{
  "best_metric": 0.07416880130767822,
  "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles7_seed2_q3_dropout_v2/checkpoint-469",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 2345,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010660980810234541,
      "grad_norm": 0.089314766228199,
      "learning_rate": 4.989339019189766e-05,
      "loss": 0.0497,
      "step": 5
    },
    {
      "epoch": 0.021321961620469083,
      "grad_norm": 0.7876163721084595,
      "learning_rate": 4.978678038379531e-05,
      "loss": 0.0134,
      "step": 10
    },
    {
      "epoch": 0.031982942430703626,
      "grad_norm": 0.07032600790262222,
      "learning_rate": 4.9680170575692967e-05,
      "loss": 0.0316,
      "step": 15
    },
    {
      "epoch": 0.042643923240938165,
      "grad_norm": 2.452097177505493,
      "learning_rate": 4.957356076759062e-05,
      "loss": 0.0382,
      "step": 20
    },
    {
      "epoch": 0.053304904051172705,
      "grad_norm": 3.198247194290161,
      "learning_rate": 4.9466950959488276e-05,
      "loss": 0.061,
      "step": 25
    },
    {
      "epoch": 0.06396588486140725,
      "grad_norm": 2.2304351329803467,
      "learning_rate": 4.936034115138593e-05,
      "loss": 0.052,
      "step": 30
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.30216485261917114,
      "learning_rate": 4.9253731343283586e-05,
      "loss": 0.0488,
      "step": 35
    },
    {
      "epoch": 0.08528784648187633,
      "grad_norm": 3.4997708797454834,
      "learning_rate": 4.914712153518124e-05,
      "loss": 0.11,
      "step": 40
    },
    {
      "epoch": 0.09594882729211088,
      "grad_norm": 8.029236793518066,
      "learning_rate": 4.904051172707889e-05,
      "loss": 0.0712,
      "step": 45
    },
    {
      "epoch": 0.10660980810234541,
      "grad_norm": 3.4754045009613037,
      "learning_rate": 4.893390191897655e-05,
      "loss": 0.0892,
      "step": 50
    },
    {
      "epoch": 0.11727078891257996,
      "grad_norm": 0.07975251227617264,
      "learning_rate": 4.88272921108742e-05,
      "loss": 0.0297,
      "step": 55
    },
    {
      "epoch": 0.1279317697228145,
      "grad_norm": 4.289177417755127,
      "learning_rate": 4.872068230277186e-05,
      "loss": 0.055,
      "step": 60
    },
    {
      "epoch": 0.13859275053304904,
      "grad_norm": 0.1036117821931839,
      "learning_rate": 4.861407249466951e-05,
      "loss": 0.0717,
      "step": 65
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 2.215583086013794,
      "learning_rate": 4.850746268656717e-05,
      "loss": 0.0578,
      "step": 70
    },
    {
      "epoch": 0.15991471215351813,
      "grad_norm": 4.49214506149292,
      "learning_rate": 4.840085287846482e-05,
      "loss": 0.0274,
      "step": 75
    },
    {
      "epoch": 0.17057569296375266,
      "grad_norm": 6.780641078948975,
      "learning_rate": 4.829424307036248e-05,
      "loss": 0.0239,
      "step": 80
    },
    {
      "epoch": 0.1812366737739872,
      "grad_norm": 7.513982772827148,
      "learning_rate": 4.8187633262260126e-05,
      "loss": 0.1082,
      "step": 85
    },
    {
      "epoch": 0.19189765458422176,
      "grad_norm": 1.3966752290725708,
      "learning_rate": 4.808102345415779e-05,
      "loss": 0.0548,
      "step": 90
    },
    {
      "epoch": 0.2025586353944563,
      "grad_norm": 0.06923407316207886,
      "learning_rate": 4.7974413646055436e-05,
      "loss": 0.089,
      "step": 95
    },
    {
      "epoch": 0.21321961620469082,
      "grad_norm": 4.584115982055664,
      "learning_rate": 4.78678038379531e-05,
      "loss": 0.0295,
      "step": 100
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 1.5590736865997314,
      "learning_rate": 4.7761194029850745e-05,
      "loss": 0.0797,
      "step": 105
    },
    {
      "epoch": 0.2345415778251599,
      "grad_norm": 5.624256134033203,
      "learning_rate": 4.765458422174841e-05,
      "loss": 0.0974,
      "step": 110
    },
    {
      "epoch": 0.24520255863539445,
      "grad_norm": 0.16852571070194244,
      "learning_rate": 4.7547974413646055e-05,
      "loss": 0.0744,
      "step": 115
    },
    {
      "epoch": 0.255863539445629,
      "grad_norm": 4.0069990158081055,
      "learning_rate": 4.7441364605543716e-05,
      "loss": 0.0327,
      "step": 120
    },
    {
      "epoch": 0.26652452025586354,
      "grad_norm": 0.05257545784115791,
      "learning_rate": 4.7334754797441364e-05,
      "loss": 0.0112,
      "step": 125
    },
    {
      "epoch": 0.2771855010660981,
      "grad_norm": 0.10090459138154984,
      "learning_rate": 4.7228144989339026e-05,
      "loss": 0.0821,
      "step": 130
    },
    {
      "epoch": 0.2878464818763326,
      "grad_norm": 1.674806833267212,
      "learning_rate": 4.7121535181236674e-05,
      "loss": 0.0113,
      "step": 135
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.8493373394012451,
      "learning_rate": 4.7014925373134335e-05,
      "loss": 0.0288,
      "step": 140
    },
    {
      "epoch": 0.3091684434968017,
      "grad_norm": 5.2708330154418945,
      "learning_rate": 4.690831556503198e-05,
      "loss": 0.0218,
      "step": 145
    },
    {
      "epoch": 0.31982942430703626,
      "grad_norm": 2.858956813812256,
      "learning_rate": 4.6801705756929645e-05,
      "loss": 0.0459,
      "step": 150
    },
    {
      "epoch": 0.3304904051172708,
      "grad_norm": 5.069364070892334,
      "learning_rate": 4.669509594882729e-05,
      "loss": 0.1695,
      "step": 155
    },
    {
      "epoch": 0.3411513859275053,
      "grad_norm": 0.938015341758728,
      "learning_rate": 4.658848614072495e-05,
      "loss": 0.1072,
      "step": 160
    },
    {
      "epoch": 0.35181236673773986,
      "grad_norm": 5.146633148193359,
      "learning_rate": 4.64818763326226e-05,
      "loss": 0.0556,
      "step": 165
    },
    {
      "epoch": 0.3624733475479744,
      "grad_norm": 3.7878060340881348,
      "learning_rate": 4.637526652452026e-05,
      "loss": 0.036,
      "step": 170
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.05061347782611847,
      "learning_rate": 4.626865671641791e-05,
      "loss": 0.1145,
      "step": 175
    },
    {
      "epoch": 0.3837953091684435,
      "grad_norm": 3.6897590160369873,
      "learning_rate": 4.6162046908315566e-05,
      "loss": 0.1092,
      "step": 180
    },
    {
      "epoch": 0.39445628997867804,
      "grad_norm": 2.005059242248535,
      "learning_rate": 4.605543710021322e-05,
      "loss": 0.0445,
      "step": 185
    },
    {
      "epoch": 0.4051172707889126,
      "grad_norm": 3.1155292987823486,
      "learning_rate": 4.5948827292110876e-05,
      "loss": 0.1087,
      "step": 190
    },
    {
      "epoch": 0.4157782515991471,
      "grad_norm": 5.733726978302002,
      "learning_rate": 4.584221748400853e-05,
      "loss": 0.0585,
      "step": 195
    },
    {
      "epoch": 0.42643923240938164,
      "grad_norm": 0.08829709142446518,
      "learning_rate": 4.5735607675906185e-05,
      "loss": 0.0108,
      "step": 200
    },
    {
      "epoch": 0.43710021321961623,
      "grad_norm": 3.1341664791107178,
      "learning_rate": 4.562899786780384e-05,
      "loss": 0.0195,
      "step": 205
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.038583576679229736,
      "learning_rate": 4.5522388059701495e-05,
      "loss": 0.0243,
      "step": 210
    },
    {
      "epoch": 0.4584221748400853,
      "grad_norm": 0.20733138918876648,
      "learning_rate": 4.541577825159915e-05,
      "loss": 0.0098,
      "step": 215
    },
    {
      "epoch": 0.4690831556503198,
      "grad_norm": 0.41318878531455994,
      "learning_rate": 4.5309168443496804e-05,
      "loss": 0.0097,
      "step": 220
    },
    {
      "epoch": 0.47974413646055436,
      "grad_norm": 1.8579902648925781,
      "learning_rate": 4.520255863539446e-05,
      "loss": 0.0544,
      "step": 225
    },
    {
      "epoch": 0.4904051172707889,
      "grad_norm": 1.3941482305526733,
      "learning_rate": 4.5095948827292114e-05,
      "loss": 0.0863,
      "step": 230
    },
    {
      "epoch": 0.5010660980810234,
      "grad_norm": 4.815855979919434,
      "learning_rate": 4.498933901918977e-05,
      "loss": 0.0827,
      "step": 235
    },
    {
      "epoch": 0.511727078891258,
      "grad_norm": 6.854145050048828,
      "learning_rate": 4.488272921108742e-05,
      "loss": 0.0434,
      "step": 240
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 7.457388877868652,
      "learning_rate": 4.477611940298508e-05,
      "loss": 0.0226,
      "step": 245
    },
    {
      "epoch": 0.5330490405117271,
      "grad_norm": 0.03205385431647301,
      "learning_rate": 4.466950959488273e-05,
      "loss": 0.0279,
      "step": 250
    },
    {
      "epoch": 0.5437100213219617,
      "grad_norm": 8.858234405517578,
      "learning_rate": 4.456289978678039e-05,
      "loss": 0.042,
      "step": 255
    },
    {
      "epoch": 0.5543710021321961,
      "grad_norm": 2.314207077026367,
      "learning_rate": 4.445628997867804e-05,
      "loss": 0.007,
      "step": 260
    },
    {
      "epoch": 0.5650319829424307,
      "grad_norm": 2.1279854774475098,
      "learning_rate": 4.43496801705757e-05,
      "loss": 0.0957,
      "step": 265
    },
    {
      "epoch": 0.5756929637526652,
      "grad_norm": 4.006611347198486,
      "learning_rate": 4.424307036247335e-05,
      "loss": 0.0407,
      "step": 270
    },
    {
      "epoch": 0.5863539445628998,
      "grad_norm": 0.37693729996681213,
      "learning_rate": 4.4136460554371006e-05,
      "loss": 0.03,
      "step": 275
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.8131709694862366,
      "learning_rate": 4.402985074626866e-05,
      "loss": 0.0339,
      "step": 280
    },
    {
      "epoch": 0.6076759061833689,
      "grad_norm": 8.656360626220703,
      "learning_rate": 4.3923240938166316e-05,
      "loss": 0.0581,
      "step": 285
    },
    {
      "epoch": 0.6183368869936035,
      "grad_norm": 1.0311282873153687,
      "learning_rate": 4.381663113006397e-05,
      "loss": 0.0487,
      "step": 290
    },
    {
      "epoch": 0.6289978678038379,
      "grad_norm": 0.040167029947042465,
      "learning_rate": 4.3710021321961625e-05,
      "loss": 0.0345,
      "step": 295
    },
    {
      "epoch": 0.6396588486140725,
      "grad_norm": 0.03608911857008934,
      "learning_rate": 4.360341151385928e-05,
      "loss": 0.0142,
      "step": 300
    },
    {
      "epoch": 0.650319829424307,
      "grad_norm": 2.757634401321411,
      "learning_rate": 4.3496801705756935e-05,
      "loss": 0.0081,
      "step": 305
    },
    {
      "epoch": 0.6609808102345416,
      "grad_norm": 0.05704599618911743,
      "learning_rate": 4.339019189765459e-05,
      "loss": 0.0086,
      "step": 310
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 1.930027723312378,
      "learning_rate": 4.328358208955224e-05,
      "loss": 0.0872,
      "step": 315
    },
    {
      "epoch": 0.6823027718550106,
      "grad_norm": 0.9900903701782227,
      "learning_rate": 4.31769722814499e-05,
      "loss": 0.0092,
      "step": 320
    },
    {
      "epoch": 0.6929637526652452,
      "grad_norm": 4.703475475311279,
      "learning_rate": 4.307036247334755e-05,
      "loss": 0.0739,
      "step": 325
    },
    {
      "epoch": 0.7036247334754797,
      "grad_norm": 0.02604069374501705,
      "learning_rate": 4.29637526652452e-05,
      "loss": 0.0679,
      "step": 330
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.032000526785850525,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.0758,
      "step": 335
    },
    {
      "epoch": 0.7249466950959488,
      "grad_norm": 0.03631070628762245,
      "learning_rate": 4.275053304904051e-05,
      "loss": 0.036,
      "step": 340
    },
    {
      "epoch": 0.7356076759061834,
      "grad_norm": 0.02731061913073063,
      "learning_rate": 4.2643923240938166e-05,
      "loss": 0.0656,
      "step": 345
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.06623335927724838,
      "learning_rate": 4.253731343283582e-05,
      "loss": 0.0233,
      "step": 350
    },
    {
      "epoch": 0.7569296375266524,
      "grad_norm": 4.656162738800049,
      "learning_rate": 4.2430703624733475e-05,
      "loss": 0.0301,
      "step": 355
    },
    {
      "epoch": 0.767590618336887,
      "grad_norm": 0.3432367742061615,
      "learning_rate": 4.232409381663113e-05,
      "loss": 0.0481,
      "step": 360
    },
    {
      "epoch": 0.7782515991471215,
      "grad_norm": 4.111856937408447,
      "learning_rate": 4.2217484008528785e-05,
      "loss": 0.016,
      "step": 365
    },
    {
      "epoch": 0.7889125799573561,
      "grad_norm": 9.072041511535645,
      "learning_rate": 4.211087420042644e-05,
      "loss": 0.0883,
      "step": 370
    },
    {
      "epoch": 0.7995735607675906,
      "grad_norm": 0.0411033108830452,
      "learning_rate": 4.2004264392324094e-05,
      "loss": 0.038,
      "step": 375
    },
    {
      "epoch": 0.8102345415778252,
      "grad_norm": 2.539846181869507,
      "learning_rate": 4.189765458422175e-05,
      "loss": 0.0595,
      "step": 380
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 3.8313751220703125,
      "learning_rate": 4.1791044776119404e-05,
      "loss": 0.1117,
      "step": 385
    },
    {
      "epoch": 0.8315565031982942,
      "grad_norm": 5.764938831329346,
      "learning_rate": 4.168443496801706e-05,
      "loss": 0.055,
      "step": 390
    },
    {
      "epoch": 0.8422174840085288,
      "grad_norm": 3.9832303524017334,
      "learning_rate": 4.157782515991471e-05,
      "loss": 0.0494,
      "step": 395
    },
    {
      "epoch": 0.8528784648187633,
      "grad_norm": 0.118993379175663,
      "learning_rate": 4.147121535181237e-05,
      "loss": 0.0314,
      "step": 400
    },
    {
      "epoch": 0.8635394456289979,
      "grad_norm": 0.9116721749305725,
      "learning_rate": 4.136460554371002e-05,
      "loss": 0.0377,
      "step": 405
    },
    {
      "epoch": 0.8742004264392325,
      "grad_norm": 0.04301665723323822,
      "learning_rate": 4.125799573560768e-05,
      "loss": 0.0183,
      "step": 410
    },
    {
      "epoch": 0.8848614072494669,
      "grad_norm": 3.0369009971618652,
      "learning_rate": 4.115138592750533e-05,
      "loss": 0.06,
      "step": 415
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.9150505661964417,
      "learning_rate": 4.104477611940299e-05,
      "loss": 0.0538,
      "step": 420
    },
    {
      "epoch": 0.906183368869936,
      "grad_norm": 2.4762611389160156,
      "learning_rate": 4.093816631130064e-05,
      "loss": 0.0642,
      "step": 425
    },
    {
      "epoch": 0.9168443496801706,
      "grad_norm": 0.35692906379699707,
      "learning_rate": 4.0831556503198296e-05,
      "loss": 0.0487,
      "step": 430
    },
    {
      "epoch": 0.9275053304904051,
      "grad_norm": 0.2154560536146164,
      "learning_rate": 4.072494669509595e-05,
      "loss": 0.0321,
      "step": 435
    },
    {
      "epoch": 0.9381663113006397,
      "grad_norm": 0.09918001294136047,
      "learning_rate": 4.0618336886993606e-05,
      "loss": 0.0283,
      "step": 440
    },
    {
      "epoch": 0.9488272921108742,
      "grad_norm": 6.488087177276611,
      "learning_rate": 4.051172707889126e-05,
      "loss": 0.0772,
      "step": 445
    },
    {
      "epoch": 0.9594882729211087,
      "grad_norm": 0.05787067487835884,
      "learning_rate": 4.0405117270788915e-05,
      "loss": 0.0523,
      "step": 450
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.1128556951880455,
      "learning_rate": 4.029850746268657e-05,
      "loss": 0.0496,
      "step": 455
    },
    {
      "epoch": 0.9808102345415778,
      "grad_norm": 2.36838698387146,
      "learning_rate": 4.0191897654584225e-05,
      "loss": 0.0603,
      "step": 460
    },
    {
      "epoch": 0.9914712153518124,
      "grad_norm": 1.358279824256897,
      "learning_rate": 4.008528784648188e-05,
      "loss": 0.0079,
      "step": 465
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9738666666666667,
      "eval_loss": 0.07416880130767822,
      "eval_runtime": 20.0489,
      "eval_samples_per_second": 187.043,
      "eval_steps_per_second": 5.886,
      "step": 469
    },
    {
      "epoch": 1.0021321961620469,
      "grad_norm": 0.22900037467479706,
      "learning_rate": 3.997867803837953e-05,
      "loss": 0.0487,
      "step": 470
    },
    {
      "epoch": 1.0127931769722816,
      "grad_norm": 1.0007952451705933,
      "learning_rate": 3.987206823027719e-05,
      "loss": 0.0084,
      "step": 475
    },
    {
      "epoch": 1.023454157782516,
      "grad_norm": 4.713870048522949,
      "learning_rate": 3.976545842217484e-05,
      "loss": 0.0295,
      "step": 480
    },
    {
      "epoch": 1.0341151385927505,
      "grad_norm": 0.24120108783245087,
      "learning_rate": 3.96588486140725e-05,
      "loss": 0.0053,
      "step": 485
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 0.295411080121994,
      "learning_rate": 3.9552238805970146e-05,
      "loss": 0.0142,
      "step": 490
    },
    {
      "epoch": 1.0554371002132197,
      "grad_norm": 0.042514558881521225,
      "learning_rate": 3.944562899786781e-05,
      "loss": 0.0178,
      "step": 495
    },
    {
      "epoch": 1.0660980810234542,
      "grad_norm": 2.7463572025299072,
      "learning_rate": 3.9339019189765456e-05,
      "loss": 0.0175,
      "step": 500
    },
    {
      "epoch": 1.0767590618336886,
      "grad_norm": 0.9280542135238647,
      "learning_rate": 3.923240938166312e-05,
      "loss": 0.0148,
      "step": 505
    },
    {
      "epoch": 1.0874200426439233,
      "grad_norm": 1.3411023616790771,
      "learning_rate": 3.9125799573560765e-05,
      "loss": 0.0055,
      "step": 510
    },
    {
      "epoch": 1.0980810234541578,
      "grad_norm": 0.221579909324646,
      "learning_rate": 3.901918976545843e-05,
      "loss": 0.0031,
      "step": 515
    },
    {
      "epoch": 1.1087420042643923,
      "grad_norm": 0.8755818009376526,
      "learning_rate": 3.8912579957356075e-05,
      "loss": 0.0068,
      "step": 520
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 4.702971458435059,
      "learning_rate": 3.8805970149253736e-05,
      "loss": 0.0139,
      "step": 525
    },
    {
      "epoch": 1.1300639658848615,
      "grad_norm": 4.165937900543213,
      "learning_rate": 3.8699360341151384e-05,
      "loss": 0.0404,
      "step": 530
    },
    {
      "epoch": 1.140724946695096,
      "grad_norm": 0.018623312935233116,
      "learning_rate": 3.8592750533049046e-05,
      "loss": 0.0041,
      "step": 535
    },
    {
      "epoch": 1.1513859275053304,
      "grad_norm": 8.683755874633789,
      "learning_rate": 3.8486140724946694e-05,
      "loss": 0.0141,
      "step": 540
    },
    {
      "epoch": 1.1620469083155651,
      "grad_norm": 0.8486212491989136,
      "learning_rate": 3.8379530916844355e-05,
      "loss": 0.0035,
      "step": 545
    },
    {
      "epoch": 1.1727078891257996,
      "grad_norm": 0.3219517767429352,
      "learning_rate": 3.8272921108742e-05,
      "loss": 0.0041,
      "step": 550
    },
    {
      "epoch": 1.183368869936034,
      "grad_norm": 0.032257046550512314,
      "learning_rate": 3.8166311300639665e-05,
      "loss": 0.0029,
      "step": 555
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 0.01878332532942295,
      "learning_rate": 3.805970149253731e-05,
      "loss": 0.0458,
      "step": 560
    },
    {
      "epoch": 1.2046908315565032,
      "grad_norm": 0.6530747413635254,
      "learning_rate": 3.7953091684434974e-05,
      "loss": 0.0459,
      "step": 565
    },
    {
      "epoch": 1.2153518123667377,
      "grad_norm": 0.03721063211560249,
      "learning_rate": 3.784648187633262e-05,
      "loss": 0.0027,
      "step": 570
    },
    {
      "epoch": 1.2260127931769722,
      "grad_norm": 0.018165690824389458,
      "learning_rate": 3.7739872068230284e-05,
      "loss": 0.0034,
      "step": 575
    },
    {
      "epoch": 1.236673773987207,
      "grad_norm": 0.025399623438715935,
      "learning_rate": 3.763326226012793e-05,
      "loss": 0.0024,
      "step": 580
    },
    {
      "epoch": 1.2473347547974414,
      "grad_norm": 0.01623561419546604,
      "learning_rate": 3.752665245202559e-05,
      "loss": 0.0024,
      "step": 585
    },
    {
      "epoch": 1.2579957356076759,
      "grad_norm": 0.1283871978521347,
      "learning_rate": 3.742004264392324e-05,
      "loss": 0.0021,
      "step": 590
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 0.027587631717324257,
      "learning_rate": 3.73134328358209e-05,
      "loss": 0.0958,
      "step": 595
    },
    {
      "epoch": 1.279317697228145,
      "grad_norm": 0.020109975710511208,
      "learning_rate": 3.720682302771855e-05,
      "loss": 0.0022,
      "step": 600
    },
    {
      "epoch": 1.2899786780383795,
      "grad_norm": 3.5448546409606934,
      "learning_rate": 3.710021321961621e-05,
      "loss": 0.0205,
      "step": 605
    },
    {
      "epoch": 1.3006396588486142,
      "grad_norm": 0.0627860501408577,
      "learning_rate": 3.699360341151386e-05,
      "loss": 0.0018,
      "step": 610
    },
    {
      "epoch": 1.3113006396588487,
      "grad_norm": 5.531160831451416,
      "learning_rate": 3.6886993603411515e-05,
      "loss": 0.029,
      "step": 615
    },
    {
      "epoch": 1.3219616204690832,
      "grad_norm": 0.2497386783361435,
      "learning_rate": 3.678038379530917e-05,
      "loss": 0.0045,
      "step": 620
    },
    {
      "epoch": 1.3326226012793176,
      "grad_norm": 1.4399714469909668,
      "learning_rate": 3.6673773987206824e-05,
      "loss": 0.0039,
      "step": 625
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 0.019945133477449417,
      "learning_rate": 3.656716417910448e-05,
      "loss": 0.0165,
      "step": 630
    },
    {
      "epoch": 1.3539445628997868,
      "grad_norm": 0.03204401955008507,
      "learning_rate": 3.6460554371002134e-05,
      "loss": 0.0309,
      "step": 635
    },
    {
      "epoch": 1.3646055437100213,
      "grad_norm": 0.02860732562839985,
      "learning_rate": 3.635394456289979e-05,
      "loss": 0.0393,
      "step": 640
    },
    {
      "epoch": 1.375266524520256,
      "grad_norm": 0.015683352947235107,
      "learning_rate": 3.624733475479744e-05,
      "loss": 0.008,
      "step": 645
    },
    {
      "epoch": 1.3859275053304905,
      "grad_norm": 4.806375503540039,
      "learning_rate": 3.61407249466951e-05,
      "loss": 0.0073,
      "step": 650
    },
    {
      "epoch": 1.396588486140725,
      "grad_norm": 1.2326937913894653,
      "learning_rate": 3.603411513859275e-05,
      "loss": 0.044,
      "step": 655
    },
    {
      "epoch": 1.4072494669509594,
      "grad_norm": 0.014211839064955711,
      "learning_rate": 3.592750533049041e-05,
      "loss": 0.0019,
      "step": 660
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 0.052608516067266464,
      "learning_rate": 3.582089552238806e-05,
      "loss": 0.0443,
      "step": 665
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.10361786186695099,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.0302,
      "step": 670
    },
    {
      "epoch": 1.439232409381663,
      "grad_norm": 0.014748088084161282,
      "learning_rate": 3.560767590618337e-05,
      "loss": 0.0758,
      "step": 675
    },
    {
      "epoch": 1.4498933901918978,
      "grad_norm": 0.536562442779541,
      "learning_rate": 3.5501066098081026e-05,
      "loss": 0.0108,
      "step": 680
    },
    {
      "epoch": 1.4605543710021323,
      "grad_norm": 1.8626726865768433,
      "learning_rate": 3.539445628997868e-05,
      "loss": 0.0321,
      "step": 685
    },
    {
      "epoch": 1.4712153518123667,
      "grad_norm": 0.017998071387410164,
      "learning_rate": 3.5287846481876336e-05,
      "loss": 0.0135,
      "step": 690
    },
    {
      "epoch": 1.4818763326226012,
      "grad_norm": 0.015704525634646416,
      "learning_rate": 3.518123667377399e-05,
      "loss": 0.002,
      "step": 695
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 7.488823413848877,
      "learning_rate": 3.5074626865671645e-05,
      "loss": 0.0513,
      "step": 700
    },
    {
      "epoch": 1.5031982942430704,
      "grad_norm": 0.01628691889345646,
      "learning_rate": 3.496801705756929e-05,
      "loss": 0.0206,
      "step": 705
    },
    {
      "epoch": 1.5138592750533049,
      "grad_norm": 0.19601155817508698,
      "learning_rate": 3.4861407249466955e-05,
      "loss": 0.0027,
      "step": 710
    },
    {
      "epoch": 1.5245202558635396,
      "grad_norm": 0.029433907940983772,
      "learning_rate": 3.47547974413646e-05,
      "loss": 0.0228,
      "step": 715
    },
    {
      "epoch": 1.535181236673774,
      "grad_norm": 0.013401803560554981,
      "learning_rate": 3.4648187633262264e-05,
      "loss": 0.0149,
      "step": 720
    },
    {
      "epoch": 1.5458422174840085,
      "grad_norm": 0.01640295423567295,
      "learning_rate": 3.454157782515991e-05,
      "loss": 0.002,
      "step": 725
    },
    {
      "epoch": 1.556503198294243,
      "grad_norm": 0.29793670773506165,
      "learning_rate": 3.4434968017057574e-05,
      "loss": 0.0216,
      "step": 730
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 12.721835136413574,
      "learning_rate": 3.432835820895522e-05,
      "loss": 0.0502,
      "step": 735
    },
    {
      "epoch": 1.5778251599147122,
      "grad_norm": 0.014269648119807243,
      "learning_rate": 3.422174840085288e-05,
      "loss": 0.0017,
      "step": 740
    },
    {
      "epoch": 1.5884861407249466,
      "grad_norm": 0.023622652515769005,
      "learning_rate": 3.411513859275053e-05,
      "loss": 0.005,
      "step": 745
    },
    {
      "epoch": 1.5991471215351813,
      "grad_norm": 0.013874546624720097,
      "learning_rate": 3.400852878464819e-05,
      "loss": 0.0017,
      "step": 750
    },
    {
      "epoch": 1.6098081023454158,
      "grad_norm": 0.016342196613550186,
      "learning_rate": 3.390191897654584e-05,
      "loss": 0.0017,
      "step": 755
    },
    {
      "epoch": 1.6204690831556503,
      "grad_norm": 0.024624589830636978,
      "learning_rate": 3.37953091684435e-05,
      "loss": 0.0473,
      "step": 760
    },
    {
      "epoch": 1.6311300639658848,
      "grad_norm": 0.0175275057554245,
      "learning_rate": 3.368869936034115e-05,
      "loss": 0.0231,
      "step": 765
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 0.08611518144607544,
      "learning_rate": 3.358208955223881e-05,
      "loss": 0.0441,
      "step": 770
    },
    {
      "epoch": 1.652452025586354,
      "grad_norm": 0.10698140412569046,
      "learning_rate": 3.347547974413646e-05,
      "loss": 0.0199,
      "step": 775
    },
    {
      "epoch": 1.6631130063965884,
      "grad_norm": 0.021477028727531433,
      "learning_rate": 3.336886993603412e-05,
      "loss": 0.0368,
      "step": 780
    },
    {
      "epoch": 1.6737739872068231,
      "grad_norm": 0.015199307352304459,
      "learning_rate": 3.326226012793177e-05,
      "loss": 0.0035,
      "step": 785
    },
    {
      "epoch": 1.6844349680170576,
      "grad_norm": 1.1268423795700073,
      "learning_rate": 3.3155650319829424e-05,
      "loss": 0.0022,
      "step": 790
    },
    {
      "epoch": 1.695095948827292,
      "grad_norm": 0.01745384931564331,
      "learning_rate": 3.304904051172708e-05,
      "loss": 0.0171,
      "step": 795
    },
    {
      "epoch": 1.7057569296375266,
      "grad_norm": 2.6572446823120117,
      "learning_rate": 3.294243070362473e-05,
      "loss": 0.0038,
      "step": 800
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 12.420907020568848,
      "learning_rate": 3.283582089552239e-05,
      "loss": 0.0153,
      "step": 805
    },
    {
      "epoch": 1.7270788912579957,
      "grad_norm": 0.48043471574783325,
      "learning_rate": 3.272921108742004e-05,
      "loss": 0.0153,
      "step": 810
    },
    {
      "epoch": 1.7377398720682304,
      "grad_norm": 0.08240003883838654,
      "learning_rate": 3.26226012793177e-05,
      "loss": 0.0027,
      "step": 815
    },
    {
      "epoch": 1.748400852878465,
      "grad_norm": 0.10537918657064438,
      "learning_rate": 3.251599147121535e-05,
      "loss": 0.0016,
      "step": 820
    },
    {
      "epoch": 1.7590618336886994,
      "grad_norm": 0.0860021635890007,
      "learning_rate": 3.240938166311301e-05,
      "loss": 0.0017,
      "step": 825
    },
    {
      "epoch": 1.7697228144989339,
      "grad_norm": 0.016195915639400482,
      "learning_rate": 3.230277185501066e-05,
      "loss": 0.0014,
      "step": 830
    },
    {
      "epoch": 1.7803837953091683,
      "grad_norm": 0.014040791429579258,
      "learning_rate": 3.2196162046908317e-05,
      "loss": 0.0093,
      "step": 835
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 0.018761547282338142,
      "learning_rate": 3.208955223880597e-05,
      "loss": 0.0015,
      "step": 840
    },
    {
      "epoch": 1.8017057569296375,
      "grad_norm": 10.541072845458984,
      "learning_rate": 3.1982942430703626e-05,
      "loss": 0.039,
      "step": 845
    },
    {
      "epoch": 1.8123667377398722,
      "grad_norm": 0.012204175814986229,
      "learning_rate": 3.187633262260128e-05,
      "loss": 0.0013,
      "step": 850
    },
    {
      "epoch": 1.8230277185501067,
      "grad_norm": 0.013040111400187016,
      "learning_rate": 3.1769722814498935e-05,
      "loss": 0.0337,
      "step": 855
    },
    {
      "epoch": 1.8336886993603412,
      "grad_norm": 0.038593094795942307,
      "learning_rate": 3.166311300639659e-05,
      "loss": 0.0016,
      "step": 860
    },
    {
      "epoch": 1.8443496801705757,
      "grad_norm": 0.013586961664259434,
      "learning_rate": 3.1556503198294245e-05,
      "loss": 0.0031,
      "step": 865
    },
    {
      "epoch": 1.8550106609808101,
      "grad_norm": 0.028022266924381256,
      "learning_rate": 3.14498933901919e-05,
      "loss": 0.0288,
      "step": 870
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 0.03445981815457344,
      "learning_rate": 3.1343283582089554e-05,
      "loss": 0.002,
      "step": 875
    },
    {
      "epoch": 1.8763326226012793,
      "grad_norm": 10.31727409362793,
      "learning_rate": 3.123667377398721e-05,
      "loss": 0.0813,
      "step": 880
    },
    {
      "epoch": 1.886993603411514,
      "grad_norm": 0.012935107573866844,
      "learning_rate": 3.1130063965884864e-05,
      "loss": 0.0173,
      "step": 885
    },
    {
      "epoch": 1.8976545842217485,
      "grad_norm": 4.1242547035217285,
      "learning_rate": 3.102345415778252e-05,
      "loss": 0.0324,
      "step": 890
    },
    {
      "epoch": 1.908315565031983,
      "grad_norm": 0.012002996169030666,
      "learning_rate": 3.0916844349680173e-05,
      "loss": 0.0015,
      "step": 895
    },
    {
      "epoch": 1.9189765458422174,
      "grad_norm": 0.10620272159576416,
      "learning_rate": 3.081023454157783e-05,
      "loss": 0.0095,
      "step": 900
    },
    {
      "epoch": 1.929637526652452,
      "grad_norm": 9.02376937866211,
      "learning_rate": 3.070362473347548e-05,
      "loss": 0.0099,
      "step": 905
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 0.016139136627316475,
      "learning_rate": 3.059701492537314e-05,
      "loss": 0.0095,
      "step": 910
    },
    {
      "epoch": 1.950959488272921,
      "grad_norm": 0.011453837156295776,
      "learning_rate": 3.0490405117270792e-05,
      "loss": 0.0294,
      "step": 915
    },
    {
      "epoch": 1.9616204690831558,
      "grad_norm": 0.011852367781102657,
      "learning_rate": 3.0383795309168444e-05,
      "loss": 0.0286,
      "step": 920
    },
    {
      "epoch": 1.9722814498933903,
      "grad_norm": 0.02334183268249035,
      "learning_rate": 3.0277185501066102e-05,
      "loss": 0.0096,
      "step": 925
    },
    {
      "epoch": 1.9829424307036247,
      "grad_norm": 0.010254183784127235,
      "learning_rate": 3.0170575692963753e-05,
      "loss": 0.0023,
      "step": 930
    },
    {
      "epoch": 1.9936034115138592,
      "grad_norm": 0.04167815297842026,
      "learning_rate": 3.006396588486141e-05,
      "loss": 0.0016,
      "step": 935
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9786666666666667,
      "eval_loss": 0.0762680321931839,
      "eval_runtime": 19.9661,
      "eval_samples_per_second": 187.819,
      "eval_steps_per_second": 5.91,
      "step": 938
    },
    {
      "epoch": 2.0042643923240937,
      "grad_norm": 0.019564837217330933,
      "learning_rate": 2.9957356076759063e-05,
      "loss": 0.0018,
      "step": 940
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 0.012155752629041672,
      "learning_rate": 2.9850746268656714e-05,
      "loss": 0.0013,
      "step": 945
    },
    {
      "epoch": 2.025586353944563,
      "grad_norm": 0.01120598055422306,
      "learning_rate": 2.9744136460554372e-05,
      "loss": 0.0014,
      "step": 950
    },
    {
      "epoch": 2.0362473347547976,
      "grad_norm": 0.010837409645318985,
      "learning_rate": 2.9637526652452023e-05,
      "loss": 0.0012,
      "step": 955
    },
    {
      "epoch": 2.046908315565032,
      "grad_norm": 0.11006344854831696,
      "learning_rate": 2.953091684434968e-05,
      "loss": 0.0014,
      "step": 960
    },
    {
      "epoch": 2.0575692963752665,
      "grad_norm": 0.010318398475646973,
      "learning_rate": 2.9424307036247333e-05,
      "loss": 0.0011,
      "step": 965
    },
    {
      "epoch": 2.068230277185501,
      "grad_norm": 0.47877737879753113,
      "learning_rate": 2.931769722814499e-05,
      "loss": 0.0023,
      "step": 970
    },
    {
      "epoch": 2.0788912579957355,
      "grad_norm": 0.8020907044410706,
      "learning_rate": 2.9211087420042642e-05,
      "loss": 0.0016,
      "step": 975
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 13.61236572265625,
      "learning_rate": 2.91044776119403e-05,
      "loss": 0.0186,
      "step": 980
    },
    {
      "epoch": 2.100213219616205,
      "grad_norm": 0.01077519915997982,
      "learning_rate": 2.8997867803837952e-05,
      "loss": 0.0013,
      "step": 985
    },
    {
      "epoch": 2.1108742004264394,
      "grad_norm": 0.013567216694355011,
      "learning_rate": 2.889125799573561e-05,
      "loss": 0.0017,
      "step": 990
    },
    {
      "epoch": 2.121535181236674,
      "grad_norm": 0.011567119508981705,
      "learning_rate": 2.878464818763326e-05,
      "loss": 0.0011,
      "step": 995
    },
    {
      "epoch": 2.1321961620469083,
      "grad_norm": 0.010799186304211617,
      "learning_rate": 2.867803837953092e-05,
      "loss": 0.0378,
      "step": 1000
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.009237868711352348,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.0025,
      "step": 1005
    },
    {
      "epoch": 2.1535181236673773,
      "grad_norm": 0.009166714735329151,
      "learning_rate": 2.846481876332623e-05,
      "loss": 0.0042,
      "step": 1010
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 0.008551483042538166,
      "learning_rate": 2.835820895522388e-05,
      "loss": 0.0014,
      "step": 1015
    },
    {
      "epoch": 2.1748400852878467,
      "grad_norm": 3.046365976333618,
      "learning_rate": 2.825159914712154e-05,
      "loss": 0.0046,
      "step": 1020
    },
    {
      "epoch": 2.185501066098081,
      "grad_norm": 0.010284727439284325,
      "learning_rate": 2.814498933901919e-05,
      "loss": 0.0348,
      "step": 1025
    },
    {
      "epoch": 2.1961620469083156,
      "grad_norm": 0.008667991496622562,
      "learning_rate": 2.8038379530916848e-05,
      "loss": 0.0013,
      "step": 1030
    },
    {
      "epoch": 2.20682302771855,
      "grad_norm": 0.008947396650910378,
      "learning_rate": 2.79317697228145e-05,
      "loss": 0.001,
      "step": 1035
    },
    {
      "epoch": 2.2174840085287846,
      "grad_norm": 0.008680208586156368,
      "learning_rate": 2.7825159914712157e-05,
      "loss": 0.0012,
      "step": 1040
    },
    {
      "epoch": 2.228144989339019,
      "grad_norm": 0.0086894566193223,
      "learning_rate": 2.771855010660981e-05,
      "loss": 0.0011,
      "step": 1045
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 0.517271876335144,
      "learning_rate": 2.7611940298507467e-05,
      "loss": 0.0022,
      "step": 1050
    },
    {
      "epoch": 2.2494669509594885,
      "grad_norm": 0.014216389507055283,
      "learning_rate": 2.7505330490405118e-05,
      "loss": 0.0025,
      "step": 1055
    },
    {
      "epoch": 2.260127931769723,
      "grad_norm": 20.532705307006836,
      "learning_rate": 2.7398720682302776e-05,
      "loss": 0.0235,
      "step": 1060
    },
    {
      "epoch": 2.2707889125799574,
      "grad_norm": 0.009220330975949764,
      "learning_rate": 2.7292110874200428e-05,
      "loss": 0.0497,
      "step": 1065
    },
    {
      "epoch": 2.281449893390192,
      "grad_norm": 0.30938297510147095,
      "learning_rate": 2.7185501066098086e-05,
      "loss": 0.0201,
      "step": 1070
    },
    {
      "epoch": 2.2921108742004264,
      "grad_norm": 0.013161562383174896,
      "learning_rate": 2.7078891257995737e-05,
      "loss": 0.0014,
      "step": 1075
    },
    {
      "epoch": 2.302771855010661,
      "grad_norm": 2.2976741790771484,
      "learning_rate": 2.6972281449893395e-05,
      "loss": 0.0424,
      "step": 1080
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 0.00953697320073843,
      "learning_rate": 2.6865671641791047e-05,
      "loss": 0.0155,
      "step": 1085
    },
    {
      "epoch": 2.3240938166311302,
      "grad_norm": 0.015441151335835457,
      "learning_rate": 2.6759061833688705e-05,
      "loss": 0.012,
      "step": 1090
    },
    {
      "epoch": 2.3347547974413647,
      "grad_norm": 0.012890947051346302,
      "learning_rate": 2.6652452025586356e-05,
      "loss": 0.0013,
      "step": 1095
    },
    {
      "epoch": 2.345415778251599,
      "grad_norm": 0.03095391020178795,
      "learning_rate": 2.6545842217484007e-05,
      "loss": 0.0011,
      "step": 1100
    },
    {
      "epoch": 2.3560767590618337,
      "grad_norm": 0.008807855658233166,
      "learning_rate": 2.6439232409381666e-05,
      "loss": 0.001,
      "step": 1105
    },
    {
      "epoch": 2.366737739872068,
      "grad_norm": 0.00888295378535986,
      "learning_rate": 2.6332622601279317e-05,
      "loss": 0.0444,
      "step": 1110
    },
    {
      "epoch": 2.3773987206823026,
      "grad_norm": 0.00881026592105627,
      "learning_rate": 2.6226012793176975e-05,
      "loss": 0.0271,
      "step": 1115
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 0.009612774476408958,
      "learning_rate": 2.6119402985074626e-05,
      "loss": 0.0014,
      "step": 1120
    },
    {
      "epoch": 2.398720682302772,
      "grad_norm": 0.012726151384413242,
      "learning_rate": 2.6012793176972285e-05,
      "loss": 0.001,
      "step": 1125
    },
    {
      "epoch": 2.4093816631130065,
      "grad_norm": 0.009462360292673111,
      "learning_rate": 2.5906183368869936e-05,
      "loss": 0.0016,
      "step": 1130
    },
    {
      "epoch": 2.420042643923241,
      "grad_norm": 0.009177952073514462,
      "learning_rate": 2.5799573560767594e-05,
      "loss": 0.0011,
      "step": 1135
    },
    {
      "epoch": 2.4307036247334755,
      "grad_norm": 0.0850144550204277,
      "learning_rate": 2.5692963752665245e-05,
      "loss": 0.0011,
      "step": 1140
    },
    {
      "epoch": 2.44136460554371,
      "grad_norm": 0.00982835702598095,
      "learning_rate": 2.5586353944562904e-05,
      "loss": 0.0454,
      "step": 1145
    },
    {
      "epoch": 2.4520255863539444,
      "grad_norm": 0.011294318363070488,
      "learning_rate": 2.5479744136460555e-05,
      "loss": 0.001,
      "step": 1150
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 0.009459765627980232,
      "learning_rate": 2.537313432835821e-05,
      "loss": 0.001,
      "step": 1155
    },
    {
      "epoch": 2.473347547974414,
      "grad_norm": 0.011211371049284935,
      "learning_rate": 2.5266524520255864e-05,
      "loss": 0.001,
      "step": 1160
    },
    {
      "epoch": 2.4840085287846483,
      "grad_norm": 0.008506160229444504,
      "learning_rate": 2.515991471215352e-05,
      "loss": 0.0014,
      "step": 1165
    },
    {
      "epoch": 2.4946695095948828,
      "grad_norm": 0.008855476975440979,
      "learning_rate": 2.5053304904051174e-05,
      "loss": 0.0177,
      "step": 1170
    },
    {
      "epoch": 2.5053304904051172,
      "grad_norm": 0.008432762697339058,
      "learning_rate": 2.494669509594883e-05,
      "loss": 0.0009,
      "step": 1175
    },
    {
      "epoch": 2.5159914712153517,
      "grad_norm": 0.011202191933989525,
      "learning_rate": 2.4840085287846483e-05,
      "loss": 0.0067,
      "step": 1180
    },
    {
      "epoch": 2.526652452025586,
      "grad_norm": 0.04894540086388588,
      "learning_rate": 2.4733475479744138e-05,
      "loss": 0.0011,
      "step": 1185
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 0.011918343603610992,
      "learning_rate": 2.4626865671641793e-05,
      "loss": 0.001,
      "step": 1190
    },
    {
      "epoch": 2.5479744136460556,
      "grad_norm": 0.01149666216224432,
      "learning_rate": 2.4520255863539444e-05,
      "loss": 0.0233,
      "step": 1195
    },
    {
      "epoch": 2.55863539445629,
      "grad_norm": 0.007119244430214167,
      "learning_rate": 2.44136460554371e-05,
      "loss": 0.0009,
      "step": 1200
    },
    {
      "epoch": 2.5692963752665245,
      "grad_norm": 0.3545883893966675,
      "learning_rate": 2.4307036247334754e-05,
      "loss": 0.001,
      "step": 1205
    },
    {
      "epoch": 2.579957356076759,
      "grad_norm": 1.0182620286941528,
      "learning_rate": 2.420042643923241e-05,
      "loss": 0.0017,
      "step": 1210
    },
    {
      "epoch": 2.5906183368869935,
      "grad_norm": 0.008021708577871323,
      "learning_rate": 2.4093816631130063e-05,
      "loss": 0.0009,
      "step": 1215
    },
    {
      "epoch": 2.6012793176972284,
      "grad_norm": 0.008916998282074928,
      "learning_rate": 2.3987206823027718e-05,
      "loss": 0.001,
      "step": 1220
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 1.931862235069275,
      "learning_rate": 2.3880597014925373e-05,
      "loss": 0.0329,
      "step": 1225
    },
    {
      "epoch": 2.6226012793176974,
      "grad_norm": 0.008216742426156998,
      "learning_rate": 2.3773987206823027e-05,
      "loss": 0.0009,
      "step": 1230
    },
    {
      "epoch": 2.633262260127932,
      "grad_norm": 0.1254955232143402,
      "learning_rate": 2.3667377398720682e-05,
      "loss": 0.0011,
      "step": 1235
    },
    {
      "epoch": 2.6439232409381663,
      "grad_norm": 0.008119771257042885,
      "learning_rate": 2.3560767590618337e-05,
      "loss": 0.0177,
      "step": 1240
    },
    {
      "epoch": 2.654584221748401,
      "grad_norm": 0.02235514111816883,
      "learning_rate": 2.345415778251599e-05,
      "loss": 0.001,
      "step": 1245
    },
    {
      "epoch": 2.6652452025586353,
      "grad_norm": 0.007437188643962145,
      "learning_rate": 2.3347547974413646e-05,
      "loss": 0.0009,
      "step": 1250
    },
    {
      "epoch": 2.6759061833688698,
      "grad_norm": 0.008978968486189842,
      "learning_rate": 2.32409381663113e-05,
      "loss": 0.0009,
      "step": 1255
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 0.035328932106494904,
      "learning_rate": 2.3134328358208956e-05,
      "loss": 0.0011,
      "step": 1260
    },
    {
      "epoch": 2.697228144989339,
      "grad_norm": 0.013226247392594814,
      "learning_rate": 2.302771855010661e-05,
      "loss": 0.0046,
      "step": 1265
    },
    {
      "epoch": 2.7078891257995736,
      "grad_norm": 0.008316855877637863,
      "learning_rate": 2.2921108742004265e-05,
      "loss": 0.001,
      "step": 1270
    },
    {
      "epoch": 2.718550106609808,
      "grad_norm": 0.00758874136954546,
      "learning_rate": 2.281449893390192e-05,
      "loss": 0.0058,
      "step": 1275
    },
    {
      "epoch": 2.7292110874200426,
      "grad_norm": 0.01189603190869093,
      "learning_rate": 2.2707889125799575e-05,
      "loss": 0.0009,
      "step": 1280
    },
    {
      "epoch": 2.739872068230277,
      "grad_norm": 0.009532790631055832,
      "learning_rate": 2.260127931769723e-05,
      "loss": 0.0009,
      "step": 1285
    },
    {
      "epoch": 2.750533049040512,
      "grad_norm": 0.019796231761574745,
      "learning_rate": 2.2494669509594884e-05,
      "loss": 0.0013,
      "step": 1290
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 0.009213976562023163,
      "learning_rate": 2.238805970149254e-05,
      "loss": 0.0077,
      "step": 1295
    },
    {
      "epoch": 2.771855010660981,
      "grad_norm": 0.01788158528506756,
      "learning_rate": 2.2281449893390194e-05,
      "loss": 0.0009,
      "step": 1300
    },
    {
      "epoch": 2.7825159914712154,
      "grad_norm": 0.007638400420546532,
      "learning_rate": 2.217484008528785e-05,
      "loss": 0.0021,
      "step": 1305
    },
    {
      "epoch": 2.79317697228145,
      "grad_norm": 0.007290796376764774,
      "learning_rate": 2.2068230277185503e-05,
      "loss": 0.0008,
      "step": 1310
    },
    {
      "epoch": 2.8038379530916844,
      "grad_norm": 0.007044069468975067,
      "learning_rate": 2.1961620469083158e-05,
      "loss": 0.0008,
      "step": 1315
    },
    {
      "epoch": 2.814498933901919,
      "grad_norm": 0.0070238965563476086,
      "learning_rate": 2.1855010660980813e-05,
      "loss": 0.0008,
      "step": 1320
    },
    {
      "epoch": 2.8251599147121533,
      "grad_norm": 0.007280718069523573,
      "learning_rate": 2.1748400852878467e-05,
      "loss": 0.005,
      "step": 1325
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 0.007587130181491375,
      "learning_rate": 2.164179104477612e-05,
      "loss": 0.0009,
      "step": 1330
    },
    {
      "epoch": 2.8464818763326227,
      "grad_norm": 0.007839350961148739,
      "learning_rate": 2.1535181236673773e-05,
      "loss": 0.0008,
      "step": 1335
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.007785406894981861,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.0008,
      "step": 1340
    },
    {
      "epoch": 2.8678038379530917,
      "grad_norm": 0.008242282085120678,
      "learning_rate": 2.1321961620469083e-05,
      "loss": 0.0225,
      "step": 1345
    },
    {
      "epoch": 2.878464818763326,
      "grad_norm": 0.036355577409267426,
      "learning_rate": 2.1215351812366738e-05,
      "loss": 0.0009,
      "step": 1350
    },
    {
      "epoch": 2.8891257995735606,
      "grad_norm": 0.021888498216867447,
      "learning_rate": 2.1108742004264392e-05,
      "loss": 0.0008,
      "step": 1355
    },
    {
      "epoch": 2.8997867803837956,
      "grad_norm": 0.014615601859986782,
      "learning_rate": 2.1002132196162047e-05,
      "loss": 0.0008,
      "step": 1360
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 0.03845909982919693,
      "learning_rate": 2.0895522388059702e-05,
      "loss": 0.0173,
      "step": 1365
    },
    {
      "epoch": 2.9211087420042645,
      "grad_norm": 0.008877183310687542,
      "learning_rate": 2.0788912579957357e-05,
      "loss": 0.0008,
      "step": 1370
    },
    {
      "epoch": 2.931769722814499,
      "grad_norm": 0.007351716980338097,
      "learning_rate": 2.068230277185501e-05,
      "loss": 0.0483,
      "step": 1375
    },
    {
      "epoch": 2.9424307036247335,
      "grad_norm": 0.00871069636195898,
      "learning_rate": 2.0575692963752666e-05,
      "loss": 0.0008,
      "step": 1380
    },
    {
      "epoch": 2.953091684434968,
      "grad_norm": 0.006365648470818996,
      "learning_rate": 2.046908315565032e-05,
      "loss": 0.0008,
      "step": 1385
    },
    {
      "epoch": 2.9637526652452024,
      "grad_norm": 0.008357345126569271,
      "learning_rate": 2.0362473347547976e-05,
      "loss": 0.046,
      "step": 1390
    },
    {
      "epoch": 2.974413646055437,
      "grad_norm": 0.007308799773454666,
      "learning_rate": 2.025586353944563e-05,
      "loss": 0.0458,
      "step": 1395
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 0.006932723335921764,
      "learning_rate": 2.0149253731343285e-05,
      "loss": 0.0008,
      "step": 1400
    },
    {
      "epoch": 2.9957356076759063,
      "grad_norm": 0.01034647598862648,
      "learning_rate": 2.004264392324094e-05,
      "loss": 0.0009,
      "step": 1405
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9794666666666667,
      "eval_loss": 0.09366792440414429,
      "eval_runtime": 20.1098,
      "eval_samples_per_second": 186.476,
      "eval_steps_per_second": 5.868,
      "step": 1407
    },
    {
      "epoch": 3.0063965884861408,
      "grad_norm": 0.013827606104314327,
      "learning_rate": 1.9936034115138594e-05,
      "loss": 0.0008,
      "step": 1410
    },
    {
      "epoch": 3.0170575692963753,
      "grad_norm": 0.02764023467898369,
      "learning_rate": 1.982942430703625e-05,
      "loss": 0.0008,
      "step": 1415
    },
    {
      "epoch": 3.0277185501066097,
      "grad_norm": 0.0076223681680858135,
      "learning_rate": 1.9722814498933904e-05,
      "loss": 0.0008,
      "step": 1420
    },
    {
      "epoch": 3.038379530916844,
      "grad_norm": 0.008151568472385406,
      "learning_rate": 1.961620469083156e-05,
      "loss": 0.0008,
      "step": 1425
    },
    {
      "epoch": 3.0490405117270787,
      "grad_norm": 0.015321053564548492,
      "learning_rate": 1.9509594882729213e-05,
      "loss": 0.0008,
      "step": 1430
    },
    {
      "epoch": 3.0597014925373136,
      "grad_norm": 0.0070846728049218655,
      "learning_rate": 1.9402985074626868e-05,
      "loss": 0.0008,
      "step": 1435
    },
    {
      "epoch": 3.070362473347548,
      "grad_norm": 0.006407500244677067,
      "learning_rate": 1.9296375266524523e-05,
      "loss": 0.0008,
      "step": 1440
    },
    {
      "epoch": 3.0810234541577826,
      "grad_norm": 0.007791341748088598,
      "learning_rate": 1.9189765458422178e-05,
      "loss": 0.0008,
      "step": 1445
    },
    {
      "epoch": 3.091684434968017,
      "grad_norm": 0.007972660474479198,
      "learning_rate": 1.9083155650319832e-05,
      "loss": 0.0008,
      "step": 1450
    },
    {
      "epoch": 3.1023454157782515,
      "grad_norm": 0.80890953540802,
      "learning_rate": 1.8976545842217487e-05,
      "loss": 0.0011,
      "step": 1455
    },
    {
      "epoch": 3.113006396588486,
      "grad_norm": 0.006264610216021538,
      "learning_rate": 1.8869936034115142e-05,
      "loss": 0.0034,
      "step": 1460
    },
    {
      "epoch": 3.1236673773987205,
      "grad_norm": 0.007522228639572859,
      "learning_rate": 1.8763326226012797e-05,
      "loss": 0.0008,
      "step": 1465
    },
    {
      "epoch": 3.1343283582089554,
      "grad_norm": 0.006920546758919954,
      "learning_rate": 1.865671641791045e-05,
      "loss": 0.0011,
      "step": 1470
    },
    {
      "epoch": 3.14498933901919,
      "grad_norm": 0.00800524652004242,
      "learning_rate": 1.8550106609808106e-05,
      "loss": 0.0132,
      "step": 1475
    },
    {
      "epoch": 3.1556503198294243,
      "grad_norm": 0.0068819657899439335,
      "learning_rate": 1.8443496801705757e-05,
      "loss": 0.0008,
      "step": 1480
    },
    {
      "epoch": 3.166311300639659,
      "grad_norm": 0.007231331430375576,
      "learning_rate": 1.8336886993603412e-05,
      "loss": 0.0087,
      "step": 1485
    },
    {
      "epoch": 3.1769722814498933,
      "grad_norm": 0.006309058517217636,
      "learning_rate": 1.8230277185501067e-05,
      "loss": 0.0007,
      "step": 1490
    },
    {
      "epoch": 3.1876332622601278,
      "grad_norm": 0.007262353785336018,
      "learning_rate": 1.812366737739872e-05,
      "loss": 0.0007,
      "step": 1495
    },
    {
      "epoch": 3.1982942430703627,
      "grad_norm": 9.726395606994629,
      "learning_rate": 1.8017057569296376e-05,
      "loss": 0.0044,
      "step": 1500
    },
    {
      "epoch": 3.208955223880597,
      "grad_norm": 0.009845616295933723,
      "learning_rate": 1.791044776119403e-05,
      "loss": 0.0008,
      "step": 1505
    },
    {
      "epoch": 3.2196162046908317,
      "grad_norm": 0.00690560694783926,
      "learning_rate": 1.7803837953091686e-05,
      "loss": 0.0007,
      "step": 1510
    },
    {
      "epoch": 3.230277185501066,
      "grad_norm": 0.008733430877327919,
      "learning_rate": 1.769722814498934e-05,
      "loss": 0.0007,
      "step": 1515
    },
    {
      "epoch": 3.2409381663113006,
      "grad_norm": 0.006656458135694265,
      "learning_rate": 1.7590618336886995e-05,
      "loss": 0.0007,
      "step": 1520
    },
    {
      "epoch": 3.251599147121535,
      "grad_norm": 0.006267220713198185,
      "learning_rate": 1.7484008528784647e-05,
      "loss": 0.0007,
      "step": 1525
    },
    {
      "epoch": 3.2622601279317696,
      "grad_norm": 0.0077451677061617374,
      "learning_rate": 1.73773987206823e-05,
      "loss": 0.0008,
      "step": 1530
    },
    {
      "epoch": 3.272921108742004,
      "grad_norm": 0.008423611521720886,
      "learning_rate": 1.7270788912579956e-05,
      "loss": 0.0007,
      "step": 1535
    },
    {
      "epoch": 3.283582089552239,
      "grad_norm": 0.006680880673229694,
      "learning_rate": 1.716417910447761e-05,
      "loss": 0.0008,
      "step": 1540
    },
    {
      "epoch": 3.2942430703624734,
      "grad_norm": 0.006745797581970692,
      "learning_rate": 1.7057569296375266e-05,
      "loss": 0.0007,
      "step": 1545
    },
    {
      "epoch": 3.304904051172708,
      "grad_norm": 0.006607946939766407,
      "learning_rate": 1.695095948827292e-05,
      "loss": 0.0451,
      "step": 1550
    },
    {
      "epoch": 3.3155650319829424,
      "grad_norm": 0.006421698722988367,
      "learning_rate": 1.6844349680170575e-05,
      "loss": 0.0007,
      "step": 1555
    },
    {
      "epoch": 3.326226012793177,
      "grad_norm": 0.0060383775271475315,
      "learning_rate": 1.673773987206823e-05,
      "loss": 0.001,
      "step": 1560
    },
    {
      "epoch": 3.3368869936034113,
      "grad_norm": 0.006617910694330931,
      "learning_rate": 1.6631130063965885e-05,
      "loss": 0.0007,
      "step": 1565
    },
    {
      "epoch": 3.3475479744136463,
      "grad_norm": 0.006189167965203524,
      "learning_rate": 1.652452025586354e-05,
      "loss": 0.0008,
      "step": 1570
    },
    {
      "epoch": 3.3582089552238807,
      "grad_norm": 1.9492641687393188,
      "learning_rate": 1.6417910447761194e-05,
      "loss": 0.0018,
      "step": 1575
    },
    {
      "epoch": 3.368869936034115,
      "grad_norm": 0.0074950577691197395,
      "learning_rate": 1.631130063965885e-05,
      "loss": 0.0007,
      "step": 1580
    },
    {
      "epoch": 3.3795309168443497,
      "grad_norm": 0.13474752008914948,
      "learning_rate": 1.6204690831556504e-05,
      "loss": 0.0008,
      "step": 1585
    },
    {
      "epoch": 3.390191897654584,
      "grad_norm": 0.008317158557474613,
      "learning_rate": 1.6098081023454158e-05,
      "loss": 0.0007,
      "step": 1590
    },
    {
      "epoch": 3.4008528784648187,
      "grad_norm": 0.006375185213983059,
      "learning_rate": 1.5991471215351813e-05,
      "loss": 0.0007,
      "step": 1595
    },
    {
      "epoch": 3.411513859275053,
      "grad_norm": 0.0072063482366502285,
      "learning_rate": 1.5884861407249468e-05,
      "loss": 0.0007,
      "step": 1600
    },
    {
      "epoch": 3.4221748400852876,
      "grad_norm": 0.006525915581732988,
      "learning_rate": 1.5778251599147122e-05,
      "loss": 0.0008,
      "step": 1605
    },
    {
      "epoch": 3.4328358208955225,
      "grad_norm": 0.011640388518571854,
      "learning_rate": 1.5671641791044777e-05,
      "loss": 0.0008,
      "step": 1610
    },
    {
      "epoch": 3.443496801705757,
      "grad_norm": 0.005998261272907257,
      "learning_rate": 1.5565031982942432e-05,
      "loss": 0.0007,
      "step": 1615
    },
    {
      "epoch": 3.4541577825159915,
      "grad_norm": 0.0062679629772901535,
      "learning_rate": 1.5458422174840087e-05,
      "loss": 0.0008,
      "step": 1620
    },
    {
      "epoch": 3.464818763326226,
      "grad_norm": 0.006080123595893383,
      "learning_rate": 1.535181236673774e-05,
      "loss": 0.0007,
      "step": 1625
    },
    {
      "epoch": 3.4754797441364604,
      "grad_norm": 0.006211758591234684,
      "learning_rate": 1.5245202558635396e-05,
      "loss": 0.0006,
      "step": 1630
    },
    {
      "epoch": 3.486140724946695,
      "grad_norm": 0.006736230570822954,
      "learning_rate": 1.5138592750533051e-05,
      "loss": 0.0424,
      "step": 1635
    },
    {
      "epoch": 3.49680170575693,
      "grad_norm": 0.006093813572078943,
      "learning_rate": 1.5031982942430706e-05,
      "loss": 0.0007,
      "step": 1640
    },
    {
      "epoch": 3.5074626865671643,
      "grad_norm": 0.006493781693279743,
      "learning_rate": 1.4925373134328357e-05,
      "loss": 0.0008,
      "step": 1645
    },
    {
      "epoch": 3.518123667377399,
      "grad_norm": 0.006536037195473909,
      "learning_rate": 1.4818763326226012e-05,
      "loss": 0.0007,
      "step": 1650
    },
    {
      "epoch": 3.5287846481876333,
      "grad_norm": 0.006346667651087046,
      "learning_rate": 1.4712153518123666e-05,
      "loss": 0.0007,
      "step": 1655
    },
    {
      "epoch": 3.5394456289978677,
      "grad_norm": 0.006048380862921476,
      "learning_rate": 1.4605543710021321e-05,
      "loss": 0.0031,
      "step": 1660
    },
    {
      "epoch": 3.550106609808102,
      "grad_norm": 0.007553269155323505,
      "learning_rate": 1.4498933901918976e-05,
      "loss": 0.0008,
      "step": 1665
    },
    {
      "epoch": 3.5607675906183367,
      "grad_norm": 0.029202468693256378,
      "learning_rate": 1.439232409381663e-05,
      "loss": 0.0011,
      "step": 1670
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.0067880465649068356,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.0009,
      "step": 1675
    },
    {
      "epoch": 3.582089552238806,
      "grad_norm": 0.009043511934578419,
      "learning_rate": 1.417910447761194e-05,
      "loss": 0.0007,
      "step": 1680
    },
    {
      "epoch": 3.5927505330490406,
      "grad_norm": 0.06923191994428635,
      "learning_rate": 1.4072494669509595e-05,
      "loss": 0.0007,
      "step": 1685
    },
    {
      "epoch": 3.603411513859275,
      "grad_norm": 0.005750897340476513,
      "learning_rate": 1.396588486140725e-05,
      "loss": 0.0007,
      "step": 1690
    },
    {
      "epoch": 3.6140724946695095,
      "grad_norm": 0.00608739722520113,
      "learning_rate": 1.3859275053304904e-05,
      "loss": 0.0008,
      "step": 1695
    },
    {
      "epoch": 3.624733475479744,
      "grad_norm": 0.006191660650074482,
      "learning_rate": 1.3752665245202559e-05,
      "loss": 0.0008,
      "step": 1700
    },
    {
      "epoch": 3.635394456289979,
      "grad_norm": 0.02501906268298626,
      "learning_rate": 1.3646055437100214e-05,
      "loss": 0.0007,
      "step": 1705
    },
    {
      "epoch": 3.6460554371002134,
      "grad_norm": 0.0060061002150177956,
      "learning_rate": 1.3539445628997869e-05,
      "loss": 0.0007,
      "step": 1710
    },
    {
      "epoch": 3.656716417910448,
      "grad_norm": 0.005908344406634569,
      "learning_rate": 1.3432835820895523e-05,
      "loss": 0.0007,
      "step": 1715
    },
    {
      "epoch": 3.6673773987206824,
      "grad_norm": 0.014983762986958027,
      "learning_rate": 1.3326226012793178e-05,
      "loss": 0.0007,
      "step": 1720
    },
    {
      "epoch": 3.678038379530917,
      "grad_norm": 0.006315229926258326,
      "learning_rate": 1.3219616204690833e-05,
      "loss": 0.0447,
      "step": 1725
    },
    {
      "epoch": 3.6886993603411513,
      "grad_norm": 0.006016280502080917,
      "learning_rate": 1.3113006396588488e-05,
      "loss": 0.0007,
      "step": 1730
    },
    {
      "epoch": 3.699360341151386,
      "grad_norm": 12.3385009765625,
      "learning_rate": 1.3006396588486142e-05,
      "loss": 0.0229,
      "step": 1735
    },
    {
      "epoch": 3.7100213219616203,
      "grad_norm": 0.006276819854974747,
      "learning_rate": 1.2899786780383797e-05,
      "loss": 0.0007,
      "step": 1740
    },
    {
      "epoch": 3.7206823027718547,
      "grad_norm": 0.00832104030996561,
      "learning_rate": 1.2793176972281452e-05,
      "loss": 0.0007,
      "step": 1745
    },
    {
      "epoch": 3.7313432835820897,
      "grad_norm": 0.006885137874633074,
      "learning_rate": 1.2686567164179105e-05,
      "loss": 0.0008,
      "step": 1750
    },
    {
      "epoch": 3.742004264392324,
      "grad_norm": 0.005737548694014549,
      "learning_rate": 1.257995735607676e-05,
      "loss": 0.0008,
      "step": 1755
    },
    {
      "epoch": 3.7526652452025586,
      "grad_norm": 0.01088783610612154,
      "learning_rate": 1.2473347547974414e-05,
      "loss": 0.0007,
      "step": 1760
    },
    {
      "epoch": 3.763326226012793,
      "grad_norm": 0.007666490972042084,
      "learning_rate": 1.2366737739872069e-05,
      "loss": 0.0007,
      "step": 1765
    },
    {
      "epoch": 3.7739872068230276,
      "grad_norm": 0.005950309801846743,
      "learning_rate": 1.2260127931769722e-05,
      "loss": 0.0007,
      "step": 1770
    },
    {
      "epoch": 3.7846481876332625,
      "grad_norm": 0.006053774151951075,
      "learning_rate": 1.2153518123667377e-05,
      "loss": 0.0007,
      "step": 1775
    },
    {
      "epoch": 3.795309168443497,
      "grad_norm": 0.006015722639858723,
      "learning_rate": 1.2046908315565032e-05,
      "loss": 0.0007,
      "step": 1780
    },
    {
      "epoch": 3.8059701492537314,
      "grad_norm": 0.007579673547297716,
      "learning_rate": 1.1940298507462686e-05,
      "loss": 0.0007,
      "step": 1785
    },
    {
      "epoch": 3.816631130063966,
      "grad_norm": 0.0063930051401257515,
      "learning_rate": 1.1833688699360341e-05,
      "loss": 0.0007,
      "step": 1790
    },
    {
      "epoch": 3.8272921108742004,
      "grad_norm": 0.006336760241538286,
      "learning_rate": 1.1727078891257996e-05,
      "loss": 0.0044,
      "step": 1795
    },
    {
      "epoch": 3.837953091684435,
      "grad_norm": 0.00632160110399127,
      "learning_rate": 1.162046908315565e-05,
      "loss": 0.0007,
      "step": 1800
    },
    {
      "epoch": 3.8486140724946694,
      "grad_norm": 0.00951367150992155,
      "learning_rate": 1.1513859275053305e-05,
      "loss": 0.0007,
      "step": 1805
    },
    {
      "epoch": 3.859275053304904,
      "grad_norm": 0.007132424972951412,
      "learning_rate": 1.140724946695096e-05,
      "loss": 0.0007,
      "step": 1810
    },
    {
      "epoch": 3.8699360341151388,
      "grad_norm": 0.006522058043628931,
      "learning_rate": 1.1300639658848615e-05,
      "loss": 0.0007,
      "step": 1815
    },
    {
      "epoch": 3.8805970149253732,
      "grad_norm": 0.00559121323749423,
      "learning_rate": 1.119402985074627e-05,
      "loss": 0.0006,
      "step": 1820
    },
    {
      "epoch": 3.8912579957356077,
      "grad_norm": 0.006240976043045521,
      "learning_rate": 1.1087420042643924e-05,
      "loss": 0.0007,
      "step": 1825
    },
    {
      "epoch": 3.901918976545842,
      "grad_norm": 0.0062920828349888325,
      "learning_rate": 1.0980810234541579e-05,
      "loss": 0.0006,
      "step": 1830
    },
    {
      "epoch": 3.9125799573560767,
      "grad_norm": 0.008827520534396172,
      "learning_rate": 1.0874200426439234e-05,
      "loss": 0.0007,
      "step": 1835
    },
    {
      "epoch": 3.923240938166311,
      "grad_norm": 0.009680266492068768,
      "learning_rate": 1.0767590618336887e-05,
      "loss": 0.0007,
      "step": 1840
    },
    {
      "epoch": 3.933901918976546,
      "grad_norm": 12.629690170288086,
      "learning_rate": 1.0660980810234541e-05,
      "loss": 0.0051,
      "step": 1845
    },
    {
      "epoch": 3.9445628997867805,
      "grad_norm": 0.0053851669654250145,
      "learning_rate": 1.0554371002132196e-05,
      "loss": 0.0006,
      "step": 1850
    },
    {
      "epoch": 3.955223880597015,
      "grad_norm": 0.005973598454147577,
      "learning_rate": 1.0447761194029851e-05,
      "loss": 0.0007,
      "step": 1855
    },
    {
      "epoch": 3.9658848614072495,
      "grad_norm": 0.006448433268815279,
      "learning_rate": 1.0341151385927506e-05,
      "loss": 0.0006,
      "step": 1860
    },
    {
      "epoch": 3.976545842217484,
      "grad_norm": 0.01381625048816204,
      "learning_rate": 1.023454157782516e-05,
      "loss": 0.0007,
      "step": 1865
    },
    {
      "epoch": 3.9872068230277184,
      "grad_norm": 0.00548245944082737,
      "learning_rate": 1.0127931769722815e-05,
      "loss": 0.0007,
      "step": 1870
    },
    {
      "epoch": 3.997867803837953,
      "grad_norm": 0.006060993764549494,
      "learning_rate": 1.002132196162047e-05,
      "loss": 0.0007,
      "step": 1875
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9784,
      "eval_loss": 0.10362313687801361,
      "eval_runtime": 20.0006,
      "eval_samples_per_second": 187.494,
      "eval_steps_per_second": 5.9,
      "step": 1876
    },
    {
      "epoch": 4.008528784648187,
      "grad_norm": 0.006188530940562487,
      "learning_rate": 9.914712153518125e-06,
      "loss": 0.0006,
      "step": 1880
    },
    {
      "epoch": 4.019189765458422,
      "grad_norm": 0.008521688170731068,
      "learning_rate": 9.80810234541578e-06,
      "loss": 0.0007,
      "step": 1885
    },
    {
      "epoch": 4.029850746268656,
      "grad_norm": 0.009525089524686337,
      "learning_rate": 9.701492537313434e-06,
      "loss": 0.0006,
      "step": 1890
    },
    {
      "epoch": 4.040511727078891,
      "grad_norm": 0.005723011679947376,
      "learning_rate": 9.594882729211089e-06,
      "loss": 0.0006,
      "step": 1895
    },
    {
      "epoch": 4.051172707889126,
      "grad_norm": 0.012250931933522224,
      "learning_rate": 9.488272921108744e-06,
      "loss": 0.0008,
      "step": 1900
    },
    {
      "epoch": 4.061833688699361,
      "grad_norm": 0.005616276524960995,
      "learning_rate": 9.381663113006398e-06,
      "loss": 0.0006,
      "step": 1905
    },
    {
      "epoch": 4.072494669509595,
      "grad_norm": 0.0070989904925227165,
      "learning_rate": 9.275053304904053e-06,
      "loss": 0.0007,
      "step": 1910
    },
    {
      "epoch": 4.08315565031983,
      "grad_norm": 0.005373702850192785,
      "learning_rate": 9.168443496801706e-06,
      "loss": 0.0006,
      "step": 1915
    },
    {
      "epoch": 4.093816631130064,
      "grad_norm": 31.406415939331055,
      "learning_rate": 9.06183368869936e-06,
      "loss": 0.0094,
      "step": 1920
    },
    {
      "epoch": 4.104477611940299,
      "grad_norm": 0.005601997021585703,
      "learning_rate": 8.955223880597016e-06,
      "loss": 0.0008,
      "step": 1925
    },
    {
      "epoch": 4.115138592750533,
      "grad_norm": 0.005039342679083347,
      "learning_rate": 8.84861407249467e-06,
      "loss": 0.0006,
      "step": 1930
    },
    {
      "epoch": 4.1257995735607675,
      "grad_norm": 0.00596946757286787,
      "learning_rate": 8.742004264392323e-06,
      "loss": 0.0006,
      "step": 1935
    },
    {
      "epoch": 4.136460554371002,
      "grad_norm": 0.005609456449747086,
      "learning_rate": 8.635394456289978e-06,
      "loss": 0.0006,
      "step": 1940
    },
    {
      "epoch": 4.1471215351812365,
      "grad_norm": 0.005230541806668043,
      "learning_rate": 8.528784648187633e-06,
      "loss": 0.0006,
      "step": 1945
    },
    {
      "epoch": 4.157782515991471,
      "grad_norm": 0.005722707603126764,
      "learning_rate": 8.422174840085288e-06,
      "loss": 0.0006,
      "step": 1950
    },
    {
      "epoch": 4.1684434968017055,
      "grad_norm": 0.007707034703344107,
      "learning_rate": 8.315565031982942e-06,
      "loss": 0.0006,
      "step": 1955
    },
    {
      "epoch": 4.17910447761194,
      "grad_norm": 0.005359646398574114,
      "learning_rate": 8.208955223880597e-06,
      "loss": 0.0007,
      "step": 1960
    },
    {
      "epoch": 4.189765458422174,
      "grad_norm": 0.005356407258659601,
      "learning_rate": 8.102345415778252e-06,
      "loss": 0.0006,
      "step": 1965
    },
    {
      "epoch": 4.20042643923241,
      "grad_norm": 0.005907480139285326,
      "learning_rate": 7.995735607675907e-06,
      "loss": 0.0006,
      "step": 1970
    },
    {
      "epoch": 4.211087420042644,
      "grad_norm": 0.005819487851113081,
      "learning_rate": 7.889125799573561e-06,
      "loss": 0.0006,
      "step": 1975
    },
    {
      "epoch": 4.221748400852879,
      "grad_norm": 0.005277236923575401,
      "learning_rate": 7.782515991471216e-06,
      "loss": 0.0006,
      "step": 1980
    },
    {
      "epoch": 4.232409381663113,
      "grad_norm": 10.111306190490723,
      "learning_rate": 7.67590618336887e-06,
      "loss": 0.0032,
      "step": 1985
    },
    {
      "epoch": 4.243070362473348,
      "grad_norm": 0.021291470155119896,
      "learning_rate": 7.5692963752665255e-06,
      "loss": 0.0006,
      "step": 1990
    },
    {
      "epoch": 4.253731343283582,
      "grad_norm": 0.0051004705019295216,
      "learning_rate": 7.4626865671641785e-06,
      "loss": 0.0006,
      "step": 1995
    },
    {
      "epoch": 4.264392324093817,
      "grad_norm": 0.005545586347579956,
      "learning_rate": 7.356076759061833e-06,
      "loss": 0.0006,
      "step": 2000
    },
    {
      "epoch": 4.275053304904051,
      "grad_norm": 0.005330509506165981,
      "learning_rate": 7.249466950959488e-06,
      "loss": 0.0032,
      "step": 2005
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.006067847833037376,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.0006,
      "step": 2010
    },
    {
      "epoch": 4.29637526652452,
      "grad_norm": 0.005352153442800045,
      "learning_rate": 7.0362473347547975e-06,
      "loss": 0.0452,
      "step": 2015
    },
    {
      "epoch": 4.3070362473347545,
      "grad_norm": 0.006050963886082172,
      "learning_rate": 6.929637526652452e-06,
      "loss": 0.0006,
      "step": 2020
    },
    {
      "epoch": 4.317697228144989,
      "grad_norm": 0.005619670730084181,
      "learning_rate": 6.823027718550107e-06,
      "loss": 0.0006,
      "step": 2025
    },
    {
      "epoch": 4.3283582089552235,
      "grad_norm": 0.004804551601409912,
      "learning_rate": 6.716417910447762e-06,
      "loss": 0.0007,
      "step": 2030
    },
    {
      "epoch": 4.339019189765459,
      "grad_norm": 0.005901563446968794,
      "learning_rate": 6.609808102345416e-06,
      "loss": 0.0006,
      "step": 2035
    },
    {
      "epoch": 4.349680170575693,
      "grad_norm": 0.005345991812646389,
      "learning_rate": 6.503198294243071e-06,
      "loss": 0.0006,
      "step": 2040
    },
    {
      "epoch": 4.360341151385928,
      "grad_norm": 0.00534421019256115,
      "learning_rate": 6.396588486140726e-06,
      "loss": 0.0006,
      "step": 2045
    },
    {
      "epoch": 4.371002132196162,
      "grad_norm": 0.006262521725147963,
      "learning_rate": 6.28997867803838e-06,
      "loss": 0.0006,
      "step": 2050
    },
    {
      "epoch": 4.381663113006397,
      "grad_norm": 0.005501191597431898,
      "learning_rate": 6.1833688699360345e-06,
      "loss": 0.0006,
      "step": 2055
    },
    {
      "epoch": 4.392324093816631,
      "grad_norm": 0.0064369793981313705,
      "learning_rate": 6.076759061833688e-06,
      "loss": 0.0006,
      "step": 2060
    },
    {
      "epoch": 4.402985074626866,
      "grad_norm": 0.005606994032859802,
      "learning_rate": 5.970149253731343e-06,
      "loss": 0.0007,
      "step": 2065
    },
    {
      "epoch": 4.4136460554371,
      "grad_norm": 0.005272231996059418,
      "learning_rate": 5.863539445628998e-06,
      "loss": 0.0006,
      "step": 2070
    },
    {
      "epoch": 4.424307036247335,
      "grad_norm": 0.006807511206716299,
      "learning_rate": 5.756929637526653e-06,
      "loss": 0.0006,
      "step": 2075
    },
    {
      "epoch": 4.434968017057569,
      "grad_norm": 0.004948188550770283,
      "learning_rate": 5.650319829424307e-06,
      "loss": 0.0006,
      "step": 2080
    },
    {
      "epoch": 4.445628997867804,
      "grad_norm": 0.009398166090250015,
      "learning_rate": 5.543710021321962e-06,
      "loss": 0.0006,
      "step": 2085
    },
    {
      "epoch": 4.456289978678038,
      "grad_norm": 0.005757671780884266,
      "learning_rate": 5.437100213219617e-06,
      "loss": 0.0006,
      "step": 2090
    },
    {
      "epoch": 4.466950959488273,
      "grad_norm": 0.006003624759614468,
      "learning_rate": 5.330490405117271e-06,
      "loss": 0.0006,
      "step": 2095
    },
    {
      "epoch": 4.477611940298507,
      "grad_norm": 0.00876291561871767,
      "learning_rate": 5.2238805970149255e-06,
      "loss": 0.0006,
      "step": 2100
    },
    {
      "epoch": 4.4882729211087415,
      "grad_norm": 0.005351038184016943,
      "learning_rate": 5.11727078891258e-06,
      "loss": 0.0006,
      "step": 2105
    },
    {
      "epoch": 4.498933901918977,
      "grad_norm": 0.005153369624167681,
      "learning_rate": 5.010660980810235e-06,
      "loss": 0.0006,
      "step": 2110
    },
    {
      "epoch": 4.509594882729211,
      "grad_norm": 0.00548686645925045,
      "learning_rate": 4.90405117270789e-06,
      "loss": 0.0006,
      "step": 2115
    },
    {
      "epoch": 4.520255863539446,
      "grad_norm": 0.00587236974388361,
      "learning_rate": 4.797441364605544e-06,
      "loss": 0.0006,
      "step": 2120
    },
    {
      "epoch": 4.53091684434968,
      "grad_norm": 0.006007646210491657,
      "learning_rate": 4.690831556503199e-06,
      "loss": 0.0006,
      "step": 2125
    },
    {
      "epoch": 4.541577825159915,
      "grad_norm": 0.0054948534816503525,
      "learning_rate": 4.584221748400853e-06,
      "loss": 0.0006,
      "step": 2130
    },
    {
      "epoch": 4.552238805970149,
      "grad_norm": 0.005331332329660654,
      "learning_rate": 4.477611940298508e-06,
      "loss": 0.0006,
      "step": 2135
    },
    {
      "epoch": 4.562899786780384,
      "grad_norm": 0.005566426552832127,
      "learning_rate": 4.371002132196162e-06,
      "loss": 0.0006,
      "step": 2140
    },
    {
      "epoch": 4.573560767590618,
      "grad_norm": 0.01621875911951065,
      "learning_rate": 4.264392324093816e-06,
      "loss": 0.0007,
      "step": 2145
    },
    {
      "epoch": 4.584221748400853,
      "grad_norm": 0.005645028781145811,
      "learning_rate": 4.157782515991471e-06,
      "loss": 0.0006,
      "step": 2150
    },
    {
      "epoch": 4.594882729211087,
      "grad_norm": 0.006122419144958258,
      "learning_rate": 4.051172707889126e-06,
      "loss": 0.0006,
      "step": 2155
    },
    {
      "epoch": 4.605543710021322,
      "grad_norm": 0.005490147043019533,
      "learning_rate": 3.944562899786781e-06,
      "loss": 0.0006,
      "step": 2160
    },
    {
      "epoch": 4.616204690831556,
      "grad_norm": 0.0057923984713852406,
      "learning_rate": 3.837953091684435e-06,
      "loss": 0.0006,
      "step": 2165
    },
    {
      "epoch": 4.6268656716417915,
      "grad_norm": 0.004970306996256113,
      "learning_rate": 3.7313432835820893e-06,
      "loss": 0.0006,
      "step": 2170
    },
    {
      "epoch": 4.637526652452026,
      "grad_norm": 0.005383606534451246,
      "learning_rate": 3.624733475479744e-06,
      "loss": 0.0006,
      "step": 2175
    },
    {
      "epoch": 4.6481876332622605,
      "grad_norm": 0.005078316666185856,
      "learning_rate": 3.5181236673773987e-06,
      "loss": 0.0006,
      "step": 2180
    },
    {
      "epoch": 4.658848614072495,
      "grad_norm": 0.005838215351104736,
      "learning_rate": 3.4115138592750535e-06,
      "loss": 0.0006,
      "step": 2185
    },
    {
      "epoch": 4.669509594882729,
      "grad_norm": 0.005395676475018263,
      "learning_rate": 3.304904051172708e-06,
      "loss": 0.0006,
      "step": 2190
    },
    {
      "epoch": 4.680170575692964,
      "grad_norm": 0.0054429667070508,
      "learning_rate": 3.198294243070363e-06,
      "loss": 0.0006,
      "step": 2195
    },
    {
      "epoch": 4.690831556503198,
      "grad_norm": 0.008534586988389492,
      "learning_rate": 3.0916844349680173e-06,
      "loss": 0.0006,
      "step": 2200
    },
    {
      "epoch": 4.701492537313433,
      "grad_norm": 0.00460290489718318,
      "learning_rate": 2.9850746268656716e-06,
      "loss": 0.0006,
      "step": 2205
    },
    {
      "epoch": 4.712153518123667,
      "grad_norm": 0.005978610832244158,
      "learning_rate": 2.8784648187633263e-06,
      "loss": 0.0006,
      "step": 2210
    },
    {
      "epoch": 4.722814498933902,
      "grad_norm": 0.005226532928645611,
      "learning_rate": 2.771855010660981e-06,
      "loss": 0.0005,
      "step": 2215
    },
    {
      "epoch": 4.733475479744136,
      "grad_norm": 0.005466914270073175,
      "learning_rate": 2.6652452025586354e-06,
      "loss": 0.0006,
      "step": 2220
    },
    {
      "epoch": 4.744136460554371,
      "grad_norm": 0.005896831396967173,
      "learning_rate": 2.55863539445629e-06,
      "loss": 0.0006,
      "step": 2225
    },
    {
      "epoch": 4.754797441364605,
      "grad_norm": 0.0056927744299173355,
      "learning_rate": 2.452025586353945e-06,
      "loss": 0.0006,
      "step": 2230
    },
    {
      "epoch": 4.76545842217484,
      "grad_norm": 0.006180021911859512,
      "learning_rate": 2.3454157782515996e-06,
      "loss": 0.0006,
      "step": 2235
    },
    {
      "epoch": 4.776119402985074,
      "grad_norm": 0.005636970512568951,
      "learning_rate": 2.238805970149254e-06,
      "loss": 0.0006,
      "step": 2240
    },
    {
      "epoch": 4.786780383795309,
      "grad_norm": 0.00810902751982212,
      "learning_rate": 2.132196162046908e-06,
      "loss": 0.0005,
      "step": 2245
    },
    {
      "epoch": 4.797441364605544,
      "grad_norm": 0.005615860689431429,
      "learning_rate": 2.025586353944563e-06,
      "loss": 0.0006,
      "step": 2250
    },
    {
      "epoch": 4.8081023454157785,
      "grad_norm": 0.02348656766116619,
      "learning_rate": 1.9189765458422177e-06,
      "loss": 0.0007,
      "step": 2255
    },
    {
      "epoch": 4.818763326226013,
      "grad_norm": 0.0202944315969944,
      "learning_rate": 1.812366737739872e-06,
      "loss": 0.0006,
      "step": 2260
    },
    {
      "epoch": 4.8294243070362475,
      "grad_norm": 0.006284301169216633,
      "learning_rate": 1.7057569296375267e-06,
      "loss": 0.0006,
      "step": 2265
    },
    {
      "epoch": 4.840085287846482,
      "grad_norm": 0.01675177924335003,
      "learning_rate": 1.5991471215351815e-06,
      "loss": 0.0007,
      "step": 2270
    },
    {
      "epoch": 4.850746268656716,
      "grad_norm": 0.005214667413383722,
      "learning_rate": 1.4925373134328358e-06,
      "loss": 0.0005,
      "step": 2275
    },
    {
      "epoch": 4.861407249466951,
      "grad_norm": 0.005733064375817776,
      "learning_rate": 1.3859275053304905e-06,
      "loss": 0.0006,
      "step": 2280
    },
    {
      "epoch": 4.872068230277185,
      "grad_norm": 0.004514030180871487,
      "learning_rate": 1.279317697228145e-06,
      "loss": 0.0005,
      "step": 2285
    },
    {
      "epoch": 4.88272921108742,
      "grad_norm": 0.005337245762348175,
      "learning_rate": 1.1727078891257998e-06,
      "loss": 0.0006,
      "step": 2290
    },
    {
      "epoch": 4.893390191897654,
      "grad_norm": 0.005524416454136372,
      "learning_rate": 1.066098081023454e-06,
      "loss": 0.0006,
      "step": 2295
    },
    {
      "epoch": 4.904051172707889,
      "grad_norm": 0.00571380415931344,
      "learning_rate": 9.594882729211088e-07,
      "loss": 0.0006,
      "step": 2300
    },
    {
      "epoch": 4.914712153518123,
      "grad_norm": 0.00518675334751606,
      "learning_rate": 8.528784648187634e-07,
      "loss": 0.0007,
      "step": 2305
    },
    {
      "epoch": 4.925373134328359,
      "grad_norm": 0.005830350797623396,
      "learning_rate": 7.462686567164179e-07,
      "loss": 0.0006,
      "step": 2310
    },
    {
      "epoch": 4.936034115138593,
      "grad_norm": 0.005985878873616457,
      "learning_rate": 6.396588486140725e-07,
      "loss": 0.0006,
      "step": 2315
    },
    {
      "epoch": 4.946695095948828,
      "grad_norm": 0.00588690722361207,
      "learning_rate": 5.33049040511727e-07,
      "loss": 0.0006,
      "step": 2320
    },
    {
      "epoch": 4.957356076759062,
      "grad_norm": 0.005412915721535683,
      "learning_rate": 4.264392324093817e-07,
      "loss": 0.0006,
      "step": 2325
    },
    {
      "epoch": 4.968017057569297,
      "grad_norm": 0.005403020419180393,
      "learning_rate": 3.1982942430703626e-07,
      "loss": 0.0006,
      "step": 2330
    },
    {
      "epoch": 4.978678038379531,
      "grad_norm": 0.006161978933960199,
      "learning_rate": 2.1321961620469084e-07,
      "loss": 0.0006,
      "step": 2335
    },
    {
      "epoch": 4.9893390191897655,
      "grad_norm": 0.005091626197099686,
      "learning_rate": 1.0660980810234542e-07,
      "loss": 0.0006,
      "step": 2340
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.011390967294573784,
      "learning_rate": 0.0,
      "loss": 0.0006,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9816,
      "eval_loss": 0.08941872417926788,
      "eval_runtime": 20.0658,
      "eval_samples_per_second": 186.885,
      "eval_steps_per_second": 5.881,
      "step": 2345
    },
    {
      "epoch": 5.0,
      "step": 2345,
      "total_flos": 5.8118992210944e+18,
      "train_loss": 0.016130604191438982,
      "train_runtime": 880.5301,
      "train_samples_per_second": 85.176,
      "train_steps_per_second": 2.663
    }
  ],
  "logging_steps": 5,
  "max_steps": 2345,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.8118992210944e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}