{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 330,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.1157,
      "step": 1
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.082,
      "step": 2
    },
    {
      "epoch": 0.03,
      "learning_rate": 6e-06,
      "loss": 1.0156,
      "step": 3
    },
    {
      "epoch": 0.04,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.7246,
      "step": 4
    },
    {
      "epoch": 0.05,
      "learning_rate": 1e-05,
      "loss": 0.3215,
      "step": 5
    },
    {
      "epoch": 0.05,
      "learning_rate": 1.2e-05,
      "loss": 0.2169,
      "step": 6
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.4e-05,
      "loss": 0.1764,
      "step": 7
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.3184,
      "step": 8
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.8e-05,
      "loss": 0.1483,
      "step": 9
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 0.1492,
      "step": 10
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.999951808959328e-05,
      "loss": 0.3608,
      "step": 11
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.9998072404820648e-05,
      "loss": 0.1768,
      "step": 12
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.9995663085020215e-05,
      "loss": 0.1752,
      "step": 13
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9992290362407232e-05,
      "loss": 0.1899,
      "step": 14
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.9987954562051724e-05,
      "loss": 0.1331,
      "step": 15
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.998265610184716e-05,
      "loss": 0.1675,
      "step": 16
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.997639549247016e-05,
      "loss": 0.1531,
      "step": 17
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.9969173337331283e-05,
      "loss": 0.1414,
      "step": 18
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9960990332516875e-05,
      "loss": 0.1227,
      "step": 19
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.995184726672197e-05,
      "loss": 0.1495,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.9941745021174284e-05,
      "loss": 0.1515,
      "step": 21
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.9930684569549265e-05,
      "loss": 0.1419,
      "step": 22
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.991866697787626e-05,
      "loss": 0.1406,
      "step": 23
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.990569340443577e-05,
      "loss": 0.1284,
      "step": 24
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.989176509964781e-05,
      "loss": 0.1436,
      "step": 25
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.9876883405951378e-05,
      "loss": 0.1414,
      "step": 26
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.9861049757675087e-05,
      "loss": 0.1323,
      "step": 27
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.9844265680898917e-05,
      "loss": 0.1447,
      "step": 28
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.982653279330712e-05,
      "loss": 0.1517,
      "step": 29
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.9807852804032306e-05,
      "loss": 0.1463,
      "step": 30
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.9788227513490724e-05,
      "loss": 0.1349,
      "step": 31
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9767658813208725e-05,
      "loss": 0.1534,
      "step": 32
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.974614868564045e-05,
      "loss": 0.1437,
      "step": 33
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9723699203976768e-05,
      "loss": 0.1225,
      "step": 34
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.9700312531945444e-05,
      "loss": 0.1322,
      "step": 35
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.96759909236026e-05,
      "loss": 0.1331,
      "step": 36
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.9650736723115476e-05,
      "loss": 0.1261,
      "step": 37
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.9624552364536472e-05,
      "loss": 0.1213,
      "step": 38
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.9597440371568576e-05,
      "loss": 0.1149,
      "step": 39
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.956940335732209e-05,
      "loss": 0.1254,
      "step": 40
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.9540444024062807e-05,
      "loss": 0.121,
      "step": 41
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.9510565162951538e-05,
      "loss": 0.1371,
      "step": 42
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.9479769653775107e-05,
      "loss": 0.1579,
      "step": 43
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.944806046466878e-05,
      "loss": 0.1309,
      "step": 44
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.941544065183021e-05,
      "loss": 0.1232,
      "step": 45
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.9381913359224844e-05,
      "loss": 0.1309,
      "step": 46
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.9347481818282927e-05,
      "loss": 0.1241,
      "step": 47
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.9312149347588035e-05,
      "loss": 0.1333,
      "step": 48
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.9275919352557242e-05,
      "loss": 0.1309,
      "step": 49
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.9238795325112867e-05,
      "loss": 0.1167,
      "step": 50
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.920078084334595e-05,
      "loss": 0.1165,
      "step": 51
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.916187957117136e-05,
      "loss": 0.1158,
      "step": 52
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.9122095257974676e-05,
      "loss": 0.1201,
      "step": 53
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.9081431738250815e-05,
      "loss": 0.1453,
      "step": 54
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.9039892931234434e-05,
      "loss": 0.1194,
      "step": 55
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.8997482840522218e-05,
      "loss": 0.1323,
      "step": 56
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.895420555368697e-05,
      "loss": 0.1325,
      "step": 57
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.891006524188368e-05,
      "loss": 0.1209,
      "step": 58
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.8865066159447468e-05,
      "loss": 0.1304,
      "step": 59
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.881921264348355e-05,
      "loss": 0.1283,
      "step": 60
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.8772509113449243e-05,
      "loss": 0.1204,
      "step": 61
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.8724960070727974e-05,
      "loss": 0.1387,
      "step": 62
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.8676570098195443e-05,
      "loss": 0.126,
      "step": 63
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.862734385977792e-05,
      "loss": 0.1206,
      "step": 64
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.8577286100002723e-05,
      "loss": 0.1086,
      "step": 65
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.8526401643540924e-05,
      "loss": 0.1212,
      "step": 66
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.8474695394742345e-05,
      "loss": 0.1266,
      "step": 67
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8422172337162865e-05,
      "loss": 0.1277,
      "step": 68
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8368837533084092e-05,
      "loss": 0.1342,
      "step": 69
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.8314696123025456e-05,
      "loss": 0.1227,
      "step": 70
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.825975332524873e-05,
      "loss": 0.1191,
      "step": 71
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.8204014435255136e-05,
      "loss": 0.11,
      "step": 72
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.8147484825274895e-05,
      "loss": 0.1202,
      "step": 73
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.1083,
      "step": 74
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.803207531480645e-05,
      "loss": 0.0953,
      "step": 75
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.797320653772707e-05,
      "loss": 0.1032,
      "step": 76
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.7913569286406606e-05,
      "loss": 0.1109,
      "step": 77
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.785316930880745e-05,
      "loss": 0.127,
      "step": 78
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.779201242640517e-05,
      "loss": 0.143,
      "step": 79
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.773010453362737e-05,
      "loss": 0.124,
      "step": 80
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.7667451597285617e-05,
      "loss": 0.1215,
      "step": 81
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.7604059656000313e-05,
      "loss": 0.1086,
      "step": 82
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.7539934819618696e-05,
      "loss": 0.1035,
      "step": 83
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.747508326862597e-05,
      "loss": 0.0967,
      "step": 84
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.7409511253549592e-05,
      "loss": 0.1171,
      "step": 85
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.7343225094356857e-05,
      "loss": 0.1216,
      "step": 86
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.727623117984575e-05,
      "loss": 0.1075,
      "step": 87
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.720853596702919e-05,
      "loss": 0.1187,
      "step": 88
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.7140145980512684e-05,
      "loss": 0.1092,
      "step": 89
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.7071067811865477e-05,
      "loss": 0.134,
      "step": 90
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.7001308118985237e-05,
      "loss": 0.1008,
      "step": 91
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.6930873625456362e-05,
      "loss": 0.1127,
      "step": 92
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.685977111990193e-05,
      "loss": 0.1024,
      "step": 93
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.678800745532942e-05,
      "loss": 0.1041,
      "step": 94
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.6715589548470187e-05,
      "loss": 0.1287,
      "step": 95
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.664252437911282e-05,
      "loss": 0.1157,
      "step": 96
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.6568818989430416e-05,
      "loss": 0.1132,
      "step": 97
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.6494480483301836e-05,
      "loss": 0.1151,
      "step": 98
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.641951602562703e-05,
      "loss": 0.1016,
      "step": 99
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.6343932841636455e-05,
      "loss": 0.1094,
      "step": 100
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.6267738216194698e-05,
      "loss": 0.1066,
      "step": 101
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.6190939493098344e-05,
      "loss": 0.1069,
      "step": 102
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.6113544074368166e-05,
      "loss": 0.1185,
      "step": 103
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.6035559419535714e-05,
      "loss": 0.1226,
      "step": 104
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.5956993044924334e-05,
      "loss": 0.1036,
      "step": 105
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.5877852522924733e-05,
      "loss": 0.1055,
      "step": 106
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.579814548126514e-05,
      "loss": 0.1236,
      "step": 107
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.5717879602276123e-05,
      "loss": 0.1119,
      "step": 108
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.5637062622150168e-05,
      "loss": 0.1162,
      "step": 109
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.5555702330196024e-05,
      "loss": 0.1254,
      "step": 110
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.547380656808797e-05,
      "loss": 0.1168,
      "step": 111
    },
    {
      "epoch": 1.02,
      "learning_rate": 1.5391383229110005e-05,
      "loss": 0.102,
      "step": 112
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.5308440257395095e-05,
      "loss": 0.1043,
      "step": 113
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.5224985647159489e-05,
      "loss": 0.0925,
      "step": 114
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5141027441932217e-05,
      "loss": 0.1136,
      "step": 115
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5056573733779848e-05,
      "loss": 0.1081,
      "step": 116
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.4971632662526545e-05,
      "loss": 0.1047,
      "step": 117
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.4886212414969551e-05,
      "loss": 0.1124,
      "step": 118
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.4800321224090114e-05,
      "loss": 0.0949,
      "step": 119
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.4713967368259981e-05,
      "loss": 0.1113,
      "step": 120
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.4627159170443504e-05,
      "loss": 0.0907,
      "step": 121
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.4539904997395468e-05,
      "loss": 0.1129,
      "step": 122
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.4452213258854684e-05,
      "loss": 0.0899,
      "step": 123
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.436409240673342e-05,
      "loss": 0.0906,
      "step": 124
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.4275550934302822e-05,
      "loss": 0.0928,
      "step": 125
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4186597375374283e-05,
      "loss": 0.118,
      "step": 126
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.4097240303476955e-05,
      "loss": 0.0967,
      "step": 127
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.4007488331031409e-05,
      "loss": 0.0887,
      "step": 128
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.391735010851956e-05,
      "loss": 0.0891,
      "step": 129
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.3826834323650899e-05,
      "loss": 0.0977,
      "step": 130
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.3735949700525164e-05,
      "loss": 0.1004,
      "step": 131
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.3644704998791501e-05,
      "loss": 0.0825,
      "step": 132
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.3553109012804162e-05,
      "loss": 0.0958,
      "step": 133
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.346117057077493e-05,
      "loss": 0.1005,
      "step": 134
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.3368898533922202e-05,
      "loss": 0.0959,
      "step": 135
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.3276301795616937e-05,
      "loss": 0.0959,
      "step": 136
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.3183389280525497e-05,
      "loss": 0.0836,
      "step": 137
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.0912,
      "step": 138
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.2996652769962567e-05,
      "loss": 0.0966,
      "step": 139
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.2902846772544625e-05,
      "loss": 0.0916,
      "step": 140
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.2808760992712923e-05,
      "loss": 0.0861,
      "step": 141
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.2714404498650743e-05,
      "loss": 0.0884,
      "step": 142
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.2619786384633374e-05,
      "loss": 0.0791,
      "step": 143
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.252491577015158e-05,
      "loss": 0.0821,
      "step": 144
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.242980179903264e-05,
      "loss": 0.1128,
      "step": 145
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.2334453638559057e-05,
      "loss": 0.0871,
      "step": 146
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.2238880478584987e-05,
      "loss": 0.0884,
      "step": 147
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.2143091530650508e-05,
      "loss": 0.1058,
      "step": 148
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.2047096027093798e-05,
      "loss": 0.092,
      "step": 149
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.1950903220161286e-05,
      "loss": 0.1069,
      "step": 150
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.185452238111591e-05,
      "loss": 0.1006,
      "step": 151
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.1757962799343548e-05,
      "loss": 0.087,
      "step": 152
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.1661233781457655e-05,
      "loss": 0.0894,
      "step": 153
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.156434465040231e-05,
      "loss": 0.1061,
      "step": 154
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.1467304744553618e-05,
      "loss": 0.1179,
      "step": 155
    },
    {
      "epoch": 1.42,
      "learning_rate": 1.1370123416819683e-05,
      "loss": 0.0969,
      "step": 156
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.1272810033739134e-05,
      "loss": 0.1108,
      "step": 157
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.1175373974578378e-05,
      "loss": 0.0968,
      "step": 158
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.1077824630427593e-05,
      "loss": 0.1031,
      "step": 159
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.098017140329561e-05,
      "loss": 0.094,
      "step": 160
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.0882423705203698e-05,
      "loss": 0.0778,
      "step": 161
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.0784590957278452e-05,
      "loss": 0.0812,
      "step": 162
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.0686682588843737e-05,
      "loss": 0.1117,
      "step": 163
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.058870803651189e-05,
      "loss": 0.1008,
      "step": 164
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.0490676743274181e-05,
      "loss": 0.092,
      "step": 165
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.0392598157590687e-05,
      "loss": 0.0959,
      "step": 166
    },
    {
      "epoch": 1.52,
      "learning_rate": 1.0294481732479635e-05,
      "loss": 0.1012,
      "step": 167
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.0196336924606282e-05,
      "loss": 0.0991,
      "step": 168
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.0098173193371498e-05,
      "loss": 0.1132,
      "step": 169
    },
    {
      "epoch": 1.55,
      "learning_rate": 1e-05,
      "loss": 0.0865,
      "step": 170
    },
    {
      "epoch": 1.55,
      "learning_rate": 9.901826806628505e-06,
      "loss": 0.0828,
      "step": 171
    },
    {
      "epoch": 1.56,
      "learning_rate": 9.80366307539372e-06,
      "loss": 0.087,
      "step": 172
    },
    {
      "epoch": 1.57,
      "learning_rate": 9.705518267520369e-06,
      "loss": 0.0967,
      "step": 173
    },
    {
      "epoch": 1.58,
      "learning_rate": 9.607401842409318e-06,
      "loss": 0.0854,
      "step": 174
    },
    {
      "epoch": 1.59,
      "learning_rate": 9.50932325672582e-06,
      "loss": 0.0903,
      "step": 175
    },
    {
      "epoch": 1.6,
      "learning_rate": 9.41129196348811e-06,
      "loss": 0.0935,
      "step": 176
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.313317411156265e-06,
      "loss": 0.0974,
      "step": 177
    },
    {
      "epoch": 1.62,
      "learning_rate": 9.215409042721553e-06,
      "loss": 0.0976,
      "step": 178
    },
    {
      "epoch": 1.63,
      "learning_rate": 9.117576294796307e-06,
      "loss": 0.0968,
      "step": 179
    },
    {
      "epoch": 1.64,
      "learning_rate": 9.019828596704394e-06,
      "loss": 0.0985,
      "step": 180
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.922175369572407e-06,
      "loss": 0.0834,
      "step": 181
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.824626025421625e-06,
      "loss": 0.1207,
      "step": 182
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.72718996626087e-06,
      "loss": 0.088,
      "step": 183
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.629876583180322e-06,
      "loss": 0.0934,
      "step": 184
    },
    {
      "epoch": 1.68,
      "learning_rate": 8.532695255446384e-06,
      "loss": 0.0891,
      "step": 185
    },
    {
      "epoch": 1.69,
      "learning_rate": 8.43565534959769e-06,
      "loss": 0.1017,
      "step": 186
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.338766218542348e-06,
      "loss": 0.0998,
      "step": 187
    },
    {
      "epoch": 1.71,
      "learning_rate": 8.242037200656455e-06,
      "loss": 0.0726,
      "step": 188
    },
    {
      "epoch": 1.72,
      "learning_rate": 8.145477618884092e-06,
      "loss": 0.091,
      "step": 189
    },
    {
      "epoch": 1.73,
      "learning_rate": 8.04909677983872e-06,
      "loss": 0.1099,
      "step": 190
    },
    {
      "epoch": 1.74,
      "learning_rate": 7.952903972906205e-06,
      "loss": 0.1021,
      "step": 191
    },
    {
      "epoch": 1.75,
      "learning_rate": 7.856908469349495e-06,
      "loss": 0.0925,
      "step": 192
    },
    {
      "epoch": 1.75,
      "learning_rate": 7.761119521415017e-06,
      "loss": 0.1089,
      "step": 193
    },
    {
      "epoch": 1.76,
      "learning_rate": 7.66554636144095e-06,
      "loss": 0.1067,
      "step": 194
    },
    {
      "epoch": 1.77,
      "learning_rate": 7.570198200967363e-06,
      "loss": 0.0988,
      "step": 195
    },
    {
      "epoch": 1.78,
      "learning_rate": 7.4750842298484205e-06,
      "loss": 0.0957,
      "step": 196
    },
    {
      "epoch": 1.79,
      "learning_rate": 7.380213615366627e-06,
      "loss": 0.1194,
      "step": 197
    },
    {
      "epoch": 1.8,
      "learning_rate": 7.285595501349259e-06,
      "loss": 0.1039,
      "step": 198
    },
    {
      "epoch": 1.81,
      "learning_rate": 7.191239007287082e-06,
      "loss": 0.0911,
      "step": 199
    },
    {
      "epoch": 1.82,
      "learning_rate": 7.097153227455379e-06,
      "loss": 0.1081,
      "step": 200
    },
    {
      "epoch": 1.83,
      "learning_rate": 7.003347230037434e-06,
      "loss": 0.1016,
      "step": 201
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.0986,
      "step": 202
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.816610719474503e-06,
      "loss": 0.1006,
      "step": 203
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.723698204383067e-06,
      "loss": 0.0978,
      "step": 204
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.631101466077801e-06,
      "loss": 0.1046,
      "step": 205
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.538829429225068e-06,
      "loss": 0.0872,
      "step": 206
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.446890987195842e-06,
      "loss": 0.0891,
      "step": 207
    },
    {
      "epoch": 1.89,
      "learning_rate": 6.355295001208504e-06,
      "loss": 0.0919,
      "step": 208
    },
    {
      "epoch": 1.9,
      "learning_rate": 6.2640502994748375e-06,
      "loss": 0.1004,
      "step": 209
    },
    {
      "epoch": 1.91,
      "learning_rate": 6.173165676349103e-06,
      "loss": 0.0837,
      "step": 210
    },
    {
      "epoch": 1.92,
      "learning_rate": 6.082649891480441e-06,
      "loss": 0.0828,
      "step": 211
    },
    {
      "epoch": 1.93,
      "learning_rate": 5.9925116689685925e-06,
      "loss": 0.0937,
      "step": 212
    },
    {
      "epoch": 1.94,
      "learning_rate": 5.902759696523046e-06,
      "loss": 0.0836,
      "step": 213
    },
    {
      "epoch": 1.95,
      "learning_rate": 5.813402624625722e-06,
      "loss": 0.1102,
      "step": 214
    },
    {
      "epoch": 1.95,
      "learning_rate": 5.724449065697182e-06,
      "loss": 0.097,
      "step": 215
    },
    {
      "epoch": 1.96,
      "learning_rate": 5.635907593266578e-06,
      "loss": 0.0809,
      "step": 216
    },
    {
      "epoch": 1.97,
      "learning_rate": 5.54778674114532e-06,
      "loss": 0.0947,
      "step": 217
    },
    {
      "epoch": 1.98,
      "learning_rate": 5.460095002604533e-06,
      "loss": 0.0814,
      "step": 218
    },
    {
      "epoch": 1.99,
      "learning_rate": 5.3728408295565e-06,
      "loss": 0.0982,
      "step": 219
    },
    {
      "epoch": 2.0,
      "learning_rate": 5.286032631740023e-06,
      "loss": 0.0726,
      "step": 220
    },
    {
      "epoch": 2.01,
      "learning_rate": 5.199678775909889e-06,
      "loss": 0.0674,
      "step": 221
    },
    {
      "epoch": 2.02,
      "learning_rate": 5.1137875850304545e-06,
      "loss": 0.079,
      "step": 222
    },
    {
      "epoch": 2.03,
      "learning_rate": 5.0283673374734546e-06,
      "loss": 0.0707,
      "step": 223
    },
    {
      "epoch": 2.04,
      "learning_rate": 4.943426266220156e-06,
      "loss": 0.0679,
      "step": 224
    },
    {
      "epoch": 2.05,
      "learning_rate": 4.858972558067784e-06,
      "loss": 0.0661,
      "step": 225
    },
    {
      "epoch": 2.05,
      "learning_rate": 4.775014352840512e-06,
      "loss": 0.0699,
      "step": 226
    },
    {
      "epoch": 2.06,
      "learning_rate": 4.691559742604906e-06,
      "loss": 0.0608,
      "step": 227
    },
    {
      "epoch": 2.07,
      "learning_rate": 4.608616770889998e-06,
      "loss": 0.0905,
      "step": 228
    },
    {
      "epoch": 2.08,
      "learning_rate": 4.526193431912038e-06,
      "loss": 0.0687,
      "step": 229
    },
    {
      "epoch": 2.09,
      "learning_rate": 4.444297669803981e-06,
      "loss": 0.0648,
      "step": 230
    },
    {
      "epoch": 2.1,
      "learning_rate": 4.362937377849832e-06,
      "loss": 0.0662,
      "step": 231
    },
    {
      "epoch": 2.11,
      "learning_rate": 4.282120397723879e-06,
      "loss": 0.0502,
      "step": 232
    },
    {
      "epoch": 2.12,
      "learning_rate": 4.2018545187348645e-06,
      "loss": 0.0625,
      "step": 233
    },
    {
      "epoch": 2.13,
      "learning_rate": 4.12214747707527e-06,
      "loss": 0.0853,
      "step": 234
    },
    {
      "epoch": 2.14,
      "learning_rate": 4.043006955075667e-06,
      "loss": 0.0774,
      "step": 235
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.964440580464286e-06,
      "loss": 0.0585,
      "step": 236
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.8864559256318375e-06,
      "loss": 0.0697,
      "step": 237
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.8090605069016596e-06,
      "loss": 0.0663,
      "step": 238
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.7322617838053066e-06,
      "loss": 0.0665,
      "step": 239
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.6560671583635467e-06,
      "loss": 0.0691,
      "step": 240
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.58048397437297e-06,
      "loss": 0.0827,
      "step": 241
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.505519516698165e-06,
      "loss": 0.0594,
      "step": 242
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.4311810105695875e-06,
      "loss": 0.0781,
      "step": 243
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.3574756208871862e-06,
      "loss": 0.0573,
      "step": 244
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.284410451529816e-06,
      "loss": 0.0809,
      "step": 245
    },
    {
      "epoch": 2.24,
      "learning_rate": 3.2119925446705824e-06,
      "loss": 0.0698,
      "step": 246
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.140228880098074e-06,
      "loss": 0.0694,
      "step": 247
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.069126374543643e-06,
      "loss": 0.0482,
      "step": 248
    },
    {
      "epoch": 2.26,
      "learning_rate": 2.998691881014765e-06,
      "loss": 0.0688,
      "step": 249
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.9289321881345257e-06,
      "loss": 0.0829,
      "step": 250
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.859854019487318e-06,
      "loss": 0.0689,
      "step": 251
    },
    {
      "epoch": 2.29,
      "learning_rate": 2.791464032970812e-06,
      "loss": 0.0615,
      "step": 252
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.723768820154251e-06,
      "loss": 0.0619,
      "step": 253
    },
    {
      "epoch": 2.31,
      "learning_rate": 2.656774905643147e-06,
      "loss": 0.0605,
      "step": 254
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.5904887464504115e-06,
      "loss": 0.0607,
      "step": 255
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.5249167313740307e-06,
      "loss": 0.0531,
      "step": 256
    },
    {
      "epoch": 2.34,
      "learning_rate": 2.4600651803813057e-06,
      "loss": 0.0531,
      "step": 257
    },
    {
      "epoch": 2.35,
      "learning_rate": 2.395940343999691e-06,
      "loss": 0.0504,
      "step": 258
    },
    {
      "epoch": 2.35,
      "learning_rate": 2.332548402714385e-06,
      "loss": 0.0983,
      "step": 259
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.26989546637263e-06,
      "loss": 0.0811,
      "step": 260
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.207987573594833e-06,
      "loss": 0.0576,
      "step": 261
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.146830691192553e-06,
      "loss": 0.0946,
      "step": 262
    },
    {
      "epoch": 2.39,
      "learning_rate": 2.086430713593397e-06,
      "loss": 0.0441,
      "step": 263
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.02679346227293e-06,
      "loss": 0.0585,
      "step": 264
    },
    {
      "epoch": 2.41,
      "learning_rate": 1.967924685193552e-06,
      "loss": 0.0527,
      "step": 265
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 0.0662,
      "step": 266
    },
    {
      "epoch": 2.43,
      "learning_rate": 1.8525151747251058e-06,
      "loss": 0.0449,
      "step": 267
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.7959855647448642e-06,
      "loss": 0.0384,
      "step": 268
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.7402466747512704e-06,
      "loss": 0.0834,
      "step": 269
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.6853038769745466e-06,
      "loss": 0.0517,
      "step": 270
    },
    {
      "epoch": 2.46,
      "learning_rate": 1.6311624669159064e-06,
      "loss": 0.0505,
      "step": 271
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.577827662837136e-06,
      "loss": 0.066,
      "step": 272
    },
    {
      "epoch": 2.48,
      "learning_rate": 1.5253046052576559e-06,
      "loss": 0.0609,
      "step": 273
    },
    {
      "epoch": 2.49,
      "learning_rate": 1.4735983564590784e-06,
      "loss": 0.0614,
      "step": 274
    },
    {
      "epoch": 2.5,
      "learning_rate": 1.4227138999972801e-06,
      "loss": 0.0618,
      "step": 275
    },
    {
      "epoch": 2.51,
      "learning_rate": 1.3726561402220818e-06,
      "loss": 0.0662,
      "step": 276
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.3234299018045615e-06,
      "loss": 0.0628,
      "step": 277
    },
    {
      "epoch": 2.53,
      "learning_rate": 1.2750399292720284e-06,
      "loss": 0.072,
      "step": 278
    },
    {
      "epoch": 2.54,
      "learning_rate": 1.2274908865507595e-06,
      "loss": 0.0593,
      "step": 279
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.1807873565164507e-06,
      "loss": 0.0758,
      "step": 280
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.1349338405525368e-06,
      "loss": 0.0627,
      "step": 281
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.0899347581163222e-06,
      "loss": 0.059,
      "step": 282
    },
    {
      "epoch": 2.57,
      "learning_rate": 1.045794446313031e-06,
      "loss": 0.0748,
      "step": 283
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.0025171594777872e-06,
      "loss": 0.0499,
      "step": 284
    },
    {
      "epoch": 2.59,
      "learning_rate": 9.601070687655667e-07,
      "loss": 0.0541,
      "step": 285
    },
    {
      "epoch": 2.6,
      "learning_rate": 9.185682617491865e-07,
      "loss": 0.0505,
      "step": 286
    },
    {
      "epoch": 2.61,
      "learning_rate": 8.779047420253239e-07,
      "loss": 0.0478,
      "step": 287
    },
    {
      "epoch": 2.62,
      "learning_rate": 8.381204288286415e-07,
      "loss": 0.0704,
      "step": 288
    },
    {
      "epoch": 2.63,
      "learning_rate": 7.992191566540519e-07,
      "loss": 0.0565,
      "step": 289
    },
    {
      "epoch": 2.64,
      "learning_rate": 7.612046748871327e-07,
      "loss": 0.0654,
      "step": 290
    },
    {
      "epoch": 2.65,
      "learning_rate": 7.240806474427598e-07,
      "loss": 0.0633,
      "step": 291
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.878506524119644e-07,
      "loss": 0.0647,
      "step": 292
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.525181817170756e-07,
      "loss": 0.077,
      "step": 293
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.180866407751595e-07,
      "loss": 0.0616,
      "step": 294
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.845593481697931e-07,
      "loss": 0.0638,
      "step": 295
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.519395353312195e-07,
      "loss": 0.0668,
      "step": 296
    },
    {
      "epoch": 2.7,
      "learning_rate": 5.20230346224897e-07,
      "loss": 0.059,
      "step": 297
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.894348370484648e-07,
      "loss": 0.0596,
      "step": 298
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.5955597593719593e-07,
      "loss": 0.0668,
      "step": 299
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.305966426779118e-07,
      "loss": 0.0503,
      "step": 300
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.025596284314259e-07,
      "loss": 0.054,
      "step": 301
    },
    {
      "epoch": 2.75,
      "learning_rate": 3.7544763546352834e-07,
      "loss": 0.0558,
      "step": 302
    },
    {
      "epoch": 2.75,
      "learning_rate": 3.492632768845261e-07,
      "loss": 0.0583,
      "step": 303
    },
    {
      "epoch": 2.76,
      "learning_rate": 3.2400907639740243e-07,
      "loss": 0.0621,
      "step": 304
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.996874680545603e-07,
      "loss": 0.0617,
      "step": 305
    },
    {
      "epoch": 2.78,
      "learning_rate": 2.7630079602323447e-07,
      "loss": 0.0547,
      "step": 306
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.5385131435955e-07,
      "loss": 0.0583,
      "step": 307
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.3234118679127615e-07,
      "loss": 0.0546,
      "step": 308
    },
    {
      "epoch": 2.81,
      "learning_rate": 2.117724865092774e-07,
      "loss": 0.0476,
      "step": 309
    },
    {
      "epoch": 2.82,
      "learning_rate": 1.921471959676957e-07,
      "loss": 0.0627,
      "step": 310
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.734672066928822e-07,
      "loss": 0.0578,
      "step": 311
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.5573431910108404e-07,
      "loss": 0.0457,
      "step": 312
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.3895024232491338e-07,
      "loss": 0.0713,
      "step": 313
    },
    {
      "epoch": 2.85,
      "learning_rate": 1.231165940486234e-07,
      "loss": 0.0668,
      "step": 314
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.0823490035218986e-07,
      "loss": 0.0524,
      "step": 315
    },
    {
      "epoch": 2.87,
      "learning_rate": 9.43065955642275e-08,
      "loss": 0.0484,
      "step": 316
    },
    {
      "epoch": 2.88,
      "learning_rate": 8.133302212373961e-08,
      "loss": 0.0517,
      "step": 317
    },
    {
      "epoch": 2.89,
      "learning_rate": 6.931543045073708e-08,
      "loss": 0.0462,
      "step": 318
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.8254978825718065e-08,
      "loss": 0.0568,
      "step": 319
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.815273327803183e-08,
      "loss": 0.0475,
      "step": 320
    },
    {
      "epoch": 2.92,
      "learning_rate": 3.900966748312862e-08,
      "loss": 0.0692,
      "step": 321
    },
    {
      "epoch": 2.93,
      "learning_rate": 3.082666266872036e-08,
      "loss": 0.0603,
      "step": 322
    },
    {
      "epoch": 2.94,
      "learning_rate": 2.3604507529843e-08,
      "loss": 0.0667,
      "step": 323
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.7343898152841765e-08,
      "loss": 0.0449,
      "step": 324
    },
    {
      "epoch": 2.95,
      "learning_rate": 1.2045437948275952e-08,
      "loss": 0.0583,
      "step": 325
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.70963759277099e-09,
      "loss": 0.0738,
      "step": 326
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.336914979787832e-09,
      "loss": 0.0706,
      "step": 327
    },
    {
      "epoch": 2.98,
      "learning_rate": 1.9275951793518154e-09,
      "loss": 0.0468,
      "step": 328
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.819104067199653e-10,
      "loss": 0.0531,
      "step": 329
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0,
      "loss": 0.0516,
      "step": 330
    },
    {
      "epoch": 3.0,
      "step": 330,
      "total_flos": 7827978977280.0,
      "train_loss": 0.10793415416370739,
      "train_runtime": 3315.1857,
      "train_samples_per_second": 6.334,
      "train_steps_per_second": 0.1
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 330,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 111,
  "total_flos": 7827978977280.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}