{
  "best_metric": 0.7930574098798397,
  "best_model_checkpoint": "videomae-base-finetuned-crema-d8/checkpoint-5952",
  "epoch": 7.123823924731183,
  "eval_steps": 500,
  "global_step": 5952,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 5.881244659423828,
      "learning_rate": 8.389261744966443e-07,
      "loss": 1.8522,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.760950088500977,
      "learning_rate": 1.6778523489932886e-06,
      "loss": 1.8565,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.432248115539551,
      "learning_rate": 2.5167785234899326e-06,
      "loss": 1.908,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.651343822479248,
      "learning_rate": 3.3557046979865773e-06,
      "loss": 1.8041,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.095512390136719,
      "learning_rate": 4.194630872483222e-06,
      "loss": 1.812,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.4584431648254395,
      "learning_rate": 5.033557046979865e-06,
      "loss": 1.8109,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.230469703674316,
      "learning_rate": 5.87248322147651e-06,
      "loss": 1.7805,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.99364185333252,
      "learning_rate": 6.7114093959731546e-06,
      "loss": 1.8088,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.67277717590332,
      "learning_rate": 7.5503355704698e-06,
      "loss": 1.8313,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.8557710647583,
      "learning_rate": 8.389261744966444e-06,
      "loss": 1.787,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.0035834312438965,
      "learning_rate": 9.228187919463089e-06,
      "loss": 1.8086,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.592473983764648,
      "learning_rate": 1.006711409395973e-05,
      "loss": 1.8229,
      "step": 120
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.967762470245361,
      "learning_rate": 1.0906040268456376e-05,
      "loss": 1.8,
      "step": 130
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.766205787658691,
      "learning_rate": 1.174496644295302e-05,
      "loss": 1.8001,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.129546165466309,
      "learning_rate": 1.2583892617449666e-05,
      "loss": 1.8243,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.669356346130371,
      "learning_rate": 1.3422818791946309e-05,
      "loss": 1.7965,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.489925384521484,
      "learning_rate": 1.4261744966442953e-05,
      "loss": 1.7891,
      "step": 170
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.921316146850586,
      "learning_rate": 1.51006711409396e-05,
      "loss": 1.8401,
      "step": 180
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.117949485778809,
      "learning_rate": 1.5939597315436243e-05,
      "loss": 1.8209,
      "step": 190
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.04592514038086,
      "learning_rate": 1.6778523489932888e-05,
      "loss": 1.8287,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.954629898071289,
      "learning_rate": 1.761744966442953e-05,
      "loss": 1.8208,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.442968368530273,
      "learning_rate": 1.8456375838926178e-05,
      "loss": 1.7883,
      "step": 220
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.213078498840332,
      "learning_rate": 1.929530201342282e-05,
      "loss": 1.7369,
      "step": 230
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.5759806632995605,
      "learning_rate": 2.013422818791946e-05,
      "loss": 1.7885,
      "step": 240
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.507256031036377,
      "learning_rate": 2.097315436241611e-05,
      "loss": 1.7416,
      "step": 250
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.662291526794434,
      "learning_rate": 2.181208053691275e-05,
      "loss": 1.7269,
      "step": 260
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.283073425292969,
      "learning_rate": 2.2651006711409396e-05,
      "loss": 1.7217,
      "step": 270
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.17520809173584,
      "learning_rate": 2.348993288590604e-05,
      "loss": 1.6601,
      "step": 280
    },
    {
      "epoch": 0.05,
      "grad_norm": 8.257431030273438,
      "learning_rate": 2.4328859060402687e-05,
      "loss": 1.7131,
      "step": 290
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.119696617126465,
      "learning_rate": 2.516778523489933e-05,
      "loss": 1.7287,
      "step": 300
    },
    {
      "epoch": 0.05,
      "grad_norm": 6.9683942794799805,
      "learning_rate": 2.6006711409395973e-05,
      "loss": 1.7874,
      "step": 310
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.026366233825684,
      "learning_rate": 2.6845637583892618e-05,
      "loss": 1.6577,
      "step": 320
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.696808338165283,
      "learning_rate": 2.7684563758389263e-05,
      "loss": 1.6542,
      "step": 330
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.984933376312256,
      "learning_rate": 2.8523489932885905e-05,
      "loss": 1.6476,
      "step": 340
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.91267728805542,
      "learning_rate": 2.936241610738255e-05,
      "loss": 1.6855,
      "step": 350
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.043091773986816,
      "learning_rate": 3.02013422818792e-05,
      "loss": 1.8219,
      "step": 360
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.10373067855835,
      "learning_rate": 3.1040268456375844e-05,
      "loss": 1.6523,
      "step": 370
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.729587078094482,
      "learning_rate": 3.1879194630872485e-05,
      "loss": 1.6402,
      "step": 380
    },
    {
      "epoch": 0.07,
      "grad_norm": 5.648786544799805,
      "learning_rate": 3.271812080536913e-05,
      "loss": 1.6723,
      "step": 390
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.498968601226807,
      "learning_rate": 3.3557046979865775e-05,
      "loss": 1.7108,
      "step": 400
    },
    {
      "epoch": 0.07,
      "grad_norm": 8.128254890441895,
      "learning_rate": 3.439597315436242e-05,
      "loss": 1.6536,
      "step": 410
    },
    {
      "epoch": 0.07,
      "grad_norm": 5.086948871612549,
      "learning_rate": 3.523489932885906e-05,
      "loss": 1.67,
      "step": 420
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.3559980392456055,
      "learning_rate": 3.607382550335571e-05,
      "loss": 1.5262,
      "step": 430
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.808907508850098,
      "learning_rate": 3.6912751677852356e-05,
      "loss": 1.5247,
      "step": 440
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.132946014404297,
      "learning_rate": 3.775167785234899e-05,
      "loss": 1.6465,
      "step": 450
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.056805610656738,
      "learning_rate": 3.859060402684564e-05,
      "loss": 1.7219,
      "step": 460
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.494626998901367,
      "learning_rate": 3.942953020134229e-05,
      "loss": 1.541,
      "step": 470
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.917198657989502,
      "learning_rate": 4.026845637583892e-05,
      "loss": 1.5328,
      "step": 480
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.69942569732666,
      "learning_rate": 4.110738255033557e-05,
      "loss": 1.5356,
      "step": 490
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.75998306274414,
      "learning_rate": 4.194630872483222e-05,
      "loss": 1.5105,
      "step": 500
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.057405948638916,
      "learning_rate": 4.278523489932886e-05,
      "loss": 1.4101,
      "step": 510
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.895244598388672,
      "learning_rate": 4.36241610738255e-05,
      "loss": 1.715,
      "step": 520
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.775789737701416,
      "learning_rate": 4.446308724832215e-05,
      "loss": 1.5758,
      "step": 530
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.584057331085205,
      "learning_rate": 4.530201342281879e-05,
      "loss": 1.5876,
      "step": 540
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.187722682952881,
      "learning_rate": 4.6140939597315434e-05,
      "loss": 1.4255,
      "step": 550
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.943014144897461,
      "learning_rate": 4.697986577181208e-05,
      "loss": 1.703,
      "step": 560
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.033366203308105,
      "learning_rate": 4.7818791946308725e-05,
      "loss": 1.573,
      "step": 570
    },
    {
      "epoch": 0.1,
      "grad_norm": 14.540526390075684,
      "learning_rate": 4.865771812080537e-05,
      "loss": 1.4141,
      "step": 580
    },
    {
      "epoch": 0.1,
      "grad_norm": 6.717778205871582,
      "learning_rate": 4.9496644295302015e-05,
      "loss": 1.4235,
      "step": 590
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.104679107666016,
      "learning_rate": 4.996265870052278e-05,
      "loss": 1.4492,
      "step": 600
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.074448585510254,
      "learning_rate": 4.986930545182973e-05,
      "loss": 1.4318,
      "step": 610
    },
    {
      "epoch": 0.1,
      "grad_norm": 8.634200096130371,
      "learning_rate": 4.977595220313667e-05,
      "loss": 1.3045,
      "step": 620
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.917612075805664,
      "learning_rate": 4.9682598954443615e-05,
      "loss": 1.3824,
      "step": 630
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.843932151794434,
      "learning_rate": 4.958924570575056e-05,
      "loss": 1.1116,
      "step": 640
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.2613525390625,
      "learning_rate": 4.949589245705751e-05,
      "loss": 1.3314,
      "step": 650
    },
    {
      "epoch": 0.11,
      "grad_norm": 6.22221565246582,
      "learning_rate": 4.9402539208364454e-05,
      "loss": 1.3183,
      "step": 660
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.652546882629395,
      "learning_rate": 4.93091859596714e-05,
      "loss": 1.3282,
      "step": 670
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.172412872314453,
      "learning_rate": 4.9215832710978346e-05,
      "loss": 1.345,
      "step": 680
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.713068008422852,
      "learning_rate": 4.912247946228529e-05,
      "loss": 1.3923,
      "step": 690
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.730290412902832,
      "learning_rate": 4.902912621359224e-05,
      "loss": 1.2178,
      "step": 700
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.1386189460754395,
      "learning_rate": 4.893577296489918e-05,
      "loss": 1.1482,
      "step": 710
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.427709579467773,
      "learning_rate": 4.884241971620613e-05,
      "loss": 1.492,
      "step": 720
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.728306770324707,
      "learning_rate": 4.874906646751307e-05,
      "loss": 1.3139,
      "step": 730
    },
    {
      "epoch": 0.12,
      "grad_norm": 7.621016979217529,
      "learning_rate": 4.8655713218820016e-05,
      "loss": 1.3492,
      "step": 740
    },
    {
      "epoch": 0.13,
      "eval_accuracy": 0.5126835781041389,
      "eval_loss": 1.3353450298309326,
      "eval_runtime": 532.6055,
      "eval_samples_per_second": 1.406,
      "eval_steps_per_second": 0.176,
      "step": 745
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.568509101867676,
      "learning_rate": 4.856235997012696e-05,
      "loss": 1.1979,
      "step": 750
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.509567260742188,
      "learning_rate": 4.846900672143391e-05,
      "loss": 1.138,
      "step": 760
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.080005645751953,
      "learning_rate": 4.8375653472740855e-05,
      "loss": 1.3821,
      "step": 770
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.756476402282715,
      "learning_rate": 4.82823002240478e-05,
      "loss": 1.3165,
      "step": 780
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.856130123138428,
      "learning_rate": 4.818894697535474e-05,
      "loss": 1.1757,
      "step": 790
    },
    {
      "epoch": 1.01,
      "grad_norm": 5.623302936553955,
      "learning_rate": 4.809559372666169e-05,
      "loss": 1.2131,
      "step": 800
    },
    {
      "epoch": 1.01,
      "grad_norm": 9.992005348205566,
      "learning_rate": 4.800224047796863e-05,
      "loss": 1.1972,
      "step": 810
    },
    {
      "epoch": 1.01,
      "grad_norm": 16.679943084716797,
      "learning_rate": 4.790888722927558e-05,
      "loss": 0.991,
      "step": 820
    },
    {
      "epoch": 1.01,
      "grad_norm": 8.39699935913086,
      "learning_rate": 4.7815533980582525e-05,
      "loss": 1.2499,
      "step": 830
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.164064407348633,
      "learning_rate": 4.772218073188947e-05,
      "loss": 1.0056,
      "step": 840
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.30912971496582,
      "learning_rate": 4.762882748319642e-05,
      "loss": 1.2122,
      "step": 850
    },
    {
      "epoch": 1.02,
      "grad_norm": 9.575423240661621,
      "learning_rate": 4.753547423450336e-05,
      "loss": 1.2111,
      "step": 860
    },
    {
      "epoch": 1.02,
      "grad_norm": 6.919304370880127,
      "learning_rate": 4.744212098581031e-05,
      "loss": 1.0568,
      "step": 870
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.245553970336914,
      "learning_rate": 4.7348767737117256e-05,
      "loss": 1.2895,
      "step": 880
    },
    {
      "epoch": 1.02,
      "grad_norm": 6.206485271453857,
      "learning_rate": 4.72554144884242e-05,
      "loss": 1.0963,
      "step": 890
    },
    {
      "epoch": 1.03,
      "grad_norm": 14.38132095336914,
      "learning_rate": 4.716206123973114e-05,
      "loss": 1.0835,
      "step": 900
    },
    {
      "epoch": 1.03,
      "grad_norm": 7.840010166168213,
      "learning_rate": 4.7068707991038094e-05,
      "loss": 1.0129,
      "step": 910
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.073530673980713,
      "learning_rate": 4.697535474234503e-05,
      "loss": 1.0401,
      "step": 920
    },
    {
      "epoch": 1.03,
      "grad_norm": 11.431855201721191,
      "learning_rate": 4.6882001493651986e-05,
      "loss": 0.9816,
      "step": 930
    },
    {
      "epoch": 1.03,
      "grad_norm": 9.11495304107666,
      "learning_rate": 4.6788648244958926e-05,
      "loss": 0.9431,
      "step": 940
    },
    {
      "epoch": 1.03,
      "grad_norm": 10.392239570617676,
      "learning_rate": 4.669529499626587e-05,
      "loss": 1.0666,
      "step": 950
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.40548324584961,
      "learning_rate": 4.660194174757282e-05,
      "loss": 1.0039,
      "step": 960
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.694117546081543,
      "learning_rate": 4.6508588498879764e-05,
      "loss": 1.019,
      "step": 970
    },
    {
      "epoch": 1.04,
      "grad_norm": 12.926886558532715,
      "learning_rate": 4.6415235250186703e-05,
      "loss": 1.1456,
      "step": 980
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.995538234710693,
      "learning_rate": 4.6321882001493656e-05,
      "loss": 1.1484,
      "step": 990
    },
    {
      "epoch": 1.04,
      "grad_norm": 11.876310348510742,
      "learning_rate": 4.6228528752800596e-05,
      "loss": 1.0705,
      "step": 1000
    },
    {
      "epoch": 1.04,
      "grad_norm": 10.454874038696289,
      "learning_rate": 4.613517550410755e-05,
      "loss": 1.0672,
      "step": 1010
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.644485473632812,
      "learning_rate": 4.604182225541449e-05,
      "loss": 1.0194,
      "step": 1020
    },
    {
      "epoch": 1.05,
      "grad_norm": 11.500072479248047,
      "learning_rate": 4.5948469006721434e-05,
      "loss": 0.9129,
      "step": 1030
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.650041580200195,
      "learning_rate": 4.585511575802838e-05,
      "loss": 0.6978,
      "step": 1040
    },
    {
      "epoch": 1.05,
      "grad_norm": 12.148303985595703,
      "learning_rate": 4.5761762509335327e-05,
      "loss": 1.08,
      "step": 1050
    },
    {
      "epoch": 1.05,
      "grad_norm": 6.243818759918213,
      "learning_rate": 4.566840926064227e-05,
      "loss": 0.7159,
      "step": 1060
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.437726020812988,
      "learning_rate": 4.557505601194922e-05,
      "loss": 1.0204,
      "step": 1070
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.030414581298828,
      "learning_rate": 4.5481702763256165e-05,
      "loss": 0.9837,
      "step": 1080
    },
    {
      "epoch": 1.06,
      "grad_norm": 12.592327117919922,
      "learning_rate": 4.538834951456311e-05,
      "loss": 0.7511,
      "step": 1090
    },
    {
      "epoch": 1.06,
      "grad_norm": 13.41678237915039,
      "learning_rate": 4.529499626587006e-05,
      "loss": 1.043,
      "step": 1100
    },
    {
      "epoch": 1.06,
      "grad_norm": 16.153676986694336,
      "learning_rate": 4.5201643017177e-05,
      "loss": 1.195,
      "step": 1110
    },
    {
      "epoch": 1.06,
      "grad_norm": 8.62240982055664,
      "learning_rate": 4.510828976848395e-05,
      "loss": 0.9355,
      "step": 1120
    },
    {
      "epoch": 1.06,
      "grad_norm": 10.290811538696289,
      "learning_rate": 4.501493651979089e-05,
      "loss": 0.7375,
      "step": 1130
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.986602783203125,
      "learning_rate": 4.492158327109784e-05,
      "loss": 0.9695,
      "step": 1140
    },
    {
      "epoch": 1.07,
      "grad_norm": 8.108384132385254,
      "learning_rate": 4.482823002240478e-05,
      "loss": 1.0521,
      "step": 1150
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.250555992126465,
      "learning_rate": 4.473487677371173e-05,
      "loss": 0.9775,
      "step": 1160
    },
    {
      "epoch": 1.07,
      "grad_norm": 6.976318359375,
      "learning_rate": 4.4641523525018674e-05,
      "loss": 0.8981,
      "step": 1170
    },
    {
      "epoch": 1.07,
      "grad_norm": 5.915456771850586,
      "learning_rate": 4.454817027632562e-05,
      "loss": 1.0194,
      "step": 1180
    },
    {
      "epoch": 1.07,
      "grad_norm": 10.186369895935059,
      "learning_rate": 4.445481702763256e-05,
      "loss": 1.1633,
      "step": 1190
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.099218368530273,
      "learning_rate": 4.436146377893951e-05,
      "loss": 0.6142,
      "step": 1200
    },
    {
      "epoch": 1.08,
      "grad_norm": 14.789891242980957,
      "learning_rate": 4.426811053024645e-05,
      "loss": 0.9571,
      "step": 1210
    },
    {
      "epoch": 1.08,
      "grad_norm": 8.024323463439941,
      "learning_rate": 4.4174757281553404e-05,
      "loss": 0.9158,
      "step": 1220
    },
    {
      "epoch": 1.08,
      "grad_norm": 9.476771354675293,
      "learning_rate": 4.4081404032860344e-05,
      "loss": 0.7556,
      "step": 1230
    },
    {
      "epoch": 1.08,
      "grad_norm": 10.900609016418457,
      "learning_rate": 4.398805078416729e-05,
      "loss": 1.0184,
      "step": 1240
    },
    {
      "epoch": 1.08,
      "grad_norm": 12.15243148803711,
      "learning_rate": 4.3894697535474236e-05,
      "loss": 1.0357,
      "step": 1250
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.207107543945312,
      "learning_rate": 4.380134428678118e-05,
      "loss": 1.31,
      "step": 1260
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.220967292785645,
      "learning_rate": 4.370799103808813e-05,
      "loss": 0.7858,
      "step": 1270
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.639920711517334,
      "learning_rate": 4.3614637789395075e-05,
      "loss": 0.8076,
      "step": 1280
    },
    {
      "epoch": 1.09,
      "grad_norm": 12.56861686706543,
      "learning_rate": 4.352128454070202e-05,
      "loss": 0.8895,
      "step": 1290
    },
    {
      "epoch": 1.09,
      "grad_norm": 11.024540901184082,
      "learning_rate": 4.342793129200897e-05,
      "loss": 0.9184,
      "step": 1300
    },
    {
      "epoch": 1.09,
      "grad_norm": 9.709234237670898,
      "learning_rate": 4.333457804331591e-05,
      "loss": 0.7979,
      "step": 1310
    },
    {
      "epoch": 1.1,
      "grad_norm": 4.263139247894287,
      "learning_rate": 4.324122479462285e-05,
      "loss": 0.7441,
      "step": 1320
    },
    {
      "epoch": 1.1,
      "grad_norm": 9.647358894348145,
      "learning_rate": 4.3147871545929805e-05,
      "loss": 0.7048,
      "step": 1330
    },
    {
      "epoch": 1.1,
      "grad_norm": 10.529332160949707,
      "learning_rate": 4.3054518297236745e-05,
      "loss": 1.0735,
      "step": 1340
    },
    {
      "epoch": 1.1,
      "grad_norm": 19.13629150390625,
      "learning_rate": 4.296116504854369e-05,
      "loss": 0.9224,
      "step": 1350
    },
    {
      "epoch": 1.1,
      "grad_norm": 11.141228675842285,
      "learning_rate": 4.286781179985064e-05,
      "loss": 0.9048,
      "step": 1360
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.264597415924072,
      "learning_rate": 4.277445855115758e-05,
      "loss": 0.7945,
      "step": 1370
    },
    {
      "epoch": 1.11,
      "grad_norm": 6.68549108505249,
      "learning_rate": 4.268110530246453e-05,
      "loss": 0.9528,
      "step": 1380
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.21309757232666,
      "learning_rate": 4.2587752053771475e-05,
      "loss": 0.8536,
      "step": 1390
    },
    {
      "epoch": 1.11,
      "grad_norm": 18.798063278198242,
      "learning_rate": 4.2494398805078415e-05,
      "loss": 0.7947,
      "step": 1400
    },
    {
      "epoch": 1.11,
      "grad_norm": 15.983850479125977,
      "learning_rate": 4.240104555638537e-05,
      "loss": 0.912,
      "step": 1410
    },
    {
      "epoch": 1.11,
      "grad_norm": 9.7501220703125,
      "learning_rate": 4.230769230769231e-05,
      "loss": 0.9997,
      "step": 1420
    },
    {
      "epoch": 1.12,
      "grad_norm": 5.131944179534912,
      "learning_rate": 4.221433905899925e-05,
      "loss": 0.7686,
      "step": 1430
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.2156805992126465,
      "learning_rate": 4.21209858103062e-05,
      "loss": 0.7812,
      "step": 1440
    },
    {
      "epoch": 1.12,
      "grad_norm": 15.262615203857422,
      "learning_rate": 4.2027632561613146e-05,
      "loss": 0.9678,
      "step": 1450
    },
    {
      "epoch": 1.12,
      "grad_norm": 11.919661521911621,
      "learning_rate": 4.193427931292009e-05,
      "loss": 0.9752,
      "step": 1460
    },
    {
      "epoch": 1.12,
      "grad_norm": 20.424823760986328,
      "learning_rate": 4.184092606422704e-05,
      "loss": 0.8504,
      "step": 1470
    },
    {
      "epoch": 1.12,
      "grad_norm": 17.759782791137695,
      "learning_rate": 4.1747572815533984e-05,
      "loss": 0.8313,
      "step": 1480
    },
    {
      "epoch": 1.13,
      "grad_norm": 38.34241485595703,
      "learning_rate": 4.165421956684093e-05,
      "loss": 0.9541,
      "step": 1490
    },
    {
      "epoch": 1.13,
      "eval_accuracy": 0.684913217623498,
      "eval_loss": 0.9003641605377197,
      "eval_runtime": 519.8999,
      "eval_samples_per_second": 1.441,
      "eval_steps_per_second": 0.181,
      "step": 1490
    },
    {
      "epoch": 2.0,
      "grad_norm": 11.179780006408691,
      "learning_rate": 4.1560866318147876e-05,
      "loss": 0.7782,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.399374961853027,
      "learning_rate": 4.1467513069454816e-05,
      "loss": 0.8835,
      "step": 1510
    },
    {
      "epoch": 2.01,
      "grad_norm": 5.939724922180176,
      "learning_rate": 4.137415982076177e-05,
      "loss": 0.6691,
      "step": 1520
    },
    {
      "epoch": 2.01,
      "grad_norm": 16.65831184387207,
      "learning_rate": 4.128080657206871e-05,
      "loss": 0.8445,
      "step": 1530
    },
    {
      "epoch": 2.01,
      "grad_norm": 20.12108039855957,
      "learning_rate": 4.118745332337566e-05,
      "loss": 0.6285,
      "step": 1540
    },
    {
      "epoch": 2.01,
      "grad_norm": 18.240739822387695,
      "learning_rate": 4.10941000746826e-05,
      "loss": 0.833,
      "step": 1550
    },
    {
      "epoch": 2.01,
      "grad_norm": 13.109556198120117,
      "learning_rate": 4.1000746825989546e-05,
      "loss": 0.8259,
      "step": 1560
    },
    {
      "epoch": 2.01,
      "grad_norm": 12.095843315124512,
      "learning_rate": 4.090739357729649e-05,
      "loss": 0.813,
      "step": 1570
    },
    {
      "epoch": 2.02,
      "grad_norm": 10.142488479614258,
      "learning_rate": 4.081404032860344e-05,
      "loss": 0.8081,
      "step": 1580
    },
    {
      "epoch": 2.02,
      "grad_norm": 18.163326263427734,
      "learning_rate": 4.072068707991038e-05,
      "loss": 0.7403,
      "step": 1590
    },
    {
      "epoch": 2.02,
      "grad_norm": 17.268943786621094,
      "learning_rate": 4.062733383121733e-05,
      "loss": 0.7279,
      "step": 1600
    },
    {
      "epoch": 2.02,
      "grad_norm": 12.311897277832031,
      "learning_rate": 4.053398058252427e-05,
      "loss": 0.8055,
      "step": 1610
    },
    {
      "epoch": 2.02,
      "grad_norm": 14.742772102355957,
      "learning_rate": 4.0440627333831223e-05,
      "loss": 0.8789,
      "step": 1620
    },
    {
      "epoch": 2.02,
      "grad_norm": 4.9054856300354,
      "learning_rate": 4.034727408513816e-05,
      "loss": 0.8909,
      "step": 1630
    },
    {
      "epoch": 2.03,
      "grad_norm": 13.565916061401367,
      "learning_rate": 4.025392083644511e-05,
      "loss": 0.8536,
      "step": 1640
    },
    {
      "epoch": 2.03,
      "grad_norm": 11.949501991271973,
      "learning_rate": 4.0160567587752055e-05,
      "loss": 0.7359,
      "step": 1650
    },
    {
      "epoch": 2.03,
      "grad_norm": 9.746699333190918,
      "learning_rate": 4.0067214339059e-05,
      "loss": 0.6584,
      "step": 1660
    },
    {
      "epoch": 2.03,
      "grad_norm": 12.663901329040527,
      "learning_rate": 3.997386109036595e-05,
      "loss": 0.7073,
      "step": 1670
    },
    {
      "epoch": 2.03,
      "grad_norm": 8.44615650177002,
      "learning_rate": 3.9880507841672894e-05,
      "loss": 0.8865,
      "step": 1680
    },
    {
      "epoch": 2.03,
      "grad_norm": 7.577113151550293,
      "learning_rate": 3.978715459297983e-05,
      "loss": 0.7457,
      "step": 1690
    },
    {
      "epoch": 2.04,
      "grad_norm": 13.037449836730957,
      "learning_rate": 3.9693801344286786e-05,
      "loss": 1.0867,
      "step": 1700
    },
    {
      "epoch": 2.04,
      "grad_norm": 7.493896961212158,
      "learning_rate": 3.9600448095593725e-05,
      "loss": 0.8625,
      "step": 1710
    },
    {
      "epoch": 2.04,
      "grad_norm": 7.936389923095703,
      "learning_rate": 3.950709484690067e-05,
      "loss": 0.7202,
      "step": 1720
    },
    {
      "epoch": 2.04,
      "grad_norm": 4.738982200622559,
      "learning_rate": 3.941374159820762e-05,
      "loss": 0.6652,
      "step": 1730
    },
    {
      "epoch": 2.04,
      "grad_norm": 9.391824722290039,
      "learning_rate": 3.9320388349514564e-05,
      "loss": 0.7956,
      "step": 1740
    },
    {
      "epoch": 2.04,
      "grad_norm": 10.760794639587402,
      "learning_rate": 3.922703510082151e-05,
      "loss": 0.6821,
      "step": 1750
    },
    {
      "epoch": 2.05,
      "grad_norm": 15.05656909942627,
      "learning_rate": 3.9133681852128456e-05,
      "loss": 0.7137,
      "step": 1760
    },
    {
      "epoch": 2.05,
      "grad_norm": 10.40505313873291,
      "learning_rate": 3.90403286034354e-05,
      "loss": 0.7702,
      "step": 1770
    },
    {
      "epoch": 2.05,
      "grad_norm": 12.18364143371582,
      "learning_rate": 3.894697535474235e-05,
      "loss": 0.9514,
      "step": 1780
    },
    {
      "epoch": 2.05,
      "grad_norm": 20.13235092163086,
      "learning_rate": 3.8853622106049294e-05,
      "loss": 0.9815,
      "step": 1790
    },
    {
      "epoch": 2.05,
      "grad_norm": 12.790679931640625,
      "learning_rate": 3.8760268857356234e-05,
      "loss": 0.7515,
      "step": 1800
    },
    {
      "epoch": 2.05,
      "grad_norm": 11.214062690734863,
      "learning_rate": 3.866691560866319e-05,
      "loss": 0.8037,
      "step": 1810
    },
    {
      "epoch": 2.06,
      "grad_norm": 22.581945419311523,
      "learning_rate": 3.8573562359970126e-05,
      "loss": 0.9517,
      "step": 1820
    },
    {
      "epoch": 2.06,
      "grad_norm": 14.781002044677734,
      "learning_rate": 3.848020911127708e-05,
      "loss": 0.8295,
      "step": 1830
    },
    {
      "epoch": 2.06,
      "grad_norm": 8.621711730957031,
      "learning_rate": 3.838685586258402e-05,
      "loss": 0.7496,
      "step": 1840
    },
    {
      "epoch": 2.06,
      "grad_norm": 13.089216232299805,
      "learning_rate": 3.8293502613890965e-05,
      "loss": 0.8355,
      "step": 1850
    },
    {
      "epoch": 2.06,
      "grad_norm": 8.406702995300293,
      "learning_rate": 3.820014936519791e-05,
      "loss": 0.678,
      "step": 1860
    },
    {
      "epoch": 2.06,
      "grad_norm": 12.511942863464355,
      "learning_rate": 3.810679611650486e-05,
      "loss": 0.6752,
      "step": 1870
    },
    {
      "epoch": 2.07,
      "grad_norm": 13.4660062789917,
      "learning_rate": 3.8013442867811796e-05,
      "loss": 0.9581,
      "step": 1880
    },
    {
      "epoch": 2.07,
      "grad_norm": 11.665181159973145,
      "learning_rate": 3.792008961911875e-05,
      "loss": 0.9555,
      "step": 1890
    },
    {
      "epoch": 2.07,
      "grad_norm": 12.686610221862793,
      "learning_rate": 3.782673637042569e-05,
      "loss": 0.6219,
      "step": 1900
    },
    {
      "epoch": 2.07,
      "grad_norm": 11.197409629821777,
      "learning_rate": 3.773338312173264e-05,
      "loss": 0.9832,
      "step": 1910
    },
    {
      "epoch": 2.07,
      "grad_norm": 10.146992683410645,
      "learning_rate": 3.764002987303958e-05,
      "loss": 0.6686,
      "step": 1920
    },
    {
      "epoch": 2.07,
      "grad_norm": 15.430662155151367,
      "learning_rate": 3.754667662434653e-05,
      "loss": 0.6632,
      "step": 1930
    },
    {
      "epoch": 2.08,
      "grad_norm": 7.6227264404296875,
      "learning_rate": 3.745332337565347e-05,
      "loss": 0.8637,
      "step": 1940
    },
    {
      "epoch": 2.08,
      "grad_norm": 8.638508796691895,
      "learning_rate": 3.735997012696042e-05,
      "loss": 0.8894,
      "step": 1950
    },
    {
      "epoch": 2.08,
      "grad_norm": 7.317731857299805,
      "learning_rate": 3.7266616878267365e-05,
      "loss": 0.7345,
      "step": 1960
    },
    {
      "epoch": 2.08,
      "grad_norm": 10.281793594360352,
      "learning_rate": 3.717326362957431e-05,
      "loss": 0.7272,
      "step": 1970
    },
    {
      "epoch": 2.08,
      "grad_norm": 10.731279373168945,
      "learning_rate": 3.707991038088126e-05,
      "loss": 0.5959,
      "step": 1980
    },
    {
      "epoch": 2.08,
      "grad_norm": 20.986186981201172,
      "learning_rate": 3.6986557132188204e-05,
      "loss": 0.9152,
      "step": 1990
    },
    {
      "epoch": 2.09,
      "grad_norm": 12.500083923339844,
      "learning_rate": 3.689320388349515e-05,
      "loss": 0.5702,
      "step": 2000
    },
    {
      "epoch": 2.09,
      "grad_norm": 19.52236557006836,
      "learning_rate": 3.679985063480209e-05,
      "loss": 0.594,
      "step": 2010
    },
    {
      "epoch": 2.09,
      "grad_norm": 10.224320411682129,
      "learning_rate": 3.670649738610904e-05,
      "loss": 0.6608,
      "step": 2020
    },
    {
      "epoch": 2.09,
      "grad_norm": 11.070257186889648,
      "learning_rate": 3.661314413741598e-05,
      "loss": 0.7784,
      "step": 2030
    },
    {
      "epoch": 2.09,
      "grad_norm": 10.710339546203613,
      "learning_rate": 3.651979088872293e-05,
      "loss": 0.8536,
      "step": 2040
    },
    {
      "epoch": 2.09,
      "grad_norm": 9.494738578796387,
      "learning_rate": 3.6426437640029874e-05,
      "loss": 0.591,
      "step": 2050
    },
    {
      "epoch": 2.1,
      "grad_norm": 3.604220151901245,
      "learning_rate": 3.633308439133682e-05,
      "loss": 0.4326,
      "step": 2060
    },
    {
      "epoch": 2.1,
      "grad_norm": 16.6247615814209,
      "learning_rate": 3.6239731142643766e-05,
      "loss": 0.4328,
      "step": 2070
    },
    {
      "epoch": 2.1,
      "grad_norm": 18.82486915588379,
      "learning_rate": 3.614637789395071e-05,
      "loss": 0.6209,
      "step": 2080
    },
    {
      "epoch": 2.1,
      "grad_norm": 9.521064758300781,
      "learning_rate": 3.605302464525765e-05,
      "loss": 0.5655,
      "step": 2090
    },
    {
      "epoch": 2.1,
      "grad_norm": 12.07824993133545,
      "learning_rate": 3.5959671396564605e-05,
      "loss": 0.8264,
      "step": 2100
    },
    {
      "epoch": 2.1,
      "grad_norm": 9.77541732788086,
      "learning_rate": 3.5866318147871544e-05,
      "loss": 0.5653,
      "step": 2110
    },
    {
      "epoch": 2.11,
      "grad_norm": 9.046191215515137,
      "learning_rate": 3.577296489917849e-05,
      "loss": 0.6976,
      "step": 2120
    },
    {
      "epoch": 2.11,
      "grad_norm": 11.601213455200195,
      "learning_rate": 3.5679611650485437e-05,
      "loss": 0.6741,
      "step": 2130
    },
    {
      "epoch": 2.11,
      "grad_norm": 13.549548149108887,
      "learning_rate": 3.558625840179238e-05,
      "loss": 0.5945,
      "step": 2140
    },
    {
      "epoch": 2.11,
      "grad_norm": 18.340423583984375,
      "learning_rate": 3.549290515309933e-05,
      "loss": 0.5536,
      "step": 2150
    },
    {
      "epoch": 2.11,
      "grad_norm": 12.24539852142334,
      "learning_rate": 3.5399551904406275e-05,
      "loss": 0.8303,
      "step": 2160
    },
    {
      "epoch": 2.11,
      "grad_norm": 23.557832717895508,
      "learning_rate": 3.530619865571322e-05,
      "loss": 0.7909,
      "step": 2170
    },
    {
      "epoch": 2.12,
      "grad_norm": 11.522802352905273,
      "learning_rate": 3.521284540702017e-05,
      "loss": 0.8725,
      "step": 2180
    },
    {
      "epoch": 2.12,
      "grad_norm": 7.0091423988342285,
      "learning_rate": 3.5119492158327113e-05,
      "loss": 0.7891,
      "step": 2190
    },
    {
      "epoch": 2.12,
      "grad_norm": 6.060676097869873,
      "learning_rate": 3.502613890963405e-05,
      "loss": 0.9306,
      "step": 2200
    },
    {
      "epoch": 2.12,
      "grad_norm": 8.596571922302246,
      "learning_rate": 3.4932785660941006e-05,
      "loss": 0.7885,
      "step": 2210
    },
    {
      "epoch": 2.12,
      "grad_norm": 8.664740562438965,
      "learning_rate": 3.4839432412247945e-05,
      "loss": 0.5937,
      "step": 2220
    },
    {
      "epoch": 2.12,
      "grad_norm": 6.942666530609131,
      "learning_rate": 3.47460791635549e-05,
      "loss": 0.7073,
      "step": 2230
    },
    {
      "epoch": 2.13,
      "eval_accuracy": 0.7236315086782377,
      "eval_loss": 0.7945846915245056,
      "eval_runtime": 527.6098,
      "eval_samples_per_second": 1.42,
      "eval_steps_per_second": 0.178,
      "step": 2235
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.050449371337891,
      "learning_rate": 3.465272591486184e-05,
      "loss": 0.886,
      "step": 2240
    },
    {
      "epoch": 3.0,
      "grad_norm": 11.89676284790039,
      "learning_rate": 3.4559372666168784e-05,
      "loss": 0.9251,
      "step": 2250
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.360501289367676,
      "learning_rate": 3.446601941747573e-05,
      "loss": 0.6314,
      "step": 2260
    },
    {
      "epoch": 3.01,
      "grad_norm": 11.796934127807617,
      "learning_rate": 3.4372666168782676e-05,
      "loss": 0.5453,
      "step": 2270
    },
    {
      "epoch": 3.01,
      "grad_norm": 16.64284324645996,
      "learning_rate": 3.427931292008962e-05,
      "loss": 0.7497,
      "step": 2280
    },
    {
      "epoch": 3.01,
      "grad_norm": 5.327428340911865,
      "learning_rate": 3.418595967139657e-05,
      "loss": 0.6692,
      "step": 2290
    },
    {
      "epoch": 3.01,
      "grad_norm": 15.887578964233398,
      "learning_rate": 3.409260642270351e-05,
      "loss": 0.7206,
      "step": 2300
    },
    {
      "epoch": 3.01,
      "grad_norm": 11.925492286682129,
      "learning_rate": 3.399925317401046e-05,
      "loss": 0.6601,
      "step": 2310
    },
    {
      "epoch": 3.01,
      "grad_norm": 10.036468505859375,
      "learning_rate": 3.39058999253174e-05,
      "loss": 0.6249,
      "step": 2320
    },
    {
      "epoch": 3.02,
      "grad_norm": 13.341933250427246,
      "learning_rate": 3.3812546676624346e-05,
      "loss": 0.759,
      "step": 2330
    },
    {
      "epoch": 3.02,
      "grad_norm": 13.791449546813965,
      "learning_rate": 3.371919342793129e-05,
      "loss": 0.7401,
      "step": 2340
    },
    {
      "epoch": 3.02,
      "grad_norm": 13.78197956085205,
      "learning_rate": 3.362584017923824e-05,
      "loss": 0.7764,
      "step": 2350
    },
    {
      "epoch": 3.02,
      "grad_norm": 10.72469711303711,
      "learning_rate": 3.3532486930545184e-05,
      "loss": 0.5269,
      "step": 2360
    },
    {
      "epoch": 3.02,
      "grad_norm": 12.453922271728516,
      "learning_rate": 3.343913368185213e-05,
      "loss": 0.8282,
      "step": 2370
    },
    {
      "epoch": 3.02,
      "grad_norm": 14.766962051391602,
      "learning_rate": 3.334578043315908e-05,
      "loss": 0.8528,
      "step": 2380
    },
    {
      "epoch": 3.03,
      "grad_norm": 14.155939102172852,
      "learning_rate": 3.325242718446602e-05,
      "loss": 0.8481,
      "step": 2390
    },
    {
      "epoch": 3.03,
      "grad_norm": 9.234591484069824,
      "learning_rate": 3.315907393577297e-05,
      "loss": 0.8334,
      "step": 2400
    },
    {
      "epoch": 3.03,
      "grad_norm": 20.89356803894043,
      "learning_rate": 3.306572068707991e-05,
      "loss": 0.7547,
      "step": 2410
    },
    {
      "epoch": 3.03,
      "grad_norm": 17.84977149963379,
      "learning_rate": 3.297236743838686e-05,
      "loss": 0.6814,
      "step": 2420
    },
    {
      "epoch": 3.03,
      "grad_norm": 18.2919864654541,
      "learning_rate": 3.28790141896938e-05,
      "loss": 0.6879,
      "step": 2430
    },
    {
      "epoch": 3.03,
      "grad_norm": 4.913272380828857,
      "learning_rate": 3.2785660941000754e-05,
      "loss": 0.5068,
      "step": 2440
    },
    {
      "epoch": 3.04,
      "grad_norm": 21.99398422241211,
      "learning_rate": 3.269230769230769e-05,
      "loss": 0.6318,
      "step": 2450
    },
    {
      "epoch": 3.04,
      "grad_norm": 17.939348220825195,
      "learning_rate": 3.259895444361464e-05,
      "loss": 0.8155,
      "step": 2460
    },
    {
      "epoch": 3.04,
      "grad_norm": 15.396060943603516,
      "learning_rate": 3.2505601194921585e-05,
      "loss": 0.6879,
      "step": 2470
    },
    {
      "epoch": 3.04,
      "grad_norm": 12.186809539794922,
      "learning_rate": 3.241224794622853e-05,
      "loss": 0.4496,
      "step": 2480
    },
    {
      "epoch": 3.04,
      "grad_norm": 8.525858879089355,
      "learning_rate": 3.231889469753547e-05,
      "loss": 0.5562,
      "step": 2490
    },
    {
      "epoch": 3.04,
      "grad_norm": 10.064820289611816,
      "learning_rate": 3.2225541448842424e-05,
      "loss": 0.3546,
      "step": 2500
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.001558542251587,
      "learning_rate": 3.213218820014936e-05,
      "loss": 0.6542,
      "step": 2510
    },
    {
      "epoch": 3.05,
      "grad_norm": 18.06378746032715,
      "learning_rate": 3.2038834951456316e-05,
      "loss": 0.603,
      "step": 2520
    },
    {
      "epoch": 3.05,
      "grad_norm": 15.484493255615234,
      "learning_rate": 3.1945481702763256e-05,
      "loss": 0.6829,
      "step": 2530
    },
    {
      "epoch": 3.05,
      "grad_norm": 14.678126335144043,
      "learning_rate": 3.18521284540702e-05,
      "loss": 0.7742,
      "step": 2540
    },
    {
      "epoch": 3.05,
      "grad_norm": 9.602285385131836,
      "learning_rate": 3.175877520537715e-05,
      "loss": 0.5638,
      "step": 2550
    },
    {
      "epoch": 3.05,
      "grad_norm": 13.426918029785156,
      "learning_rate": 3.1665421956684094e-05,
      "loss": 0.5461,
      "step": 2560
    },
    {
      "epoch": 3.06,
      "grad_norm": 15.853680610656738,
      "learning_rate": 3.157206870799104e-05,
      "loss": 0.7363,
      "step": 2570
    },
    {
      "epoch": 3.06,
      "grad_norm": 9.000801086425781,
      "learning_rate": 3.1478715459297986e-05,
      "loss": 0.4888,
      "step": 2580
    },
    {
      "epoch": 3.06,
      "grad_norm": 7.437619686126709,
      "learning_rate": 3.138536221060493e-05,
      "loss": 0.6189,
      "step": 2590
    },
    {
      "epoch": 3.06,
      "grad_norm": 7.136799335479736,
      "learning_rate": 3.129200896191188e-05,
      "loss": 0.5689,
      "step": 2600
    },
    {
      "epoch": 3.06,
      "grad_norm": 12.685044288635254,
      "learning_rate": 3.1198655713218825e-05,
      "loss": 0.5861,
      "step": 2610
    },
    {
      "epoch": 3.06,
      "grad_norm": 11.239699363708496,
      "learning_rate": 3.1105302464525764e-05,
      "loss": 0.5326,
      "step": 2620
    },
    {
      "epoch": 3.07,
      "grad_norm": 10.661036491394043,
      "learning_rate": 3.101194921583272e-05,
      "loss": 0.5519,
      "step": 2630
    },
    {
      "epoch": 3.07,
      "grad_norm": 17.59286880493164,
      "learning_rate": 3.0918595967139656e-05,
      "loss": 0.8041,
      "step": 2640
    },
    {
      "epoch": 3.07,
      "grad_norm": 10.675395011901855,
      "learning_rate": 3.08252427184466e-05,
      "loss": 0.7223,
      "step": 2650
    },
    {
      "epoch": 3.07,
      "grad_norm": 15.262356758117676,
      "learning_rate": 3.073188946975355e-05,
      "loss": 0.5381,
      "step": 2660
    },
    {
      "epoch": 3.07,
      "grad_norm": 10.153656005859375,
      "learning_rate": 3.0638536221060495e-05,
      "loss": 0.708,
      "step": 2670
    },
    {
      "epoch": 3.07,
      "grad_norm": 11.261735916137695,
      "learning_rate": 3.054518297236744e-05,
      "loss": 0.5022,
      "step": 2680
    },
    {
      "epoch": 3.08,
      "grad_norm": 12.410257339477539,
      "learning_rate": 3.0451829723674384e-05,
      "loss": 0.7562,
      "step": 2690
    },
    {
      "epoch": 3.08,
      "grad_norm": 14.27526569366455,
      "learning_rate": 3.035847647498133e-05,
      "loss": 0.5866,
      "step": 2700
    },
    {
      "epoch": 3.08,
      "grad_norm": 9.704520225524902,
      "learning_rate": 3.0265123226288276e-05,
      "loss": 0.5984,
      "step": 2710
    },
    {
      "epoch": 3.08,
      "grad_norm": 5.257163047790527,
      "learning_rate": 3.0171769977595222e-05,
      "loss": 0.5833,
      "step": 2720
    },
    {
      "epoch": 3.08,
      "grad_norm": 10.275954246520996,
      "learning_rate": 3.0078416728902165e-05,
      "loss": 0.5731,
      "step": 2730
    },
    {
      "epoch": 3.08,
      "grad_norm": 6.9410529136657715,
      "learning_rate": 2.9985063480209115e-05,
      "loss": 0.4525,
      "step": 2740
    },
    {
      "epoch": 3.09,
      "grad_norm": 13.865569114685059,
      "learning_rate": 2.9891710231516057e-05,
      "loss": 0.6132,
      "step": 2750
    },
    {
      "epoch": 3.09,
      "grad_norm": 3.045743942260742,
      "learning_rate": 2.9798356982823007e-05,
      "loss": 0.5469,
      "step": 2760
    },
    {
      "epoch": 3.09,
      "grad_norm": 11.01517391204834,
      "learning_rate": 2.970500373412995e-05,
      "loss": 0.6552,
      "step": 2770
    },
    {
      "epoch": 3.09,
      "grad_norm": 6.852537631988525,
      "learning_rate": 2.9611650485436892e-05,
      "loss": 0.5166,
      "step": 2780
    },
    {
      "epoch": 3.09,
      "grad_norm": 10.36206340789795,
      "learning_rate": 2.9518297236743842e-05,
      "loss": 0.5781,
      "step": 2790
    },
    {
      "epoch": 3.09,
      "grad_norm": 23.712926864624023,
      "learning_rate": 2.9424943988050785e-05,
      "loss": 0.8148,
      "step": 2800
    },
    {
      "epoch": 3.1,
      "grad_norm": 12.161064147949219,
      "learning_rate": 2.9331590739357734e-05,
      "loss": 0.8599,
      "step": 2810
    },
    {
      "epoch": 3.1,
      "grad_norm": 13.12808895111084,
      "learning_rate": 2.9238237490664677e-05,
      "loss": 0.6179,
      "step": 2820
    },
    {
      "epoch": 3.1,
      "grad_norm": 14.208728790283203,
      "learning_rate": 2.914488424197162e-05,
      "loss": 0.6316,
      "step": 2830
    },
    {
      "epoch": 3.1,
      "grad_norm": 8.262664794921875,
      "learning_rate": 2.905153099327857e-05,
      "loss": 0.6265,
      "step": 2840
    },
    {
      "epoch": 3.1,
      "grad_norm": 10.486581802368164,
      "learning_rate": 2.8958177744585512e-05,
      "loss": 0.6077,
      "step": 2850
    },
    {
      "epoch": 3.11,
      "grad_norm": 14.470271110534668,
      "learning_rate": 2.8864824495892455e-05,
      "loss": 0.6055,
      "step": 2860
    },
    {
      "epoch": 3.11,
      "grad_norm": 15.240158081054688,
      "learning_rate": 2.8771471247199404e-05,
      "loss": 0.4665,
      "step": 2870
    },
    {
      "epoch": 3.11,
      "grad_norm": 16.40061378479004,
      "learning_rate": 2.8678117998506347e-05,
      "loss": 0.6428,
      "step": 2880
    },
    {
      "epoch": 3.11,
      "grad_norm": 11.96904468536377,
      "learning_rate": 2.8584764749813297e-05,
      "loss": 0.5504,
      "step": 2890
    },
    {
      "epoch": 3.11,
      "grad_norm": 4.584710121154785,
      "learning_rate": 2.849141150112024e-05,
      "loss": 0.5748,
      "step": 2900
    },
    {
      "epoch": 3.11,
      "grad_norm": 12.811570167541504,
      "learning_rate": 2.8398058252427186e-05,
      "loss": 0.5264,
      "step": 2910
    },
    {
      "epoch": 3.12,
      "grad_norm": 21.13889503479004,
      "learning_rate": 2.8304705003734132e-05,
      "loss": 0.6081,
      "step": 2920
    },
    {
      "epoch": 3.12,
      "grad_norm": 19.0864315032959,
      "learning_rate": 2.8211351755041078e-05,
      "loss": 0.5696,
      "step": 2930
    },
    {
      "epoch": 3.12,
      "grad_norm": 6.2545647621154785,
      "learning_rate": 2.811799850634802e-05,
      "loss": 0.4955,
      "step": 2940
    },
    {
      "epoch": 3.12,
      "grad_norm": 6.386425018310547,
      "learning_rate": 2.802464525765497e-05,
      "loss": 0.5312,
      "step": 2950
    },
    {
      "epoch": 3.12,
      "grad_norm": 15.030627250671387,
      "learning_rate": 2.7931292008961913e-05,
      "loss": 0.5301,
      "step": 2960
    },
    {
      "epoch": 3.12,
      "grad_norm": 10.548439025878906,
      "learning_rate": 2.7837938760268863e-05,
      "loss": 0.6695,
      "step": 2970
    },
    {
      "epoch": 3.13,
      "grad_norm": 43.06850051879883,
      "learning_rate": 2.7744585511575805e-05,
      "loss": 0.8417,
      "step": 2980
    },
    {
      "epoch": 3.13,
      "eval_accuracy": 0.6875834445927904,
      "eval_loss": 0.851558268070221,
      "eval_runtime": 455.6708,
      "eval_samples_per_second": 1.644,
      "eval_steps_per_second": 0.206,
      "step": 2980
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.430088996887207,
      "learning_rate": 2.7651232262882748e-05,
      "loss": 0.5937,
      "step": 2990
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.522638320922852,
      "learning_rate": 2.7557879014189698e-05,
      "loss": 0.5213,
      "step": 3000
    },
    {
      "epoch": 4.01,
      "grad_norm": 20.326101303100586,
      "learning_rate": 2.746452576549664e-05,
      "loss": 0.71,
      "step": 3010
    },
    {
      "epoch": 4.01,
      "grad_norm": 15.900239944458008,
      "learning_rate": 2.7371172516803583e-05,
      "loss": 0.7077,
      "step": 3020
    },
    {
      "epoch": 4.01,
      "grad_norm": 7.540781021118164,
      "learning_rate": 2.7277819268110533e-05,
      "loss": 0.4259,
      "step": 3030
    },
    {
      "epoch": 4.01,
      "grad_norm": 13.728072166442871,
      "learning_rate": 2.7184466019417475e-05,
      "loss": 0.5173,
      "step": 3040
    },
    {
      "epoch": 4.01,
      "grad_norm": 14.119942665100098,
      "learning_rate": 2.7091112770724425e-05,
      "loss": 0.6718,
      "step": 3050
    },
    {
      "epoch": 4.01,
      "grad_norm": 6.129566192626953,
      "learning_rate": 2.6997759522031368e-05,
      "loss": 0.525,
      "step": 3060
    },
    {
      "epoch": 4.02,
      "grad_norm": 20.588115692138672,
      "learning_rate": 2.690440627333831e-05,
      "loss": 0.5695,
      "step": 3070
    },
    {
      "epoch": 4.02,
      "grad_norm": 11.256523132324219,
      "learning_rate": 2.681105302464526e-05,
      "loss": 0.6944,
      "step": 3080
    },
    {
      "epoch": 4.02,
      "grad_norm": 17.56951904296875,
      "learning_rate": 2.6717699775952203e-05,
      "loss": 0.4976,
      "step": 3090
    },
    {
      "epoch": 4.02,
      "grad_norm": 11.125219345092773,
      "learning_rate": 2.662434652725915e-05,
      "loss": 0.5969,
      "step": 3100
    },
    {
      "epoch": 4.02,
      "grad_norm": 9.810901641845703,
      "learning_rate": 2.6530993278566095e-05,
      "loss": 0.3574,
      "step": 3110
    },
    {
      "epoch": 4.02,
      "grad_norm": 8.470976829528809,
      "learning_rate": 2.643764002987304e-05,
      "loss": 0.5709,
      "step": 3120
    },
    {
      "epoch": 4.03,
      "grad_norm": 9.978863716125488,
      "learning_rate": 2.6344286781179987e-05,
      "loss": 0.5249,
      "step": 3130
    },
    {
      "epoch": 4.03,
      "grad_norm": 13.830111503601074,
      "learning_rate": 2.6250933532486934e-05,
      "loss": 0.51,
      "step": 3140
    },
    {
      "epoch": 4.03,
      "grad_norm": 10.391007423400879,
      "learning_rate": 2.6157580283793876e-05,
      "loss": 0.2823,
      "step": 3150
    },
    {
      "epoch": 4.03,
      "grad_norm": 13.137083053588867,
      "learning_rate": 2.6064227035100826e-05,
      "loss": 0.6245,
      "step": 3160
    },
    {
      "epoch": 4.03,
      "grad_norm": 10.020318984985352,
      "learning_rate": 2.597087378640777e-05,
      "loss": 0.5146,
      "step": 3170
    },
    {
      "epoch": 4.03,
      "grad_norm": 3.6848785877227783,
      "learning_rate": 2.587752053771471e-05,
      "loss": 0.4606,
      "step": 3180
    },
    {
      "epoch": 4.04,
      "grad_norm": 4.760696887969971,
      "learning_rate": 2.578416728902166e-05,
      "loss": 0.5904,
      "step": 3190
    },
    {
      "epoch": 4.04,
      "grad_norm": 14.180034637451172,
      "learning_rate": 2.5690814040328604e-05,
      "loss": 0.6278,
      "step": 3200
    },
    {
      "epoch": 4.04,
      "grad_norm": 4.619647026062012,
      "learning_rate": 2.5597460791635553e-05,
      "loss": 0.5799,
      "step": 3210
    },
    {
      "epoch": 4.04,
      "grad_norm": 6.640144348144531,
      "learning_rate": 2.5504107542942496e-05,
      "loss": 0.5911,
      "step": 3220
    },
    {
      "epoch": 4.04,
      "grad_norm": 10.515678405761719,
      "learning_rate": 2.541075429424944e-05,
      "loss": 0.5745,
      "step": 3230
    },
    {
      "epoch": 4.04,
      "grad_norm": 6.645077228546143,
      "learning_rate": 2.531740104555639e-05,
      "loss": 0.3824,
      "step": 3240
    },
    {
      "epoch": 4.05,
      "grad_norm": 13.339990615844727,
      "learning_rate": 2.522404779686333e-05,
      "loss": 0.482,
      "step": 3250
    },
    {
      "epoch": 4.05,
      "grad_norm": 8.3532075881958,
      "learning_rate": 2.5130694548170274e-05,
      "loss": 0.6492,
      "step": 3260
    },
    {
      "epoch": 4.05,
      "grad_norm": 8.647318840026855,
      "learning_rate": 2.5037341299477223e-05,
      "loss": 0.4287,
      "step": 3270
    },
    {
      "epoch": 4.05,
      "grad_norm": 8.45876407623291,
      "learning_rate": 2.4943988050784166e-05,
      "loss": 0.4266,
      "step": 3280
    },
    {
      "epoch": 4.05,
      "grad_norm": 3.776923418045044,
      "learning_rate": 2.4850634802091112e-05,
      "loss": 0.3847,
      "step": 3290
    },
    {
      "epoch": 4.05,
      "grad_norm": 6.598992824554443,
      "learning_rate": 2.475728155339806e-05,
      "loss": 0.474,
      "step": 3300
    },
    {
      "epoch": 4.06,
      "grad_norm": 8.877213478088379,
      "learning_rate": 2.4663928304705005e-05,
      "loss": 0.4641,
      "step": 3310
    },
    {
      "epoch": 4.06,
      "grad_norm": 20.836896896362305,
      "learning_rate": 2.4570575056011947e-05,
      "loss": 0.6016,
      "step": 3320
    },
    {
      "epoch": 4.06,
      "grad_norm": 8.716022491455078,
      "learning_rate": 2.4477221807318894e-05,
      "loss": 0.5888,
      "step": 3330
    },
    {
      "epoch": 4.06,
      "grad_norm": 14.989657402038574,
      "learning_rate": 2.4383868558625843e-05,
      "loss": 0.6509,
      "step": 3340
    },
    {
      "epoch": 4.06,
      "grad_norm": 9.637127876281738,
      "learning_rate": 2.429051530993279e-05,
      "loss": 0.5601,
      "step": 3350
    },
    {
      "epoch": 4.06,
      "grad_norm": 7.190096855163574,
      "learning_rate": 2.4197162061239732e-05,
      "loss": 0.593,
      "step": 3360
    },
    {
      "epoch": 4.07,
      "grad_norm": 11.497958183288574,
      "learning_rate": 2.4103808812546678e-05,
      "loss": 0.5872,
      "step": 3370
    },
    {
      "epoch": 4.07,
      "grad_norm": 11.88388729095459,
      "learning_rate": 2.4010455563853624e-05,
      "loss": 0.7672,
      "step": 3380
    },
    {
      "epoch": 4.07,
      "grad_norm": 14.398612022399902,
      "learning_rate": 2.391710231516057e-05,
      "loss": 0.4177,
      "step": 3390
    },
    {
      "epoch": 4.07,
      "grad_norm": 15.993851661682129,
      "learning_rate": 2.3823749066467517e-05,
      "loss": 0.7173,
      "step": 3400
    },
    {
      "epoch": 4.07,
      "grad_norm": 13.425286293029785,
      "learning_rate": 2.373039581777446e-05,
      "loss": 0.5585,
      "step": 3410
    },
    {
      "epoch": 4.07,
      "grad_norm": 10.236127853393555,
      "learning_rate": 2.3637042569081406e-05,
      "loss": 0.4752,
      "step": 3420
    },
    {
      "epoch": 4.08,
      "grad_norm": 23.924036026000977,
      "learning_rate": 2.3543689320388352e-05,
      "loss": 0.5999,
      "step": 3430
    },
    {
      "epoch": 4.08,
      "grad_norm": 13.102322578430176,
      "learning_rate": 2.3450336071695298e-05,
      "loss": 0.659,
      "step": 3440
    },
    {
      "epoch": 4.08,
      "grad_norm": 12.305949211120605,
      "learning_rate": 2.335698282300224e-05,
      "loss": 0.6002,
      "step": 3450
    },
    {
      "epoch": 4.08,
      "grad_norm": 5.7384352684021,
      "learning_rate": 2.3263629574309187e-05,
      "loss": 0.5025,
      "step": 3460
    },
    {
      "epoch": 4.08,
      "grad_norm": 9.778236389160156,
      "learning_rate": 2.3170276325616133e-05,
      "loss": 0.4095,
      "step": 3470
    },
    {
      "epoch": 4.08,
      "grad_norm": 11.141785621643066,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.5927,
      "step": 3480
    },
    {
      "epoch": 4.09,
      "grad_norm": 16.052156448364258,
      "learning_rate": 2.2983569828230022e-05,
      "loss": 0.4957,
      "step": 3490
    },
    {
      "epoch": 4.09,
      "grad_norm": 12.158838272094727,
      "learning_rate": 2.2890216579536968e-05,
      "loss": 0.434,
      "step": 3500
    },
    {
      "epoch": 4.09,
      "grad_norm": 15.396862030029297,
      "learning_rate": 2.2796863330843914e-05,
      "loss": 0.522,
      "step": 3510
    },
    {
      "epoch": 4.09,
      "grad_norm": 18.10469627380371,
      "learning_rate": 2.270351008215086e-05,
      "loss": 0.5488,
      "step": 3520
    },
    {
      "epoch": 4.09,
      "grad_norm": 5.236746788024902,
      "learning_rate": 2.2610156833457803e-05,
      "loss": 0.7389,
      "step": 3530
    },
    {
      "epoch": 4.09,
      "grad_norm": 2.8653571605682373,
      "learning_rate": 2.251680358476475e-05,
      "loss": 0.2953,
      "step": 3540
    },
    {
      "epoch": 4.1,
      "grad_norm": 9.715225219726562,
      "learning_rate": 2.2423450336071695e-05,
      "loss": 0.4078,
      "step": 3550
    },
    {
      "epoch": 4.1,
      "grad_norm": 11.62974739074707,
      "learning_rate": 2.233009708737864e-05,
      "loss": 0.5635,
      "step": 3560
    },
    {
      "epoch": 4.1,
      "grad_norm": 6.261245250701904,
      "learning_rate": 2.2236743838685588e-05,
      "loss": 0.6294,
      "step": 3570
    },
    {
      "epoch": 4.1,
      "grad_norm": 22.738595962524414,
      "learning_rate": 2.2143390589992534e-05,
      "loss": 0.6094,
      "step": 3580
    },
    {
      "epoch": 4.1,
      "grad_norm": 11.334650993347168,
      "learning_rate": 2.205003734129948e-05,
      "loss": 0.4386,
      "step": 3590
    },
    {
      "epoch": 4.1,
      "grad_norm": 8.969924926757812,
      "learning_rate": 2.1956684092606426e-05,
      "loss": 0.6962,
      "step": 3600
    },
    {
      "epoch": 4.11,
      "grad_norm": 20.473119735717773,
      "learning_rate": 2.186333084391337e-05,
      "loss": 0.5695,
      "step": 3610
    },
    {
      "epoch": 4.11,
      "grad_norm": 8.890480995178223,
      "learning_rate": 2.1769977595220315e-05,
      "loss": 0.9129,
      "step": 3620
    },
    {
      "epoch": 4.11,
      "grad_norm": 10.379404067993164,
      "learning_rate": 2.167662434652726e-05,
      "loss": 0.5648,
      "step": 3630
    },
    {
      "epoch": 4.11,
      "grad_norm": 10.535677909851074,
      "learning_rate": 2.1583271097834207e-05,
      "loss": 0.6503,
      "step": 3640
    },
    {
      "epoch": 4.11,
      "grad_norm": 11.505346298217773,
      "learning_rate": 2.148991784914115e-05,
      "loss": 0.5491,
      "step": 3650
    },
    {
      "epoch": 4.11,
      "grad_norm": 9.04132080078125,
      "learning_rate": 2.1396564600448096e-05,
      "loss": 0.3545,
      "step": 3660
    },
    {
      "epoch": 4.12,
      "grad_norm": 10.061311721801758,
      "learning_rate": 2.1303211351755042e-05,
      "loss": 0.6648,
      "step": 3670
    },
    {
      "epoch": 4.12,
      "grad_norm": 7.3036017417907715,
      "learning_rate": 2.120985810306199e-05,
      "loss": 0.5207,
      "step": 3680
    },
    {
      "epoch": 4.12,
      "grad_norm": 9.250767707824707,
      "learning_rate": 2.111650485436893e-05,
      "loss": 0.3503,
      "step": 3690
    },
    {
      "epoch": 4.12,
      "grad_norm": 22.527040481567383,
      "learning_rate": 2.1023151605675877e-05,
      "loss": 0.4452,
      "step": 3700
    },
    {
      "epoch": 4.12,
      "grad_norm": 16.892024993896484,
      "learning_rate": 2.0929798356982824e-05,
      "loss": 0.4881,
      "step": 3710
    },
    {
      "epoch": 4.12,
      "grad_norm": 10.745135307312012,
      "learning_rate": 2.083644510828977e-05,
      "loss": 0.3899,
      "step": 3720
    },
    {
      "epoch": 4.13,
      "eval_accuracy": 0.7449933244325768,
      "eval_loss": 0.7319404482841492,
      "eval_runtime": 452.6279,
      "eval_samples_per_second": 1.655,
      "eval_steps_per_second": 0.208,
      "step": 3725
    },
    {
      "epoch": 5.0,
      "grad_norm": 18.635610580444336,
      "learning_rate": 2.0743091859596713e-05,
      "loss": 0.4602,
      "step": 3730
    },
    {
      "epoch": 5.0,
      "grad_norm": 7.6850056648254395,
      "learning_rate": 2.064973861090366e-05,
      "loss": 0.412,
      "step": 3740
    },
    {
      "epoch": 5.0,
      "grad_norm": 24.052745819091797,
      "learning_rate": 2.0556385362210605e-05,
      "loss": 0.4903,
      "step": 3750
    },
    {
      "epoch": 5.01,
      "grad_norm": 6.334225654602051,
      "learning_rate": 2.046303211351755e-05,
      "loss": 0.5595,
      "step": 3760
    },
    {
      "epoch": 5.01,
      "grad_norm": 10.684706687927246,
      "learning_rate": 2.0369678864824497e-05,
      "loss": 0.4819,
      "step": 3770
    },
    {
      "epoch": 5.01,
      "grad_norm": 7.236117362976074,
      "learning_rate": 2.0276325616131443e-05,
      "loss": 0.4287,
      "step": 3780
    },
    {
      "epoch": 5.01,
      "grad_norm": 17.454057693481445,
      "learning_rate": 2.018297236743839e-05,
      "loss": 0.4321,
      "step": 3790
    },
    {
      "epoch": 5.01,
      "grad_norm": 9.47890853881836,
      "learning_rate": 2.0089619118745336e-05,
      "loss": 0.4784,
      "step": 3800
    },
    {
      "epoch": 5.01,
      "grad_norm": 18.901548385620117,
      "learning_rate": 1.999626587005228e-05,
      "loss": 0.6126,
      "step": 3810
    },
    {
      "epoch": 5.02,
      "grad_norm": 10.916252136230469,
      "learning_rate": 1.9902912621359225e-05,
      "loss": 0.4197,
      "step": 3820
    },
    {
      "epoch": 5.02,
      "grad_norm": 5.3163743019104,
      "learning_rate": 1.980955937266617e-05,
      "loss": 0.4934,
      "step": 3830
    },
    {
      "epoch": 5.02,
      "grad_norm": 16.681852340698242,
      "learning_rate": 1.9716206123973117e-05,
      "loss": 0.3536,
      "step": 3840
    },
    {
      "epoch": 5.02,
      "grad_norm": 6.279463768005371,
      "learning_rate": 1.962285287528006e-05,
      "loss": 0.464,
      "step": 3850
    },
    {
      "epoch": 5.02,
      "grad_norm": 9.089925765991211,
      "learning_rate": 1.9529499626587006e-05,
      "loss": 0.4698,
      "step": 3860
    },
    {
      "epoch": 5.02,
      "grad_norm": 6.502386093139648,
      "learning_rate": 1.9436146377893952e-05,
      "loss": 0.3453,
      "step": 3870
    },
    {
      "epoch": 5.03,
      "grad_norm": 6.205687522888184,
      "learning_rate": 1.9342793129200898e-05,
      "loss": 0.3907,
      "step": 3880
    },
    {
      "epoch": 5.03,
      "grad_norm": 2.179413080215454,
      "learning_rate": 1.924943988050784e-05,
      "loss": 0.6915,
      "step": 3890
    },
    {
      "epoch": 5.03,
      "grad_norm": 4.713302135467529,
      "learning_rate": 1.9156086631814787e-05,
      "loss": 0.4342,
      "step": 3900
    },
    {
      "epoch": 5.03,
      "grad_norm": 12.052186012268066,
      "learning_rate": 1.9062733383121733e-05,
      "loss": 0.5518,
      "step": 3910
    },
    {
      "epoch": 5.03,
      "grad_norm": 1.5166425704956055,
      "learning_rate": 1.896938013442868e-05,
      "loss": 0.3804,
      "step": 3920
    },
    {
      "epoch": 5.03,
      "grad_norm": 11.109734535217285,
      "learning_rate": 1.8876026885735622e-05,
      "loss": 0.4746,
      "step": 3930
    },
    {
      "epoch": 5.04,
      "grad_norm": 4.877770900726318,
      "learning_rate": 1.8782673637042568e-05,
      "loss": 0.5263,
      "step": 3940
    },
    {
      "epoch": 5.04,
      "grad_norm": 7.948832035064697,
      "learning_rate": 1.8689320388349514e-05,
      "loss": 0.4336,
      "step": 3950
    },
    {
      "epoch": 5.04,
      "grad_norm": 1.1953182220458984,
      "learning_rate": 1.859596713965646e-05,
      "loss": 0.3529,
      "step": 3960
    },
    {
      "epoch": 5.04,
      "grad_norm": 9.760388374328613,
      "learning_rate": 1.8502613890963407e-05,
      "loss": 0.486,
      "step": 3970
    },
    {
      "epoch": 5.04,
      "grad_norm": 14.954423904418945,
      "learning_rate": 1.8409260642270353e-05,
      "loss": 0.6088,
      "step": 3980
    },
    {
      "epoch": 5.04,
      "grad_norm": 23.564476013183594,
      "learning_rate": 1.83159073935773e-05,
      "loss": 0.2852,
      "step": 3990
    },
    {
      "epoch": 5.05,
      "grad_norm": 17.43026351928711,
      "learning_rate": 1.8222554144884245e-05,
      "loss": 0.4271,
      "step": 4000
    },
    {
      "epoch": 5.05,
      "grad_norm": 7.904264450073242,
      "learning_rate": 1.812920089619119e-05,
      "loss": 0.5144,
      "step": 4010
    },
    {
      "epoch": 5.05,
      "grad_norm": 22.415292739868164,
      "learning_rate": 1.8035847647498134e-05,
      "loss": 0.5235,
      "step": 4020
    },
    {
      "epoch": 5.05,
      "grad_norm": 19.2883358001709,
      "learning_rate": 1.794249439880508e-05,
      "loss": 0.6248,
      "step": 4030
    },
    {
      "epoch": 5.05,
      "grad_norm": 6.104731559753418,
      "learning_rate": 1.7849141150112026e-05,
      "loss": 0.4512,
      "step": 4040
    },
    {
      "epoch": 5.05,
      "grad_norm": 12.515997886657715,
      "learning_rate": 1.7755787901418973e-05,
      "loss": 0.4439,
      "step": 4050
    },
    {
      "epoch": 5.06,
      "grad_norm": 7.608014106750488,
      "learning_rate": 1.7662434652725915e-05,
      "loss": 0.5307,
      "step": 4060
    },
    {
      "epoch": 5.06,
      "grad_norm": 5.3149638175964355,
      "learning_rate": 1.756908140403286e-05,
      "loss": 0.3428,
      "step": 4070
    },
    {
      "epoch": 5.06,
      "grad_norm": 9.341575622558594,
      "learning_rate": 1.7475728155339808e-05,
      "loss": 0.3774,
      "step": 4080
    },
    {
      "epoch": 5.06,
      "grad_norm": 8.19787883758545,
      "learning_rate": 1.7382374906646754e-05,
      "loss": 0.4128,
      "step": 4090
    },
    {
      "epoch": 5.06,
      "grad_norm": 13.791829109191895,
      "learning_rate": 1.7289021657953697e-05,
      "loss": 0.4106,
      "step": 4100
    },
    {
      "epoch": 5.06,
      "grad_norm": 8.926873207092285,
      "learning_rate": 1.7195668409260643e-05,
      "loss": 0.5541,
      "step": 4110
    },
    {
      "epoch": 5.07,
      "grad_norm": 15.664535522460938,
      "learning_rate": 1.710231516056759e-05,
      "loss": 0.5746,
      "step": 4120
    },
    {
      "epoch": 5.07,
      "grad_norm": 21.65399932861328,
      "learning_rate": 1.7008961911874535e-05,
      "loss": 0.5296,
      "step": 4130
    },
    {
      "epoch": 5.07,
      "grad_norm": 11.411762237548828,
      "learning_rate": 1.6915608663181478e-05,
      "loss": 0.5531,
      "step": 4140
    },
    {
      "epoch": 5.07,
      "grad_norm": 3.2497332096099854,
      "learning_rate": 1.6822255414488424e-05,
      "loss": 0.4139,
      "step": 4150
    },
    {
      "epoch": 5.07,
      "grad_norm": 17.887279510498047,
      "learning_rate": 1.672890216579537e-05,
      "loss": 0.5703,
      "step": 4160
    },
    {
      "epoch": 5.07,
      "grad_norm": 10.63172435760498,
      "learning_rate": 1.6635548917102316e-05,
      "loss": 0.5189,
      "step": 4170
    },
    {
      "epoch": 5.08,
      "grad_norm": 6.9324140548706055,
      "learning_rate": 1.654219566840926e-05,
      "loss": 0.5173,
      "step": 4180
    },
    {
      "epoch": 5.08,
      "grad_norm": 14.986353874206543,
      "learning_rate": 1.6448842419716205e-05,
      "loss": 0.5219,
      "step": 4190
    },
    {
      "epoch": 5.08,
      "grad_norm": 8.2908935546875,
      "learning_rate": 1.635548917102315e-05,
      "loss": 0.3294,
      "step": 4200
    },
    {
      "epoch": 5.08,
      "grad_norm": 14.855971336364746,
      "learning_rate": 1.6262135922330097e-05,
      "loss": 0.4804,
      "step": 4210
    },
    {
      "epoch": 5.08,
      "grad_norm": 5.925484657287598,
      "learning_rate": 1.6168782673637044e-05,
      "loss": 0.3888,
      "step": 4220
    },
    {
      "epoch": 5.08,
      "grad_norm": 20.477813720703125,
      "learning_rate": 1.607542942494399e-05,
      "loss": 0.4624,
      "step": 4230
    },
    {
      "epoch": 5.09,
      "grad_norm": 13.53854751586914,
      "learning_rate": 1.5982076176250936e-05,
      "loss": 0.2529,
      "step": 4240
    },
    {
      "epoch": 5.09,
      "grad_norm": 24.013643264770508,
      "learning_rate": 1.5888722927557882e-05,
      "loss": 0.4825,
      "step": 4250
    },
    {
      "epoch": 5.09,
      "grad_norm": 15.870683670043945,
      "learning_rate": 1.5795369678864825e-05,
      "loss": 0.4754,
      "step": 4260
    },
    {
      "epoch": 5.09,
      "grad_norm": 7.974236965179443,
      "learning_rate": 1.570201643017177e-05,
      "loss": 0.4369,
      "step": 4270
    },
    {
      "epoch": 5.09,
      "grad_norm": 4.172085285186768,
      "learning_rate": 1.5608663181478717e-05,
      "loss": 0.463,
      "step": 4280
    },
    {
      "epoch": 5.09,
      "grad_norm": 13.204963684082031,
      "learning_rate": 1.5515309932785663e-05,
      "loss": 0.4392,
      "step": 4290
    },
    {
      "epoch": 5.1,
      "grad_norm": 9.802266120910645,
      "learning_rate": 1.5421956684092606e-05,
      "loss": 0.6392,
      "step": 4300
    },
    {
      "epoch": 5.1,
      "grad_norm": 8.404006004333496,
      "learning_rate": 1.5328603435399552e-05,
      "loss": 0.3287,
      "step": 4310
    },
    {
      "epoch": 5.1,
      "grad_norm": 1.507799744606018,
      "learning_rate": 1.5235250186706498e-05,
      "loss": 0.4654,
      "step": 4320
    },
    {
      "epoch": 5.1,
      "grad_norm": 14.254951477050781,
      "learning_rate": 1.5141896938013444e-05,
      "loss": 0.314,
      "step": 4330
    },
    {
      "epoch": 5.1,
      "grad_norm": 14.278926849365234,
      "learning_rate": 1.5048543689320387e-05,
      "loss": 0.4093,
      "step": 4340
    },
    {
      "epoch": 5.11,
      "grad_norm": 17.676769256591797,
      "learning_rate": 1.4955190440627333e-05,
      "loss": 0.4405,
      "step": 4350
    },
    {
      "epoch": 5.11,
      "grad_norm": 12.228628158569336,
      "learning_rate": 1.486183719193428e-05,
      "loss": 0.4333,
      "step": 4360
    },
    {
      "epoch": 5.11,
      "grad_norm": 7.0753254890441895,
      "learning_rate": 1.4768483943241226e-05,
      "loss": 0.5138,
      "step": 4370
    },
    {
      "epoch": 5.11,
      "grad_norm": 3.453059434890747,
      "learning_rate": 1.467513069454817e-05,
      "loss": 0.4102,
      "step": 4380
    },
    {
      "epoch": 5.11,
      "grad_norm": 13.31628131866455,
      "learning_rate": 1.4581777445855116e-05,
      "loss": 0.6762,
      "step": 4390
    },
    {
      "epoch": 5.11,
      "grad_norm": 19.06345558166504,
      "learning_rate": 1.4488424197162062e-05,
      "loss": 0.5631,
      "step": 4400
    },
    {
      "epoch": 5.12,
      "grad_norm": 13.416067123413086,
      "learning_rate": 1.4395070948469009e-05,
      "loss": 0.4472,
      "step": 4410
    },
    {
      "epoch": 5.12,
      "grad_norm": 18.242368698120117,
      "learning_rate": 1.4301717699775951e-05,
      "loss": 0.4103,
      "step": 4420
    },
    {
      "epoch": 5.12,
      "grad_norm": 14.489863395690918,
      "learning_rate": 1.4208364451082898e-05,
      "loss": 0.3765,
      "step": 4430
    },
    {
      "epoch": 5.12,
      "grad_norm": 8.337303161621094,
      "learning_rate": 1.4115011202389844e-05,
      "loss": 0.3556,
      "step": 4440
    },
    {
      "epoch": 5.12,
      "grad_norm": 14.795426368713379,
      "learning_rate": 1.402165795369679e-05,
      "loss": 0.421,
      "step": 4450
    },
    {
      "epoch": 5.12,
      "grad_norm": 24.15338134765625,
      "learning_rate": 1.3928304705003734e-05,
      "loss": 0.4476,
      "step": 4460
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.14645187556743622,
      "learning_rate": 1.383495145631068e-05,
      "loss": 0.3669,
      "step": 4470
    },
    {
      "epoch": 5.13,
      "eval_accuracy": 0.7489986648865153,
      "eval_loss": 0.7200458645820618,
      "eval_runtime": 435.4052,
      "eval_samples_per_second": 1.72,
      "eval_steps_per_second": 0.216,
      "step": 4470
    },
    {
      "epoch": 6.0,
      "grad_norm": 14.630892753601074,
      "learning_rate": 1.3741598207617627e-05,
      "loss": 0.3999,
      "step": 4480
    },
    {
      "epoch": 6.0,
      "grad_norm": 15.605926513671875,
      "learning_rate": 1.3648244958924573e-05,
      "loss": 0.4216,
      "step": 4490
    },
    {
      "epoch": 6.01,
      "grad_norm": 7.535991668701172,
      "learning_rate": 1.3554891710231516e-05,
      "loss": 0.3838,
      "step": 4500
    },
    {
      "epoch": 6.01,
      "grad_norm": 1.3880966901779175,
      "learning_rate": 1.3461538461538462e-05,
      "loss": 0.348,
      "step": 4510
    },
    {
      "epoch": 6.01,
      "grad_norm": 15.70007038116455,
      "learning_rate": 1.3368185212845408e-05,
      "loss": 0.4697,
      "step": 4520
    },
    {
      "epoch": 6.01,
      "grad_norm": 21.03390121459961,
      "learning_rate": 1.3274831964152354e-05,
      "loss": 0.419,
      "step": 4530
    },
    {
      "epoch": 6.01,
      "grad_norm": 8.029413223266602,
      "learning_rate": 1.3181478715459297e-05,
      "loss": 0.5879,
      "step": 4540
    },
    {
      "epoch": 6.01,
      "grad_norm": 22.235984802246094,
      "learning_rate": 1.3088125466766243e-05,
      "loss": 0.604,
      "step": 4550
    },
    {
      "epoch": 6.02,
      "grad_norm": 12.360527992248535,
      "learning_rate": 1.2994772218073189e-05,
      "loss": 0.4202,
      "step": 4560
    },
    {
      "epoch": 6.02,
      "grad_norm": 3.968439817428589,
      "learning_rate": 1.2901418969380135e-05,
      "loss": 0.2472,
      "step": 4570
    },
    {
      "epoch": 6.02,
      "grad_norm": 11.081033706665039,
      "learning_rate": 1.2808065720687081e-05,
      "loss": 0.303,
      "step": 4580
    },
    {
      "epoch": 6.02,
      "grad_norm": 8.187359809875488,
      "learning_rate": 1.2714712471994026e-05,
      "loss": 0.5667,
      "step": 4590
    },
    {
      "epoch": 6.02,
      "grad_norm": 5.036584854125977,
      "learning_rate": 1.2621359223300972e-05,
      "loss": 0.4209,
      "step": 4600
    },
    {
      "epoch": 6.02,
      "grad_norm": 17.97156524658203,
      "learning_rate": 1.2528005974607918e-05,
      "loss": 0.509,
      "step": 4610
    },
    {
      "epoch": 6.03,
      "grad_norm": 3.768230438232422,
      "learning_rate": 1.2434652725914863e-05,
      "loss": 0.3965,
      "step": 4620
    },
    {
      "epoch": 6.03,
      "grad_norm": 19.785255432128906,
      "learning_rate": 1.2341299477221809e-05,
      "loss": 0.2529,
      "step": 4630
    },
    {
      "epoch": 6.03,
      "grad_norm": 12.09164047241211,
      "learning_rate": 1.2247946228528753e-05,
      "loss": 0.5038,
      "step": 4640
    },
    {
      "epoch": 6.03,
      "grad_norm": 25.119037628173828,
      "learning_rate": 1.21545929798357e-05,
      "loss": 0.385,
      "step": 4650
    },
    {
      "epoch": 6.03,
      "grad_norm": 8.757922172546387,
      "learning_rate": 1.2061239731142644e-05,
      "loss": 0.5379,
      "step": 4660
    },
    {
      "epoch": 6.03,
      "grad_norm": 21.042190551757812,
      "learning_rate": 1.196788648244959e-05,
      "loss": 0.4624,
      "step": 4670
    },
    {
      "epoch": 6.04,
      "grad_norm": 5.383489608764648,
      "learning_rate": 1.1874533233756534e-05,
      "loss": 0.4029,
      "step": 4680
    },
    {
      "epoch": 6.04,
      "grad_norm": 6.265172004699707,
      "learning_rate": 1.178117998506348e-05,
      "loss": 0.4337,
      "step": 4690
    },
    {
      "epoch": 6.04,
      "grad_norm": 5.571873664855957,
      "learning_rate": 1.1687826736370427e-05,
      "loss": 0.3015,
      "step": 4700
    },
    {
      "epoch": 6.04,
      "grad_norm": 2.3241143226623535,
      "learning_rate": 1.1594473487677373e-05,
      "loss": 0.2434,
      "step": 4710
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.4731499254703522,
      "learning_rate": 1.1501120238984317e-05,
      "loss": 0.4196,
      "step": 4720
    },
    {
      "epoch": 6.04,
      "grad_norm": 5.826499938964844,
      "learning_rate": 1.1407766990291263e-05,
      "loss": 0.3686,
      "step": 4730
    },
    {
      "epoch": 6.05,
      "grad_norm": 16.62709617614746,
      "learning_rate": 1.1314413741598208e-05,
      "loss": 0.3008,
      "step": 4740
    },
    {
      "epoch": 6.05,
      "grad_norm": 9.524761199951172,
      "learning_rate": 1.1221060492905154e-05,
      "loss": 0.6079,
      "step": 4750
    },
    {
      "epoch": 6.05,
      "grad_norm": 3.8646399974823,
      "learning_rate": 1.1127707244212099e-05,
      "loss": 0.3928,
      "step": 4760
    },
    {
      "epoch": 6.05,
      "grad_norm": 13.331137657165527,
      "learning_rate": 1.1034353995519045e-05,
      "loss": 0.3135,
      "step": 4770
    },
    {
      "epoch": 6.05,
      "grad_norm": 13.564393043518066,
      "learning_rate": 1.094100074682599e-05,
      "loss": 0.4167,
      "step": 4780
    },
    {
      "epoch": 6.05,
      "grad_norm": 20.557571411132812,
      "learning_rate": 1.0847647498132935e-05,
      "loss": 0.2962,
      "step": 4790
    },
    {
      "epoch": 6.06,
      "grad_norm": 10.7247953414917,
      "learning_rate": 1.0754294249439881e-05,
      "loss": 0.4288,
      "step": 4800
    },
    {
      "epoch": 6.06,
      "grad_norm": 14.957727432250977,
      "learning_rate": 1.0660941000746828e-05,
      "loss": 0.3464,
      "step": 4810
    },
    {
      "epoch": 6.06,
      "grad_norm": 13.989706039428711,
      "learning_rate": 1.0567587752053772e-05,
      "loss": 0.2492,
      "step": 4820
    },
    {
      "epoch": 6.06,
      "grad_norm": 16.13142967224121,
      "learning_rate": 1.0474234503360718e-05,
      "loss": 0.195,
      "step": 4830
    },
    {
      "epoch": 6.06,
      "grad_norm": 9.801595687866211,
      "learning_rate": 1.0380881254667663e-05,
      "loss": 0.3412,
      "step": 4840
    },
    {
      "epoch": 6.06,
      "grad_norm": 19.950620651245117,
      "learning_rate": 1.0287528005974609e-05,
      "loss": 0.5206,
      "step": 4850
    },
    {
      "epoch": 6.07,
      "grad_norm": 20.337913513183594,
      "learning_rate": 1.0194174757281553e-05,
      "loss": 0.3611,
      "step": 4860
    },
    {
      "epoch": 6.07,
      "grad_norm": 5.3233771324157715,
      "learning_rate": 1.01008215085885e-05,
      "loss": 0.4832,
      "step": 4870
    },
    {
      "epoch": 6.07,
      "grad_norm": 20.44731903076172,
      "learning_rate": 1.0007468259895444e-05,
      "loss": 0.2491,
      "step": 4880
    },
    {
      "epoch": 6.07,
      "grad_norm": 18.423215866088867,
      "learning_rate": 9.91411501120239e-06,
      "loss": 0.3625,
      "step": 4890
    },
    {
      "epoch": 6.07,
      "grad_norm": 22.186674118041992,
      "learning_rate": 9.820761762509336e-06,
      "loss": 0.2885,
      "step": 4900
    },
    {
      "epoch": 6.07,
      "grad_norm": 5.159964561462402,
      "learning_rate": 9.727408513816282e-06,
      "loss": 0.4301,
      "step": 4910
    },
    {
      "epoch": 6.08,
      "grad_norm": 14.131031036376953,
      "learning_rate": 9.634055265123227e-06,
      "loss": 0.2741,
      "step": 4920
    },
    {
      "epoch": 6.08,
      "grad_norm": 5.518490314483643,
      "learning_rate": 9.540702016430173e-06,
      "loss": 0.4064,
      "step": 4930
    },
    {
      "epoch": 6.08,
      "grad_norm": 25.204734802246094,
      "learning_rate": 9.447348767737117e-06,
      "loss": 0.4274,
      "step": 4940
    },
    {
      "epoch": 6.08,
      "grad_norm": 9.591646194458008,
      "learning_rate": 9.353995519044064e-06,
      "loss": 0.3826,
      "step": 4950
    },
    {
      "epoch": 6.08,
      "grad_norm": 15.941423416137695,
      "learning_rate": 9.260642270351008e-06,
      "loss": 0.3005,
      "step": 4960
    },
    {
      "epoch": 6.08,
      "grad_norm": 17.492700576782227,
      "learning_rate": 9.167289021657954e-06,
      "loss": 0.2674,
      "step": 4970
    },
    {
      "epoch": 6.09,
      "grad_norm": 17.261919021606445,
      "learning_rate": 9.073935772964899e-06,
      "loss": 0.5547,
      "step": 4980
    },
    {
      "epoch": 6.09,
      "grad_norm": 13.222940444946289,
      "learning_rate": 8.980582524271845e-06,
      "loss": 0.355,
      "step": 4990
    },
    {
      "epoch": 6.09,
      "grad_norm": 21.947750091552734,
      "learning_rate": 8.88722927557879e-06,
      "loss": 0.3682,
      "step": 5000
    },
    {
      "epoch": 6.09,
      "grad_norm": 18.26040267944336,
      "learning_rate": 8.793876026885735e-06,
      "loss": 0.2092,
      "step": 5010
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.5199944376945496,
      "learning_rate": 8.700522778192682e-06,
      "loss": 0.2366,
      "step": 5020
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.4301724135875702,
      "learning_rate": 8.607169529499628e-06,
      "loss": 0.6277,
      "step": 5030
    },
    {
      "epoch": 6.1,
      "grad_norm": 13.109355926513672,
      "learning_rate": 8.513816280806572e-06,
      "loss": 0.2838,
      "step": 5040
    },
    {
      "epoch": 6.1,
      "grad_norm": 21.817338943481445,
      "learning_rate": 8.420463032113518e-06,
      "loss": 0.4725,
      "step": 5050
    },
    {
      "epoch": 6.1,
      "grad_norm": 7.111277103424072,
      "learning_rate": 8.327109783420463e-06,
      "loss": 0.3472,
      "step": 5060
    },
    {
      "epoch": 6.1,
      "grad_norm": 28.631160736083984,
      "learning_rate": 8.233756534727409e-06,
      "loss": 0.4734,
      "step": 5070
    },
    {
      "epoch": 6.1,
      "grad_norm": 16.499799728393555,
      "learning_rate": 8.140403286034353e-06,
      "loss": 0.5065,
      "step": 5080
    },
    {
      "epoch": 6.1,
      "grad_norm": 8.197222709655762,
      "learning_rate": 8.0470500373413e-06,
      "loss": 0.3103,
      "step": 5090
    },
    {
      "epoch": 6.11,
      "grad_norm": 15.519693374633789,
      "learning_rate": 7.953696788648244e-06,
      "loss": 0.4342,
      "step": 5100
    },
    {
      "epoch": 6.11,
      "grad_norm": 26.50041389465332,
      "learning_rate": 7.86034353995519e-06,
      "loss": 0.5016,
      "step": 5110
    },
    {
      "epoch": 6.11,
      "grad_norm": 17.380603790283203,
      "learning_rate": 7.766990291262136e-06,
      "loss": 0.3865,
      "step": 5120
    },
    {
      "epoch": 6.11,
      "grad_norm": 17.17337989807129,
      "learning_rate": 7.673637042569082e-06,
      "loss": 0.4256,
      "step": 5130
    },
    {
      "epoch": 6.11,
      "grad_norm": 9.048469543457031,
      "learning_rate": 7.580283793876028e-06,
      "loss": 0.2988,
      "step": 5140
    },
    {
      "epoch": 6.11,
      "grad_norm": 20.14968490600586,
      "learning_rate": 7.486930545182973e-06,
      "loss": 0.2609,
      "step": 5150
    },
    {
      "epoch": 6.12,
      "grad_norm": 13.907716751098633,
      "learning_rate": 7.393577296489919e-06,
      "loss": 0.2558,
      "step": 5160
    },
    {
      "epoch": 6.12,
      "grad_norm": 11.087940216064453,
      "learning_rate": 7.300224047796864e-06,
      "loss": 0.4919,
      "step": 5170
    },
    {
      "epoch": 6.12,
      "grad_norm": 5.2559685707092285,
      "learning_rate": 7.20687079910381e-06,
      "loss": 0.3199,
      "step": 5180
    },
    {
      "epoch": 6.12,
      "grad_norm": 8.782883644104004,
      "learning_rate": 7.113517550410754e-06,
      "loss": 0.3154,
      "step": 5190
    },
    {
      "epoch": 6.12,
      "grad_norm": 21.707958221435547,
      "learning_rate": 7.0201643017177005e-06,
      "loss": 0.3547,
      "step": 5200
    },
    {
      "epoch": 6.12,
      "grad_norm": 12.79423713684082,
      "learning_rate": 6.926811053024646e-06,
      "loss": 0.5429,
      "step": 5210
    },
    {
      "epoch": 6.13,
      "eval_accuracy": 0.7863818424566088,
      "eval_loss": 0.6304479241371155,
      "eval_runtime": 437.2454,
      "eval_samples_per_second": 1.713,
      "eval_steps_per_second": 0.215,
      "step": 5215
    },
    {
      "epoch": 7.0,
      "grad_norm": 12.861729621887207,
      "learning_rate": 6.833457804331592e-06,
      "loss": 0.3125,
      "step": 5220
    },
    {
      "epoch": 7.0,
      "grad_norm": 16.386682510375977,
      "learning_rate": 6.740104555638536e-06,
      "loss": 0.3342,
      "step": 5230
    },
    {
      "epoch": 7.0,
      "grad_norm": 14.012133598327637,
      "learning_rate": 6.6467513069454825e-06,
      "loss": 0.3253,
      "step": 5240
    },
    {
      "epoch": 7.01,
      "grad_norm": 3.292506456375122,
      "learning_rate": 6.553398058252427e-06,
      "loss": 0.3425,
      "step": 5250
    },
    {
      "epoch": 7.01,
      "grad_norm": 7.2448201179504395,
      "learning_rate": 6.460044809559373e-06,
      "loss": 0.248,
      "step": 5260
    },
    {
      "epoch": 7.01,
      "grad_norm": 24.461353302001953,
      "learning_rate": 6.3666915608663185e-06,
      "loss": 0.6045,
      "step": 5270
    },
    {
      "epoch": 7.01,
      "grad_norm": 13.742588996887207,
      "learning_rate": 6.273338312173265e-06,
      "loss": 0.3643,
      "step": 5280
    },
    {
      "epoch": 7.01,
      "grad_norm": 4.182211875915527,
      "learning_rate": 6.179985063480209e-06,
      "loss": 0.2685,
      "step": 5290
    },
    {
      "epoch": 7.01,
      "grad_norm": 9.004427909851074,
      "learning_rate": 6.086631814787154e-06,
      "loss": 0.2518,
      "step": 5300
    },
    {
      "epoch": 7.02,
      "grad_norm": 2.794624090194702,
      "learning_rate": 5.9932785660941005e-06,
      "loss": 0.3494,
      "step": 5310
    },
    {
      "epoch": 7.02,
      "grad_norm": 15.006449699401855,
      "learning_rate": 5.899925317401046e-06,
      "loss": 0.3517,
      "step": 5320
    },
    {
      "epoch": 7.02,
      "grad_norm": 3.5482723712921143,
      "learning_rate": 5.806572068707991e-06,
      "loss": 0.2967,
      "step": 5330
    },
    {
      "epoch": 7.02,
      "grad_norm": 20.5437068939209,
      "learning_rate": 5.7132188200149364e-06,
      "loss": 0.3352,
      "step": 5340
    },
    {
      "epoch": 7.02,
      "grad_norm": 11.396394729614258,
      "learning_rate": 5.619865571321882e-06,
      "loss": 0.3083,
      "step": 5350
    },
    {
      "epoch": 7.02,
      "grad_norm": 21.825687408447266,
      "learning_rate": 5.526512322628828e-06,
      "loss": 0.3418,
      "step": 5360
    },
    {
      "epoch": 7.03,
      "grad_norm": 1.4731627702713013,
      "learning_rate": 5.433159073935773e-06,
      "loss": 0.2518,
      "step": 5370
    },
    {
      "epoch": 7.03,
      "grad_norm": 29.813344955444336,
      "learning_rate": 5.3398058252427185e-06,
      "loss": 0.4244,
      "step": 5380
    },
    {
      "epoch": 7.03,
      "grad_norm": 9.344219207763672,
      "learning_rate": 5.246452576549664e-06,
      "loss": 0.2505,
      "step": 5390
    },
    {
      "epoch": 7.03,
      "grad_norm": 23.584760665893555,
      "learning_rate": 5.153099327856609e-06,
      "loss": 0.3485,
      "step": 5400
    },
    {
      "epoch": 7.03,
      "grad_norm": 3.628610372543335,
      "learning_rate": 5.0597460791635544e-06,
      "loss": 0.3664,
      "step": 5410
    },
    {
      "epoch": 7.03,
      "grad_norm": 18.990201950073242,
      "learning_rate": 4.966392830470501e-06,
      "loss": 0.2745,
      "step": 5420
    },
    {
      "epoch": 7.04,
      "grad_norm": 1.5586326122283936,
      "learning_rate": 4.873039581777447e-06,
      "loss": 0.3675,
      "step": 5430
    },
    {
      "epoch": 7.04,
      "grad_norm": 5.46973991394043,
      "learning_rate": 4.779686333084392e-06,
      "loss": 0.4131,
      "step": 5440
    },
    {
      "epoch": 7.04,
      "grad_norm": 32.40712356567383,
      "learning_rate": 4.686333084391337e-06,
      "loss": 0.3872,
      "step": 5450
    },
    {
      "epoch": 7.04,
      "grad_norm": 8.046935081481934,
      "learning_rate": 4.592979835698283e-06,
      "loss": 0.3497,
      "step": 5460
    },
    {
      "epoch": 7.04,
      "grad_norm": 2.787285089492798,
      "learning_rate": 4.499626587005228e-06,
      "loss": 0.3917,
      "step": 5470
    },
    {
      "epoch": 7.04,
      "grad_norm": 1.2981932163238525,
      "learning_rate": 4.406273338312174e-06,
      "loss": 0.2249,
      "step": 5480
    },
    {
      "epoch": 7.05,
      "grad_norm": 24.799388885498047,
      "learning_rate": 4.312920089619119e-06,
      "loss": 0.4986,
      "step": 5490
    },
    {
      "epoch": 7.05,
      "grad_norm": 18.070804595947266,
      "learning_rate": 4.219566840926065e-06,
      "loss": 0.4264,
      "step": 5500
    },
    {
      "epoch": 7.05,
      "grad_norm": 18.745569229125977,
      "learning_rate": 4.12621359223301e-06,
      "loss": 0.2922,
      "step": 5510
    },
    {
      "epoch": 7.05,
      "grad_norm": 20.306697845458984,
      "learning_rate": 4.032860343539955e-06,
      "loss": 0.3761,
      "step": 5520
    },
    {
      "epoch": 7.05,
      "grad_norm": 3.735856533050537,
      "learning_rate": 3.9395070948469015e-06,
      "loss": 0.3175,
      "step": 5530
    },
    {
      "epoch": 7.05,
      "grad_norm": 11.890303611755371,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.2791,
      "step": 5540
    },
    {
      "epoch": 7.06,
      "grad_norm": 15.098954200744629,
      "learning_rate": 3.752800597460792e-06,
      "loss": 0.2752,
      "step": 5550
    },
    {
      "epoch": 7.06,
      "grad_norm": 16.720169067382812,
      "learning_rate": 3.6594473487677374e-06,
      "loss": 0.2833,
      "step": 5560
    },
    {
      "epoch": 7.06,
      "grad_norm": 14.178203582763672,
      "learning_rate": 3.566094100074683e-06,
      "loss": 0.248,
      "step": 5570
    },
    {
      "epoch": 7.06,
      "grad_norm": 10.082883834838867,
      "learning_rate": 3.4727408513816284e-06,
      "loss": 0.434,
      "step": 5580
    },
    {
      "epoch": 7.06,
      "grad_norm": 9.096635818481445,
      "learning_rate": 3.3793876026885737e-06,
      "loss": 0.3905,
      "step": 5590
    },
    {
      "epoch": 7.06,
      "grad_norm": 13.701728820800781,
      "learning_rate": 3.2860343539955195e-06,
      "loss": 0.2409,
      "step": 5600
    },
    {
      "epoch": 7.07,
      "grad_norm": 17.59809112548828,
      "learning_rate": 3.1926811053024648e-06,
      "loss": 0.285,
      "step": 5610
    },
    {
      "epoch": 7.07,
      "grad_norm": 11.866859436035156,
      "learning_rate": 3.09932785660941e-06,
      "loss": 0.2237,
      "step": 5620
    },
    {
      "epoch": 7.07,
      "grad_norm": 26.126977920532227,
      "learning_rate": 3.005974607916356e-06,
      "loss": 0.3001,
      "step": 5630
    },
    {
      "epoch": 7.07,
      "grad_norm": 5.371022701263428,
      "learning_rate": 2.912621359223301e-06,
      "loss": 0.2168,
      "step": 5640
    },
    {
      "epoch": 7.07,
      "grad_norm": 9.820216178894043,
      "learning_rate": 2.819268110530247e-06,
      "loss": 0.223,
      "step": 5650
    },
    {
      "epoch": 7.07,
      "grad_norm": 17.462440490722656,
      "learning_rate": 2.725914861837192e-06,
      "loss": 0.3205,
      "step": 5660
    },
    {
      "epoch": 7.08,
      "grad_norm": 23.553321838378906,
      "learning_rate": 2.6325616131441375e-06,
      "loss": 0.3914,
      "step": 5670
    },
    {
      "epoch": 7.08,
      "grad_norm": 12.580613136291504,
      "learning_rate": 2.539208364451083e-06,
      "loss": 0.2615,
      "step": 5680
    },
    {
      "epoch": 7.08,
      "grad_norm": 1.5106370449066162,
      "learning_rate": 2.4458551157580285e-06,
      "loss": 0.1088,
      "step": 5690
    },
    {
      "epoch": 7.08,
      "grad_norm": 15.336580276489258,
      "learning_rate": 2.352501867064974e-06,
      "loss": 0.256,
      "step": 5700
    },
    {
      "epoch": 7.08,
      "grad_norm": 8.88757610321045,
      "learning_rate": 2.2591486183719195e-06,
      "loss": 0.5456,
      "step": 5710
    },
    {
      "epoch": 7.08,
      "grad_norm": 1.397916555404663,
      "learning_rate": 2.165795369678865e-06,
      "loss": 0.4549,
      "step": 5720
    },
    {
      "epoch": 7.09,
      "grad_norm": 18.283489227294922,
      "learning_rate": 2.0724421209858106e-06,
      "loss": 0.3268,
      "step": 5730
    },
    {
      "epoch": 7.09,
      "grad_norm": 13.778904914855957,
      "learning_rate": 1.979088872292756e-06,
      "loss": 0.3873,
      "step": 5740
    },
    {
      "epoch": 7.09,
      "grad_norm": 16.9188175201416,
      "learning_rate": 1.8857356235997014e-06,
      "loss": 0.323,
      "step": 5750
    },
    {
      "epoch": 7.09,
      "grad_norm": 5.98646879196167,
      "learning_rate": 1.7923823749066467e-06,
      "loss": 0.15,
      "step": 5760
    },
    {
      "epoch": 7.09,
      "grad_norm": 12.303747177124023,
      "learning_rate": 1.6990291262135922e-06,
      "loss": 0.225,
      "step": 5770
    },
    {
      "epoch": 7.09,
      "grad_norm": 1.0051525831222534,
      "learning_rate": 1.6056758775205377e-06,
      "loss": 0.2016,
      "step": 5780
    },
    {
      "epoch": 7.1,
      "grad_norm": 4.382647514343262,
      "learning_rate": 1.5123226288274833e-06,
      "loss": 0.192,
      "step": 5790
    },
    {
      "epoch": 7.1,
      "grad_norm": 8.962074279785156,
      "learning_rate": 1.4189693801344288e-06,
      "loss": 0.4438,
      "step": 5800
    },
    {
      "epoch": 7.1,
      "grad_norm": 10.814083099365234,
      "learning_rate": 1.3256161314413743e-06,
      "loss": 0.2024,
      "step": 5810
    },
    {
      "epoch": 7.1,
      "grad_norm": 28.174787521362305,
      "learning_rate": 1.2322628827483198e-06,
      "loss": 0.2938,
      "step": 5820
    },
    {
      "epoch": 7.1,
      "grad_norm": 5.699619770050049,
      "learning_rate": 1.1389096340552653e-06,
      "loss": 0.1617,
      "step": 5830
    },
    {
      "epoch": 7.11,
      "grad_norm": 19.78471565246582,
      "learning_rate": 1.0455563853622106e-06,
      "loss": 0.5498,
      "step": 5840
    },
    {
      "epoch": 7.11,
      "grad_norm": 3.480914354324341,
      "learning_rate": 9.522031366691561e-07,
      "loss": 0.2463,
      "step": 5850
    },
    {
      "epoch": 7.11,
      "grad_norm": 2.4700801372528076,
      "learning_rate": 8.588498879761017e-07,
      "loss": 0.2854,
      "step": 5860
    },
    {
      "epoch": 7.11,
      "grad_norm": 2.976410388946533,
      "learning_rate": 7.654966392830471e-07,
      "loss": 0.3037,
      "step": 5870
    },
    {
      "epoch": 7.11,
      "grad_norm": 12.807668685913086,
      "learning_rate": 6.721433905899926e-07,
      "loss": 0.2541,
      "step": 5880
    },
    {
      "epoch": 7.11,
      "grad_norm": 8.027728080749512,
      "learning_rate": 5.78790141896938e-07,
      "loss": 0.298,
      "step": 5890
    },
    {
      "epoch": 7.12,
      "grad_norm": 14.356344223022461,
      "learning_rate": 4.854368932038835e-07,
      "loss": 0.3442,
      "step": 5900
    },
    {
      "epoch": 7.12,
      "grad_norm": 11.988846778869629,
      "learning_rate": 3.9208364451082904e-07,
      "loss": 0.1241,
      "step": 5910
    },
    {
      "epoch": 7.12,
      "grad_norm": 11.10311508178711,
      "learning_rate": 2.987303958177745e-07,
      "loss": 0.3483,
      "step": 5920
    },
    {
      "epoch": 7.12,
      "grad_norm": 19.410511016845703,
      "learning_rate": 2.0537714712471997e-07,
      "loss": 0.3085,
      "step": 5930
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.2636525332927704,
      "learning_rate": 1.1202389843166542e-07,
      "loss": 0.1939,
      "step": 5940
    },
    {
      "epoch": 7.12,
      "grad_norm": 9.14077091217041,
      "learning_rate": 1.8670649738610906e-08,
      "loss": 0.2831,
      "step": 5950
    },
    {
      "epoch": 7.12,
      "eval_accuracy": 0.7930574098798397,
      "eval_loss": 0.6373482346534729,
      "eval_runtime": 432.7751,
      "eval_samples_per_second": 1.731,
      "eval_steps_per_second": 0.217,
      "step": 5952
    },
    {
      "epoch": 7.12,
      "step": 5952,
      "total_flos": 5.927366869559732e+19,
      "train_loss": 0.7148157978049849,
      "train_runtime": 70229.849,
      "train_samples_per_second": 0.678,
      "train_steps_per_second": 0.085
    },
    {
      "epoch": 7.12,
      "eval_accuracy": 0.7643908969210174,
      "eval_loss": 0.7377305626869202,
      "eval_runtime": 456.3132,
      "eval_samples_per_second": 1.637,
      "eval_steps_per_second": 0.206,
      "step": 5952
    },
    {
      "epoch": 7.12,
      "eval_accuracy": 0.7643908969210174,
      "eval_loss": 0.7377305626869202,
      "eval_runtime": 450.9291,
      "eval_samples_per_second": 1.657,
      "eval_steps_per_second": 0.208,
      "step": 5952
    }
  ],
  "logging_steps": 10,
  "max_steps": 5952,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 5.927366869559732e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}