{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 625,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 60.66320037841797,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 6.1881,
      "step": 2
    },
    {
      "epoch": 0.0064,
      "grad_norm": 42.748294830322266,
      "learning_rate": 6.349206349206349e-06,
      "loss": 5.7668,
      "step": 4
    },
    {
      "epoch": 0.0096,
      "grad_norm": 13.887235641479492,
      "learning_rate": 9.523809523809523e-06,
      "loss": 5.0841,
      "step": 6
    },
    {
      "epoch": 0.0128,
      "grad_norm": 12.488709449768066,
      "learning_rate": 1.2698412698412699e-05,
      "loss": 4.7192,
      "step": 8
    },
    {
      "epoch": 0.016,
      "grad_norm": 7.208029270172119,
      "learning_rate": 1.5873015873015872e-05,
      "loss": 4.5679,
      "step": 10
    },
    {
      "epoch": 0.0192,
      "grad_norm": 7.34755277633667,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 4.368,
      "step": 12
    },
    {
      "epoch": 0.0224,
      "grad_norm": 7.489256858825684,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 4.1398,
      "step": 14
    },
    {
      "epoch": 0.0256,
      "grad_norm": 4.802424430847168,
      "learning_rate": 2.5396825396825397e-05,
      "loss": 4.0532,
      "step": 16
    },
    {
      "epoch": 0.0288,
      "grad_norm": 4.138615131378174,
      "learning_rate": 2.857142857142857e-05,
      "loss": 3.7559,
      "step": 18
    },
    {
      "epoch": 0.032,
      "grad_norm": 5.362161636352539,
      "learning_rate": 3.1746031746031745e-05,
      "loss": 3.4854,
      "step": 20
    },
    {
      "epoch": 0.0352,
      "grad_norm": 3.878138303756714,
      "learning_rate": 3.492063492063492e-05,
      "loss": 3.5869,
      "step": 22
    },
    {
      "epoch": 0.0384,
      "grad_norm": 3.480282783508301,
      "learning_rate": 3.809523809523809e-05,
      "loss": 3.2734,
      "step": 24
    },
    {
      "epoch": 0.0416,
      "grad_norm": 3.527137279510498,
      "learning_rate": 4.126984126984127e-05,
      "loss": 3.1484,
      "step": 26
    },
    {
      "epoch": 0.0448,
      "grad_norm": 3.094705820083618,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 3.0356,
      "step": 28
    },
    {
      "epoch": 0.048,
      "grad_norm": 4.064608573913574,
      "learning_rate": 4.761904761904762e-05,
      "loss": 3.0157,
      "step": 30
    },
    {
      "epoch": 0.0512,
      "grad_norm": 3.166187047958374,
      "learning_rate": 5.0793650793650794e-05,
      "loss": 2.9229,
      "step": 32
    },
    {
      "epoch": 0.0544,
      "grad_norm": 2.909856081008911,
      "learning_rate": 5.396825396825397e-05,
      "loss": 2.8386,
      "step": 34
    },
    {
      "epoch": 0.0576,
      "grad_norm": 3.2195472717285156,
      "learning_rate": 5.714285714285714e-05,
      "loss": 2.7819,
      "step": 36
    },
    {
      "epoch": 0.0608,
      "grad_norm": 2.603515625,
      "learning_rate": 6.0317460317460316e-05,
      "loss": 2.7116,
      "step": 38
    },
    {
      "epoch": 0.064,
      "grad_norm": 3.353161096572876,
      "learning_rate": 6.349206349206349e-05,
      "loss": 2.7137,
      "step": 40
    },
    {
      "epoch": 0.0672,
      "grad_norm": 2.7278943061828613,
      "learning_rate": 6.666666666666667e-05,
      "loss": 2.6213,
      "step": 42
    },
    {
      "epoch": 0.0704,
      "grad_norm": 3.275580883026123,
      "learning_rate": 6.984126984126984e-05,
      "loss": 2.5585,
      "step": 44
    },
    {
      "epoch": 0.0736,
      "grad_norm": 2.934300422668457,
      "learning_rate": 7.301587301587302e-05,
      "loss": 2.6126,
      "step": 46
    },
    {
      "epoch": 0.0768,
      "grad_norm": 2.483461380004883,
      "learning_rate": 7.619047619047618e-05,
      "loss": 2.4616,
      "step": 48
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.5167787075042725,
      "learning_rate": 7.936507936507937e-05,
      "loss": 2.4316,
      "step": 50
    },
    {
      "epoch": 0.0832,
      "grad_norm": 2.211185932159424,
      "learning_rate": 8.253968253968255e-05,
      "loss": 2.63,
      "step": 52
    },
    {
      "epoch": 0.0864,
      "grad_norm": 3.2666075229644775,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.4892,
      "step": 54
    },
    {
      "epoch": 0.0896,
      "grad_norm": 2.849605083465576,
      "learning_rate": 8.888888888888889e-05,
      "loss": 2.4584,
      "step": 56
    },
    {
      "epoch": 0.0928,
      "grad_norm": 3.0103588104248047,
      "learning_rate": 9.206349206349206e-05,
      "loss": 2.2766,
      "step": 58
    },
    {
      "epoch": 0.096,
      "grad_norm": 2.306534767150879,
      "learning_rate": 9.523809523809524e-05,
      "loss": 2.4357,
      "step": 60
    },
    {
      "epoch": 0.0992,
      "grad_norm": 2.3400485515594482,
      "learning_rate": 9.841269841269841e-05,
      "loss": 2.4118,
      "step": 62
    },
    {
      "epoch": 0.1024,
      "grad_norm": 2.583407163619995,
      "learning_rate": 9.99998248790669e-05,
      "loss": 2.2326,
      "step": 64
    },
    {
      "epoch": 0.1056,
      "grad_norm": 2.57265567779541,
      "learning_rate": 9.999842391896222e-05,
      "loss": 2.2923,
      "step": 66
    },
    {
      "epoch": 0.1088,
      "grad_norm": 2.307471990585327,
      "learning_rate": 9.999562203800676e-05,
      "loss": 2.2235,
      "step": 68
    },
    {
      "epoch": 0.112,
      "grad_norm": 2.357084035873413,
      "learning_rate": 9.999141931470729e-05,
      "loss": 2.2896,
      "step": 70
    },
    {
      "epoch": 0.1152,
      "grad_norm": 2.6181581020355225,
      "learning_rate": 9.998581586682116e-05,
      "loss": 2.3015,
      "step": 72
    },
    {
      "epoch": 0.1184,
      "grad_norm": 2.253117799758911,
      "learning_rate": 9.997881185135307e-05,
      "loss": 2.1824,
      "step": 74
    },
    {
      "epoch": 0.1216,
      "grad_norm": 2.546729326248169,
      "learning_rate": 9.997040746455062e-05,
      "loss": 2.1502,
      "step": 76
    },
    {
      "epoch": 0.1248,
      "grad_norm": 2.4144699573516846,
      "learning_rate": 9.996060294189887e-05,
      "loss": 2.3715,
      "step": 78
    },
    {
      "epoch": 0.128,
      "grad_norm": 2.3093016147613525,
      "learning_rate": 9.994939855811362e-05,
      "loss": 2.2753,
      "step": 80
    },
    {
      "epoch": 0.1312,
      "grad_norm": 2.5628409385681152,
      "learning_rate": 9.993679462713395e-05,
      "loss": 2.3152,
      "step": 82
    },
    {
      "epoch": 0.1344,
      "grad_norm": 2.549136161804199,
      "learning_rate": 9.992279150211314e-05,
      "loss": 2.1171,
      "step": 84
    },
    {
      "epoch": 0.1376,
      "grad_norm": 2.4570610523223877,
      "learning_rate": 9.990738957540896e-05,
      "loss": 2.2414,
      "step": 86
    },
    {
      "epoch": 0.1408,
      "grad_norm": 2.256564140319824,
      "learning_rate": 9.989058927857263e-05,
      "loss": 2.1324,
      "step": 88
    },
    {
      "epoch": 0.144,
      "grad_norm": 2.818751811981201,
      "learning_rate": 9.987239108233668e-05,
      "loss": 2.184,
      "step": 90
    },
    {
      "epoch": 0.1472,
      "grad_norm": 2.432871103286743,
      "learning_rate": 9.985279549660185e-05,
      "loss": 2.1899,
      "step": 92
    },
    {
      "epoch": 0.1504,
      "grad_norm": 2.1021323204040527,
      "learning_rate": 9.983180307042274e-05,
      "loss": 2.1064,
      "step": 94
    },
    {
      "epoch": 0.1536,
      "grad_norm": 2.7487058639526367,
      "learning_rate": 9.980941439199246e-05,
      "loss": 2.2197,
      "step": 96
    },
    {
      "epoch": 0.1568,
      "grad_norm": 2.82835054397583,
      "learning_rate": 9.97856300886261e-05,
      "loss": 2.2048,
      "step": 98
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.25872802734375,
      "learning_rate": 9.976045082674319e-05,
      "loss": 2.1002,
      "step": 100
    },
    {
      "epoch": 0.1632,
      "grad_norm": 2.040614366531372,
      "learning_rate": 9.973387731184902e-05,
      "loss": 2.1031,
      "step": 102
    },
    {
      "epoch": 0.1664,
      "grad_norm": 2.437248706817627,
      "learning_rate": 9.97059102885149e-05,
      "loss": 2.1416,
      "step": 104
    },
    {
      "epoch": 0.1696,
      "grad_norm": 2.0928878784179688,
      "learning_rate": 9.967655054035727e-05,
      "loss": 2.1576,
      "step": 106
    },
    {
      "epoch": 0.1728,
      "grad_norm": 2.2243545055389404,
      "learning_rate": 9.964579889001569e-05,
      "loss": 1.9863,
      "step": 108
    },
    {
      "epoch": 0.176,
      "grad_norm": 2.1860439777374268,
      "learning_rate": 9.961365619912989e-05,
      "loss": 2.0016,
      "step": 110
    },
    {
      "epoch": 0.1792,
      "grad_norm": 2.527122735977173,
      "learning_rate": 9.95801233683156e-05,
      "loss": 2.1272,
      "step": 112
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.8613876104354858,
      "learning_rate": 9.954520133713924e-05,
      "loss": 2.2001,
      "step": 114
    },
    {
      "epoch": 0.1856,
      "grad_norm": 2.115910530090332,
      "learning_rate": 9.950889108409172e-05,
      "loss": 2.0871,
      "step": 116
    },
    {
      "epoch": 0.1888,
      "grad_norm": 2.361309051513672,
      "learning_rate": 9.947119362656092e-05,
      "loss": 2.017,
      "step": 118
    },
    {
      "epoch": 0.192,
      "grad_norm": 2.09470272064209,
      "learning_rate": 9.94321100208032e-05,
      "loss": 2.1847,
      "step": 120
    },
    {
      "epoch": 0.1952,
      "grad_norm": 1.9747451543807983,
      "learning_rate": 9.939164136191384e-05,
      "loss": 2.324,
      "step": 122
    },
    {
      "epoch": 0.1984,
      "grad_norm": 1.8229223489761353,
      "learning_rate": 9.934978878379636e-05,
      "loss": 2.1454,
      "step": 124
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.9113378524780273,
      "learning_rate": 9.930655345913071e-05,
      "loss": 2.0096,
      "step": 126
    },
    {
      "epoch": 0.2048,
      "grad_norm": 2.385289192199707,
      "learning_rate": 9.926193659934043e-05,
      "loss": 2.1029,
      "step": 128
    },
    {
      "epoch": 0.208,
      "grad_norm": 2.101463794708252,
      "learning_rate": 9.921593945455869e-05,
      "loss": 2.0172,
      "step": 130
    },
    {
      "epoch": 0.2112,
      "grad_norm": 2.2676024436950684,
      "learning_rate": 9.916856331359335e-05,
      "loss": 1.9966,
      "step": 132
    },
    {
      "epoch": 0.2144,
      "grad_norm": 2.0168168544769287,
      "learning_rate": 9.911980950389067e-05,
      "loss": 2.1807,
      "step": 134
    },
    {
      "epoch": 0.2176,
      "grad_norm": 2.1054186820983887,
      "learning_rate": 9.906967939149831e-05,
      "loss": 1.9759,
      "step": 136
    },
    {
      "epoch": 0.2208,
      "grad_norm": 2.3354573249816895,
      "learning_rate": 9.901817438102695e-05,
      "loss": 1.995,
      "step": 138
    },
    {
      "epoch": 0.224,
      "grad_norm": 2.2721822261810303,
      "learning_rate": 9.896529591561093e-05,
      "loss": 2.2239,
      "step": 140
    },
    {
      "epoch": 0.2272,
      "grad_norm": 1.9209738969802856,
      "learning_rate": 9.891104547686782e-05,
      "loss": 2.0051,
      "step": 142
    },
    {
      "epoch": 0.2304,
      "grad_norm": 1.978259801864624,
      "learning_rate": 9.8855424584857e-05,
      "loss": 2.0367,
      "step": 144
    },
    {
      "epoch": 0.2336,
      "grad_norm": 1.9169765710830688,
      "learning_rate": 9.879843479803691e-05,
      "loss": 2.1009,
      "step": 146
    },
    {
      "epoch": 0.2368,
      "grad_norm": 1.8380109071731567,
      "learning_rate": 9.874007771322151e-05,
      "loss": 2.1456,
      "step": 148
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.1143693923950195,
      "learning_rate": 9.868035496553546e-05,
      "loss": 1.925,
      "step": 150
    },
    {
      "epoch": 0.2432,
      "grad_norm": 1.8774141073226929,
      "learning_rate": 9.86192682283684e-05,
      "loss": 1.9616,
      "step": 152
    },
    {
      "epoch": 0.2464,
      "grad_norm": 2.3532581329345703,
      "learning_rate": 9.855681921332793e-05,
      "loss": 2.0289,
      "step": 154
    },
    {
      "epoch": 0.2496,
      "grad_norm": 2.1421797275543213,
      "learning_rate": 9.849300967019175e-05,
      "loss": 2.0153,
      "step": 156
    },
    {
      "epoch": 0.2528,
      "grad_norm": 2.0029852390289307,
      "learning_rate": 9.84278413868586e-05,
      "loss": 2.0726,
      "step": 158
    },
    {
      "epoch": 0.256,
      "grad_norm": 2.0344998836517334,
      "learning_rate": 9.836131618929819e-05,
      "loss": 2.0215,
      "step": 160
    },
    {
      "epoch": 0.2592,
      "grad_norm": 1.8781356811523438,
      "learning_rate": 9.82934359415e-05,
      "loss": 2.0622,
      "step": 162
    },
    {
      "epoch": 0.2624,
      "grad_norm": 1.9795514345169067,
      "learning_rate": 9.822420254542108e-05,
      "loss": 2.0249,
      "step": 164
    },
    {
      "epoch": 0.2656,
      "grad_norm": 2.012881278991699,
      "learning_rate": 9.815361794093272e-05,
      "loss": 1.9815,
      "step": 166
    },
    {
      "epoch": 0.2688,
      "grad_norm": 2.264941453933716,
      "learning_rate": 9.808168410576617e-05,
      "loss": 2.0232,
      "step": 168
    },
    {
      "epoch": 0.272,
      "grad_norm": 2.4006729125976562,
      "learning_rate": 9.800840305545715e-05,
      "loss": 2.0844,
      "step": 170
    },
    {
      "epoch": 0.2752,
      "grad_norm": 2.0443308353424072,
      "learning_rate": 9.793377684328939e-05,
      "loss": 2.2302,
      "step": 172
    },
    {
      "epoch": 0.2784,
      "grad_norm": 2.164515972137451,
      "learning_rate": 9.785780756023714e-05,
      "loss": 1.9808,
      "step": 174
    },
    {
      "epoch": 0.2816,
      "grad_norm": 1.9512875080108643,
      "learning_rate": 9.778049733490655e-05,
      "loss": 2.0968,
      "step": 176
    },
    {
      "epoch": 0.2848,
      "grad_norm": 1.9964834451675415,
      "learning_rate": 9.770184833347606e-05,
      "loss": 1.9889,
      "step": 178
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.9380826950073242,
      "learning_rate": 9.762186275963563e-05,
      "loss": 1.9766,
      "step": 180
    },
    {
      "epoch": 0.2912,
      "grad_norm": 1.943260669708252,
      "learning_rate": 9.754054285452506e-05,
      "loss": 1.9298,
      "step": 182
    },
    {
      "epoch": 0.2944,
      "grad_norm": 2.1821844577789307,
      "learning_rate": 9.745789089667121e-05,
      "loss": 2.1202,
      "step": 184
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.7526299953460693,
      "learning_rate": 9.737390920192408e-05,
      "loss": 2.0635,
      "step": 186
    },
    {
      "epoch": 0.3008,
      "grad_norm": 2.229520082473755,
      "learning_rate": 9.7288600123392e-05,
      "loss": 1.9582,
      "step": 188
    },
    {
      "epoch": 0.304,
      "grad_norm": 2.3614768981933594,
      "learning_rate": 9.720196605137565e-05,
      "loss": 2.0278,
      "step": 190
    },
    {
      "epoch": 0.3072,
      "grad_norm": 2.1270534992218018,
      "learning_rate": 9.71140094133011e-05,
      "loss": 2.1036,
      "step": 192
    },
    {
      "epoch": 0.3104,
      "grad_norm": 2.2983131408691406,
      "learning_rate": 9.702473267365182e-05,
      "loss": 2.0558,
      "step": 194
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.9561504125595093,
      "learning_rate": 9.693413833389956e-05,
      "loss": 1.9173,
      "step": 196
    },
    {
      "epoch": 0.3168,
      "grad_norm": 2.234160900115967,
      "learning_rate": 9.684222893243431e-05,
      "loss": 2.1188,
      "step": 198
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.883965015411377,
      "learning_rate": 9.674900704449324e-05,
      "loss": 1.9584,
      "step": 200
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.7237235307693481,
      "learning_rate": 9.665447528208836e-05,
      "loss": 1.9351,
      "step": 202
    },
    {
      "epoch": 0.3264,
      "grad_norm": 2.0437498092651367,
      "learning_rate": 9.655863629393351e-05,
      "loss": 1.9079,
      "step": 204
    },
    {
      "epoch": 0.3296,
      "grad_norm": 2.014540195465088,
      "learning_rate": 9.64614927653701e-05,
      "loss": 1.8612,
      "step": 206
    },
    {
      "epoch": 0.3328,
      "grad_norm": 2.379439115524292,
      "learning_rate": 9.636304741829181e-05,
      "loss": 1.9976,
      "step": 208
    },
    {
      "epoch": 0.336,
      "grad_norm": 1.962538242340088,
      "learning_rate": 9.626330301106837e-05,
      "loss": 1.932,
      "step": 210
    },
    {
      "epoch": 0.3392,
      "grad_norm": 1.862244725227356,
      "learning_rate": 9.616226233846828e-05,
      "loss": 1.8992,
      "step": 212
    },
    {
      "epoch": 0.3424,
      "grad_norm": 1.7304776906967163,
      "learning_rate": 9.605992823158046e-05,
      "loss": 2.0777,
      "step": 214
    },
    {
      "epoch": 0.3456,
      "grad_norm": 2.2403054237365723,
      "learning_rate": 9.595630355773501e-05,
      "loss": 1.8658,
      "step": 216
    },
    {
      "epoch": 0.3488,
      "grad_norm": 3.3899903297424316,
      "learning_rate": 9.585139122042274e-05,
      "loss": 1.9963,
      "step": 218
    },
    {
      "epoch": 0.352,
      "grad_norm": 2.261810064315796,
      "learning_rate": 9.574519415921396e-05,
      "loss": 1.947,
      "step": 220
    },
    {
      "epoch": 0.3552,
      "grad_norm": 2.2053134441375732,
      "learning_rate": 9.5637715349676e-05,
      "loss": 2.0544,
      "step": 222
    },
    {
      "epoch": 0.3584,
      "grad_norm": 1.871773362159729,
      "learning_rate": 9.552895780328987e-05,
      "loss": 1.8976,
      "step": 224
    },
    {
      "epoch": 0.3616,
      "grad_norm": 1.6700202226638794,
      "learning_rate": 9.541892456736595e-05,
      "loss": 2.1166,
      "step": 226
    },
    {
      "epoch": 0.3648,
      "grad_norm": 1.9986639022827148,
      "learning_rate": 9.530761872495849e-05,
      "loss": 1.9311,
      "step": 228
    },
    {
      "epoch": 0.368,
      "grad_norm": 2.288973331451416,
      "learning_rate": 9.519504339477932e-05,
      "loss": 1.98,
      "step": 230
    },
    {
      "epoch": 0.3712,
      "grad_norm": 2.177896738052368,
      "learning_rate": 9.508120173111039e-05,
      "loss": 1.862,
      "step": 232
    },
    {
      "epoch": 0.3744,
      "grad_norm": 1.9860484600067139,
      "learning_rate": 9.496609692371548e-05,
      "loss": 1.9192,
      "step": 234
    },
    {
      "epoch": 0.3776,
      "grad_norm": 1.924127221107483,
      "learning_rate": 9.484973219775074e-05,
      "loss": 1.871,
      "step": 236
    },
    {
      "epoch": 0.3808,
      "grad_norm": 1.9022867679595947,
      "learning_rate": 9.473211081367436e-05,
      "loss": 1.9067,
      "step": 238
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.7447446584701538,
      "learning_rate": 9.46132360671552e-05,
      "loss": 1.8984,
      "step": 240
    },
    {
      "epoch": 0.3872,
      "grad_norm": 2.809067487716675,
      "learning_rate": 9.449311128898049e-05,
      "loss": 1.8327,
      "step": 242
    },
    {
      "epoch": 0.3904,
      "grad_norm": 1.9946494102478027,
      "learning_rate": 9.437173984496246e-05,
      "loss": 1.9735,
      "step": 244
    },
    {
      "epoch": 0.3936,
      "grad_norm": 1.8834348917007446,
      "learning_rate": 9.424912513584401e-05,
      "loss": 2.0294,
      "step": 246
    },
    {
      "epoch": 0.3968,
      "grad_norm": 1.9426389932632446,
      "learning_rate": 9.412527059720352e-05,
      "loss": 1.9919,
      "step": 248
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.823935627937317,
      "learning_rate": 9.400017969935848e-05,
      "loss": 1.8907,
      "step": 250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 2.1048786640167236,
      "learning_rate": 9.387385594726829e-05,
      "loss": 1.8855,
      "step": 252
    },
    {
      "epoch": 0.4064,
      "grad_norm": 1.9253580570220947,
      "learning_rate": 9.374630288043614e-05,
      "loss": 2.0577,
      "step": 254
    },
    {
      "epoch": 0.4096,
      "grad_norm": 1.785396695137024,
      "learning_rate": 9.361752407280965e-05,
      "loss": 1.9675,
      "step": 256
    },
    {
      "epoch": 0.4128,
      "grad_norm": 1.9203846454620361,
      "learning_rate": 9.348752313268093e-05,
      "loss": 1.8934,
      "step": 258
    },
    {
      "epoch": 0.416,
      "grad_norm": 1.986392855644226,
      "learning_rate": 9.335630370258533e-05,
      "loss": 1.9838,
      "step": 260
    },
    {
      "epoch": 0.4192,
      "grad_norm": 1.953905463218689,
      "learning_rate": 9.322386945919946e-05,
      "loss": 1.7604,
      "step": 262
    },
    {
      "epoch": 0.4224,
      "grad_norm": 1.7314627170562744,
      "learning_rate": 9.309022411323816e-05,
      "loss": 2.0328,
      "step": 264
    },
    {
      "epoch": 0.4256,
      "grad_norm": 1.6745048761367798,
      "learning_rate": 9.295537140935049e-05,
      "loss": 1.9734,
      "step": 266
    },
    {
      "epoch": 0.4288,
      "grad_norm": 1.8622961044311523,
      "learning_rate": 9.281931512601485e-05,
      "loss": 1.9509,
      "step": 268
    },
    {
      "epoch": 0.432,
      "grad_norm": 2.014514684677124,
      "learning_rate": 9.26820590754331e-05,
      "loss": 1.8272,
      "step": 270
    },
    {
      "epoch": 0.4352,
      "grad_norm": 2.118647336959839,
      "learning_rate": 9.254360710342371e-05,
      "loss": 1.8347,
      "step": 272
    },
    {
      "epoch": 0.4384,
      "grad_norm": 2.04239821434021,
      "learning_rate": 9.240396308931407e-05,
      "loss": 1.8675,
      "step": 274
    },
    {
      "epoch": 0.4416,
      "grad_norm": 1.951341152191162,
      "learning_rate": 9.226313094583173e-05,
      "loss": 1.9559,
      "step": 276
    },
    {
      "epoch": 0.4448,
      "grad_norm": 1.7053275108337402,
      "learning_rate": 9.212111461899479e-05,
      "loss": 2.0715,
      "step": 278
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.7789607048034668,
      "learning_rate": 9.197791808800135e-05,
      "loss": 1.89,
      "step": 280
    },
    {
      "epoch": 0.4512,
      "grad_norm": 1.8625364303588867,
      "learning_rate": 9.183354536511803e-05,
      "loss": 1.9809,
      "step": 282
    },
    {
      "epoch": 0.4544,
      "grad_norm": 1.6965309381484985,
      "learning_rate": 9.168800049556747e-05,
      "loss": 1.8365,
      "step": 284
    },
    {
      "epoch": 0.4576,
      "grad_norm": 2.1207497119903564,
      "learning_rate": 9.154128755741509e-05,
      "loss": 1.8314,
      "step": 286
    },
    {
      "epoch": 0.4608,
      "grad_norm": 1.8182010650634766,
      "learning_rate": 9.139341066145472e-05,
      "loss": 1.8906,
      "step": 288
    },
    {
      "epoch": 0.464,
      "grad_norm": 1.977777361869812,
      "learning_rate": 9.124437395109353e-05,
      "loss": 1.8562,
      "step": 290
    },
    {
      "epoch": 0.4672,
      "grad_norm": 1.9953404664993286,
      "learning_rate": 9.109418160223585e-05,
      "loss": 1.8364,
      "step": 292
    },
    {
      "epoch": 0.4704,
      "grad_norm": 1.9941433668136597,
      "learning_rate": 9.094283782316619e-05,
      "loss": 1.7585,
      "step": 294
    },
    {
      "epoch": 0.4736,
      "grad_norm": 1.9799609184265137,
      "learning_rate": 9.079034685443133e-05,
      "loss": 1.8669,
      "step": 296
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.755238652229309,
      "learning_rate": 9.063671296872149e-05,
      "loss": 1.8001,
      "step": 298
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.059305429458618,
      "learning_rate": 9.048194047075069e-05,
      "loss": 1.9259,
      "step": 300
    },
    {
      "epoch": 0.4832,
      "grad_norm": 1.7116378545761108,
      "learning_rate": 9.032603369713596e-05,
      "loss": 1.6954,
      "step": 302
    },
    {
      "epoch": 0.4864,
      "grad_norm": 2.472815990447998,
      "learning_rate": 9.016899701627604e-05,
      "loss": 1.8413,
      "step": 304
    },
    {
      "epoch": 0.4896,
      "grad_norm": 1.8934400081634521,
      "learning_rate": 9.00108348282288e-05,
      "loss": 1.9545,
      "step": 306
    },
    {
      "epoch": 0.4928,
      "grad_norm": 2.147753953933716,
      "learning_rate": 8.985155156458811e-05,
      "loss": 1.7679,
      "step": 308
    },
    {
      "epoch": 0.496,
      "grad_norm": 2.2302675247192383,
      "learning_rate": 8.969115168835954e-05,
      "loss": 1.8257,
      "step": 310
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.6578640937805176,
      "learning_rate": 8.952963969383538e-05,
      "loss": 1.7151,
      "step": 312
    },
    {
      "epoch": 0.5024,
      "grad_norm": 1.754835844039917,
      "learning_rate": 8.93670201064687e-05,
      "loss": 2.0074,
      "step": 314
    },
    {
      "epoch": 0.5056,
      "grad_norm": 2.130150556564331,
      "learning_rate": 8.920329748274649e-05,
      "loss": 1.8657,
      "step": 316
    },
    {
      "epoch": 0.5088,
      "grad_norm": 1.7068381309509277,
      "learning_rate": 8.903847641006218e-05,
      "loss": 1.8955,
      "step": 318
    },
    {
      "epoch": 0.512,
      "grad_norm": 2.0879528522491455,
      "learning_rate": 8.887256150658684e-05,
      "loss": 1.7092,
      "step": 320
    },
    {
      "epoch": 0.5152,
      "grad_norm": 1.8985047340393066,
      "learning_rate": 8.870555742113998e-05,
      "loss": 1.8091,
      "step": 322
    },
    {
      "epoch": 0.5184,
      "grad_norm": 1.7577992677688599,
      "learning_rate": 8.85374688330592e-05,
      "loss": 1.8895,
      "step": 324
    },
    {
      "epoch": 0.5216,
      "grad_norm": 1.8277013301849365,
      "learning_rate": 8.836830045206911e-05,
      "loss": 1.8192,
      "step": 326
    },
    {
      "epoch": 0.5248,
      "grad_norm": 1.8492199182510376,
      "learning_rate": 8.81980570181494e-05,
      "loss": 2.0282,
      "step": 328
    },
    {
      "epoch": 0.528,
      "grad_norm": 1.8850246667861938,
      "learning_rate": 8.802674330140192e-05,
      "loss": 1.7955,
      "step": 330
    },
    {
      "epoch": 0.5312,
      "grad_norm": 1.7965402603149414,
      "learning_rate": 8.785436410191714e-05,
      "loss": 1.8271,
      "step": 332
    },
    {
      "epoch": 0.5344,
      "grad_norm": 2.0495541095733643,
      "learning_rate": 8.76809242496396e-05,
      "loss": 1.9308,
      "step": 334
    },
    {
      "epoch": 0.5376,
      "grad_norm": 1.8388515710830688,
      "learning_rate": 8.750642860423262e-05,
      "loss": 1.8831,
      "step": 336
    },
    {
      "epoch": 0.5408,
      "grad_norm": 2.2101669311523438,
      "learning_rate": 8.733088205494205e-05,
      "loss": 1.9837,
      "step": 338
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.7564021348953247,
      "learning_rate": 8.715428952045936e-05,
      "loss": 2.0114,
      "step": 340
    },
    {
      "epoch": 0.5472,
      "grad_norm": 2.0515785217285156,
      "learning_rate": 8.697665594878382e-05,
      "loss": 1.7574,
      "step": 342
    },
    {
      "epoch": 0.5504,
      "grad_norm": 2.1503772735595703,
      "learning_rate": 8.679798631708375e-05,
      "loss": 1.9549,
      "step": 344
    },
    {
      "epoch": 0.5536,
      "grad_norm": 1.6707327365875244,
      "learning_rate": 8.661828563155727e-05,
      "loss": 1.9318,
      "step": 346
    },
    {
      "epoch": 0.5568,
      "grad_norm": 1.9014642238616943,
      "learning_rate": 8.643755892729179e-05,
      "loss": 1.9853,
      "step": 348
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.9820547103881836,
      "learning_rate": 8.625581126812312e-05,
      "loss": 1.8178,
      "step": 350
    },
    {
      "epoch": 0.5632,
      "grad_norm": 2.810029983520508,
      "learning_rate": 8.607304774649349e-05,
      "loss": 2.0081,
      "step": 352
    },
    {
      "epoch": 0.5664,
      "grad_norm": 1.8511972427368164,
      "learning_rate": 8.588927348330887e-05,
      "loss": 1.7794,
      "step": 354
    },
    {
      "epoch": 0.5696,
      "grad_norm": 1.954455852508545,
      "learning_rate": 8.57044936277955e-05,
      "loss": 1.9215,
      "step": 356
    },
    {
      "epoch": 0.5728,
      "grad_norm": 1.8836822509765625,
      "learning_rate": 8.551871335735565e-05,
      "loss": 1.7449,
      "step": 358
    },
    {
      "epoch": 0.576,
      "grad_norm": 1.8966975212097168,
      "learning_rate": 8.533193787742251e-05,
      "loss": 1.7689,
      "step": 360
    },
    {
      "epoch": 0.5792,
      "grad_norm": 1.7771093845367432,
      "learning_rate": 8.51441724213143e-05,
      "loss": 1.8151,
      "step": 362
    },
    {
      "epoch": 0.5824,
      "grad_norm": 1.880419135093689,
      "learning_rate": 8.495542225008771e-05,
      "loss": 1.805,
      "step": 364
    },
    {
      "epoch": 0.5856,
      "grad_norm": 1.820349097251892,
      "learning_rate": 8.476569265239046e-05,
      "loss": 1.758,
      "step": 366
    },
    {
      "epoch": 0.5888,
      "grad_norm": 1.984392523765564,
      "learning_rate": 8.457498894431311e-05,
      "loss": 1.7321,
      "step": 368
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.710229516029358,
      "learning_rate": 8.438331646924013e-05,
      "loss": 1.7819,
      "step": 370
    },
    {
      "epoch": 0.5952,
      "grad_norm": 1.736141324043274,
      "learning_rate": 8.419068059770011e-05,
      "loss": 1.8351,
      "step": 372
    },
    {
      "epoch": 0.5984,
      "grad_norm": 1.6661279201507568,
      "learning_rate": 8.399708672721539e-05,
      "loss": 1.803,
      "step": 374
    },
    {
      "epoch": 0.6016,
      "grad_norm": 4.828789710998535,
      "learning_rate": 8.380254028215076e-05,
      "loss": 1.8539,
      "step": 376
    },
    {
      "epoch": 0.6048,
      "grad_norm": 2.078886032104492,
      "learning_rate": 8.360704671356145e-05,
      "loss": 1.7976,
      "step": 378
    },
    {
      "epoch": 0.608,
      "grad_norm": 1.720009684562683,
      "learning_rate": 8.341061149904045e-05,
      "loss": 1.9524,
      "step": 380
    },
    {
      "epoch": 0.6112,
      "grad_norm": 1.935594081878662,
      "learning_rate": 8.321324014256504e-05,
      "loss": 1.8671,
      "step": 382
    },
    {
      "epoch": 0.6144,
      "grad_norm": 1.868320345878601,
      "learning_rate": 8.30149381743425e-05,
      "loss": 1.8896,
      "step": 384
    },
    {
      "epoch": 0.6176,
      "grad_norm": 2.0732314586639404,
      "learning_rate": 8.28157111506552e-05,
      "loss": 1.8446,
      "step": 386
    },
    {
      "epoch": 0.6208,
      "grad_norm": 1.5798280239105225,
      "learning_rate": 8.261556465370493e-05,
      "loss": 1.9207,
      "step": 388
    },
    {
      "epoch": 0.624,
      "grad_norm": 1.6934467554092407,
      "learning_rate": 8.24145042914565e-05,
      "loss": 1.7548,
      "step": 390
    },
    {
      "epoch": 0.6272,
      "grad_norm": 1.7732023000717163,
      "learning_rate": 8.221253569748055e-05,
      "loss": 1.7041,
      "step": 392
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.9565222263336182,
      "learning_rate": 8.200966453079575e-05,
      "loss": 1.8865,
      "step": 394
    },
    {
      "epoch": 0.6336,
      "grad_norm": 1.7031235694885254,
      "learning_rate": 8.180589647571023e-05,
      "loss": 2.0219,
      "step": 396
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.8705931901931763,
      "learning_rate": 8.16012372416623e-05,
      "loss": 1.7774,
      "step": 398
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7355400323867798,
      "learning_rate": 8.13956925630605e-05,
      "loss": 1.7273,
      "step": 400
    },
    {
      "epoch": 0.6432,
      "grad_norm": 1.7146542072296143,
      "learning_rate": 8.118926819912287e-05,
      "loss": 1.8275,
      "step": 402
    },
    {
      "epoch": 0.6464,
      "grad_norm": 1.8502819538116455,
      "learning_rate": 8.098196993371565e-05,
      "loss": 1.856,
      "step": 404
    },
    {
      "epoch": 0.6496,
      "grad_norm": 1.6460517644882202,
      "learning_rate": 8.077380357519115e-05,
      "loss": 1.7826,
      "step": 406
    },
    {
      "epoch": 0.6528,
      "grad_norm": 1.6977733373641968,
      "learning_rate": 8.056477495622511e-05,
      "loss": 2.0396,
      "step": 408
    },
    {
      "epoch": 0.656,
      "grad_norm": 2.395606756210327,
      "learning_rate": 8.035488993365312e-05,
      "loss": 1.755,
      "step": 410
    },
    {
      "epoch": 0.6592,
      "grad_norm": 1.6800931692123413,
      "learning_rate": 8.014415438830667e-05,
      "loss": 1.9174,
      "step": 412
    },
    {
      "epoch": 0.6624,
      "grad_norm": 1.940741777420044,
      "learning_rate": 7.993257422484826e-05,
      "loss": 1.7259,
      "step": 414
    },
    {
      "epoch": 0.6656,
      "grad_norm": 1.6088985204696655,
      "learning_rate": 7.972015537160602e-05,
      "loss": 1.9236,
      "step": 416
    },
    {
      "epoch": 0.6688,
      "grad_norm": 1.77496337890625,
      "learning_rate": 7.950690378040758e-05,
      "loss": 1.9956,
      "step": 418
    },
    {
      "epoch": 0.672,
      "grad_norm": 2.08013653755188,
      "learning_rate": 7.929282542641325e-05,
      "loss": 1.71,
      "step": 420
    },
    {
      "epoch": 0.6752,
      "grad_norm": 1.9645555019378662,
      "learning_rate": 7.907792630794876e-05,
      "loss": 1.6586,
      "step": 422
    },
    {
      "epoch": 0.6784,
      "grad_norm": 2.035111904144287,
      "learning_rate": 7.886221244633703e-05,
      "loss": 1.8481,
      "step": 424
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.617519736289978,
      "learning_rate": 7.864568988572947e-05,
      "loss": 1.8787,
      "step": 426
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.9266173839569092,
      "learning_rate": 7.842836469293673e-05,
      "loss": 1.7332,
      "step": 428
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.6716456413269043,
      "learning_rate": 7.821024295725865e-05,
      "loss": 1.8147,
      "step": 430
    },
    {
      "epoch": 0.6912,
      "grad_norm": 1.9675475358963013,
      "learning_rate": 7.79913307903136e-05,
      "loss": 1.77,
      "step": 432
    },
    {
      "epoch": 0.6944,
      "grad_norm": 2.048152208328247,
      "learning_rate": 7.777163432586734e-05,
      "loss": 1.7438,
      "step": 434
    },
    {
      "epoch": 0.6976,
      "grad_norm": 1.7210822105407715,
      "learning_rate": 7.755115971966104e-05,
      "loss": 1.7988,
      "step": 436
    },
    {
      "epoch": 0.7008,
      "grad_norm": 2.126711845397949,
      "learning_rate": 7.732991314923891e-05,
      "loss": 1.7376,
      "step": 438
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.7960891723632812,
      "learning_rate": 7.710790081377502e-05,
      "loss": 1.7875,
      "step": 440
    },
    {
      "epoch": 0.7072,
      "grad_norm": 1.6610071659088135,
      "learning_rate": 7.688512893389964e-05,
      "loss": 1.7334,
      "step": 442
    },
    {
      "epoch": 0.7104,
      "grad_norm": 1.6998896598815918,
      "learning_rate": 7.666160375152496e-05,
      "loss": 1.886,
      "step": 444
    },
    {
      "epoch": 0.7136,
      "grad_norm": 1.6629440784454346,
      "learning_rate": 7.643733152967019e-05,
      "loss": 1.786,
      "step": 446
    },
    {
      "epoch": 0.7168,
      "grad_norm": 1.6910452842712402,
      "learning_rate": 7.621231855228604e-05,
      "loss": 2.0343,
      "step": 448
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.9952099323272705,
      "learning_rate": 7.598657112407865e-05,
      "loss": 1.7571,
      "step": 450
    },
    {
      "epoch": 0.7232,
      "grad_norm": 1.7345885038375854,
      "learning_rate": 7.576009557033304e-05,
      "loss": 2.0908,
      "step": 452
    },
    {
      "epoch": 0.7264,
      "grad_norm": 1.6344877481460571,
      "learning_rate": 7.553289823673568e-05,
      "loss": 1.8395,
      "step": 454
    },
    {
      "epoch": 0.7296,
      "grad_norm": 2.138115406036377,
      "learning_rate": 7.530498548919693e-05,
      "loss": 1.7072,
      "step": 456
    },
    {
      "epoch": 0.7328,
      "grad_norm": 1.9216474294662476,
      "learning_rate": 7.507636371367246e-05,
      "loss": 1.6516,
      "step": 458
    },
    {
      "epoch": 0.736,
      "grad_norm": 1.4932810068130493,
      "learning_rate": 7.484703931598445e-05,
      "loss": 1.9351,
      "step": 460
    },
    {
      "epoch": 0.7392,
      "grad_norm": 1.8183472156524658,
      "learning_rate": 7.461701872164204e-05,
      "loss": 1.8441,
      "step": 462
    },
    {
      "epoch": 0.7424,
      "grad_norm": 1.5970336198806763,
      "learning_rate": 7.438630837566133e-05,
      "loss": 1.8145,
      "step": 464
    },
    {
      "epoch": 0.7456,
      "grad_norm": 1.7351387739181519,
      "learning_rate": 7.415491474238475e-05,
      "loss": 1.8858,
      "step": 466
    },
    {
      "epoch": 0.7488,
      "grad_norm": 1.6989448070526123,
      "learning_rate": 7.39228443053e-05,
      "loss": 1.8566,
      "step": 468
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.8217098712921143,
      "learning_rate": 7.369010356685833e-05,
      "loss": 1.692,
      "step": 470
    },
    {
      "epoch": 0.7552,
      "grad_norm": 1.7833845615386963,
      "learning_rate": 7.345669904829237e-05,
      "loss": 1.8145,
      "step": 472
    },
    {
      "epoch": 0.7584,
      "grad_norm": 1.7113256454467773,
      "learning_rate": 7.32226372894334e-05,
      "loss": 1.907,
      "step": 474
    },
    {
      "epoch": 0.7616,
      "grad_norm": 1.66838800907135,
      "learning_rate": 7.298792484852808e-05,
      "loss": 1.8243,
      "step": 476
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.8057668209075928,
      "learning_rate": 7.27525683020548e-05,
      "loss": 1.6788,
      "step": 478
    },
    {
      "epoch": 0.768,
      "grad_norm": 1.7563303709030151,
      "learning_rate": 7.251657424453928e-05,
      "loss": 2.0148,
      "step": 480
    },
    {
      "epoch": 0.7712,
      "grad_norm": 1.75275719165802,
      "learning_rate": 7.227994928836988e-05,
      "loss": 1.7584,
      "step": 482
    },
    {
      "epoch": 0.7744,
      "grad_norm": 1.6364191770553589,
      "learning_rate": 7.204270006361228e-05,
      "loss": 1.9348,
      "step": 484
    },
    {
      "epoch": 0.7776,
      "grad_norm": 1.7930974960327148,
      "learning_rate": 7.180483321782374e-05,
      "loss": 1.9014,
      "step": 486
    },
    {
      "epoch": 0.7808,
      "grad_norm": 1.8914506435394287,
      "learning_rate": 7.156635541586682e-05,
      "loss": 1.7977,
      "step": 488
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.7024521827697754,
      "learning_rate": 7.132727333972265e-05,
      "loss": 1.6993,
      "step": 490
    },
    {
      "epoch": 0.7872,
      "grad_norm": 1.7870112657546997,
      "learning_rate": 7.108759368830371e-05,
      "loss": 1.6965,
      "step": 492
    },
    {
      "epoch": 0.7904,
      "grad_norm": 1.763691782951355,
      "learning_rate": 7.084732317726611e-05,
      "loss": 1.7948,
      "step": 494
    },
    {
      "epoch": 0.7936,
      "grad_norm": 1.683468222618103,
      "learning_rate": 7.060646853882145e-05,
      "loss": 1.9145,
      "step": 496
    },
    {
      "epoch": 0.7968,
      "grad_norm": 1.9888768196105957,
      "learning_rate": 7.036503652154812e-05,
      "loss": 1.8192,
      "step": 498
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5705928802490234,
      "learning_rate": 7.012303389020234e-05,
      "loss": 1.7831,
      "step": 500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 1.860660433769226,
      "learning_rate": 6.988046742552845e-05,
      "loss": 1.7904,
      "step": 502
    },
    {
      "epoch": 0.8064,
      "grad_norm": 1.8895405530929565,
      "learning_rate": 6.963734392406907e-05,
      "loss": 1.8645,
      "step": 504
    },
    {
      "epoch": 0.8096,
      "grad_norm": 1.74190354347229,
      "learning_rate": 6.93936701979746e-05,
      "loss": 1.8455,
      "step": 506
    },
    {
      "epoch": 0.8128,
      "grad_norm": 1.9230369329452515,
      "learning_rate": 6.914945307481228e-05,
      "loss": 1.8388,
      "step": 508
    },
    {
      "epoch": 0.816,
      "grad_norm": 1.5093566179275513,
      "learning_rate": 6.890469939737506e-05,
      "loss": 1.752,
      "step": 510
    },
    {
      "epoch": 0.8192,
      "grad_norm": 1.5916728973388672,
      "learning_rate": 6.865941602348966e-05,
      "loss": 1.7105,
      "step": 512
    },
    {
      "epoch": 0.8224,
      "grad_norm": 1.7378982305526733,
      "learning_rate": 6.841360982582463e-05,
      "loss": 1.9789,
      "step": 514
    },
    {
      "epoch": 0.8256,
      "grad_norm": 1.7520698308944702,
      "learning_rate": 6.816728769169757e-05,
      "loss": 1.7566,
      "step": 516
    },
    {
      "epoch": 0.8288,
      "grad_norm": 1.8129826784133911,
      "learning_rate": 6.792045652288234e-05,
      "loss": 1.8551,
      "step": 518
    },
    {
      "epoch": 0.832,
      "grad_norm": 1.9102818965911865,
      "learning_rate": 6.767312323541555e-05,
      "loss": 1.7726,
      "step": 520
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.5088154077529907,
      "learning_rate": 6.742529475940284e-05,
      "loss": 1.6381,
      "step": 522
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.7010055780410767,
      "learning_rate": 6.717697803882467e-05,
      "loss": 1.8741,
      "step": 524
    },
    {
      "epoch": 0.8416,
      "grad_norm": 1.6840184926986694,
      "learning_rate": 6.692818003134184e-05,
      "loss": 1.8617,
      "step": 526
    },
    {
      "epoch": 0.8448,
      "grad_norm": 1.7205629348754883,
      "learning_rate": 6.667890770810035e-05,
      "loss": 1.7349,
      "step": 528
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.520727515220642,
      "learning_rate": 6.64291680535363e-05,
      "loss": 1.749,
      "step": 530
    },
    {
      "epoch": 0.8512,
      "grad_norm": 1.5941743850708008,
      "learning_rate": 6.617896806518005e-05,
      "loss": 1.7076,
      "step": 532
    },
    {
      "epoch": 0.8544,
      "grad_norm": 1.7745941877365112,
      "learning_rate": 6.592831475346018e-05,
      "loss": 1.792,
      "step": 534
    },
    {
      "epoch": 0.8576,
      "grad_norm": 1.5072052478790283,
      "learning_rate": 6.56772151415071e-05,
      "loss": 1.6149,
      "step": 536
    },
    {
      "epoch": 0.8608,
      "grad_norm": 1.6202104091644287,
      "learning_rate": 6.542567626495619e-05,
      "loss": 1.756,
      "step": 538
    },
    {
      "epoch": 0.864,
      "grad_norm": 1.4974113702774048,
      "learning_rate": 6.517370517175081e-05,
      "loss": 1.7919,
      "step": 540
    },
    {
      "epoch": 0.8672,
      "grad_norm": 1.653824806213379,
      "learning_rate": 6.492130892194461e-05,
      "loss": 2.0103,
      "step": 542
    },
    {
      "epoch": 0.8704,
      "grad_norm": 1.683524489402771,
      "learning_rate": 6.466849458750394e-05,
      "loss": 2.0337,
      "step": 544
    },
    {
      "epoch": 0.8736,
      "grad_norm": 1.5982547998428345,
      "learning_rate": 6.441526925210949e-05,
      "loss": 1.8919,
      "step": 546
    },
    {
      "epoch": 0.8768,
      "grad_norm": 1.838497519493103,
      "learning_rate": 6.416164001095799e-05,
      "loss": 1.7648,
      "step": 548
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.524348258972168,
      "learning_rate": 6.390761397056328e-05,
      "loss": 1.6804,
      "step": 550
    },
    {
      "epoch": 0.8832,
      "grad_norm": 1.6498512029647827,
      "learning_rate": 6.365319824855727e-05,
      "loss": 1.6334,
      "step": 552
    },
    {
      "epoch": 0.8864,
      "grad_norm": 1.5689668655395508,
      "learning_rate": 6.339839997349045e-05,
      "loss": 1.9048,
      "step": 554
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.7050296068191528,
      "learning_rate": 6.314322628463219e-05,
      "loss": 1.6864,
      "step": 556
    },
    {
      "epoch": 0.8928,
      "grad_norm": 2.038351535797119,
      "learning_rate": 6.288768433177068e-05,
      "loss": 1.7531,
      "step": 558
    },
    {
      "epoch": 0.896,
      "grad_norm": 1.7489795684814453,
      "learning_rate": 6.26317812750126e-05,
      "loss": 1.8467,
      "step": 560
    },
    {
      "epoch": 0.8992,
      "grad_norm": 1.6068861484527588,
      "learning_rate": 6.237552428458256e-05,
      "loss": 1.8459,
      "step": 562
    },
    {
      "epoch": 0.9024,
      "grad_norm": 1.616613745689392,
      "learning_rate": 6.21189205406221e-05,
      "loss": 1.8173,
      "step": 564
    },
    {
      "epoch": 0.9056,
      "grad_norm": 1.6885602474212646,
      "learning_rate": 6.186197723298855e-05,
      "loss": 1.8358,
      "step": 566
    },
    {
      "epoch": 0.9088,
      "grad_norm": 1.688711404800415,
      "learning_rate": 6.160470156105362e-05,
      "loss": 1.6996,
      "step": 568
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.74298894405365,
      "learning_rate": 6.134710073350156e-05,
      "loss": 1.722,
      "step": 570
    },
    {
      "epoch": 0.9152,
      "grad_norm": 1.6249070167541504,
      "learning_rate": 6.108918196812734e-05,
      "loss": 1.7909,
      "step": 572
    },
    {
      "epoch": 0.9184,
      "grad_norm": 1.659416675567627,
      "learning_rate": 6.083095249163424e-05,
      "loss": 1.6625,
      "step": 574
    },
    {
      "epoch": 0.9216,
      "grad_norm": 1.6332143545150757,
      "learning_rate": 6.057241953943154e-05,
      "loss": 1.8297,
      "step": 576
    },
    {
      "epoch": 0.9248,
      "grad_norm": 1.6717133522033691,
      "learning_rate": 6.031359035543158e-05,
      "loss": 2.0601,
      "step": 578
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.8736896514892578,
      "learning_rate": 6.005447219184702e-05,
      "loss": 1.8117,
      "step": 580
    },
    {
      "epoch": 0.9312,
      "grad_norm": 1.6602182388305664,
      "learning_rate": 5.9795072308987485e-05,
      "loss": 1.7275,
      "step": 582
    },
    {
      "epoch": 0.9344,
      "grad_norm": 1.6776071786880493,
      "learning_rate": 5.9535397975056154e-05,
      "loss": 1.8988,
      "step": 584
    },
    {
      "epoch": 0.9376,
      "grad_norm": 1.588109016418457,
      "learning_rate": 5.927545646594617e-05,
      "loss": 1.7716,
      "step": 586
    },
    {
      "epoch": 0.9408,
      "grad_norm": 1.6331814527511597,
      "learning_rate": 5.901525506503668e-05,
      "loss": 1.8081,
      "step": 588
    },
    {
      "epoch": 0.944,
      "grad_norm": 1.7309777736663818,
      "learning_rate": 5.87548010629889e-05,
      "loss": 1.8243,
      "step": 590
    },
    {
      "epoch": 0.9472,
      "grad_norm": 1.6374008655548096,
      "learning_rate": 5.8494101757541676e-05,
      "loss": 1.7065,
      "step": 592
    },
    {
      "epoch": 0.9504,
      "grad_norm": 1.6291025876998901,
      "learning_rate": 5.8233164453307156e-05,
      "loss": 1.8013,
      "step": 594
    },
    {
      "epoch": 0.9536,
      "grad_norm": 1.7339948415756226,
      "learning_rate": 5.797199646156596e-05,
      "loss": 1.7998,
      "step": 596
    },
    {
      "epoch": 0.9568,
      "grad_norm": 1.581697940826416,
      "learning_rate": 5.7710605100062485e-05,
      "loss": 1.645,
      "step": 598
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.713205337524414,
      "learning_rate": 5.7448997692799764e-05,
      "loss": 1.9092,
      "step": 600
    },
    {
      "epoch": 0.9632,
      "grad_norm": 1.7023775577545166,
      "learning_rate": 5.718718156983428e-05,
      "loss": 1.7403,
      "step": 602
    },
    {
      "epoch": 0.9664,
      "grad_norm": 1.606632947921753,
      "learning_rate": 5.69251640670706e-05,
      "loss": 1.679,
      "step": 604
    },
    {
      "epoch": 0.9696,
      "grad_norm": 1.5328476428985596,
      "learning_rate": 5.6662952526055793e-05,
      "loss": 1.7899,
      "step": 606
    },
    {
      "epoch": 0.9728,
      "grad_norm": 1.5965962409973145,
      "learning_rate": 5.6400554293773744e-05,
      "loss": 1.7776,
      "step": 608
    },
    {
      "epoch": 0.976,
      "grad_norm": 1.5174623727798462,
      "learning_rate": 5.61379767224393e-05,
      "loss": 1.6602,
      "step": 610
    },
    {
      "epoch": 0.9792,
      "grad_norm": 1.6876877546310425,
      "learning_rate": 5.587522716929228e-05,
      "loss": 1.6656,
      "step": 612
    },
    {
      "epoch": 0.9824,
      "grad_norm": 1.5483810901641846,
      "learning_rate": 5.561231299639127e-05,
      "loss": 1.6531,
      "step": 614
    },
    {
      "epoch": 0.9856,
      "grad_norm": 1.464625597000122,
      "learning_rate": 5.534924157040745e-05,
      "loss": 1.8967,
      "step": 616
    },
    {
      "epoch": 0.9888,
      "grad_norm": 1.7587417364120483,
      "learning_rate": 5.508602026241807e-05,
      "loss": 1.6637,
      "step": 618
    },
    {
      "epoch": 0.992,
      "grad_norm": 1.5783720016479492,
      "learning_rate": 5.482265644769998e-05,
      "loss": 1.7628,
      "step": 620
    },
    {
      "epoch": 0.9952,
      "grad_norm": 1.602127194404602,
      "learning_rate": 5.4559157505522985e-05,
      "loss": 1.7458,
      "step": 622
    },
    {
      "epoch": 0.9984,
      "grad_norm": 1.8909751176834106,
      "learning_rate": 5.429553081894304e-05,
      "loss": 1.6952,
      "step": 624
    }
  ],
  "logging_steps": 2,
  "max_steps": 1250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 625,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0566821953614643e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}