{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 9.582864761352539,
      "learning_rate": 3.1746031746031746e-06,
      "loss": 4.5696,
      "step": 2
    },
    {
      "epoch": 0.0064,
      "grad_norm": 8.04224967956543,
      "learning_rate": 6.349206349206349e-06,
      "loss": 4.6008,
      "step": 4
    },
    {
      "epoch": 0.0096,
      "grad_norm": 5.386164665222168,
      "learning_rate": 9.523809523809523e-06,
      "loss": 4.4741,
      "step": 6
    },
    {
      "epoch": 0.0128,
      "grad_norm": 6.327683925628662,
      "learning_rate": 1.2698412698412699e-05,
      "loss": 4.3362,
      "step": 8
    },
    {
      "epoch": 0.016,
      "grad_norm": 4.981070518493652,
      "learning_rate": 1.5873015873015872e-05,
      "loss": 4.3088,
      "step": 10
    },
    {
      "epoch": 0.0192,
      "grad_norm": 5.140510559082031,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 4.1842,
      "step": 12
    },
    {
      "epoch": 0.0224,
      "grad_norm": 4.240128993988037,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 4.0743,
      "step": 14
    },
    {
      "epoch": 0.0256,
      "grad_norm": 4.0903096199035645,
      "learning_rate": 2.5396825396825397e-05,
      "loss": 4.0638,
      "step": 16
    },
    {
      "epoch": 0.0288,
      "grad_norm": 3.6583104133605957,
      "learning_rate": 2.857142857142857e-05,
      "loss": 3.8422,
      "step": 18
    },
    {
      "epoch": 0.032,
      "grad_norm": 4.4048027992248535,
      "learning_rate": 3.1746031746031745e-05,
      "loss": 3.8378,
      "step": 20
    },
    {
      "epoch": 0.0352,
      "grad_norm": 2.8964569568634033,
      "learning_rate": 3.492063492063492e-05,
      "loss": 3.8245,
      "step": 22
    },
    {
      "epoch": 0.0384,
      "grad_norm": 4.253384113311768,
      "learning_rate": 3.809523809523809e-05,
      "loss": 3.7656,
      "step": 24
    },
    {
      "epoch": 0.0416,
      "grad_norm": 2.827834367752075,
      "learning_rate": 4.126984126984127e-05,
      "loss": 3.4934,
      "step": 26
    },
    {
      "epoch": 0.0448,
      "grad_norm": 3.3544514179229736,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 3.5784,
      "step": 28
    },
    {
      "epoch": 0.048,
      "grad_norm": 3.2228829860687256,
      "learning_rate": 4.761904761904762e-05,
      "loss": 3.5281,
      "step": 30
    },
    {
      "epoch": 0.0512,
      "grad_norm": 2.926682472229004,
      "learning_rate": 5.0793650793650794e-05,
      "loss": 3.3777,
      "step": 32
    },
    {
      "epoch": 0.0544,
      "grad_norm": 3.137467861175537,
      "learning_rate": 5.396825396825397e-05,
      "loss": 3.333,
      "step": 34
    },
    {
      "epoch": 0.0576,
      "grad_norm": 3.6552820205688477,
      "learning_rate": 5.714285714285714e-05,
      "loss": 3.3148,
      "step": 36
    },
    {
      "epoch": 0.0608,
      "grad_norm": 2.9889962673187256,
      "learning_rate": 6.0317460317460316e-05,
      "loss": 3.2377,
      "step": 38
    },
    {
      "epoch": 0.064,
      "grad_norm": 2.781177520751953,
      "learning_rate": 6.349206349206349e-05,
      "loss": 3.2271,
      "step": 40
    },
    {
      "epoch": 0.0672,
      "grad_norm": 3.2072460651397705,
      "learning_rate": 6.666666666666667e-05,
      "loss": 3.1445,
      "step": 42
    },
    {
      "epoch": 0.0704,
      "grad_norm": 2.859445333480835,
      "learning_rate": 6.984126984126984e-05,
      "loss": 3.0457,
      "step": 44
    },
    {
      "epoch": 0.0736,
      "grad_norm": 4.333473205566406,
      "learning_rate": 7.301587301587302e-05,
      "loss": 3.1162,
      "step": 46
    },
    {
      "epoch": 0.0768,
      "grad_norm": 3.8664698600769043,
      "learning_rate": 7.619047619047618e-05,
      "loss": 2.9101,
      "step": 48
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.0567679405212402,
      "learning_rate": 7.936507936507937e-05,
      "loss": 2.9774,
      "step": 50
    },
    {
      "epoch": 0.0832,
      "grad_norm": 2.5626678466796875,
      "learning_rate": 8.253968253968255e-05,
      "loss": 3.0596,
      "step": 52
    },
    {
      "epoch": 0.0864,
      "grad_norm": 3.4151525497436523,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.9537,
      "step": 54
    },
    {
      "epoch": 0.0896,
      "grad_norm": 3.9550039768218994,
      "learning_rate": 8.888888888888889e-05,
      "loss": 2.8869,
      "step": 56
    },
    {
      "epoch": 0.0928,
      "grad_norm": 3.560119867324829,
      "learning_rate": 9.206349206349206e-05,
      "loss": 2.8154,
      "step": 58
    },
    {
      "epoch": 0.096,
      "grad_norm": 2.9746081829071045,
      "learning_rate": 9.523809523809524e-05,
      "loss": 2.841,
      "step": 60
    },
    {
      "epoch": 0.0992,
      "grad_norm": 2.8667147159576416,
      "learning_rate": 9.841269841269841e-05,
      "loss": 2.777,
      "step": 62
    },
    {
      "epoch": 0.1024,
      "grad_norm": 2.7489001750946045,
      "learning_rate": 9.99998248790669e-05,
      "loss": 2.7027,
      "step": 64
    },
    {
      "epoch": 0.1056,
      "grad_norm": 2.734243392944336,
      "learning_rate": 9.999842391896222e-05,
      "loss": 2.7284,
      "step": 66
    },
    {
      "epoch": 0.1088,
      "grad_norm": 2.6667640209198,
      "learning_rate": 9.999562203800676e-05,
      "loss": 2.6692,
      "step": 68
    },
    {
      "epoch": 0.112,
      "grad_norm": 3.080273389816284,
      "learning_rate": 9.999141931470729e-05,
      "loss": 2.6861,
      "step": 70
    },
    {
      "epoch": 0.1152,
      "grad_norm": 3.324246406555176,
      "learning_rate": 9.998581586682116e-05,
      "loss": 2.7227,
      "step": 72
    },
    {
      "epoch": 0.1184,
      "grad_norm": 2.7069027423858643,
      "learning_rate": 9.997881185135307e-05,
      "loss": 2.5695,
      "step": 74
    },
    {
      "epoch": 0.1216,
      "grad_norm": 2.781511068344116,
      "learning_rate": 9.997040746455062e-05,
      "loss": 2.5662,
      "step": 76
    },
    {
      "epoch": 0.1248,
      "grad_norm": 2.453489303588867,
      "learning_rate": 9.996060294189887e-05,
      "loss": 2.7143,
      "step": 78
    },
    {
      "epoch": 0.128,
      "grad_norm": 2.8613131046295166,
      "learning_rate": 9.994939855811362e-05,
      "loss": 2.6042,
      "step": 80
    },
    {
      "epoch": 0.1312,
      "grad_norm": 2.4560399055480957,
      "learning_rate": 9.993679462713395e-05,
      "loss": 2.6329,
      "step": 82
    },
    {
      "epoch": 0.1344,
      "grad_norm": 3.461649179458618,
      "learning_rate": 9.992279150211314e-05,
      "loss": 2.5056,
      "step": 84
    },
    {
      "epoch": 0.1376,
      "grad_norm": 2.901437997817993,
      "learning_rate": 9.990738957540896e-05,
      "loss": 2.5626,
      "step": 86
    },
    {
      "epoch": 0.1408,
      "grad_norm": 2.5126583576202393,
      "learning_rate": 9.989058927857263e-05,
      "loss": 2.4684,
      "step": 88
    },
    {
      "epoch": 0.144,
      "grad_norm": 3.184246778488159,
      "learning_rate": 9.987239108233668e-05,
      "loss": 2.4756,
      "step": 90
    },
    {
      "epoch": 0.1472,
      "grad_norm": 3.1928048133850098,
      "learning_rate": 9.985279549660185e-05,
      "loss": 2.4961,
      "step": 92
    },
    {
      "epoch": 0.1504,
      "grad_norm": 2.3420567512512207,
      "learning_rate": 9.983180307042274e-05,
      "loss": 2.3984,
      "step": 94
    },
    {
      "epoch": 0.1536,
      "grad_norm": 3.066572427749634,
      "learning_rate": 9.980941439199246e-05,
      "loss": 2.5094,
      "step": 96
    },
    {
      "epoch": 0.1568,
      "grad_norm": 2.7460713386535645,
      "learning_rate": 9.97856300886261e-05,
      "loss": 2.4763,
      "step": 98
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.51401424407959,
      "learning_rate": 9.976045082674319e-05,
      "loss": 2.3846,
      "step": 100
    },
    {
      "epoch": 0.1632,
      "grad_norm": 2.2191078662872314,
      "learning_rate": 9.973387731184902e-05,
      "loss": 2.36,
      "step": 102
    },
    {
      "epoch": 0.1664,
      "grad_norm": 2.5087525844573975,
      "learning_rate": 9.97059102885149e-05,
      "loss": 2.4398,
      "step": 104
    },
    {
      "epoch": 0.1696,
      "grad_norm": 2.396519184112549,
      "learning_rate": 9.967655054035727e-05,
      "loss": 2.459,
      "step": 106
    },
    {
      "epoch": 0.1728,
      "grad_norm": 2.6870367527008057,
      "learning_rate": 9.964579889001569e-05,
      "loss": 2.3118,
      "step": 108
    },
    {
      "epoch": 0.176,
      "grad_norm": 2.317108392715454,
      "learning_rate": 9.961365619912989e-05,
      "loss": 2.3219,
      "step": 110
    },
    {
      "epoch": 0.1792,
      "grad_norm": 2.2667131423950195,
      "learning_rate": 9.95801233683156e-05,
      "loss": 2.4226,
      "step": 112
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.9936238527297974,
      "learning_rate": 9.954520133713924e-05,
      "loss": 2.4241,
      "step": 114
    },
    {
      "epoch": 0.1856,
      "grad_norm": 2.6429572105407715,
      "learning_rate": 9.950889108409172e-05,
      "loss": 2.3513,
      "step": 116
    },
    {
      "epoch": 0.1888,
      "grad_norm": 2.2465708255767822,
      "learning_rate": 9.947119362656092e-05,
      "loss": 2.2819,
      "step": 118
    },
    {
      "epoch": 0.192,
      "grad_norm": 3.0118119716644287,
      "learning_rate": 9.94321100208032e-05,
      "loss": 2.4245,
      "step": 120
    },
    {
      "epoch": 0.1952,
      "grad_norm": 2.177351236343384,
      "learning_rate": 9.939164136191384e-05,
      "loss": 2.5335,
      "step": 122
    },
    {
      "epoch": 0.1984,
      "grad_norm": 2.1183319091796875,
      "learning_rate": 9.934978878379636e-05,
      "loss": 2.3807,
      "step": 124
    },
    {
      "epoch": 0.2016,
      "grad_norm": 2.0215203762054443,
      "learning_rate": 9.930655345913071e-05,
      "loss": 2.228,
      "step": 126
    },
    {
      "epoch": 0.2048,
      "grad_norm": 2.589059591293335,
      "learning_rate": 9.926193659934043e-05,
      "loss": 2.3367,
      "step": 128
    },
    {
      "epoch": 0.208,
      "grad_norm": 2.5615551471710205,
      "learning_rate": 9.921593945455869e-05,
      "loss": 2.2462,
      "step": 130
    },
    {
      "epoch": 0.2112,
      "grad_norm": 2.5255088806152344,
      "learning_rate": 9.916856331359335e-05,
      "loss": 2.2562,
      "step": 132
    },
    {
      "epoch": 0.2144,
      "grad_norm": 2.201739549636841,
      "learning_rate": 9.911980950389067e-05,
      "loss": 2.3823,
      "step": 134
    },
    {
      "epoch": 0.2176,
      "grad_norm": 2.3065388202667236,
      "learning_rate": 9.906967939149831e-05,
      "loss": 2.2231,
      "step": 136
    },
    {
      "epoch": 0.2208,
      "grad_norm": 2.7823500633239746,
      "learning_rate": 9.901817438102695e-05,
      "loss": 2.2203,
      "step": 138
    },
    {
      "epoch": 0.224,
      "grad_norm": 2.521483898162842,
      "learning_rate": 9.896529591561093e-05,
      "loss": 2.4067,
      "step": 140
    },
    {
      "epoch": 0.2272,
      "grad_norm": 2.114596128463745,
      "learning_rate": 9.891104547686782e-05,
      "loss": 2.2335,
      "step": 142
    },
    {
      "epoch": 0.2304,
      "grad_norm": 2.151817560195923,
      "learning_rate": 9.8855424584857e-05,
      "loss": 2.2483,
      "step": 144
    },
    {
      "epoch": 0.2336,
      "grad_norm": 2.050886392593384,
      "learning_rate": 9.879843479803691e-05,
      "loss": 2.2825,
      "step": 146
    },
    {
      "epoch": 0.2368,
      "grad_norm": 2.388864278793335,
      "learning_rate": 9.874007771322151e-05,
      "loss": 2.3412,
      "step": 148
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.22013258934021,
      "learning_rate": 9.868035496553546e-05,
      "loss": 2.1599,
      "step": 150
    },
    {
      "epoch": 0.2432,
      "grad_norm": 2.0260703563690186,
      "learning_rate": 9.86192682283684e-05,
      "loss": 2.1551,
      "step": 152
    },
    {
      "epoch": 0.2464,
      "grad_norm": 2.3314321041107178,
      "learning_rate": 9.855681921332793e-05,
      "loss": 2.2231,
      "step": 154
    },
    {
      "epoch": 0.2496,
      "grad_norm": 2.6237032413482666,
      "learning_rate": 9.849300967019175e-05,
      "loss": 2.2275,
      "step": 156
    },
    {
      "epoch": 0.2528,
      "grad_norm": 2.053867816925049,
      "learning_rate": 9.84278413868586e-05,
      "loss": 2.2417,
      "step": 158
    },
    {
      "epoch": 0.256,
      "grad_norm": 2.3856852054595947,
      "learning_rate": 9.836131618929819e-05,
      "loss": 2.2179,
      "step": 160
    },
    {
      "epoch": 0.2592,
      "grad_norm": 1.919459342956543,
      "learning_rate": 9.82934359415e-05,
      "loss": 2.2195,
      "step": 162
    },
    {
      "epoch": 0.2624,
      "grad_norm": 2.281283378601074,
      "learning_rate": 9.822420254542108e-05,
      "loss": 2.211,
      "step": 164
    },
    {
      "epoch": 0.2656,
      "grad_norm": 1.97679603099823,
      "learning_rate": 9.815361794093272e-05,
      "loss": 2.1523,
      "step": 166
    },
    {
      "epoch": 0.2688,
      "grad_norm": 2.190410852432251,
      "learning_rate": 9.808168410576617e-05,
      "loss": 2.179,
      "step": 168
    },
    {
      "epoch": 0.272,
      "grad_norm": 2.4369115829467773,
      "learning_rate": 9.800840305545715e-05,
      "loss": 2.2628,
      "step": 170
    },
    {
      "epoch": 0.2752,
      "grad_norm": 1.9605038166046143,
      "learning_rate": 9.793377684328939e-05,
      "loss": 2.3508,
      "step": 172
    },
    {
      "epoch": 0.2784,
      "grad_norm": 2.3224270343780518,
      "learning_rate": 9.785780756023714e-05,
      "loss": 2.1655,
      "step": 174
    },
    {
      "epoch": 0.2816,
      "grad_norm": 2.045945167541504,
      "learning_rate": 9.778049733490655e-05,
      "loss": 2.2663,
      "step": 176
    },
    {
      "epoch": 0.2848,
      "grad_norm": 2.2260525226593018,
      "learning_rate": 9.770184833347606e-05,
      "loss": 2.1685,
      "step": 178
    },
    {
      "epoch": 0.288,
      "grad_norm": 2.0497450828552246,
      "learning_rate": 9.762186275963563e-05,
      "loss": 2.1108,
      "step": 180
    },
    {
      "epoch": 0.2912,
      "grad_norm": 2.2114014625549316,
      "learning_rate": 9.754054285452506e-05,
      "loss": 2.1031,
      "step": 182
    },
    {
      "epoch": 0.2944,
      "grad_norm": 2.2554216384887695,
      "learning_rate": 9.745789089667121e-05,
      "loss": 2.2677,
      "step": 184
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.8373414278030396,
      "learning_rate": 9.737390920192408e-05,
      "loss": 2.2118,
      "step": 186
    },
    {
      "epoch": 0.3008,
      "grad_norm": 2.319699764251709,
      "learning_rate": 9.7288600123392e-05,
      "loss": 2.1008,
      "step": 188
    },
    {
      "epoch": 0.304,
      "grad_norm": 2.0968804359436035,
      "learning_rate": 9.720196605137565e-05,
      "loss": 2.1803,
      "step": 190
    },
    {
      "epoch": 0.3072,
      "grad_norm": 2.0866589546203613,
      "learning_rate": 9.71140094133011e-05,
      "loss": 2.2203,
      "step": 192
    },
    {
      "epoch": 0.3104,
      "grad_norm": 2.236036777496338,
      "learning_rate": 9.702473267365182e-05,
      "loss": 2.2041,
      "step": 194
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.8478028774261475,
      "learning_rate": 9.693413833389956e-05,
      "loss": 2.0953,
      "step": 196
    },
    {
      "epoch": 0.3168,
      "grad_norm": 2.449432849884033,
      "learning_rate": 9.684222893243431e-05,
      "loss": 2.2525,
      "step": 198
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.9800256490707397,
      "learning_rate": 9.674900704449324e-05,
      "loss": 2.0962,
      "step": 200
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.9219201803207397,
      "learning_rate": 9.665447528208836e-05,
      "loss": 2.0759,
      "step": 202
    },
    {
      "epoch": 0.3264,
      "grad_norm": 2.214970350265503,
      "learning_rate": 9.655863629393351e-05,
      "loss": 2.0526,
      "step": 204
    },
    {
      "epoch": 0.3296,
      "grad_norm": 2.141845226287842,
      "learning_rate": 9.64614927653701e-05,
      "loss": 2.0032,
      "step": 206
    },
    {
      "epoch": 0.3328,
      "grad_norm": 2.117124319076538,
      "learning_rate": 9.636304741829181e-05,
      "loss": 2.1147,
      "step": 208
    },
    {
      "epoch": 0.336,
      "grad_norm": 2.2099063396453857,
      "learning_rate": 9.626330301106837e-05,
      "loss": 2.0526,
      "step": 210
    },
    {
      "epoch": 0.3392,
      "grad_norm": 2.162337064743042,
      "learning_rate": 9.616226233846828e-05,
      "loss": 2.0378,
      "step": 212
    },
    {
      "epoch": 0.3424,
      "grad_norm": 1.9509609937667847,
      "learning_rate": 9.605992823158046e-05,
      "loss": 2.2311,
      "step": 214
    },
    {
      "epoch": 0.3456,
      "grad_norm": 2.105393171310425,
      "learning_rate": 9.595630355773501e-05,
      "loss": 2.0036,
      "step": 216
    },
    {
      "epoch": 0.3488,
      "grad_norm": 2.210383176803589,
      "learning_rate": 9.585139122042274e-05,
      "loss": 2.1254,
      "step": 218
    },
    {
      "epoch": 0.352,
      "grad_norm": 2.3816072940826416,
      "learning_rate": 9.574519415921396e-05,
      "loss": 2.078,
      "step": 220
    },
    {
      "epoch": 0.3552,
      "grad_norm": 2.019214153289795,
      "learning_rate": 9.5637715349676e-05,
      "loss": 2.1598,
      "step": 222
    },
    {
      "epoch": 0.3584,
      "grad_norm": 2.09932804107666,
      "learning_rate": 9.552895780328987e-05,
      "loss": 2.0171,
      "step": 224
    },
    {
      "epoch": 0.3616,
      "grad_norm": 1.85040283203125,
      "learning_rate": 9.541892456736595e-05,
      "loss": 2.2289,
      "step": 226
    },
    {
      "epoch": 0.3648,
      "grad_norm": 2.067749261856079,
      "learning_rate": 9.530761872495849e-05,
      "loss": 2.064,
      "step": 228
    },
    {
      "epoch": 0.368,
      "grad_norm": 2.581606388092041,
      "learning_rate": 9.519504339477932e-05,
      "loss": 2.1191,
      "step": 230
    },
    {
      "epoch": 0.3712,
      "grad_norm": 2.451923131942749,
      "learning_rate": 9.508120173111039e-05,
      "loss": 1.9523,
      "step": 232
    },
    {
      "epoch": 0.3744,
      "grad_norm": 1.9244287014007568,
      "learning_rate": 9.496609692371548e-05,
      "loss": 2.0501,
      "step": 234
    },
    {
      "epoch": 0.3776,
      "grad_norm": 2.272866725921631,
      "learning_rate": 9.484973219775074e-05,
      "loss": 1.9952,
      "step": 236
    },
    {
      "epoch": 0.3808,
      "grad_norm": 2.396249294281006,
      "learning_rate": 9.473211081367436e-05,
      "loss": 2.0369,
      "step": 238
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.967800259590149,
      "learning_rate": 9.46132360671552e-05,
      "loss": 2.0272,
      "step": 240
    },
    {
      "epoch": 0.3872,
      "grad_norm": 2.252021312713623,
      "learning_rate": 9.449311128898049e-05,
      "loss": 1.9491,
      "step": 242
    },
    {
      "epoch": 0.3904,
      "grad_norm": 2.425724506378174,
      "learning_rate": 9.437173984496246e-05,
      "loss": 2.076,
      "step": 244
    },
    {
      "epoch": 0.3936,
      "grad_norm": 2.0590147972106934,
      "learning_rate": 9.424912513584401e-05,
      "loss": 2.1357,
      "step": 246
    },
    {
      "epoch": 0.3968,
      "grad_norm": 2.0368545055389404,
      "learning_rate": 9.412527059720352e-05,
      "loss": 2.0673,
      "step": 248
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.0116140842437744,
      "learning_rate": 9.400017969935848e-05,
      "loss": 1.9939,
      "step": 250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 1.872152328491211,
      "learning_rate": 9.387385594726829e-05,
      "loss": 1.9777,
      "step": 252
    },
    {
      "epoch": 0.4064,
      "grad_norm": 2.1459996700286865,
      "learning_rate": 9.374630288043614e-05,
      "loss": 2.1791,
      "step": 254
    },
    {
      "epoch": 0.4096,
      "grad_norm": 2.0004026889801025,
      "learning_rate": 9.361752407280965e-05,
      "loss": 2.069,
      "step": 256
    },
    {
      "epoch": 0.4128,
      "grad_norm": 1.9468940496444702,
      "learning_rate": 9.348752313268093e-05,
      "loss": 1.9954,
      "step": 258
    },
    {
      "epoch": 0.416,
      "grad_norm": 1.9674925804138184,
      "learning_rate": 9.335630370258533e-05,
      "loss": 2.0868,
      "step": 260
    },
    {
      "epoch": 0.4192,
      "grad_norm": 1.8100794553756714,
      "learning_rate": 9.322386945919946e-05,
      "loss": 1.8544,
      "step": 262
    },
    {
      "epoch": 0.4224,
      "grad_norm": 1.7924424409866333,
      "learning_rate": 9.309022411323816e-05,
      "loss": 2.1337,
      "step": 264
    },
    {
      "epoch": 0.4256,
      "grad_norm": 2.407942771911621,
      "learning_rate": 9.295537140935049e-05,
      "loss": 2.0707,
      "step": 266
    },
    {
      "epoch": 0.4288,
      "grad_norm": 3.1549575328826904,
      "learning_rate": 9.281931512601485e-05,
      "loss": 2.0362,
      "step": 268
    },
    {
      "epoch": 0.432,
      "grad_norm": 2.299870491027832,
      "learning_rate": 9.26820590754331e-05,
      "loss": 1.9683,
      "step": 270
    },
    {
      "epoch": 0.4352,
      "grad_norm": 2.0636281967163086,
      "learning_rate": 9.254360710342371e-05,
      "loss": 1.9565,
      "step": 272
    },
    {
      "epoch": 0.4384,
      "grad_norm": 2.0548760890960693,
      "learning_rate": 9.240396308931407e-05,
      "loss": 1.9686,
      "step": 274
    },
    {
      "epoch": 0.4416,
      "grad_norm": 2.189924955368042,
      "learning_rate": 9.226313094583173e-05,
      "loss": 2.0544,
      "step": 276
    },
    {
      "epoch": 0.4448,
      "grad_norm": 1.7277733087539673,
      "learning_rate": 9.212111461899479e-05,
      "loss": 2.1443,
      "step": 278
    },
    {
      "epoch": 0.448,
      "grad_norm": 2.1883883476257324,
      "learning_rate": 9.197791808800135e-05,
      "loss": 1.9916,
      "step": 280
    },
    {
      "epoch": 0.4512,
      "grad_norm": 2.1574831008911133,
      "learning_rate": 9.183354536511803e-05,
      "loss": 2.0649,
      "step": 282
    },
    {
      "epoch": 0.4544,
      "grad_norm": 1.9589487314224243,
      "learning_rate": 9.168800049556747e-05,
      "loss": 1.9466,
      "step": 284
    },
    {
      "epoch": 0.4576,
      "grad_norm": 2.365525007247925,
      "learning_rate": 9.154128755741509e-05,
      "loss": 1.9666,
      "step": 286
    },
    {
      "epoch": 0.4608,
      "grad_norm": 1.946046233177185,
      "learning_rate": 9.139341066145472e-05,
      "loss": 1.9854,
      "step": 288
    },
    {
      "epoch": 0.464,
      "grad_norm": 2.5430924892425537,
      "learning_rate": 9.124437395109353e-05,
      "loss": 1.9625,
      "step": 290
    },
    {
      "epoch": 0.4672,
      "grad_norm": 2.154215097427368,
      "learning_rate": 9.109418160223585e-05,
      "loss": 1.9279,
      "step": 292
    },
    {
      "epoch": 0.4704,
      "grad_norm": 2.0101840496063232,
      "learning_rate": 9.094283782316619e-05,
      "loss": 1.8481,
      "step": 294
    },
    {
      "epoch": 0.4736,
      "grad_norm": 2.403233051300049,
      "learning_rate": 9.079034685443133e-05,
      "loss": 1.9778,
      "step": 296
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.8693294525146484,
      "learning_rate": 9.063671296872149e-05,
      "loss": 1.8872,
      "step": 298
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8005255460739136,
      "learning_rate": 9.048194047075069e-05,
      "loss": 1.9892,
      "step": 300
    },
    {
      "epoch": 0.4832,
      "grad_norm": 1.8485512733459473,
      "learning_rate": 9.032603369713596e-05,
      "loss": 1.7879,
      "step": 302
    },
    {
      "epoch": 0.4864,
      "grad_norm": 2.034468650817871,
      "learning_rate": 9.016899701627604e-05,
      "loss": 1.9269,
      "step": 304
    },
    {
      "epoch": 0.4896,
      "grad_norm": 1.8233674764633179,
      "learning_rate": 9.00108348282288e-05,
      "loss": 2.0422,
      "step": 306
    },
    {
      "epoch": 0.4928,
      "grad_norm": 2.109724998474121,
      "learning_rate": 8.985155156458811e-05,
      "loss": 1.8479,
      "step": 308
    },
    {
      "epoch": 0.496,
      "grad_norm": 2.107543468475342,
      "learning_rate": 8.969115168835954e-05,
      "loss": 1.9216,
      "step": 310
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.83890700340271,
      "learning_rate": 8.952963969383538e-05,
      "loss": 1.8112,
      "step": 312
    },
    {
      "epoch": 0.5024,
      "grad_norm": 1.8299955129623413,
      "learning_rate": 8.93670201064687e-05,
      "loss": 2.0874,
      "step": 314
    },
    {
      "epoch": 0.5056,
      "grad_norm": 1.7529444694519043,
      "learning_rate": 8.920329748274649e-05,
      "loss": 1.9491,
      "step": 316
    },
    {
      "epoch": 0.5088,
      "grad_norm": 2.1236982345581055,
      "learning_rate": 8.903847641006218e-05,
      "loss": 1.9849,
      "step": 318
    },
    {
      "epoch": 0.512,
      "grad_norm": 1.9501066207885742,
      "learning_rate": 8.887256150658684e-05,
      "loss": 1.8132,
      "step": 320
    },
    {
      "epoch": 0.5152,
      "grad_norm": 2.057251214981079,
      "learning_rate": 8.870555742113998e-05,
      "loss": 1.8913,
      "step": 322
    },
    {
      "epoch": 0.5184,
      "grad_norm": 1.5886375904083252,
      "learning_rate": 8.85374688330592e-05,
      "loss": 1.9817,
      "step": 324
    },
    {
      "epoch": 0.5216,
      "grad_norm": 2.109905242919922,
      "learning_rate": 8.836830045206911e-05,
      "loss": 1.908,
      "step": 326
    },
    {
      "epoch": 0.5248,
      "grad_norm": 1.9238070249557495,
      "learning_rate": 8.81980570181494e-05,
      "loss": 2.1003,
      "step": 328
    },
    {
      "epoch": 0.528,
      "grad_norm": 1.7516288757324219,
      "learning_rate": 8.802674330140192e-05,
      "loss": 1.8825,
      "step": 330
    },
    {
      "epoch": 0.5312,
      "grad_norm": 1.7113988399505615,
      "learning_rate": 8.785436410191714e-05,
      "loss": 1.8977,
      "step": 332
    },
    {
      "epoch": 0.5344,
      "grad_norm": 1.919970154762268,
      "learning_rate": 8.76809242496396e-05,
      "loss": 2.0056,
      "step": 334
    },
    {
      "epoch": 0.5376,
      "grad_norm": 1.8441989421844482,
      "learning_rate": 8.750642860423262e-05,
      "loss": 1.9522,
      "step": 336
    },
    {
      "epoch": 0.5408,
      "grad_norm": 2.240172863006592,
      "learning_rate": 8.733088205494205e-05,
      "loss": 2.0764,
      "step": 338
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.7431479692459106,
      "learning_rate": 8.715428952045936e-05,
      "loss": 2.0673,
      "step": 340
    },
    {
      "epoch": 0.5472,
      "grad_norm": 1.7653355598449707,
      "learning_rate": 8.697665594878382e-05,
      "loss": 1.8442,
      "step": 342
    },
    {
      "epoch": 0.5504,
      "grad_norm": 2.4551286697387695,
      "learning_rate": 8.679798631708375e-05,
      "loss": 2.0315,
      "step": 344
    },
    {
      "epoch": 0.5536,
      "grad_norm": 1.743158221244812,
      "learning_rate": 8.661828563155727e-05,
      "loss": 2.0135,
      "step": 346
    },
    {
      "epoch": 0.5568,
      "grad_norm": 1.7268255949020386,
      "learning_rate": 8.643755892729179e-05,
      "loss": 2.0524,
      "step": 348
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.032898426055908,
      "learning_rate": 8.625581126812312e-05,
      "loss": 1.9173,
      "step": 350
    },
    {
      "epoch": 0.5632,
      "grad_norm": 2.1082491874694824,
      "learning_rate": 8.607304774649349e-05,
      "loss": 2.0753,
      "step": 352
    },
    {
      "epoch": 0.5664,
      "grad_norm": 1.8048080205917358,
      "learning_rate": 8.588927348330887e-05,
      "loss": 1.8533,
      "step": 354
    },
    {
      "epoch": 0.5696,
      "grad_norm": 1.8337701559066772,
      "learning_rate": 8.57044936277955e-05,
      "loss": 1.9847,
      "step": 356
    },
    {
      "epoch": 0.5728,
      "grad_norm": 1.898160457611084,
      "learning_rate": 8.551871335735565e-05,
      "loss": 1.8207,
      "step": 358
    },
    {
      "epoch": 0.576,
      "grad_norm": 1.672937273979187,
      "learning_rate": 8.533193787742251e-05,
      "loss": 1.8368,
      "step": 360
    },
    {
      "epoch": 0.5792,
      "grad_norm": 1.9867445230484009,
      "learning_rate": 8.51441724213143e-05,
      "loss": 1.9168,
      "step": 362
    },
    {
      "epoch": 0.5824,
      "grad_norm": 1.8571895360946655,
      "learning_rate": 8.495542225008771e-05,
      "loss": 1.8886,
      "step": 364
    },
    {
      "epoch": 0.5856,
      "grad_norm": 1.8292155265808105,
      "learning_rate": 8.476569265239046e-05,
      "loss": 1.8527,
      "step": 366
    },
    {
      "epoch": 0.5888,
      "grad_norm": 1.817756175994873,
      "learning_rate": 8.457498894431311e-05,
      "loss": 1.843,
      "step": 368
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.9946959018707275,
      "learning_rate": 8.438331646924013e-05,
      "loss": 1.8689,
      "step": 370
    },
    {
      "epoch": 0.5952,
      "grad_norm": 1.6774402856826782,
      "learning_rate": 8.419068059770011e-05,
      "loss": 1.9314,
      "step": 372
    },
    {
      "epoch": 0.5984,
      "grad_norm": 1.7767459154129028,
      "learning_rate": 8.399708672721539e-05,
      "loss": 1.8765,
      "step": 374
    },
    {
      "epoch": 0.6016,
      "grad_norm": 2.1188535690307617,
      "learning_rate": 8.380254028215076e-05,
      "loss": 1.9367,
      "step": 376
    },
    {
      "epoch": 0.6048,
      "grad_norm": 1.9462740421295166,
      "learning_rate": 8.360704671356145e-05,
      "loss": 1.8991,
      "step": 378
    },
    {
      "epoch": 0.608,
      "grad_norm": 1.850533127784729,
      "learning_rate": 8.341061149904045e-05,
      "loss": 2.0175,
      "step": 380
    },
    {
      "epoch": 0.6112,
      "grad_norm": 1.957470178604126,
      "learning_rate": 8.321324014256504e-05,
      "loss": 1.9628,
      "step": 382
    },
    {
      "epoch": 0.6144,
      "grad_norm": 1.9237133264541626,
      "learning_rate": 8.30149381743425e-05,
      "loss": 1.9568,
      "step": 384
    },
    {
      "epoch": 0.6176,
      "grad_norm": 2.103835105895996,
      "learning_rate": 8.28157111506552e-05,
      "loss": 1.932,
      "step": 386
    },
    {
      "epoch": 0.6208,
      "grad_norm": 1.685325264930725,
      "learning_rate": 8.261556465370493e-05,
      "loss": 2.0071,
      "step": 388
    },
    {
      "epoch": 0.624,
      "grad_norm": 1.9392411708831787,
      "learning_rate": 8.24145042914565e-05,
      "loss": 1.8254,
      "step": 390
    },
    {
      "epoch": 0.6272,
      "grad_norm": 1.759429931640625,
      "learning_rate": 8.221253569748055e-05,
      "loss": 1.7943,
      "step": 392
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.9010182619094849,
      "learning_rate": 8.200966453079575e-05,
      "loss": 1.9736,
      "step": 394
    },
    {
      "epoch": 0.6336,
      "grad_norm": 1.859298586845398,
      "learning_rate": 8.180589647571023e-05,
      "loss": 2.0887,
      "step": 396
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.9668726921081543,
      "learning_rate": 8.16012372416623e-05,
      "loss": 1.8353,
      "step": 398
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.8924115896224976,
      "learning_rate": 8.13956925630605e-05,
      "loss": 1.8227,
      "step": 400
    },
    {
      "epoch": 0.6432,
      "grad_norm": 2.260485887527466,
      "learning_rate": 8.118926819912287e-05,
      "loss": 1.9317,
      "step": 402
    },
    {
      "epoch": 0.6464,
      "grad_norm": 1.7944382429122925,
      "learning_rate": 8.098196993371565e-05,
      "loss": 1.92,
      "step": 404
    },
    {
      "epoch": 0.6496,
      "grad_norm": 1.6864358186721802,
      "learning_rate": 8.077380357519115e-05,
      "loss": 1.8648,
      "step": 406
    },
    {
      "epoch": 0.6528,
      "grad_norm": 1.7029087543487549,
      "learning_rate": 8.056477495622511e-05,
      "loss": 2.1089,
      "step": 408
    },
    {
      "epoch": 0.656,
      "grad_norm": 2.3869247436523438,
      "learning_rate": 8.035488993365312e-05,
      "loss": 1.8285,
      "step": 410
    },
    {
      "epoch": 0.6592,
      "grad_norm": 1.8005602359771729,
      "learning_rate": 8.014415438830667e-05,
      "loss": 1.9826,
      "step": 412
    },
    {
      "epoch": 0.6624,
      "grad_norm": 2.076176881790161,
      "learning_rate": 7.993257422484826e-05,
      "loss": 1.8091,
      "step": 414
    },
    {
      "epoch": 0.6656,
      "grad_norm": 1.6236470937728882,
      "learning_rate": 7.972015537160602e-05,
      "loss": 1.9846,
      "step": 416
    },
    {
      "epoch": 0.6688,
      "grad_norm": 1.7574564218521118,
      "learning_rate": 7.950690378040758e-05,
      "loss": 2.0633,
      "step": 418
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.8413279056549072,
      "learning_rate": 7.929282542641325e-05,
      "loss": 1.7691,
      "step": 420
    },
    {
      "epoch": 0.6752,
      "grad_norm": 1.750153660774231,
      "learning_rate": 7.907792630794876e-05,
      "loss": 1.724,
      "step": 422
    },
    {
      "epoch": 0.6784,
      "grad_norm": 1.9424185752868652,
      "learning_rate": 7.886221244633703e-05,
      "loss": 1.8961,
      "step": 424
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.8231415748596191,
      "learning_rate": 7.864568988572947e-05,
      "loss": 1.9455,
      "step": 426
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.9073402881622314,
      "learning_rate": 7.842836469293673e-05,
      "loss": 1.7979,
      "step": 428
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.8199973106384277,
      "learning_rate": 7.821024295725865e-05,
      "loss": 1.8819,
      "step": 430
    },
    {
      "epoch": 0.6912,
      "grad_norm": 2.1633074283599854,
      "learning_rate": 7.79913307903136e-05,
      "loss": 1.8375,
      "step": 432
    },
    {
      "epoch": 0.6944,
      "grad_norm": 2.62290620803833,
      "learning_rate": 7.777163432586734e-05,
      "loss": 1.8191,
      "step": 434
    },
    {
      "epoch": 0.6976,
      "grad_norm": 1.9004080295562744,
      "learning_rate": 7.755115971966104e-05,
      "loss": 1.8747,
      "step": 436
    },
    {
      "epoch": 0.7008,
      "grad_norm": 2.1421079635620117,
      "learning_rate": 7.732991314923891e-05,
      "loss": 1.804,
      "step": 438
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.7792892456054688,
      "learning_rate": 7.710790081377502e-05,
      "loss": 1.8574,
      "step": 440
    },
    {
      "epoch": 0.7072,
      "grad_norm": 1.7586500644683838,
      "learning_rate": 7.688512893389964e-05,
      "loss": 1.7981,
      "step": 442
    },
    {
      "epoch": 0.7104,
      "grad_norm": 1.8390204906463623,
      "learning_rate": 7.666160375152496e-05,
      "loss": 1.9504,
      "step": 444
    },
    {
      "epoch": 0.7136,
      "grad_norm": 1.8994488716125488,
      "learning_rate": 7.643733152967019e-05,
      "loss": 1.8502,
      "step": 446
    },
    {
      "epoch": 0.7168,
      "grad_norm": 1.869563102722168,
      "learning_rate": 7.621231855228604e-05,
      "loss": 2.0965,
      "step": 448
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.0518429279327393,
      "learning_rate": 7.598657112407865e-05,
      "loss": 1.821,
      "step": 450
    },
    {
      "epoch": 0.7232,
      "grad_norm": 1.7781238555908203,
      "learning_rate": 7.576009557033304e-05,
      "loss": 2.1547,
      "step": 452
    },
    {
      "epoch": 0.7264,
      "grad_norm": 1.897527813911438,
      "learning_rate": 7.553289823673568e-05,
      "loss": 1.905,
      "step": 454
    },
    {
      "epoch": 0.7296,
      "grad_norm": 1.7613269090652466,
      "learning_rate": 7.530498548919693e-05,
      "loss": 1.7703,
      "step": 456
    },
    {
      "epoch": 0.7328,
      "grad_norm": 2.0733559131622314,
      "learning_rate": 7.507636371367246e-05,
      "loss": 1.7406,
      "step": 458
    },
    {
      "epoch": 0.736,
      "grad_norm": 1.6685259342193604,
      "learning_rate": 7.484703931598445e-05,
      "loss": 2.008,
      "step": 460
    },
    {
      "epoch": 0.7392,
      "grad_norm": 1.7001707553863525,
      "learning_rate": 7.461701872164204e-05,
      "loss": 1.8958,
      "step": 462
    },
    {
      "epoch": 0.7424,
      "grad_norm": 1.6040148735046387,
      "learning_rate": 7.438630837566133e-05,
      "loss": 1.8792,
      "step": 464
    },
    {
      "epoch": 0.7456,
      "grad_norm": 1.8001751899719238,
      "learning_rate": 7.415491474238475e-05,
      "loss": 1.9369,
      "step": 466
    },
    {
      "epoch": 0.7488,
      "grad_norm": 1.7213135957717896,
      "learning_rate": 7.39228443053e-05,
      "loss": 1.9091,
      "step": 468
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.8896820545196533,
      "learning_rate": 7.369010356685833e-05,
      "loss": 1.7765,
      "step": 470
    },
    {
      "epoch": 0.7552,
      "grad_norm": 1.9900362491607666,
      "learning_rate": 7.345669904829237e-05,
      "loss": 1.8875,
      "step": 472
    },
    {
      "epoch": 0.7584,
      "grad_norm": 1.7322590351104736,
      "learning_rate": 7.32226372894334e-05,
      "loss": 1.9698,
      "step": 474
    },
    {
      "epoch": 0.7616,
      "grad_norm": 1.7330385446548462,
      "learning_rate": 7.298792484852808e-05,
      "loss": 1.8868,
      "step": 476
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.7039341926574707,
      "learning_rate": 7.27525683020548e-05,
      "loss": 1.7435,
      "step": 478
    },
    {
      "epoch": 0.768,
      "grad_norm": 2.421879529953003,
      "learning_rate": 7.251657424453928e-05,
      "loss": 2.065,
      "step": 480
    },
    {
      "epoch": 0.7712,
      "grad_norm": 1.754511833190918,
      "learning_rate": 7.227994928836988e-05,
      "loss": 1.8281,
      "step": 482
    },
    {
      "epoch": 0.7744,
      "grad_norm": 1.6815738677978516,
      "learning_rate": 7.204270006361228e-05,
      "loss": 1.9892,
      "step": 484
    },
    {
      "epoch": 0.7776,
      "grad_norm": 1.944504976272583,
      "learning_rate": 7.180483321782374e-05,
      "loss": 1.936,
      "step": 486
    },
    {
      "epoch": 0.7808,
      "grad_norm": 1.8897745609283447,
      "learning_rate": 7.156635541586682e-05,
      "loss": 1.8541,
      "step": 488
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.7814854383468628,
      "learning_rate": 7.132727333972265e-05,
      "loss": 1.7648,
      "step": 490
    },
    {
      "epoch": 0.7872,
      "grad_norm": 2.232374906539917,
      "learning_rate": 7.108759368830371e-05,
      "loss": 1.7671,
      "step": 492
    },
    {
      "epoch": 0.7904,
      "grad_norm": 1.7056297063827515,
      "learning_rate": 7.084732317726611e-05,
      "loss": 1.837,
      "step": 494
    },
    {
      "epoch": 0.7936,
      "grad_norm": 1.651081919670105,
      "learning_rate": 7.060646853882145e-05,
      "loss": 1.9683,
      "step": 496
    },
    {
      "epoch": 0.7968,
      "grad_norm": 1.8357526063919067,
      "learning_rate": 7.036503652154812e-05,
      "loss": 1.8938,
      "step": 498
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5519391298294067,
      "learning_rate": 7.012303389020234e-05,
      "loss": 1.8466,
      "step": 500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 1.894633412361145,
      "learning_rate": 6.988046742552845e-05,
      "loss": 1.8514,
      "step": 502
    },
    {
      "epoch": 0.8064,
      "grad_norm": 1.754236102104187,
      "learning_rate": 6.963734392406907e-05,
      "loss": 1.935,
      "step": 504
    },
    {
      "epoch": 0.8096,
      "grad_norm": 1.702377438545227,
      "learning_rate": 6.93936701979746e-05,
      "loss": 1.8913,
      "step": 506
    },
    {
      "epoch": 0.8128,
      "grad_norm": 2.0201923847198486,
      "learning_rate": 6.914945307481228e-05,
      "loss": 1.893,
      "step": 508
    },
    {
      "epoch": 0.816,
      "grad_norm": 1.64589524269104,
      "learning_rate": 6.890469939737506e-05,
      "loss": 1.8145,
      "step": 510
    },
    {
      "epoch": 0.8192,
      "grad_norm": 1.6376056671142578,
      "learning_rate": 6.865941602348966e-05,
      "loss": 1.7757,
      "step": 512
    },
    {
      "epoch": 0.8224,
      "grad_norm": 1.746893286705017,
      "learning_rate": 6.841360982582463e-05,
      "loss": 2.0359,
      "step": 514
    },
    {
      "epoch": 0.8256,
      "grad_norm": 1.8241147994995117,
      "learning_rate": 6.816728769169757e-05,
      "loss": 1.8158,
      "step": 516
    },
    {
      "epoch": 0.8288,
      "grad_norm": 1.8170113563537598,
      "learning_rate": 6.792045652288234e-05,
      "loss": 1.9136,
      "step": 518
    },
    {
      "epoch": 0.832,
      "grad_norm": 1.9374558925628662,
      "learning_rate": 6.767312323541555e-05,
      "loss": 1.8295,
      "step": 520
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.7372517585754395,
      "learning_rate": 6.742529475940284e-05,
      "loss": 1.7011,
      "step": 522
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.785727858543396,
      "learning_rate": 6.717697803882467e-05,
      "loss": 1.9221,
      "step": 524
    },
    {
      "epoch": 0.8416,
      "grad_norm": 1.8377923965454102,
      "learning_rate": 6.692818003134184e-05,
      "loss": 1.9293,
      "step": 526
    },
    {
      "epoch": 0.8448,
      "grad_norm": 1.8372453451156616,
      "learning_rate": 6.667890770810035e-05,
      "loss": 1.7977,
      "step": 528
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.6308479309082031,
      "learning_rate": 6.64291680535363e-05,
      "loss": 1.8136,
      "step": 530
    },
    {
      "epoch": 0.8512,
      "grad_norm": 1.6363590955734253,
      "learning_rate": 6.617896806518005e-05,
      "loss": 1.766,
      "step": 532
    },
    {
      "epoch": 0.8544,
      "grad_norm": 1.8008592128753662,
      "learning_rate": 6.592831475346018e-05,
      "loss": 1.8626,
      "step": 534
    },
    {
      "epoch": 0.8576,
      "grad_norm": 1.5823259353637695,
      "learning_rate": 6.56772151415071e-05,
      "loss": 1.6812,
      "step": 536
    },
    {
      "epoch": 0.8608,
      "grad_norm": 1.6105380058288574,
      "learning_rate": 6.542567626495619e-05,
      "loss": 1.8237,
      "step": 538
    },
    {
      "epoch": 0.864,
      "grad_norm": 1.5150834321975708,
      "learning_rate": 6.517370517175081e-05,
      "loss": 1.8367,
      "step": 540
    },
    {
      "epoch": 0.8672,
      "grad_norm": 1.796002984046936,
      "learning_rate": 6.492130892194461e-05,
      "loss": 2.0477,
      "step": 542
    },
    {
      "epoch": 0.8704,
      "grad_norm": 1.647668480873108,
      "learning_rate": 6.466849458750394e-05,
      "loss": 2.0732,
      "step": 544
    },
    {
      "epoch": 0.8736,
      "grad_norm": 1.5862977504730225,
      "learning_rate": 6.441526925210949e-05,
      "loss": 1.9539,
      "step": 546
    },
    {
      "epoch": 0.8768,
      "grad_norm": 2.1107430458068848,
      "learning_rate": 6.416164001095799e-05,
      "loss": 1.8033,
      "step": 548
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.886512041091919,
      "learning_rate": 6.390761397056328e-05,
      "loss": 1.7506,
      "step": 550
    },
    {
      "epoch": 0.8832,
      "grad_norm": 1.6387286186218262,
      "learning_rate": 6.365319824855727e-05,
      "loss": 1.7166,
      "step": 552
    },
    {
      "epoch": 0.8864,
      "grad_norm": 1.720944881439209,
      "learning_rate": 6.339839997349045e-05,
      "loss": 1.955,
      "step": 554
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.8100652694702148,
      "learning_rate": 6.314322628463219e-05,
      "loss": 1.7356,
      "step": 556
    },
    {
      "epoch": 0.8928,
      "grad_norm": 1.7271335124969482,
      "learning_rate": 6.288768433177068e-05,
      "loss": 1.8124,
      "step": 558
    },
    {
      "epoch": 0.896,
      "grad_norm": 1.8463622331619263,
      "learning_rate": 6.26317812750126e-05,
      "loss": 1.8924,
      "step": 560
    },
    {
      "epoch": 0.8992,
      "grad_norm": 1.612862467765808,
      "learning_rate": 6.237552428458256e-05,
      "loss": 1.8984,
      "step": 562
    },
    {
      "epoch": 0.9024,
      "grad_norm": 1.618046760559082,
      "learning_rate": 6.21189205406221e-05,
      "loss": 1.8639,
      "step": 564
    },
    {
      "epoch": 0.9056,
      "grad_norm": 1.8677492141723633,
      "learning_rate": 6.186197723298855e-05,
      "loss": 1.9007,
      "step": 566
    },
    {
      "epoch": 0.9088,
      "grad_norm": 1.5078552961349487,
      "learning_rate": 6.160470156105362e-05,
      "loss": 1.7474,
      "step": 568
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.8413493633270264,
      "learning_rate": 6.134710073350156e-05,
      "loss": 1.7687,
      "step": 570
    },
    {
      "epoch": 0.9152,
      "grad_norm": 1.6803293228149414,
      "learning_rate": 6.108918196812734e-05,
      "loss": 1.8477,
      "step": 572
    },
    {
      "epoch": 0.9184,
      "grad_norm": 1.6897441148757935,
      "learning_rate": 6.083095249163424e-05,
      "loss": 1.7068,
      "step": 574
    },
    {
      "epoch": 0.9216,
      "grad_norm": 1.6542898416519165,
      "learning_rate": 6.057241953943154e-05,
      "loss": 1.8733,
      "step": 576
    },
    {
      "epoch": 0.9248,
      "grad_norm": 1.6907879114151,
      "learning_rate": 6.031359035543158e-05,
      "loss": 2.1063,
      "step": 578
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.8007756471633911,
      "learning_rate": 6.005447219184702e-05,
      "loss": 1.8521,
      "step": 580
    },
    {
      "epoch": 0.9312,
      "grad_norm": 1.7871769666671753,
      "learning_rate": 5.9795072308987485e-05,
      "loss": 1.7819,
      "step": 582
    },
    {
      "epoch": 0.9344,
      "grad_norm": 1.678968071937561,
      "learning_rate": 5.9535397975056154e-05,
      "loss": 1.9387,
      "step": 584
    },
    {
      "epoch": 0.9376,
      "grad_norm": 1.6597601175308228,
      "learning_rate": 5.927545646594617e-05,
      "loss": 1.825,
      "step": 586
    },
    {
      "epoch": 0.9408,
      "grad_norm": 1.8375909328460693,
      "learning_rate": 5.901525506503668e-05,
      "loss": 1.8603,
      "step": 588
    },
    {
      "epoch": 0.944,
      "grad_norm": 1.7722874879837036,
      "learning_rate": 5.87548010629889e-05,
      "loss": 1.88,
      "step": 590
    },
    {
      "epoch": 0.9472,
      "grad_norm": 1.6075280904769897,
      "learning_rate": 5.8494101757541676e-05,
      "loss": 1.7607,
      "step": 592
    },
    {
      "epoch": 0.9504,
      "grad_norm": 1.7452820539474487,
      "learning_rate": 5.8233164453307156e-05,
      "loss": 1.8466,
      "step": 594
    },
    {
      "epoch": 0.9536,
      "grad_norm": 1.6466399431228638,
      "learning_rate": 5.797199646156596e-05,
      "loss": 1.8487,
      "step": 596
    },
    {
      "epoch": 0.9568,
      "grad_norm": 1.5579028129577637,
      "learning_rate": 5.7710605100062485e-05,
      "loss": 1.6981,
      "step": 598
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.6615177392959595,
      "learning_rate": 5.7448997692799764e-05,
      "loss": 1.9517,
      "step": 600
    },
    {
      "epoch": 0.9632,
      "grad_norm": 1.6678462028503418,
      "learning_rate": 5.718718156983428e-05,
      "loss": 1.7882,
      "step": 602
    },
    {
      "epoch": 0.9664,
      "grad_norm": 1.4700486660003662,
      "learning_rate": 5.69251640670706e-05,
      "loss": 1.7402,
      "step": 604
    },
    {
      "epoch": 0.9696,
      "grad_norm": 1.5739563703536987,
      "learning_rate": 5.6662952526055793e-05,
      "loss": 1.843,
      "step": 606
    },
    {
      "epoch": 0.9728,
      "grad_norm": 1.6076737642288208,
      "learning_rate": 5.6400554293773744e-05,
      "loss": 1.8319,
      "step": 608
    },
    {
      "epoch": 0.976,
      "grad_norm": 1.5903165340423584,
      "learning_rate": 5.61379767224393e-05,
      "loss": 1.7108,
      "step": 610
    },
    {
      "epoch": 0.9792,
      "grad_norm": 1.6222702264785767,
      "learning_rate": 5.587522716929228e-05,
      "loss": 1.7096,
      "step": 612
    },
    {
      "epoch": 0.9824,
      "grad_norm": 1.4682416915893555,
      "learning_rate": 5.561231299639127e-05,
      "loss": 1.7172,
      "step": 614
    },
    {
      "epoch": 0.9856,
      "grad_norm": 1.542601466178894,
      "learning_rate": 5.534924157040745e-05,
      "loss": 1.9472,
      "step": 616
    },
    {
      "epoch": 0.9888,
      "grad_norm": 1.862457513809204,
      "learning_rate": 5.508602026241807e-05,
      "loss": 1.7044,
      "step": 618
    },
    {
      "epoch": 0.992,
      "grad_norm": 1.7092889547348022,
      "learning_rate": 5.482265644769998e-05,
      "loss": 1.8067,
      "step": 620
    },
    {
      "epoch": 0.9952,
      "grad_norm": 1.7380741834640503,
      "learning_rate": 5.4559157505522985e-05,
      "loss": 1.7857,
      "step": 622
    },
    {
      "epoch": 0.9984,
      "grad_norm": 2.111462116241455,
      "learning_rate": 5.429553081894304e-05,
      "loss": 1.7447,
      "step": 624
    },
    {
      "epoch": 1.0016,
      "grad_norm": 1.511717677116394,
      "learning_rate": 5.4031783774595455e-05,
      "loss": 1.6521,
      "step": 626
    },
    {
      "epoch": 1.0048,
      "grad_norm": 1.4466677904129028,
      "learning_rate": 5.3767923762487824e-05,
      "loss": 1.5505,
      "step": 628
    },
    {
      "epoch": 1.008,
      "grad_norm": 1.4657745361328125,
      "learning_rate": 5.3503958175793055e-05,
      "loss": 1.5376,
      "step": 630
    },
    {
      "epoch": 1.0112,
      "grad_norm": 1.5471076965332031,
      "learning_rate": 5.323989441064216e-05,
      "loss": 1.6454,
      "step": 632
    },
    {
      "epoch": 1.0144,
      "grad_norm": 1.7591346502304077,
      "learning_rate": 5.2975739865917074e-05,
      "loss": 1.6523,
      "step": 634
    },
    {
      "epoch": 1.0176,
      "grad_norm": 1.6937857866287231,
      "learning_rate": 5.271150194304326e-05,
      "loss": 1.6116,
      "step": 636
    },
    {
      "epoch": 1.0208,
      "grad_norm": 1.6300266981124878,
      "learning_rate": 5.244718804578246e-05,
      "loss": 1.5488,
      "step": 638
    },
    {
      "epoch": 1.024,
      "grad_norm": 1.612653136253357,
      "learning_rate": 5.218280558002506e-05,
      "loss": 1.543,
      "step": 640
    },
    {
      "epoch": 1.0272,
      "grad_norm": 1.6032813787460327,
      "learning_rate": 5.191836195358278e-05,
      "loss": 1.5711,
      "step": 642
    },
    {
      "epoch": 1.0304,
      "grad_norm": 1.6160064935684204,
      "learning_rate": 5.165386457598099e-05,
      "loss": 1.5716,
      "step": 644
    },
    {
      "epoch": 1.0336,
      "grad_norm": 1.5153204202651978,
      "learning_rate": 5.13893208582511e-05,
      "loss": 1.5672,
      "step": 646
    },
    {
      "epoch": 1.0368,
      "grad_norm": 1.5654356479644775,
      "learning_rate": 5.1124738212722966e-05,
      "loss": 1.4573,
      "step": 648
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.4795836210250854,
      "learning_rate": 5.086012405281717e-05,
      "loss": 1.428,
      "step": 650
    },
    {
      "epoch": 1.0432,
      "grad_norm": 1.5888456106185913,
      "learning_rate": 5.0595485792837305e-05,
      "loss": 1.6168,
      "step": 652
    },
    {
      "epoch": 1.0464,
      "grad_norm": 1.6040419340133667,
      "learning_rate": 5.033083084776222e-05,
      "loss": 1.5238,
      "step": 654
    },
    {
      "epoch": 1.0496,
      "grad_norm": 1.7781437635421753,
      "learning_rate": 5.0066166633038305e-05,
      "loss": 1.5852,
      "step": 656
    },
    {
      "epoch": 1.0528,
      "grad_norm": 1.7350730895996094,
      "learning_rate": 4.980150056437163e-05,
      "loss": 1.5622,
      "step": 658
    },
    {
      "epoch": 1.056,
      "grad_norm": 1.4517589807510376,
      "learning_rate": 4.9536840057520224e-05,
      "loss": 1.5135,
      "step": 660
    },
    {
      "epoch": 1.0592,
      "grad_norm": 1.7444967031478882,
      "learning_rate": 4.927219252808631e-05,
      "loss": 1.5136,
      "step": 662
    },
    {
      "epoch": 1.0624,
      "grad_norm": 1.6197596788406372,
      "learning_rate": 4.900756539130846e-05,
      "loss": 1.506,
      "step": 664
    },
    {
      "epoch": 1.0656,
      "grad_norm": 1.5582484006881714,
      "learning_rate": 4.874296606185387e-05,
      "loss": 1.6418,
      "step": 666
    },
    {
      "epoch": 1.0688,
      "grad_norm": 1.6707253456115723,
      "learning_rate": 4.847840195361058e-05,
      "loss": 1.6327,
      "step": 668
    },
    {
      "epoch": 1.072,
      "grad_norm": 2.6252167224884033,
      "learning_rate": 4.821388047947979e-05,
      "loss": 1.5304,
      "step": 670
    },
    {
      "epoch": 1.0752,
      "grad_norm": 1.5231362581253052,
      "learning_rate": 4.7949409051168085e-05,
      "loss": 1.5779,
      "step": 672
    },
    {
      "epoch": 1.0784,
      "grad_norm": 1.5844740867614746,
      "learning_rate": 4.768499507897981e-05,
      "loss": 1.5455,
      "step": 674
    },
    {
      "epoch": 1.0816,
      "grad_norm": 1.4232511520385742,
      "learning_rate": 4.742064597160948e-05,
      "loss": 1.5498,
      "step": 676
    },
    {
      "epoch": 1.0848,
      "grad_norm": 1.492144227027893,
      "learning_rate": 4.715636913593404e-05,
      "loss": 1.6364,
      "step": 678
    },
    {
      "epoch": 1.088,
      "grad_norm": 1.543325662612915,
      "learning_rate": 4.689217197680554e-05,
      "loss": 1.5252,
      "step": 680
    },
    {
      "epoch": 1.0912,
      "grad_norm": 1.6142204999923706,
      "learning_rate": 4.6628061896843474e-05,
      "loss": 1.5794,
      "step": 682
    },
    {
      "epoch": 1.0944,
      "grad_norm": 1.5524206161499023,
      "learning_rate": 4.6364046296227484e-05,
      "loss": 1.5516,
      "step": 684
    },
    {
      "epoch": 1.0976,
      "grad_norm": 1.7153761386871338,
      "learning_rate": 4.6100132572489915e-05,
      "loss": 1.7075,
      "step": 686
    },
    {
      "epoch": 1.1008,
      "grad_norm": 1.5478098392486572,
      "learning_rate": 4.5836328120308674e-05,
      "loss": 1.6028,
      "step": 688
    },
    {
      "epoch": 1.104,
      "grad_norm": 1.7050093412399292,
      "learning_rate": 4.5572640331299875e-05,
      "loss": 1.5922,
      "step": 690
    },
    {
      "epoch": 1.1072,
      "grad_norm": 1.399055004119873,
      "learning_rate": 4.530907659381086e-05,
      "loss": 1.4775,
      "step": 692
    },
    {
      "epoch": 1.1104,
      "grad_norm": 1.51021146774292,
      "learning_rate": 4.504564429271311e-05,
      "loss": 1.498,
      "step": 694
    },
    {
      "epoch": 1.1136,
      "grad_norm": 1.3979012966156006,
      "learning_rate": 4.478235080919536e-05,
      "loss": 1.5185,
      "step": 696
    },
    {
      "epoch": 1.1168,
      "grad_norm": 1.7461631298065186,
      "learning_rate": 4.451920352055678e-05,
      "loss": 1.554,
      "step": 698
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.5486574172973633,
      "learning_rate": 4.425620980000026e-05,
      "loss": 1.6174,
      "step": 700
    },
    {
      "epoch": 1.1232,
      "grad_norm": 1.5407499074935913,
      "learning_rate": 4.39933770164258e-05,
      "loss": 1.5983,
      "step": 702
    },
    {
      "epoch": 1.1264,
      "grad_norm": 1.5229569673538208,
      "learning_rate": 4.373071253422408e-05,
      "loss": 1.6455,
      "step": 704
    },
    {
      "epoch": 1.1296,
      "grad_norm": 1.599682092666626,
      "learning_rate": 4.346822371307009e-05,
      "loss": 1.6372,
      "step": 706
    },
    {
      "epoch": 1.1328,
      "grad_norm": 1.7229453325271606,
      "learning_rate": 4.320591790771691e-05,
      "loss": 1.5438,
      "step": 708
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 1.4003872871398926,
      "learning_rate": 4.294380246778966e-05,
      "loss": 1.5351,
      "step": 710
    },
    {
      "epoch": 1.1392,
      "grad_norm": 1.8065009117126465,
      "learning_rate": 4.2681884737579524e-05,
      "loss": 1.5148,
      "step": 712
    },
    {
      "epoch": 1.1424,
      "grad_norm": 1.3923487663269043,
      "learning_rate": 4.242017205583805e-05,
      "loss": 1.5805,
      "step": 714
    },
    {
      "epoch": 1.1456,
      "grad_norm": 1.542261004447937,
      "learning_rate": 4.215867175557142e-05,
      "loss": 1.515,
      "step": 716
    },
    {
      "epoch": 1.1488,
      "grad_norm": 1.5241355895996094,
      "learning_rate": 4.189739116383506e-05,
      "loss": 1.5733,
      "step": 718
    },
    {
      "epoch": 1.152,
      "grad_norm": 1.5812512636184692,
      "learning_rate": 4.163633760152834e-05,
      "loss": 1.4757,
      "step": 720
    },
    {
      "epoch": 1.1552,
      "grad_norm": 1.6010973453521729,
      "learning_rate": 4.137551838318936e-05,
      "loss": 1.5595,
      "step": 722
    },
    {
      "epoch": 1.1584,
      "grad_norm": 1.429641604423523,
      "learning_rate": 4.1114940816790135e-05,
      "loss": 1.5003,
      "step": 724
    },
    {
      "epoch": 1.1616,
      "grad_norm": 1.619784951210022,
      "learning_rate": 4.08546122035317e-05,
      "loss": 1.5059,
      "step": 726
    },
    {
      "epoch": 1.1648,
      "grad_norm": 1.6332809925079346,
      "learning_rate": 4.059453983763967e-05,
      "loss": 1.5229,
      "step": 728
    },
    {
      "epoch": 1.168,
      "grad_norm": 1.6049484014511108,
      "learning_rate": 4.03347310061597e-05,
      "loss": 1.4133,
      "step": 730
    },
    {
      "epoch": 1.1712,
      "grad_norm": 1.547785997390747,
      "learning_rate": 4.007519298875347e-05,
      "loss": 1.5042,
      "step": 732
    },
    {
      "epoch": 1.1743999999999999,
      "grad_norm": 1.6845433712005615,
      "learning_rate": 3.98159330574946e-05,
      "loss": 1.507,
      "step": 734
    },
    {
      "epoch": 1.1776,
      "grad_norm": 1.6059722900390625,
      "learning_rate": 3.955695847666494e-05,
      "loss": 1.5038,
      "step": 736
    },
    {
      "epoch": 1.1808,
      "grad_norm": 1.7839707136154175,
      "learning_rate": 3.929827650255104e-05,
      "loss": 1.6538,
      "step": 738
    },
    {
      "epoch": 1.184,
      "grad_norm": 1.4794530868530273,
      "learning_rate": 3.903989438324077e-05,
      "loss": 1.4546,
      "step": 740
    },
    {
      "epoch": 1.1872,
      "grad_norm": 1.6607145071029663,
      "learning_rate": 3.878181935842033e-05,
      "loss": 1.5601,
      "step": 742
    },
    {
      "epoch": 1.1904,
      "grad_norm": 1.596228003501892,
      "learning_rate": 3.85240586591713e-05,
      "loss": 1.5132,
      "step": 744
    },
    {
      "epoch": 1.1936,
      "grad_norm": 1.5321195125579834,
      "learning_rate": 3.8266619507768126e-05,
      "loss": 1.498,
      "step": 746
    },
    {
      "epoch": 1.1968,
      "grad_norm": 1.4525630474090576,
      "learning_rate": 3.800950911747565e-05,
      "loss": 1.462,
      "step": 748
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.4488592147827148,
      "learning_rate": 3.775273469234712e-05,
      "loss": 1.627,
      "step": 750
    },
    {
      "epoch": 1.2032,
      "grad_norm": 1.6878618001937866,
      "learning_rate": 3.749630342702221e-05,
      "loss": 1.5679,
      "step": 752
    },
    {
      "epoch": 1.2064,
      "grad_norm": 1.440504789352417,
      "learning_rate": 3.724022250652557e-05,
      "loss": 1.5279,
      "step": 754
    },
    {
      "epoch": 1.2096,
      "grad_norm": 1.5879031419754028,
      "learning_rate": 3.698449910606536e-05,
      "loss": 1.5393,
      "step": 756
    },
    {
      "epoch": 1.2128,
      "grad_norm": 1.3794442415237427,
      "learning_rate": 3.672914039083233e-05,
      "loss": 1.5278,
      "step": 758
    },
    {
      "epoch": 1.216,
      "grad_norm": 1.4098923206329346,
      "learning_rate": 3.6474153515799e-05,
      "loss": 1.5186,
      "step": 760
    },
    {
      "epoch": 1.2192,
      "grad_norm": 1.6515835523605347,
      "learning_rate": 3.6219545625519145e-05,
      "loss": 1.5804,
      "step": 762
    },
    {
      "epoch": 1.2224,
      "grad_norm": 1.8193235397338867,
      "learning_rate": 3.596532385392772e-05,
      "loss": 1.699,
      "step": 764
    },
    {
      "epoch": 1.2256,
      "grad_norm": 1.4131522178649902,
      "learning_rate": 3.5711495324140845e-05,
      "loss": 1.5142,
      "step": 766
    },
    {
      "epoch": 1.2288000000000001,
      "grad_norm": 1.5099951028823853,
      "learning_rate": 3.54580671482563e-05,
      "loss": 1.5414,
      "step": 768
    },
    {
      "epoch": 1.232,
      "grad_norm": 1.5891356468200684,
      "learning_rate": 3.520504642715424e-05,
      "loss": 1.6367,
      "step": 770
    },
    {
      "epoch": 1.2352,
      "grad_norm": 1.4419711828231812,
      "learning_rate": 3.495244025029822e-05,
      "loss": 1.548,
      "step": 772
    },
    {
      "epoch": 1.2384,
      "grad_norm": 1.5342023372650146,
      "learning_rate": 3.470025569553653e-05,
      "loss": 1.5482,
      "step": 774
    },
    {
      "epoch": 1.2416,
      "grad_norm": 1.5894769430160522,
      "learning_rate": 3.444849982890393e-05,
      "loss": 1.4836,
      "step": 776
    },
    {
      "epoch": 1.2448,
      "grad_norm": 1.583613395690918,
      "learning_rate": 3.4197179704423656e-05,
      "loss": 1.5122,
      "step": 778
    },
    {
      "epoch": 1.248,
      "grad_norm": 1.524994969367981,
      "learning_rate": 3.39463023639097e-05,
      "loss": 1.5183,
      "step": 780
    },
    {
      "epoch": 1.2511999999999999,
      "grad_norm": 1.6434054374694824,
      "learning_rate": 3.36958748367696e-05,
      "loss": 1.5171,
      "step": 782
    },
    {
      "epoch": 1.2544,
      "grad_norm": 1.715895175933838,
      "learning_rate": 3.3445904139807405e-05,
      "loss": 1.5995,
      "step": 784
    },
    {
      "epoch": 1.2576,
      "grad_norm": 1.4911595582962036,
      "learning_rate": 3.319639727702716e-05,
      "loss": 1.5388,
      "step": 786
    },
    {
      "epoch": 1.2608,
      "grad_norm": 1.6039537191390991,
      "learning_rate": 3.2947361239436525e-05,
      "loss": 1.4187,
      "step": 788
    },
    {
      "epoch": 1.264,
      "grad_norm": 2.159885883331299,
      "learning_rate": 3.2698803004851026e-05,
      "loss": 1.5726,
      "step": 790
    },
    {
      "epoch": 1.2671999999999999,
      "grad_norm": 1.4904849529266357,
      "learning_rate": 3.245072953769844e-05,
      "loss": 1.5846,
      "step": 792
    },
    {
      "epoch": 1.2704,
      "grad_norm": 1.4704859256744385,
      "learning_rate": 3.2203147788823764e-05,
      "loss": 1.4765,
      "step": 794
    },
    {
      "epoch": 1.2736,
      "grad_norm": 1.6482725143432617,
      "learning_rate": 3.1956064695294305e-05,
      "loss": 1.5398,
      "step": 796
    },
    {
      "epoch": 1.2768,
      "grad_norm": 1.734695553779602,
      "learning_rate": 3.170948718020546e-05,
      "loss": 1.505,
      "step": 798
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.5135869979858398,
      "learning_rate": 3.1463422152486674e-05,
      "loss": 1.5202,
      "step": 800
    },
    {
      "epoch": 1.2832,
      "grad_norm": 1.6760534048080444,
      "learning_rate": 3.121787650670781e-05,
      "loss": 1.5748,
      "step": 802
    },
    {
      "epoch": 1.2864,
      "grad_norm": 1.482445478439331,
      "learning_rate": 3.097285712288605e-05,
      "loss": 1.4619,
      "step": 804
    },
    {
      "epoch": 1.2896,
      "grad_norm": 1.4474997520446777,
      "learning_rate": 3.072837086629305e-05,
      "loss": 1.5023,
      "step": 806
    },
    {
      "epoch": 1.2928,
      "grad_norm": 1.3969132900238037,
      "learning_rate": 3.0484424587262655e-05,
      "loss": 1.4479,
      "step": 808
    },
    {
      "epoch": 1.296,
      "grad_norm": 1.4808533191680908,
      "learning_rate": 3.024102512099889e-05,
      "loss": 1.5258,
      "step": 810
    },
    {
      "epoch": 1.2992,
      "grad_norm": 1.7867789268493652,
      "learning_rate": 2.9998179287384485e-05,
      "loss": 1.6572,
      "step": 812
    },
    {
      "epoch": 1.3024,
      "grad_norm": 1.6131802797317505,
      "learning_rate": 2.975589389078975e-05,
      "loss": 1.5263,
      "step": 814
    },
    {
      "epoch": 1.3056,
      "grad_norm": 1.4337478876113892,
      "learning_rate": 2.9514175719881966e-05,
      "loss": 1.5051,
      "step": 816
    },
    {
      "epoch": 1.3088,
      "grad_norm": 1.460150957107544,
      "learning_rate": 2.9273031547435114e-05,
      "loss": 1.4839,
      "step": 818
    },
    {
      "epoch": 1.312,
      "grad_norm": 1.5884913206100464,
      "learning_rate": 2.9032468130140168e-05,
      "loss": 1.5642,
      "step": 820
    },
    {
      "epoch": 1.3152,
      "grad_norm": 1.5978442430496216,
      "learning_rate": 2.8792492208415768e-05,
      "loss": 1.589,
      "step": 822
    },
    {
      "epoch": 1.3184,
      "grad_norm": 1.4344308376312256,
      "learning_rate": 2.8553110506219283e-05,
      "loss": 1.5009,
      "step": 824
    },
    {
      "epoch": 1.3216,
      "grad_norm": 1.5077252388000488,
      "learning_rate": 2.831432973085848e-05,
      "loss": 1.512,
      "step": 826
    },
    {
      "epoch": 1.3248,
      "grad_norm": 1.5031547546386719,
      "learning_rate": 2.8076156572803635e-05,
      "loss": 1.5223,
      "step": 828
    },
    {
      "epoch": 1.328,
      "grad_norm": 1.5477045774459839,
      "learning_rate": 2.783859770549996e-05,
      "loss": 1.516,
      "step": 830
    },
    {
      "epoch": 1.3312,
      "grad_norm": 1.4744420051574707,
      "learning_rate": 2.760165978518067e-05,
      "loss": 1.4627,
      "step": 832
    },
    {
      "epoch": 1.3344,
      "grad_norm": 1.5096094608306885,
      "learning_rate": 2.7365349450680466e-05,
      "loss": 1.5211,
      "step": 834
    },
    {
      "epoch": 1.3376000000000001,
      "grad_norm": 1.7147303819656372,
      "learning_rate": 2.7129673323249604e-05,
      "loss": 1.5509,
      "step": 836
    },
    {
      "epoch": 1.3408,
      "grad_norm": 1.5401475429534912,
      "learning_rate": 2.689463800636824e-05,
      "loss": 1.5445,
      "step": 838
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 1.4286330938339233,
      "learning_rate": 2.6660250085561457e-05,
      "loss": 1.5123,
      "step": 840
    },
    {
      "epoch": 1.3472,
      "grad_norm": 1.6168473958969116,
      "learning_rate": 2.6426516128214807e-05,
      "loss": 1.5716,
      "step": 842
    },
    {
      "epoch": 1.3504,
      "grad_norm": 1.4487000703811646,
      "learning_rate": 2.619344268339021e-05,
      "loss": 1.5159,
      "step": 844
    },
    {
      "epoch": 1.3536000000000001,
      "grad_norm": 1.3211297988891602,
      "learning_rate": 2.5961036281642493e-05,
      "loss": 1.4555,
      "step": 846
    },
    {
      "epoch": 1.3568,
      "grad_norm": 1.3641369342803955,
      "learning_rate": 2.572930343483637e-05,
      "loss": 1.4302,
      "step": 848
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 1.3713055849075317,
      "learning_rate": 2.54982506359641e-05,
      "loss": 1.5058,
      "step": 850
    },
    {
      "epoch": 1.3632,
      "grad_norm": 1.3032983541488647,
      "learning_rate": 2.526788435896339e-05,
      "loss": 1.5108,
      "step": 852
    },
    {
      "epoch": 1.3664,
      "grad_norm": 1.345499038696289,
      "learning_rate": 2.5038211058536133e-05,
      "loss": 1.495,
      "step": 854
    },
    {
      "epoch": 1.3696,
      "grad_norm": 1.6281661987304688,
      "learning_rate": 2.4809237169967458e-05,
      "loss": 1.6172,
      "step": 856
    },
    {
      "epoch": 1.3728,
      "grad_norm": 1.5278024673461914,
      "learning_rate": 2.4580969108945533e-05,
      "loss": 1.513,
      "step": 858
    },
    {
      "epoch": 1.376,
      "grad_norm": 1.4942526817321777,
      "learning_rate": 2.435341327138168e-05,
      "loss": 1.4874,
      "step": 860
    },
    {
      "epoch": 1.3792,
      "grad_norm": 1.4423761367797852,
      "learning_rate": 2.4126576033231208e-05,
      "loss": 1.4102,
      "step": 862
    },
    {
      "epoch": 1.3824,
      "grad_norm": 1.5540028810501099,
      "learning_rate": 2.3900463750314834e-05,
      "loss": 1.5373,
      "step": 864
    },
    {
      "epoch": 1.3856,
      "grad_norm": 1.473035454750061,
      "learning_rate": 2.3675082758140475e-05,
      "loss": 1.577,
      "step": 866
    },
    {
      "epoch": 1.3888,
      "grad_norm": 1.3895673751831055,
      "learning_rate": 2.3450439371725825e-05,
      "loss": 1.5744,
      "step": 868
    },
    {
      "epoch": 1.392,
      "grad_norm": 1.4731401205062866,
      "learning_rate": 2.3226539885421343e-05,
      "loss": 1.4737,
      "step": 870
    },
    {
      "epoch": 1.3952,
      "grad_norm": 1.4738645553588867,
      "learning_rate": 2.3003390572734006e-05,
      "loss": 1.5292,
      "step": 872
    },
    {
      "epoch": 1.3984,
      "grad_norm": 1.343965768814087,
      "learning_rate": 2.2780997686151378e-05,
      "loss": 1.4885,
      "step": 874
    },
    {
      "epoch": 1.4016,
      "grad_norm": 1.3527370691299438,
      "learning_rate": 2.255936745696652e-05,
      "loss": 1.4582,
      "step": 876
    },
    {
      "epoch": 1.4048,
      "grad_norm": 1.3842884302139282,
      "learning_rate": 2.2338506095103334e-05,
      "loss": 1.4873,
      "step": 878
    },
    {
      "epoch": 1.408,
      "grad_norm": 1.4450650215148926,
      "learning_rate": 2.2118419788942672e-05,
      "loss": 1.4849,
      "step": 880
    },
    {
      "epoch": 1.4112,
      "grad_norm": 1.4666897058486938,
      "learning_rate": 2.189911470514881e-05,
      "loss": 1.5189,
      "step": 882
    },
    {
      "epoch": 1.4144,
      "grad_norm": 1.598429560661316,
      "learning_rate": 2.1680596988496705e-05,
      "loss": 1.5301,
      "step": 884
    },
    {
      "epoch": 1.4176,
      "grad_norm": 1.6322098970413208,
      "learning_rate": 2.1462872761699905e-05,
      "loss": 1.5321,
      "step": 886
    },
    {
      "epoch": 1.4208,
      "grad_norm": 1.56040620803833,
      "learning_rate": 2.1245948125238867e-05,
      "loss": 1.4492,
      "step": 888
    },
    {
      "epoch": 1.424,
      "grad_norm": 1.481416940689087,
      "learning_rate": 2.1029829157190117e-05,
      "loss": 1.5002,
      "step": 890
    },
    {
      "epoch": 1.4272,
      "grad_norm": 1.3432137966156006,
      "learning_rate": 2.081452191305587e-05,
      "loss": 1.5041,
      "step": 892
    },
    {
      "epoch": 1.4304000000000001,
      "grad_norm": 1.5008589029312134,
      "learning_rate": 2.06000324255945e-05,
      "loss": 1.5247,
      "step": 894
    },
    {
      "epoch": 1.4336,
      "grad_norm": 1.4487085342407227,
      "learning_rate": 2.0386366704651315e-05,
      "loss": 1.5005,
      "step": 896
    },
    {
      "epoch": 1.4368,
      "grad_norm": 1.394859790802002,
      "learning_rate": 2.0173530736990304e-05,
      "loss": 1.4009,
      "step": 898
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.3933418989181519,
      "learning_rate": 1.9961530486126327e-05,
      "loss": 1.4702,
      "step": 900
    },
    {
      "epoch": 1.4432,
      "grad_norm": 2.8447952270507812,
      "learning_rate": 1.9750371892158103e-05,
      "loss": 1.6987,
      "step": 902
    },
    {
      "epoch": 1.4464000000000001,
      "grad_norm": 1.4291117191314697,
      "learning_rate": 1.9540060871601646e-05,
      "loss": 1.4847,
      "step": 904
    },
    {
      "epoch": 1.4496,
      "grad_norm": 1.394317865371704,
      "learning_rate": 1.933060331722457e-05,
      "loss": 1.626,
      "step": 906
    },
    {
      "epoch": 1.4527999999999999,
      "grad_norm": 1.4183759689331055,
      "learning_rate": 1.9122005097881014e-05,
      "loss": 1.5301,
      "step": 908
    },
    {
      "epoch": 1.456,
      "grad_norm": 1.5104777812957764,
      "learning_rate": 1.8914272058347088e-05,
      "loss": 1.582,
      "step": 910
    },
    {
      "epoch": 1.4592,
      "grad_norm": 1.45592200756073,
      "learning_rate": 1.8707410019157196e-05,
      "loss": 1.4969,
      "step": 912
    },
    {
      "epoch": 1.4624,
      "grad_norm": 1.4659614562988281,
      "learning_rate": 1.8501424776440907e-05,
      "loss": 1.4836,
      "step": 914
    },
    {
      "epoch": 1.4656,
      "grad_norm": 1.3881189823150635,
      "learning_rate": 1.829632210176061e-05,
      "loss": 1.5318,
      "step": 916
    },
    {
      "epoch": 1.4687999999999999,
      "grad_norm": 1.4221471548080444,
      "learning_rate": 1.809210774194971e-05,
      "loss": 1.5534,
      "step": 918
    },
    {
      "epoch": 1.472,
      "grad_norm": 1.4791103601455688,
      "learning_rate": 1.7888787418951645e-05,
      "loss": 1.5574,
      "step": 920
    },
    {
      "epoch": 1.4752,
      "grad_norm": 1.387031078338623,
      "learning_rate": 1.7686366829659628e-05,
      "loss": 1.516,
      "step": 922
    },
    {
      "epoch": 1.4784,
      "grad_norm": 1.4815185070037842,
      "learning_rate": 1.74848516457569e-05,
      "loss": 1.5312,
      "step": 924
    },
    {
      "epoch": 1.4816,
      "grad_norm": 1.495384931564331,
      "learning_rate": 1.72842475135579e-05,
      "loss": 1.5245,
      "step": 926
    },
    {
      "epoch": 1.4848,
      "grad_norm": 1.5397976636886597,
      "learning_rate": 1.7084560053850024e-05,
      "loss": 1.575,
      "step": 928
    },
    {
      "epoch": 1.488,
      "grad_norm": 1.613402247428894,
      "learning_rate": 1.6885794861736183e-05,
      "loss": 1.4387,
      "step": 930
    },
    {
      "epoch": 1.4912,
      "grad_norm": 1.502987027168274,
      "learning_rate": 1.668795750647796e-05,
      "loss": 1.4356,
      "step": 932
    },
    {
      "epoch": 1.4944,
      "grad_norm": 1.4056270122528076,
      "learning_rate": 1.6491053531339607e-05,
      "loss": 1.5072,
      "step": 934
    },
    {
      "epoch": 1.4976,
      "grad_norm": 1.5031766891479492,
      "learning_rate": 1.62950884534327e-05,
      "loss": 1.4466,
      "step": 936
    },
    {
      "epoch": 1.5008,
      "grad_norm": 1.5224567651748657,
      "learning_rate": 1.6100067763561626e-05,
      "loss": 1.426,
      "step": 938
    },
    {
      "epoch": 1.504,
      "grad_norm": 1.4938695430755615,
      "learning_rate": 1.5905996926069628e-05,
      "loss": 1.5229,
      "step": 940
    },
    {
      "epoch": 1.5072,
      "grad_norm": 1.3707947731018066,
      "learning_rate": 1.5712881378685755e-05,
      "loss": 1.4859,
      "step": 942
    },
    {
      "epoch": 1.5104,
      "grad_norm": 1.4591916799545288,
      "learning_rate": 1.5520726532372537e-05,
      "loss": 1.4418,
      "step": 944
    },
    {
      "epoch": 1.5135999999999998,
      "grad_norm": 1.4584499597549438,
      "learning_rate": 1.532953777117429e-05,
      "loss": 1.5641,
      "step": 946
    },
    {
      "epoch": 1.5168,
      "grad_norm": 1.4583097696304321,
      "learning_rate": 1.5139320452066313e-05,
      "loss": 1.5063,
      "step": 948
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.4694300889968872,
      "learning_rate": 1.4950079904804759e-05,
      "loss": 1.5389,
      "step": 950
    },
    {
      "epoch": 1.5232,
      "grad_norm": 1.5485879182815552,
      "learning_rate": 1.4761821431777373e-05,
      "loss": 1.5044,
      "step": 952
    },
    {
      "epoch": 1.5264,
      "grad_norm": 1.4218132495880127,
      "learning_rate": 1.4574550307854817e-05,
      "loss": 1.5158,
      "step": 954
    },
    {
      "epoch": 1.5295999999999998,
      "grad_norm": 1.3866400718688965,
      "learning_rate": 1.4388271780242929e-05,
      "loss": 1.5014,
      "step": 956
    },
    {
      "epoch": 1.5328,
      "grad_norm": 1.4190049171447754,
      "learning_rate": 1.4202991068335697e-05,
      "loss": 1.4995,
      "step": 958
    },
    {
      "epoch": 1.536,
      "grad_norm": 1.4268289804458618,
      "learning_rate": 1.4018713363569035e-05,
      "loss": 1.5815,
      "step": 960
    },
    {
      "epoch": 1.5392000000000001,
      "grad_norm": 1.4949264526367188,
      "learning_rate": 1.3835443829275268e-05,
      "loss": 1.481,
      "step": 962
    },
    {
      "epoch": 1.5424,
      "grad_norm": 1.4129952192306519,
      "learning_rate": 1.365318760053848e-05,
      "loss": 1.4871,
      "step": 964
    },
    {
      "epoch": 1.5455999999999999,
      "grad_norm": 1.3220092058181763,
      "learning_rate": 1.3471949784050702e-05,
      "loss": 1.3881,
      "step": 966
    },
    {
      "epoch": 1.5488,
      "grad_norm": 1.435729742050171,
      "learning_rate": 1.3291735457968701e-05,
      "loss": 1.4646,
      "step": 968
    },
    {
      "epoch": 1.552,
      "grad_norm": 1.391000747680664,
      "learning_rate": 1.3112549671771796e-05,
      "loss": 1.4582,
      "step": 970
    },
    {
      "epoch": 1.5552000000000001,
      "grad_norm": 1.457863211631775,
      "learning_rate": 1.2934397446120306e-05,
      "loss": 1.4722,
      "step": 972
    },
    {
      "epoch": 1.5584,
      "grad_norm": 1.42721426486969,
      "learning_rate": 1.2757283772714957e-05,
      "loss": 1.4576,
      "step": 974
    },
    {
      "epoch": 1.5615999999999999,
      "grad_norm": 1.3265637159347534,
      "learning_rate": 1.2581213614156928e-05,
      "loss": 1.402,
      "step": 976
    },
    {
      "epoch": 1.5648,
      "grad_norm": 1.5157214403152466,
      "learning_rate": 1.2406191903808844e-05,
      "loss": 1.5254,
      "step": 978
    },
    {
      "epoch": 1.568,
      "grad_norm": 1.4206684827804565,
      "learning_rate": 1.2232223545656552e-05,
      "loss": 1.5294,
      "step": 980
    },
    {
      "epoch": 1.5712000000000002,
      "grad_norm": 1.4457650184631348,
      "learning_rate": 1.205931341417173e-05,
      "loss": 1.5149,
      "step": 982
    },
    {
      "epoch": 1.5744,
      "grad_norm": 1.3689544200897217,
      "learning_rate": 1.1887466354175253e-05,
      "loss": 1.4844,
      "step": 984
    },
    {
      "epoch": 1.5776,
      "grad_norm": 1.382386326789856,
      "learning_rate": 1.1716687180701474e-05,
      "loss": 1.4861,
      "step": 986
    },
    {
      "epoch": 1.5808,
      "grad_norm": 1.5220627784729004,
      "learning_rate": 1.1546980678863361e-05,
      "loss": 1.4846,
      "step": 988
    },
    {
      "epoch": 1.584,
      "grad_norm": 1.4781343936920166,
      "learning_rate": 1.1378351603718312e-05,
      "loss": 1.4562,
      "step": 990
    },
    {
      "epoch": 1.5872000000000002,
      "grad_norm": 1.4625208377838135,
      "learning_rate": 1.1210804680135022e-05,
      "loss": 1.3833,
      "step": 992
    },
    {
      "epoch": 1.5904,
      "grad_norm": 1.4238309860229492,
      "learning_rate": 1.1044344602661034e-05,
      "loss": 1.4929,
      "step": 994
    },
    {
      "epoch": 1.5936,
      "grad_norm": 1.3859175443649292,
      "learning_rate": 1.0878976035391252e-05,
      "loss": 1.4077,
      "step": 996
    },
    {
      "epoch": 1.5968,
      "grad_norm": 1.3726041316986084,
      "learning_rate": 1.0714703611837201e-05,
      "loss": 1.461,
      "step": 998
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.4159475564956665,
      "learning_rate": 1.0551531934797243e-05,
      "loss": 1.3846,
      "step": 1000
    },
    {
      "epoch": 1.6032,
      "grad_norm": 1.5738614797592163,
      "learning_rate": 1.0389465576227558e-05,
      "loss": 1.5087,
      "step": 1002
    },
    {
      "epoch": 1.6064,
      "grad_norm": 1.2960219383239746,
      "learning_rate": 1.0228509077114146e-05,
      "loss": 1.545,
      "step": 1004
    },
    {
      "epoch": 1.6096,
      "grad_norm": 1.3137331008911133,
      "learning_rate": 1.0068666947345456e-05,
      "loss": 1.4164,
      "step": 1006
    },
    {
      "epoch": 1.6128,
      "grad_norm": 1.4141883850097656,
      "learning_rate": 9.909943665586102e-06,
      "loss": 1.4566,
      "step": 1008
    },
    {
      "epoch": 1.616,
      "grad_norm": 1.3438924551010132,
      "learning_rate": 9.752343679151399e-06,
      "loss": 1.4601,
      "step": 1010
    },
    {
      "epoch": 1.6192,
      "grad_norm": 1.4983160495758057,
      "learning_rate": 9.595871403882661e-06,
      "loss": 1.5657,
      "step": 1012
    },
    {
      "epoch": 1.6223999999999998,
      "grad_norm": 1.4750056266784668,
      "learning_rate": 9.440531224023552e-06,
      "loss": 1.4662,
      "step": 1014
    },
    {
      "epoch": 1.6256,
      "grad_norm": 1.4642064571380615,
      "learning_rate": 9.286327492097196e-06,
      "loss": 1.3794,
      "step": 1016
    },
    {
      "epoch": 1.6288,
      "grad_norm": 1.467745304107666,
      "learning_rate": 9.133264528784274e-06,
      "loss": 1.4913,
      "step": 1018
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 1.493890404701233,
      "learning_rate": 8.981346622801905e-06,
      "loss": 1.5614,
      "step": 1020
    },
    {
      "epoch": 1.6352,
      "grad_norm": 1.4288145303726196,
      "learning_rate": 8.830578030783493e-06,
      "loss": 1.5431,
      "step": 1022
    },
    {
      "epoch": 1.6383999999999999,
      "grad_norm": 2.091399908065796,
      "learning_rate": 8.680962977159502e-06,
      "loss": 1.5166,
      "step": 1024
    },
    {
      "epoch": 1.6416,
      "grad_norm": 1.6347687244415283,
      "learning_rate": 8.53250565403903e-06,
      "loss": 1.5904,
      "step": 1026
    },
    {
      "epoch": 1.6448,
      "grad_norm": 1.4560258388519287,
      "learning_rate": 8.385210221092382e-06,
      "loss": 1.4901,
      "step": 1028
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 1.4900085926055908,
      "learning_rate": 8.239080805434513e-06,
      "loss": 1.5286,
      "step": 1030
    },
    {
      "epoch": 1.6512,
      "grad_norm": 1.4146417379379272,
      "learning_rate": 8.094121501509399e-06,
      "loss": 1.4933,
      "step": 1032
    },
    {
      "epoch": 1.6543999999999999,
      "grad_norm": 1.3483434915542603,
      "learning_rate": 7.950336370975304e-06,
      "loss": 1.4414,
      "step": 1034
    },
    {
      "epoch": 1.6576,
      "grad_norm": 1.3271894454956055,
      "learning_rate": 7.80772944259096e-06,
      "loss": 1.5141,
      "step": 1036
    },
    {
      "epoch": 1.6608,
      "grad_norm": 1.4688785076141357,
      "learning_rate": 7.666304712102695e-06,
      "loss": 1.4516,
      "step": 1038
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 1.3282010555267334,
      "learning_rate": 7.526066142132521e-06,
      "loss": 1.5037,
      "step": 1040
    },
    {
      "epoch": 1.6672,
      "grad_norm": 1.3577277660369873,
      "learning_rate": 7.3870176620670194e-06,
      "loss": 1.4295,
      "step": 1042
    },
    {
      "epoch": 1.6703999999999999,
      "grad_norm": 1.3745217323303223,
      "learning_rate": 7.249163167947287e-06,
      "loss": 1.4165,
      "step": 1044
    },
    {
      "epoch": 1.6736,
      "grad_norm": 1.3305115699768066,
      "learning_rate": 7.1125065223598076e-06,
      "loss": 1.4388,
      "step": 1046
    },
    {
      "epoch": 1.6768,
      "grad_norm": 1.3705346584320068,
      "learning_rate": 6.9770515543281455e-06,
      "loss": 1.5005,
      "step": 1048
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 1.4601020812988281,
      "learning_rate": 6.842802059205727e-06,
      "loss": 1.4845,
      "step": 1050
    },
    {
      "epoch": 1.6832,
      "grad_norm": 1.2712180614471436,
      "learning_rate": 6.709761798569442e-06,
      "loss": 1.464,
      "step": 1052
    },
    {
      "epoch": 1.6864,
      "grad_norm": 1.4009325504302979,
      "learning_rate": 6.577934500114335e-06,
      "loss": 1.485,
      "step": 1054
    },
    {
      "epoch": 1.6896,
      "grad_norm": 1.3810163736343384,
      "learning_rate": 6.44732385754902e-06,
      "loss": 1.4643,
      "step": 1056
    },
    {
      "epoch": 1.6928,
      "grad_norm": 1.5157772302627563,
      "learning_rate": 6.3179335304923095e-06,
      "loss": 1.4778,
      "step": 1058
    },
    {
      "epoch": 1.696,
      "grad_norm": 1.4355120658874512,
      "learning_rate": 6.189767144370645e-06,
      "loss": 1.4453,
      "step": 1060
    },
    {
      "epoch": 1.6992,
      "grad_norm": 1.5161240100860596,
      "learning_rate": 6.062828290316469e-06,
      "loss": 1.475,
      "step": 1062
    },
    {
      "epoch": 1.7024,
      "grad_norm": 1.448968529701233,
      "learning_rate": 5.937120525067641e-06,
      "loss": 1.4076,
      "step": 1064
    },
    {
      "epoch": 1.7056,
      "grad_norm": 1.3712458610534668,
      "learning_rate": 5.812647370867763e-06,
      "loss": 1.4497,
      "step": 1066
    },
    {
      "epoch": 1.7088,
      "grad_norm": 1.3936611413955688,
      "learning_rate": 5.689412315367543e-06,
      "loss": 1.4947,
      "step": 1068
    },
    {
      "epoch": 1.712,
      "grad_norm": 1.315384030342102,
      "learning_rate": 5.567418811526981e-06,
      "loss": 1.4324,
      "step": 1070
    },
    {
      "epoch": 1.7151999999999998,
      "grad_norm": 1.4205055236816406,
      "learning_rate": 5.4466702775186785e-06,
      "loss": 1.4313,
      "step": 1072
    },
    {
      "epoch": 1.7184,
      "grad_norm": 1.3843483924865723,
      "learning_rate": 5.327170096632089e-06,
      "loss": 1.494,
      "step": 1074
    },
    {
      "epoch": 1.7216,
      "grad_norm": 1.2861850261688232,
      "learning_rate": 5.208921617178641e-06,
      "loss": 1.5304,
      "step": 1076
    },
    {
      "epoch": 1.7248,
      "grad_norm": 1.3764268159866333,
      "learning_rate": 5.091928152397984e-06,
      "loss": 1.4531,
      "step": 1078
    },
    {
      "epoch": 1.728,
      "grad_norm": 1.3605912923812866,
      "learning_rate": 4.976192980365124e-06,
      "loss": 1.471,
      "step": 1080
    },
    {
      "epoch": 1.7311999999999999,
      "grad_norm": 1.4176418781280518,
      "learning_rate": 4.861719343898613e-06,
      "loss": 1.4616,
      "step": 1082
    },
    {
      "epoch": 1.7344,
      "grad_norm": 1.472076416015625,
      "learning_rate": 4.748510450469623e-06,
      "loss": 1.5043,
      "step": 1084
    },
    {
      "epoch": 1.7376,
      "grad_norm": 1.3761297464370728,
      "learning_rate": 4.63656947211214e-06,
      "loss": 1.488,
      "step": 1086
    },
    {
      "epoch": 1.7408000000000001,
      "grad_norm": 1.3362466096878052,
      "learning_rate": 4.525899545334023e-06,
      "loss": 1.3892,
      "step": 1088
    },
    {
      "epoch": 1.744,
      "grad_norm": 1.8346335887908936,
      "learning_rate": 4.416503771029201e-06,
      "loss": 1.691,
      "step": 1090
    },
    {
      "epoch": 1.7471999999999999,
      "grad_norm": 1.4666078090667725,
      "learning_rate": 4.308385214390709e-06,
      "loss": 1.4893,
      "step": 1092
    },
    {
      "epoch": 1.7504,
      "grad_norm": 1.3750196695327759,
      "learning_rate": 4.2015469048248375e-06,
      "loss": 1.4558,
      "step": 1094
    },
    {
      "epoch": 1.7536,
      "grad_norm": 1.4097230434417725,
      "learning_rate": 4.095991835866275e-06,
      "loss": 1.3775,
      "step": 1096
    },
    {
      "epoch": 1.7568000000000001,
      "grad_norm": 1.3481018543243408,
      "learning_rate": 3.99172296509418e-06,
      "loss": 1.4988,
      "step": 1098
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.3261443376541138,
      "learning_rate": 3.888743214049346e-06,
      "loss": 1.511,
      "step": 1100
    },
    {
      "epoch": 1.7631999999999999,
      "grad_norm": 1.3297051191329956,
      "learning_rate": 3.7870554681523287e-06,
      "loss": 1.4489,
      "step": 1102
    },
    {
      "epoch": 1.7664,
      "grad_norm": 1.3215663433074951,
      "learning_rate": 3.6866625766226293e-06,
      "loss": 1.4078,
      "step": 1104
    },
    {
      "epoch": 1.7696,
      "grad_norm": 1.3305385112762451,
      "learning_rate": 3.587567352398796e-06,
      "loss": 1.5099,
      "step": 1106
    },
    {
      "epoch": 1.7728000000000002,
      "grad_norm": 1.411366581916809,
      "learning_rate": 3.489772572059674e-06,
      "loss": 1.5017,
      "step": 1108
    },
    {
      "epoch": 1.776,
      "grad_norm": 1.309384822845459,
      "learning_rate": 3.393280975746588e-06,
      "loss": 1.3576,
      "step": 1110
    },
    {
      "epoch": 1.7792,
      "grad_norm": 1.322572946548462,
      "learning_rate": 3.2980952670865317e-06,
      "loss": 1.4796,
      "step": 1112
    },
    {
      "epoch": 1.7824,
      "grad_norm": 1.3316235542297363,
      "learning_rate": 3.2042181131164528e-06,
      "loss": 1.4801,
      "step": 1114
    },
    {
      "epoch": 1.7856,
      "grad_norm": 1.2216185331344604,
      "learning_rate": 3.11165214420851e-06,
      "loss": 1.4122,
      "step": 1116
    },
    {
      "epoch": 1.7888,
      "grad_norm": 1.4241148233413696,
      "learning_rate": 3.020399953996389e-06,
      "loss": 1.5404,
      "step": 1118
    },
    {
      "epoch": 1.792,
      "grad_norm": 1.4014073610305786,
      "learning_rate": 2.9304640993025988e-06,
      "loss": 1.4881,
      "step": 1120
    },
    {
      "epoch": 1.7952,
      "grad_norm": 1.2601819038391113,
      "learning_rate": 2.8418471000668523e-06,
      "loss": 1.4285,
      "step": 1122
    },
    {
      "epoch": 1.7984,
      "grad_norm": 1.4038496017456055,
      "learning_rate": 2.7545514392754437e-06,
      "loss": 1.5409,
      "step": 1124
    },
    {
      "epoch": 1.8016,
      "grad_norm": 1.3646516799926758,
      "learning_rate": 2.6685795628917266e-06,
      "loss": 1.3933,
      "step": 1126
    },
    {
      "epoch": 1.8048,
      "grad_norm": 1.5789858102798462,
      "learning_rate": 2.5839338797875036e-06,
      "loss": 1.4977,
      "step": 1128
    },
    {
      "epoch": 1.808,
      "grad_norm": 1.5458101034164429,
      "learning_rate": 2.500616761675578e-06,
      "loss": 1.6548,
      "step": 1130
    },
    {
      "epoch": 1.8112,
      "grad_norm": 1.3782880306243896,
      "learning_rate": 2.41863054304331e-06,
      "loss": 1.5003,
      "step": 1132
    },
    {
      "epoch": 1.8144,
      "grad_norm": 1.3550211191177368,
      "learning_rate": 2.3379775210871648e-06,
      "loss": 1.5406,
      "step": 1134
    },
    {
      "epoch": 1.8176,
      "grad_norm": 1.3220856189727783,
      "learning_rate": 2.2586599556483734e-06,
      "loss": 1.4868,
      "step": 1136
    },
    {
      "epoch": 1.8208,
      "grad_norm": 1.37287437915802,
      "learning_rate": 2.180680069149621e-06,
      "loss": 1.4207,
      "step": 1138
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 1.3784503936767578,
      "learning_rate": 2.104040046532768e-06,
      "loss": 1.4947,
      "step": 1140
    },
    {
      "epoch": 1.8272,
      "grad_norm": 1.222406029701233,
      "learning_rate": 2.0287420351976063e-06,
      "loss": 1.4748,
      "step": 1142
    },
    {
      "epoch": 1.8304,
      "grad_norm": 1.3098535537719727,
      "learning_rate": 1.954788144941727e-06,
      "loss": 1.4453,
      "step": 1144
    },
    {
      "epoch": 1.8336000000000001,
      "grad_norm": 1.2869049310684204,
      "learning_rate": 1.8821804479013772e-06,
      "loss": 1.3639,
      "step": 1146
    },
    {
      "epoch": 1.8368,
      "grad_norm": 1.316622257232666,
      "learning_rate": 1.81092097849343e-06,
      "loss": 1.5108,
      "step": 1148
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 1.3417192697525024,
      "learning_rate": 1.7410117333583498e-06,
      "loss": 1.4509,
      "step": 1150
    },
    {
      "epoch": 1.8432,
      "grad_norm": 1.4544188976287842,
      "learning_rate": 1.6724546713042577e-06,
      "loss": 1.4119,
      "step": 1152
    },
    {
      "epoch": 1.8464,
      "grad_norm": 1.4499175548553467,
      "learning_rate": 1.6052517132520651e-06,
      "loss": 1.5089,
      "step": 1154
    },
    {
      "epoch": 1.8496000000000001,
      "grad_norm": 1.4253344535827637,
      "learning_rate": 1.5394047421816327e-06,
      "loss": 1.4798,
      "step": 1156
    },
    {
      "epoch": 1.8528,
      "grad_norm": 1.6616299152374268,
      "learning_rate": 1.4749156030790024e-06,
      "loss": 1.5351,
      "step": 1158
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 1.3056614398956299,
      "learning_rate": 1.4117861028847267e-06,
      "loss": 1.4418,
      "step": 1160
    },
    {
      "epoch": 1.8592,
      "grad_norm": 1.3334782123565674,
      "learning_rate": 1.3500180104432325e-06,
      "loss": 1.4205,
      "step": 1162
    },
    {
      "epoch": 1.8624,
      "grad_norm": 1.2939821481704712,
      "learning_rate": 1.2896130564532427e-06,
      "loss": 1.489,
      "step": 1164
    },
    {
      "epoch": 1.8656000000000001,
      "grad_norm": 1.4086552858352661,
      "learning_rate": 1.2305729334192994e-06,
      "loss": 1.5408,
      "step": 1166
    },
    {
      "epoch": 1.8688,
      "grad_norm": 1.4459562301635742,
      "learning_rate": 1.1728992956043238e-06,
      "loss": 1.5132,
      "step": 1168
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 2.8484818935394287,
      "learning_rate": 1.1165937589833087e-06,
      "loss": 1.5685,
      "step": 1170
    },
    {
      "epoch": 1.8752,
      "grad_norm": 1.4481736421585083,
      "learning_rate": 1.061657901197971e-06,
      "loss": 1.4605,
      "step": 1172
    },
    {
      "epoch": 1.8784,
      "grad_norm": 1.2909802198410034,
      "learning_rate": 1.008093261512616e-06,
      "loss": 1.4924,
      "step": 1174
    },
    {
      "epoch": 1.8816000000000002,
      "grad_norm": 1.3690600395202637,
      "learning_rate": 9.559013407709595e-07,
      "loss": 1.4364,
      "step": 1176
    },
    {
      "epoch": 1.8848,
      "grad_norm": 1.381736159324646,
      "learning_rate": 9.050836013541009e-07,
      "loss": 1.512,
      "step": 1178
    },
    {
      "epoch": 1.888,
      "grad_norm": 1.4545315504074097,
      "learning_rate": 8.55641467139534e-07,
      "loss": 1.4628,
      "step": 1180
    },
    {
      "epoch": 1.8912,
      "grad_norm": 1.321023941040039,
      "learning_rate": 8.075763234612622e-07,
      "loss": 1.4768,
      "step": 1182
    },
    {
      "epoch": 1.8944,
      "grad_norm": 1.485850214958191,
      "learning_rate": 7.60889517070984e-07,
      "loss": 1.5397,
      "step": 1184
    },
    {
      "epoch": 1.8976,
      "grad_norm": 1.3739385604858398,
      "learning_rate": 7.155823561003361e-07,
      "loss": 1.4546,
      "step": 1186
    },
    {
      "epoch": 1.9008,
      "grad_norm": 1.4102145433425903,
      "learning_rate": 6.716561100242658e-07,
      "loss": 1.4662,
      "step": 1188
    },
    {
      "epoch": 1.904,
      "grad_norm": 1.2747293710708618,
      "learning_rate": 6.291120096254433e-07,
      "loss": 1.402,
      "step": 1190
    },
    {
      "epoch": 1.9072,
      "grad_norm": 1.453570008277893,
      "learning_rate": 5.879512469598058e-07,
      "loss": 1.4189,
      "step": 1192
    },
    {
      "epoch": 1.9104,
      "grad_norm": 1.3107861280441284,
      "learning_rate": 5.481749753231124e-07,
      "loss": 1.4495,
      "step": 1194
    },
    {
      "epoch": 1.9136,
      "grad_norm": 1.4166679382324219,
      "learning_rate": 5.097843092186583e-07,
      "loss": 1.5498,
      "step": 1196
    },
    {
      "epoch": 1.9167999999999998,
      "grad_norm": 1.384834885597229,
      "learning_rate": 4.7278032432604425e-07,
      "loss": 1.4262,
      "step": 1198
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.2457374334335327,
      "learning_rate": 4.371640574710345e-07,
      "loss": 1.4426,
      "step": 1200
    },
    {
      "epoch": 1.9232,
      "grad_norm": 1.4037002325057983,
      "learning_rate": 4.0293650659650184e-07,
      "loss": 1.423,
      "step": 1202
    },
    {
      "epoch": 1.9264000000000001,
      "grad_norm": 1.3311501741409302,
      "learning_rate": 3.7009863073446673e-07,
      "loss": 1.4888,
      "step": 1204
    },
    {
      "epoch": 1.9296,
      "grad_norm": 1.4098093509674072,
      "learning_rate": 3.386513499792354e-07,
      "loss": 1.4965,
      "step": 1206
    },
    {
      "epoch": 1.9327999999999999,
      "grad_norm": 1.371372103691101,
      "learning_rate": 3.0859554546160965e-07,
      "loss": 1.3423,
      "step": 1208
    },
    {
      "epoch": 1.936,
      "grad_norm": 1.3136426210403442,
      "learning_rate": 2.7993205932420053e-07,
      "loss": 1.4077,
      "step": 1210
    },
    {
      "epoch": 1.9392,
      "grad_norm": 1.3438588380813599,
      "learning_rate": 2.5266169469783105e-07,
      "loss": 1.5836,
      "step": 1212
    },
    {
      "epoch": 1.9424000000000001,
      "grad_norm": 1.4070448875427246,
      "learning_rate": 2.2678521567903176e-07,
      "loss": 1.4745,
      "step": 1214
    },
    {
      "epoch": 1.9456,
      "grad_norm": 1.4340542554855347,
      "learning_rate": 2.023033473086411e-07,
      "loss": 1.4898,
      "step": 1216
    },
    {
      "epoch": 1.9487999999999999,
      "grad_norm": 1.4534060955047607,
      "learning_rate": 1.7921677555147177e-07,
      "loss": 1.495,
      "step": 1218
    },
    {
      "epoch": 1.952,
      "grad_norm": 1.5039774179458618,
      "learning_rate": 1.5752614727712057e-07,
      "loss": 1.4469,
      "step": 1220
    },
    {
      "epoch": 1.9552,
      "grad_norm": 1.3171782493591309,
      "learning_rate": 1.3723207024180507e-07,
      "loss": 1.4974,
      "step": 1222
    },
    {
      "epoch": 1.9584000000000001,
      "grad_norm": 1.4314095973968506,
      "learning_rate": 1.1833511307136613e-07,
      "loss": 1.4934,
      "step": 1224
    },
    {
      "epoch": 1.9616,
      "grad_norm": 1.449654459953308,
      "learning_rate": 1.0083580524531955e-07,
      "loss": 1.5354,
      "step": 1226
    },
    {
      "epoch": 1.9647999999999999,
      "grad_norm": 1.3446357250213623,
      "learning_rate": 8.473463708202345e-08,
      "loss": 1.4772,
      "step": 1228
    },
    {
      "epoch": 1.968,
      "grad_norm": 1.3610376119613647,
      "learning_rate": 7.003205972494486e-08,
      "loss": 1.3527,
      "step": 1230
    },
    {
      "epoch": 1.9712,
      "grad_norm": 1.3653186559677124,
      "learning_rate": 5.672848513000873e-08,
      "loss": 1.444,
      "step": 1232
    },
    {
      "epoch": 1.9744000000000002,
      "grad_norm": 2.1673245429992676,
      "learning_rate": 4.482428605407374e-08,
      "loss": 1.442,
      "step": 1234
    },
    {
      "epoch": 1.9776,
      "grad_norm": 1.4029662609100342,
      "learning_rate": 3.431979604445745e-08,
      "loss": 1.5326,
      "step": 1236
    },
    {
      "epoch": 1.9808,
      "grad_norm": 1.3570245504379272,
      "learning_rate": 2.521530942962702e-08,
      "loss": 1.4865,
      "step": 1238
    },
    {
      "epoch": 1.984,
      "grad_norm": 1.5136021375656128,
      "learning_rate": 1.7511081310922495e-08,
      "loss": 1.4449,
      "step": 1240
    },
    {
      "epoch": 1.9872,
      "grad_norm": 1.3614838123321533,
      "learning_rate": 1.1207327555429192e-08,
      "loss": 1.3968,
      "step": 1242
    },
    {
      "epoch": 1.9904,
      "grad_norm": 1.3964884281158447,
      "learning_rate": 6.304224789910329e-09,
      "loss": 1.4793,
      "step": 1244
    },
    {
      "epoch": 1.9936,
      "grad_norm": 1.2995529174804688,
      "learning_rate": 2.801910395877627e-09,
      "loss": 1.373,
      "step": 1246
    },
    {
      "epoch": 1.9968,
      "grad_norm": 1.4436957836151123,
      "learning_rate": 7.004825057277398e-10,
      "loss": 1.5181,
      "step": 1248
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.342529535293579,
      "learning_rate": 0.0,
      "loss": 1.3802,
      "step": 1250
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1133643907229286e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}