
      "loss": 0.2454,
      "step": 21
    },
    {
      "epoch": 0.1134020618556701,
      "grad_norm": 4.175164045570657,
      "learning_rate": 3.5593220338983053e-06,
      "loss": 0.5676,
      "step": 22
    },
    {
      "epoch": 0.11855670103092783,
      "grad_norm": 3.2745984458679542,
      "learning_rate": 3.7288135593220342e-06,
      "loss": 0.3725,
      "step": 23
    },
    {
      "epoch": 0.12371134020618557,
      "grad_norm": 2.630482014005189,
      "learning_rate": 3.898305084745763e-06,
      "loss": 0.284,
      "step": 24
    },
    {
      "epoch": 0.12886597938144329,
      "grad_norm": 2.5818523655885657,
      "learning_rate": 4.067796610169492e-06,
      "loss": 0.3914,
      "step": 25
    },
    {
      "epoch": 0.13402061855670103,
      "grad_norm": 2.1862545024847733,
      "learning_rate": 4.23728813559322e-06,
      "loss": 0.3245,
      "step": 26
    },
    {
      "epoch": 0.13917525773195877,
      "grad_norm": 1.8029553321125102,
      "learning_rate": 4.40677966101695e-06,
      "loss": 0.1906,
      "step": 27
    },
    {
      "epoch": 0.14432989690721648,
      "grad_norm": 2.1274587742776228,
      "learning_rate": 4.576271186440678e-06,
      "loss": 0.3129,
      "step": 28
    },
    {
      "epoch": 0.14948453608247422,
      "grad_norm": 2.0553069128433985,
      "learning_rate": 4.745762711864408e-06,
      "loss": 0.362,
      "step": 29
    },
    {
      "epoch": 0.15463917525773196,
      "grad_norm": 2.2636243270640812,
      "learning_rate": 4.915254237288136e-06,
      "loss": 0.3487,
      "step": 30
    },
    {
      "epoch": 0.15979381443298968,
      "grad_norm": 2.0627428648390023,
      "learning_rate": 5.084745762711865e-06,
      "loss": 0.3017,
      "step": 31
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 1.707507687529141,
      "learning_rate": 5.254237288135594e-06,
      "loss": 0.2379,
      "step": 32
    },
    {
      "epoch": 0.17010309278350516,
      "grad_norm": 1.6433474151177994,
      "learning_rate": 5.423728813559323e-06,
      "loss": 0.2785,
      "step": 33
    },
    {
      "epoch": 0.17525773195876287,
      "grad_norm": 1.8049400111968015,
      "learning_rate": 5.593220338983051e-06,
      "loss": 0.2046,
      "step": 34
    },
    {
      "epoch": 0.18041237113402062,
      "grad_norm": 1.956624509674235,
      "learning_rate": 5.7627118644067805e-06,
      "loss": 0.2924,
      "step": 35
    },
    {
      "epoch": 0.18556701030927836,
      "grad_norm": 2.043442546886233,
      "learning_rate": 5.932203389830509e-06,
      "loss": 0.2517,
      "step": 36
    },
    {
      "epoch": 0.19072164948453607,
      "grad_norm": 1.7809412898084382,
      "learning_rate": 6.1016949152542385e-06,
      "loss": 0.214,
      "step": 37
    },
    {
      "epoch": 0.1958762886597938,
      "grad_norm": 1.0871959933984234,
      "learning_rate": 6.271186440677966e-06,
      "loss": 0.0794,
      "step": 38
    },
    {
      "epoch": 0.20103092783505155,
      "grad_norm": 1.6627701820242757,
      "learning_rate": 6.440677966101695e-06,
      "loss": 0.1529,
      "step": 39
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 2.233281327693517,
      "learning_rate": 6.610169491525424e-06,
      "loss": 0.3918,
      "step": 40
    },
    {
      "epoch": 0.211340206185567,
      "grad_norm": 2.4810527246233143,
      "learning_rate": 6.779661016949153e-06,
      "loss": 0.3276,
      "step": 41
    },
    {
      "epoch": 0.21649484536082475,
      "grad_norm": 2.138313718866784,
      "learning_rate": 6.949152542372882e-06,
      "loss": 0.1874,
      "step": 42
    },
    {
      "epoch": 0.22164948453608246,
      "grad_norm": 2.3207138440634445,
      "learning_rate": 7.1186440677966106e-06,
      "loss": 0.143,
      "step": 43
    },
    {
      "epoch": 0.2268041237113402,
      "grad_norm": 1.806845922939825,
      "learning_rate": 7.288135593220339e-06,
      "loss": 0.2639,
      "step": 44
    },
    {
      "epoch": 0.23195876288659795,
      "grad_norm": 1.6856465858879643,
      "learning_rate": 7.4576271186440685e-06,
      "loss": 0.1803,
      "step": 45
    },
    {
      "epoch": 0.23711340206185566,
      "grad_norm": 1.8626644402225192,
      "learning_rate": 7.627118644067797e-06,
      "loss": 0.2046,
      "step": 46
    },
    {
      "epoch": 0.2422680412371134,
      "grad_norm": 1.7027465927641543,
      "learning_rate": 7.796610169491526e-06,
      "loss": 0.124,
      "step": 47
    },
    {
      "epoch": 0.24742268041237114,
      "grad_norm": 1.9041870166779817,
      "learning_rate": 7.966101694915255e-06,
      "loss": 0.1982,
      "step": 48
    },
    {
      "epoch": 0.25257731958762886,
      "grad_norm": 1.770069442877448,
      "learning_rate": 8.135593220338983e-06,
      "loss": 0.2813,
      "step": 49
    },
    {
      "epoch": 0.25773195876288657,
      "grad_norm": 1.953365961357983,
      "learning_rate": 8.305084745762712e-06,
      "loss": 0.2251,
      "step": 50
    },
    {
      "epoch": 0.26288659793814434,
      "grad_norm": 2.002309230380239,
      "learning_rate": 8.47457627118644e-06,
      "loss": 0.2785,
      "step": 51
    },
    {
      "epoch": 0.26804123711340205,
      "grad_norm": 1.5522236927921595,
      "learning_rate": 8.64406779661017e-06,
      "loss": 0.2388,
      "step": 52
    },
    {
      "epoch": 0.27319587628865977,
      "grad_norm": 1.4065957481275269,
      "learning_rate": 8.8135593220339e-06,
      "loss": 0.1592,
      "step": 53
    },
    {
      "epoch": 0.27835051546391754,
      "grad_norm": 1.344843484282986,
      "learning_rate": 8.983050847457628e-06,
      "loss": 0.184,
      "step": 54
    },
    {
      "epoch": 0.28350515463917525,
      "grad_norm": 1.8809824575076326,
      "learning_rate": 9.152542372881356e-06,
      "loss": 0.1559,
      "step": 55
    },
    {
      "epoch": 0.28865979381443296,
      "grad_norm": 1.554740329023379,
      "learning_rate": 9.322033898305085e-06,
      "loss": 0.1945,
      "step": 56
    },
    {
      "epoch": 0.29381443298969073,
      "grad_norm": 1.7706197688983198,
      "learning_rate": 9.491525423728815e-06,
      "loss": 0.2799,
      "step": 57
    },
    {
      "epoch": 0.29896907216494845,
      "grad_norm": 1.339143828253002,
      "learning_rate": 9.661016949152544e-06,
      "loss": 0.1094,
      "step": 58
    },
    {
      "epoch": 0.30412371134020616,
      "grad_norm": 2.1220725285103086,
      "learning_rate": 9.830508474576272e-06,
      "loss": 0.3211,
      "step": 59
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 1.3971936674913341,
      "learning_rate": 1e-05,
      "loss": 0.2081,
      "step": 60
    },
    {
      "epoch": 0.31443298969072164,
      "grad_norm": 1.8165010693502677,
      "learning_rate": 9.999909794073715e-06,
      "loss": 0.191,
      "step": 61
    },
    {
      "epoch": 0.31958762886597936,
      "grad_norm": 1.8206712917879821,
      "learning_rate": 9.999639179549699e-06,
      "loss": 0.1817,
      "step": 62
    },
    {
      "epoch": 0.3247422680412371,
      "grad_norm": 2.1904370888969704,
      "learning_rate": 9.999188166192368e-06,
      "loss": 0.1933,
      "step": 63
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 1.356123942482486,
      "learning_rate": 9.998556770275351e-06,
      "loss": 0.1529,
      "step": 64
    },
    {
      "epoch": 0.33505154639175255,
      "grad_norm": 1.4167068373050817,
      "learning_rate": 9.997745014580912e-06,
      "loss": 0.2182,
      "step": 65
    },
    {
      "epoch": 0.3402061855670103,
      "grad_norm": 1.2778114297184384,
      "learning_rate": 9.996752928399121e-06,
      "loss": 0.1231,
      "step": 66
    },
    {
      "epoch": 0.34536082474226804,
      "grad_norm": 1.4693549048926278,
      "learning_rate": 9.995580547526798e-06,
      "loss": 0.2222,
      "step": 67
    },
    {
      "epoch": 0.35051546391752575,
      "grad_norm": 1.4349642188073422,
      "learning_rate": 9.994227914266222e-06,
      "loss": 0.1934,
      "step": 68
    },
    {
      "epoch": 0.3556701030927835,
      "grad_norm": 1.3639498444936125,
      "learning_rate": 9.992695077423609e-06,
      "loss": 0.1707,
      "step": 69
    },
    {
      "epoch": 0.36082474226804123,
      "grad_norm": 1.3943030751682663,
      "learning_rate": 9.990982092307347e-06,
      "loss": 0.1716,
      "step": 70
    },
    {
      "epoch": 0.36597938144329895,
      "grad_norm": 1.4616151628165952,
      "learning_rate": 9.989089020725999e-06,
      "loss": 0.2363,
      "step": 71
    },
    {
      "epoch": 0.3711340206185567,
      "grad_norm": 1.7297954033170302,
      "learning_rate": 9.987015930986074e-06,
      "loss": 0.0612,
      "step": 72
    },
    {
      "epoch": 0.37628865979381443,
      "grad_norm": 1.4130551566794556,
      "learning_rate": 9.984762897889568e-06,
      "loss": 0.2438,
      "step": 73
    },
    {
      "epoch": 0.38144329896907214,
      "grad_norm": 1.4316589582684316,
      "learning_rate": 9.98233000273125e-06,
      "loss": 0.2529,
      "step": 74
    },
    {
      "epoch": 0.3865979381443299,
      "grad_norm": 1.4921368045815033,
      "learning_rate": 9.97971733329575e-06,
      "loss": 0.2033,
      "step": 75
    },
    {
      "epoch": 0.3917525773195876,
      "grad_norm": 1.6884913833837993,
      "learning_rate": 9.97692498385437e-06,
      "loss": 0.3251,
      "step": 76
    },
    {
      "epoch": 0.39690721649484534,
      "grad_norm": 1.2618275142031115,
      "learning_rate": 9.973953055161702e-06,
      "loss": 0.2173,
      "step": 77
    },
    {
      "epoch": 0.4020618556701031,
      "grad_norm": 1.2058891275808017,
      "learning_rate": 9.970801654451974e-06,
      "loss": 0.1337,
      "step": 78
    },
    {
      "epoch": 0.4072164948453608,
      "grad_norm": 1.3014010097389481,
      "learning_rate": 9.967470895435197e-06,
      "loss": 0.2217,
      "step": 79
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 1.0907558669955588,
      "learning_rate": 9.963960898293049e-06,
      "loss": 0.1166,
      "step": 80
    },
    {
      "epoch": 0.4175257731958763,
      "grad_norm": 1.3508925279475854,
      "learning_rate": 9.96027178967455e-06,
      "loss": 0.2185,
      "step": 81
    },
    {
      "epoch": 0.422680412371134,
      "grad_norm": 1.8405638710013943,
      "learning_rate": 9.956403702691482e-06,
      "loss": 0.2981,
      "step": 82
    },
    {
      "epoch": 0.42783505154639173,
      "grad_norm": 0.9981831988732881,
      "learning_rate": 9.952356776913594e-06,
      "loss": 0.1471,
      "step": 83
    },
    {
      "epoch": 0.4329896907216495,
      "grad_norm": 1.5438578399853147,
      "learning_rate": 9.948131158363564e-06,
      "loss": 0.2219,
      "step": 84
    },
    {
      "epoch": 0.4381443298969072,
      "grad_norm": 1.373967099765683,
      "learning_rate": 9.943726999511721e-06,
      "loss": 0.2625,
      "step": 85
    },
    {
      "epoch": 0.44329896907216493,
      "grad_norm": 1.5045665327970632,
      "learning_rate": 9.939144459270557e-06,
      "loss": 0.2378,
      "step": 86
    },
    {
      "epoch": 0.4484536082474227,
      "grad_norm": 1.2348801218707532,
      "learning_rate": 9.934383702988992e-06,
      "loss": 0.1299,
      "step": 87
    },
    {
      "epoch": 0.4536082474226804,
      "grad_norm": 1.3457854200982862,
      "learning_rate": 9.929444902446392e-06,
      "loss": 0.2193,
      "step": 88
    },
    {
      "epoch": 0.4587628865979381,
      "grad_norm": 1.5974941630503592,
      "learning_rate": 9.924328235846393e-06,
      "loss": 0.3447,
      "step": 89
    },
    {
      "epoch": 0.4639175257731959,
      "grad_norm": 1.7921462524347982,
      "learning_rate": 9.919033887810451e-06,
      "loss": 0.3081,
      "step": 90
    },
    {
      "epoch": 0.4690721649484536,
      "grad_norm": 1.0544333428105994,
      "learning_rate": 9.913562049371196e-06,
      "loss": 0.1723,
      "step": 91
    },
    {
      "epoch": 0.4742268041237113,
      "grad_norm": 1.173180659355975,
      "learning_rate": 9.90791291796553e-06,
      "loss": 0.1816,
      "step": 92
    },
    {
      "epoch": 0.4793814432989691,
      "grad_norm": 1.2343501847799747,
      "learning_rate": 9.902086697427504e-06,
      "loss": 0.2048,
      "step": 93
    },
    {
      "epoch": 0.4845360824742268,
      "grad_norm": 1.4166315003510848,
      "learning_rate": 9.896083597980968e-06,
      "loss": 0.1611,
      "step": 94
    },
    {
      "epoch": 0.4896907216494845,
      "grad_norm": 0.9316258247144923,
      "learning_rate": 9.88990383623198e-06,
      "loss": 0.1307,
      "step": 95
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 1.1284196528102999,
      "learning_rate": 9.883547635160991e-06,
      "loss": 0.1816,
      "step": 96
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9136263992805103,
      "learning_rate": 9.877015224114806e-06,
      "loss": 0.1173,
      "step": 97
    },
    {
      "epoch": 0.5051546391752577,
      "grad_norm": 1.5233730471599805,
      "learning_rate": 9.870306838798299e-06,
      "loss": 0.2239,
      "step": 98
    },
    {
      "epoch": 0.5103092783505154,
      "grad_norm": 1.4031394060036828,
      "learning_rate": 9.863422721265913e-06,
      "loss": 0.2398,
      "step": 99
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 1.12765997685746,
      "learning_rate": 9.856363119912931e-06,
      "loss": 0.1318,
      "step": 100
    },
    {
      "epoch": 0.520618556701031,
      "grad_norm": 0.8420145818838017,
      "learning_rate": 9.849128289466503e-06,
      "loss": 0.1246,
      "step": 101
    },
    {
      "epoch": 0.5257731958762887,
      "grad_norm": 0.946095131905409,
      "learning_rate": 9.841718490976461e-06,
      "loss": 0.1241,
      "step": 102
    },
    {
      "epoch": 0.5309278350515464,
      "grad_norm": 0.9796883673872017,
      "learning_rate": 9.8341339918059e-06,
      "loss": 0.0955,
      "step": 103
    },
    {
      "epoch": 0.5360824742268041,
      "grad_norm": 1.1930460993410585,
      "learning_rate": 9.826375065621533e-06,
      "loss": 0.1646,
      "step": 104
    },
    {
      "epoch": 0.5412371134020618,
      "grad_norm": 1.27651192729961,
      "learning_rate": 9.818441992383802e-06,
      "loss": 0.1566,
      "step": 105
    },
    {
      "epoch": 0.5463917525773195,
      "grad_norm": 1.1315939745886057,
      "learning_rate": 9.810335058336801e-06,
      "loss": 0.1786,
      "step": 106
    },
    {
      "epoch": 0.5515463917525774,
      "grad_norm": 0.9355015058700618,
      "learning_rate": 9.802054555997927e-06,
      "loss": 0.0945,
      "step": 107
    },
    {
      "epoch": 0.5567010309278351,
      "grad_norm": 1.7639349345635655,
      "learning_rate": 9.79360078414733e-06,
      "loss": 0.254,
      "step": 108
    },
    {
      "epoch": 0.5618556701030928,
      "grad_norm": 1.3663266258683646,
      "learning_rate": 9.784974047817142e-06,
      "loss": 0.1953,
      "step": 109
    },
    {
      "epoch": 0.5670103092783505,
      "grad_norm": 1.441702754636908,
      "learning_rate": 9.776174658280458e-06,
      "loss": 0.1841,
      "step": 110
    },
    {
      "epoch": 0.5721649484536082,
      "grad_norm": 1.1499799893613811,
      "learning_rate": 9.767202933040111e-06,
      "loss": 0.1997,
      "step": 111
    },
    {
      "epoch": 0.5773195876288659,
      "grad_norm": 1.2076330399094353,
      "learning_rate": 9.758059195817216e-06,
      "loss": 0.1768,
      "step": 112
    },
    {
      "epoch": 0.5824742268041238,
      "grad_norm": 0.878277090541755,
      "learning_rate": 9.748743776539489e-06,
      "loss": 0.1173,
      "step": 113
    },
    {
      "epoch": 0.5876288659793815,
      "grad_norm": 1.0228670375234183,
      "learning_rate": 9.739257011329336e-06,
      "loss": 0.1361,
      "step": 114
    },
    {
      "epoch": 0.5927835051546392,
      "grad_norm": 1.4761049929099612,
      "learning_rate": 9.729599242491738e-06,
      "loss": 0.2792,
      "step": 115
    },
    {
      "epoch": 0.5979381443298969,
      "grad_norm": 1.0296746075367031,
      "learning_rate": 9.719770818501885e-06,
      "loss": 0.1,
      "step": 116
    },
    {
      "epoch": 0.6030927835051546,
      "grad_norm": 0.9546648084132865,
      "learning_rate": 9.709772093992619e-06,
      "loss": 0.1072,
      "step": 117
    },
    {
      "epoch": 0.6082474226804123,
      "grad_norm": 0.9359977344927911,
      "learning_rate": 9.699603429741615e-06,
      "loss": 0.1171,
      "step": 118
    },
    {
      "epoch": 0.6134020618556701,
      "grad_norm": 1.2912642891989952,
      "learning_rate": 9.689265192658387e-06,
      "loss": 0.1783,
      "step": 119
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 1.1169115856391039,
      "learning_rate": 9.67875775577104e-06,
      "loss": 0.2051,
      "step": 120
    },
    {
      "epoch": 0.6237113402061856,
      "grad_norm": 1.4787464277701932,
      "learning_rate": 9.668081498212799e-06,
      "loss": 0.2377,
      "step": 121
    },
    {
      "epoch": 0.6288659793814433,
      "grad_norm": 0.9403827546416947,
      "learning_rate": 9.657236805208347e-06,
      "loss": 0.1219,
      "step": 122
    },
    {
      "epoch": 0.634020618556701,
      "grad_norm": 0.7358413148522787,
      "learning_rate": 9.646224068059917e-06,
      "loss": 0.0895,
      "step": 123
    },
    {
      "epoch": 0.6391752577319587,
      "grad_norm": 1.1530465641868872,
      "learning_rate": 9.63504368413317e-06,
      "loss": 0.2149,
      "step": 124
    },
    {
      "epoch": 0.6443298969072165,
      "grad_norm": 1.4217244985998034,
      "learning_rate": 9.62369605684286e-06,
      "loss": 0.197,
      "step": 125
    },
    {
      "epoch": 0.6494845360824743,
      "grad_norm": 1.910749720256578,
      "learning_rate": 9.612181595638279e-06,
      "loss": 0.2508,
      "step": 126
    },
    {
      "epoch": 0.654639175257732,
      "grad_norm": 1.2757659427730983,
      "learning_rate": 9.600500715988486e-06,
      "loss": 0.2809,
      "step": 127
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 1.0891754027329807,
      "learning_rate": 9.588653839367304e-06,
      "loss": 0.165,
      "step": 128
    },
    {
      "epoch": 0.6649484536082474,
      "grad_norm": 1.293429113847632,
      "learning_rate": 9.576641393238129e-06,
      "loss": 0.0855,
      "step": 129
    },
    {
      "epoch": 0.6701030927835051,
      "grad_norm": 1.4599213966596118,
      "learning_rate": 9.564463811038489e-06,
      "loss": 0.2503,
      "step": 130
    },
    {
      "epoch": 0.6752577319587629,
      "grad_norm": 1.3648021125385577,
      "learning_rate": 9.55212153216442e-06,
      "loss": 0.1794,
      "step": 131
    },
    {
      "epoch": 0.6804123711340206,
      "grad_norm": 0.9809648930976671,
      "learning_rate": 9.5396150019546e-06,
      "loss": 0.0749,
      "step": 132
    },
    {
      "epoch": 0.6855670103092784,
      "grad_norm": 1.2479101653995877,
      "learning_rate": 9.526944671674287e-06,
      "loss": 0.1705,
      "step": 133
    },
    {
      "epoch": 0.6907216494845361,
      "grad_norm": 1.180767703263976,
      "learning_rate": 9.514110998499032e-06,
      "loss": 0.2419,
      "step": 134
    },
    {
      "epoch": 0.6958762886597938,
      "grad_norm": 0.9973862175809532,
      "learning_rate": 9.501114445498183e-06,
      "loss": 0.1545,
      "step": 135
    },
    {
      "epoch": 0.7010309278350515,
      "grad_norm": 0.9580969711134073,
      "learning_rate": 9.487955481618184e-06,
      "loss": 0.1563,
      "step": 136
    },
    {
      "epoch": 0.7061855670103093,
      "grad_norm": 1.0576903096823542,
      "learning_rate": 9.474634581665645e-06,
      "loss": 0.128,
      "step": 137
    },
    {
      "epoch": 0.711340206185567,
      "grad_norm": 1.1608177511918483,
      "learning_rate": 9.461152226290212e-06,
      "loss": 0.2324,
      "step": 138
    },
    {
      "epoch": 0.7164948453608248,
      "grad_norm": 0.7247855062304921,
      "learning_rate": 9.44750890196723e-06,
      "loss": 0.0696,
      "step": 139
    },
    {
      "epoch": 0.7216494845360825,
      "grad_norm": 1.294941334572682,
      "learning_rate": 9.43370510098018e-06,
      "loss": 0.2068,
      "step": 140
    },
    {
      "epoch": 0.7268041237113402,
      "grad_norm": 1.5307061314891508,
      "learning_rate": 9.419741321402923e-06,
      "loss": 0.134,
      "step": 141
    },
    {
      "epoch": 0.7319587628865979,
      "grad_norm": 1.724248389396898,
      "learning_rate": 9.405618067081729e-06,
      "loss": 0.3587,
      "step": 142
    },
    {
      "epoch": 0.7371134020618557,
      "grad_norm": 0.7880104208217124,
      "learning_rate": 9.391335847617093e-06,
      "loss": 0.1081,
      "step": 143
    },
    {
      "epoch": 0.7422680412371134,
      "grad_norm": 1.176218642717777,
      "learning_rate": 9.37689517834535e-06,
      "loss": 0.2552,
      "step": 144
    },
    {
      "epoch": 0.7474226804123711,
      "grad_norm": 0.9007810641742303,
      "learning_rate": 9.362296580320078e-06,
      "loss": 0.1411,
      "step": 145
    },
    {
      "epoch": 0.7525773195876289,
      "grad_norm": 1.0987174332049452,
      "learning_rate": 9.347540580293301e-06,
      "loss": 0.1984,
      "step": 146
    },
    {
      "epoch": 0.7577319587628866,
      "grad_norm": 0.9593956667801066,
      "learning_rate": 9.332627710696477e-06,
      "loss": 0.1403,
      "step": 147
    },
    {
      "epoch": 0.7628865979381443,
      "grad_norm": 1.5203788468340629,
      "learning_rate": 9.317558509621297e-06,
      "loss": 0.3035,
      "step": 148
    },
    {
      "epoch": 0.7680412371134021,
      "grad_norm": 0.9687515693670804,
      "learning_rate": 9.302333520800253e-06,
      "loss": 0.1366,
      "step": 149
    },
    {
      "epoch": 0.7731958762886598,
      "grad_norm": 0.7832678330612435,
      "learning_rate": 9.286953293587035e-06,
      "loss": 0.11,
      "step": 150
    },
    {
      "epoch": 0.7783505154639175,
      "grad_norm": 0.9289801978987015,
      "learning_rate": 9.271418382936697e-06,
      "loss": 0.0988,
      "step": 151
    },
    {
      "epoch": 0.7835051546391752,
      "grad_norm": 1.1356540696916577,
      "learning_rate": 9.255729349385645e-06,
      "loss": 0.1375,
      "step": 152
    },
    {
      "epoch": 0.788659793814433,
      "grad_norm": 1.3495379294095782,
      "learning_rate": 9.239886759031399e-06,
      "loss": 0.3033,
      "step": 153
    },
    {
      "epoch": 0.7938144329896907,
      "grad_norm": 1.2612818337900333,
      "learning_rate": 9.223891183512174e-06,
      "loss": 0.1364,
      "step": 154
    },
    {
      "epoch": 0.7989690721649485,
      "grad_norm": 1.1325852394462053,
      "learning_rate": 9.207743199986252e-06,
      "loss": 0.1559,
      "step": 155
    },
    {
      "epoch": 0.8041237113402062,
      "grad_norm": 0.6629329908683664,
      "learning_rate": 9.191443391111157e-06,
      "loss": 0.0678,
      "step": 156
    },
    {
      "epoch": 0.8092783505154639,
      "grad_norm": 1.184280051526035,
      "learning_rate": 9.174992345022636e-06,
      "loss": 0.1762,
      "step": 157
    },
    {
      "epoch": 0.8144329896907216,
      "grad_norm": 1.4363833078261825,
      "learning_rate": 9.158390655313422e-06,
      "loss": 0.2509,
      "step": 158
    },
    {
      "epoch": 0.8195876288659794,
      "grad_norm": 1.4851296678460735,
      "learning_rate": 9.141638921011842e-06,
      "loss": 0.3053,
      "step": 159
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 0.9939630731717258,
      "learning_rate": 9.124737746560175e-06,
      "loss": 0.0952,
      "step": 160
    },
    {
      "epoch": 0.8298969072164949,
      "grad_norm": 1.2169101845107388,
      "learning_rate": 9.107687741792863e-06,
      "loss": 0.1494,
      "step": 161
    },
    {
      "epoch": 0.8350515463917526,
      "grad_norm": 1.3295674637794006,
      "learning_rate": 9.090489521914492e-06,
      "loss": 0.2761,
      "step": 162
    },
    {
      "epoch": 0.8402061855670103,
      "grad_norm": 1.2793707854001373,
      "learning_rate": 9.073143707477607e-06,
      "loss": 0.1433,
      "step": 163
    },
    {
      "epoch": 0.845360824742268,
      "grad_norm": 1.2039862398086603,
      "learning_rate": 9.055650924360308e-06,
      "loss": 0.2154,
      "step": 164
    },
    {
      "epoch": 0.8505154639175257,
      "grad_norm": 1.1250244597985437,
      "learning_rate": 9.038011803743679e-06,
      "loss": 0.1817,
      "step": 165
    },
    {
      "epoch": 0.8556701030927835,
      "grad_norm": 1.1063531458560925,
      "learning_rate": 9.020226982089005e-06,
      "loss": 0.1564,
      "step": 166
    },
    {
      "epoch": 0.8608247422680413,
      "grad_norm": 1.1306904530717008,
      "learning_rate": 9.002297101114813e-06,
      "loss": 0.1349,
      "step": 167
    },
    {
      "epoch": 0.865979381443299,
      "grad_norm": 1.0467505323127504,
      "learning_rate": 8.984222807773707e-06,
      "loss": 0.1813,
      "step": 168
    },
    {
      "epoch": 0.8711340206185567,
      "grad_norm": 1.715578103301857,
      "learning_rate": 8.966004754229037e-06,
      "loss": 0.3016,
      "step": 169
    },
    {
      "epoch": 0.8762886597938144,
      "grad_norm": 0.9646361196653148,
      "learning_rate": 8.947643597831365e-06,
      "loss": 0.0727,
      "step": 170
    },
    {
      "epoch": 0.8814432989690721,
      "grad_norm": 0.9218462690369367,
      "learning_rate": 8.929140001094734e-06,
      "loss": 0.1584,
      "step": 171
    },
    {
      "epoch": 0.8865979381443299,
      "grad_norm": 0.9892292920559456,
      "learning_rate": 8.910494631672783e-06,
      "loss": 0.1449,
      "step": 172
    },
    {
      "epoch": 0.8917525773195877,
      "grad_norm": 1.0579410641324796,
      "learning_rate": 8.891708162334635e-06,
      "loss": 0.156,
      "step": 173
    },
    {
      "epoch": 0.8969072164948454,
      "grad_norm": 0.7647207044009656,
      "learning_rate": 8.87278127094064e-06,
      "loss": 0.1107,
      "step": 174
    },
    {
      "epoch": 0.9020618556701031,
      "grad_norm": 1.1371579318359795,
      "learning_rate": 8.853714640417906e-06,
      "loss": 0.1844,
      "step": 175
    },
    {
      "epoch": 0.9072164948453608,
      "grad_norm": 0.7724480629708496,
      "learning_rate": 8.834508958735656e-06,
      "loss": 0.0826,
      "step": 176
    },
    {
      "epoch": 0.9123711340206185,
      "grad_norm": 1.0272763923631063,
      "learning_rate": 8.815164918880418e-06,
      "loss": 0.1678,
      "step": 177
    },
    {
      "epoch": 0.9175257731958762,
      "grad_norm": 1.2865546631779095,
      "learning_rate": 8.795683218831002e-06,
      "loss": 0.2714,
      "step": 178
    },
    {
      "epoch": 0.9226804123711341,
      "grad_norm": 1.8028031060416778,
      "learning_rate": 8.776064561533329e-06,
      "loss": 0.2045,
      "step": 179
    },
    {
      "epoch": 0.9278350515463918,
      "grad_norm": 1.140464084681919,
      "learning_rate": 8.756309654875059e-06,
      "loss": 0.3075,
      "step": 180
    },
    {
      "epoch": 0.9329896907216495,
      "grad_norm": 1.1867468428900187,
      "learning_rate": 8.736419211660054e-06,
      "loss": 0.2445,
      "step": 181
    },
    {
      "epoch": 0.9381443298969072,
      "grad_norm": 1.2294904183688709,
      "learning_rate": 8.716393949582656e-06,
      "loss": 0.2416,
      "step": 182
    },
    {
      "epoch": 0.9432989690721649,
      "grad_norm": 0.7483697175175423,
      "learning_rate": 8.696234591201793e-06,
      "loss": 0.1126,
      "step": 183
    },
    {
      "epoch": 0.9484536082474226,
      "grad_norm": 0.9739357119214708,
      "learning_rate": 8.6759418639149e-06,
      "loss": 0.167,
      "step": 184
    },
    {
      "epoch": 0.9536082474226805,
      "grad_norm": 1.4251293577602404,
      "learning_rate": 8.655516499931684e-06,
      "loss": 0.3314,
      "step": 185
    },
    {
      "epoch": 0.9587628865979382,
      "grad_norm": 1.140569606339371,
      "learning_rate": 8.634959236247695e-06,
      "loss": 0.2206,
      "step": 186
    },
    {
      "epoch": 0.9639175257731959,
      "grad_norm": 0.9214436194181765,
      "learning_rate": 8.61427081461774e-06,
      "loss": 0.1622,
      "step": 187
    },
    {
      "epoch": 0.9690721649484536,
      "grad_norm": 1.4025795614332346,
      "learning_rate": 8.593451981529109e-06,
      "loss": 0.2202,
      "step": 188
    },
    {
      "epoch": 0.9742268041237113,
      "grad_norm": 0.7327985656704634,
      "learning_rate": 8.572503488174655e-06,
      "loss": 0.0798,
      "step": 189
    },
    {
      "epoch": 0.979381443298969,
      "grad_norm": 1.1641896784338794,
      "learning_rate": 8.551426090425678e-06,
      "loss": 0.1978,
      "step": 190
    },
    {
      "epoch": 0.9845360824742269,
      "grad_norm": 1.086213303235639,
      "learning_rate": 8.53022054880465e-06,
      "loss": 0.1558,
      "step": 191
    },
    {
      "epoch": 0.9896907216494846,
      "grad_norm": 1.0472347005656282,
      "learning_rate": 8.508887628457783e-06,
      "loss": 0.2053,
      "step": 192
    },
    {
      "epoch": 0.9948453608247423,
      "grad_norm": 0.7910662337183756,
      "learning_rate": 8.487428099127411e-06,
      "loss": 0.0927,
      "step": 193
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.1006430197444728,
      "learning_rate": 8.465842735124224e-06,
      "loss": 0.1894,
      "step": 194
    }
  ],
  "logging_steps": 1,
  "max_steps": 582,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 64209932976128.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}