{
  "best_metric": 5.546974182128906,
  "best_model_checkpoint": "/users/hr1171/scratch/T5LA/checkpoint-100000",
  "epoch": 2.1069,
  "eval_steps": 1000,
  "global_step": 100000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005,
      "grad_norm": 0.06483318656682968,
      "learning_rate": 4.975e-05,
      "loss": 10.0248,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.07472355663776398,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.4056,
      "step": 1000
    },
    {
      "epoch": 0.01,
      "eval_accuracy": 0.043492843878108624,
      "eval_loss": 9.121541023254395,
      "eval_runtime": 110.4799,
      "eval_samples_per_second": 32.513,
      "eval_steps_per_second": 2.037,
      "step": 1000
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.08399348706007004,
      "learning_rate": 4.9250000000000004e-05,
      "loss": 8.8698,
      "step": 1500
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.08570190519094467,
      "learning_rate": 4.9e-05,
      "loss": 8.4062,
      "step": 2000
    },
    {
      "epoch": 0.02,
      "eval_accuracy": 0.04429728167514647,
      "eval_loss": 8.193856239318848,
      "eval_runtime": 112.0891,
      "eval_samples_per_second": 32.046,
      "eval_steps_per_second": 2.007,
      "step": 2000
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.08781777322292328,
      "learning_rate": 4.875e-05,
      "loss": 8.0175,
      "step": 2500
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.07474139332771301,
      "learning_rate": 4.85e-05,
      "loss": 7.7307,
      "step": 3000
    },
    {
      "epoch": 0.03,
      "eval_accuracy": 0.04438749518317016,
      "eval_loss": 7.602397918701172,
      "eval_runtime": 112.6714,
      "eval_samples_per_second": 31.88,
      "eval_steps_per_second": 1.997,
      "step": 3000
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.07133983075618744,
      "learning_rate": 4.825e-05,
      "loss": 7.5203,
      "step": 3500
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.04410657659173012,
      "learning_rate": 4.8e-05,
      "loss": 7.39,
      "step": 4000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.0443981085370553,
      "eval_loss": 7.333784103393555,
      "eval_runtime": 111.8336,
      "eval_samples_per_second": 32.119,
      "eval_steps_per_second": 2.012,
      "step": 4000
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.03503479063510895,
      "learning_rate": 4.775e-05,
      "loss": 7.3043,
      "step": 4500
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.039906181395053864,
      "learning_rate": 4.75e-05,
      "loss": 7.2546,
      "step": 5000
    },
    {
      "epoch": 0.05,
      "eval_accuracy": 0.04405276632986957,
      "eval_loss": 7.245168685913086,
      "eval_runtime": 113.4504,
      "eval_samples_per_second": 31.661,
      "eval_steps_per_second": 1.983,
      "step": 5000
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.04767516627907753,
      "learning_rate": 4.7249999999999997e-05,
      "loss": 7.2409,
      "step": 5500
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.07150296866893768,
      "learning_rate": 4.7e-05,
      "loss": 7.1985,
      "step": 6000
    },
    {
      "epoch": 0.06,
      "eval_accuracy": 0.03687500952480477,
      "eval_loss": 7.168196201324463,
      "eval_runtime": 110.8688,
      "eval_samples_per_second": 32.399,
      "eval_steps_per_second": 2.029,
      "step": 6000
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.10620597004890442,
      "learning_rate": 4.6750000000000005e-05,
      "loss": 7.1475,
      "step": 6500
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.09856470674276352,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 7.1009,
      "step": 7000
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.03457599379091584,
      "eval_loss": 7.071776866912842,
      "eval_runtime": 115.3179,
      "eval_samples_per_second": 31.149,
      "eval_steps_per_second": 1.951,
      "step": 7000
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.12105035036802292,
      "learning_rate": 4.6250000000000006e-05,
      "loss": 7.0493,
      "step": 7500
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.1322629749774933,
      "learning_rate": 4.600000000000001e-05,
      "loss": 7.004,
      "step": 8000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.03315856677269135,
      "eval_loss": 6.977820873260498,
      "eval_runtime": 110.4899,
      "eval_samples_per_second": 32.51,
      "eval_steps_per_second": 2.036,
      "step": 8000
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.16380475461483002,
      "learning_rate": 4.575e-05,
      "loss": 6.9535,
      "step": 8500
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.1642669439315796,
      "learning_rate": 4.55e-05,
      "loss": 6.9159,
      "step": 9000
    },
    {
      "epoch": 0.09,
      "eval_accuracy": 0.032452642670689945,
      "eval_loss": 6.896429061889648,
      "eval_runtime": 112.221,
      "eval_samples_per_second": 32.008,
      "eval_steps_per_second": 2.005,
      "step": 9000
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.2253541350364685,
      "learning_rate": 4.525e-05,
      "loss": 6.8866,
      "step": 9500
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.20757266879081726,
      "learning_rate": 4.5e-05,
      "loss": 6.8548,
      "step": 10000
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.03252571153012995,
      "eval_loss": 6.830728054046631,
      "eval_runtime": 112.9087,
      "eval_samples_per_second": 31.813,
      "eval_steps_per_second": 1.993,
      "step": 10000
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.21346400678157806,
      "learning_rate": 4.4750000000000004e-05,
      "loss": 6.8302,
      "step": 10500
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.24308614432811737,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 6.7833,
      "step": 11000
    },
    {
      "epoch": 0.11,
      "eval_accuracy": 0.032619326754142475,
      "eval_loss": 6.77016544342041,
      "eval_runtime": 112.3436,
      "eval_samples_per_second": 31.973,
      "eval_steps_per_second": 2.003,
      "step": 11000
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.27737605571746826,
      "learning_rate": 4.4250000000000005e-05,
      "loss": 6.7775,
      "step": 11500
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3234957158565521,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 6.7376,
      "step": 12000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.03369168370246034,
      "eval_loss": 6.716267108917236,
      "eval_runtime": 111.859,
      "eval_samples_per_second": 32.112,
      "eval_steps_per_second": 2.011,
      "step": 12000
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.27591678500175476,
      "learning_rate": 4.375e-05,
      "loss": 6.7188,
      "step": 12500
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.31338784098625183,
      "learning_rate": 4.35005e-05,
      "loss": 6.6821,
      "step": 13000
    },
    {
      "epoch": 0.13,
      "eval_accuracy": 0.03455980162280902,
      "eval_loss": 6.661470413208008,
      "eval_runtime": 120.7993,
      "eval_samples_per_second": 29.735,
      "eval_steps_per_second": 1.863,
      "step": 13000
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.513124406337738,
      "learning_rate": 4.32505e-05,
      "loss": 6.6565,
      "step": 13500
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.40841469168663025,
      "learning_rate": 4.30005e-05,
      "loss": 6.6373,
      "step": 14000
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.03493249362654492,
      "eval_loss": 6.608607292175293,
      "eval_runtime": 111.8763,
      "eval_samples_per_second": 32.107,
      "eval_steps_per_second": 2.011,
      "step": 14000
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.3946859538555145,
      "learning_rate": 4.2750500000000003e-05,
      "loss": 6.6188,
      "step": 14500
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.48516473174095154,
      "learning_rate": 4.2501000000000005e-05,
      "loss": 6.5895,
      "step": 15000
    },
    {
      "epoch": 0.15,
      "eval_accuracy": 0.0343823681168318,
      "eval_loss": 6.556947231292725,
      "eval_runtime": 112.7205,
      "eval_samples_per_second": 31.866,
      "eval_steps_per_second": 1.996,
      "step": 15000
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.543297529220581,
      "learning_rate": 4.22515e-05,
      "loss": 6.5555,
      "step": 15500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.46242478489875793,
      "learning_rate": 4.200150000000001e-05,
      "loss": 6.5421,
      "step": 16000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.035396487687420944,
      "eval_loss": 6.511915683746338,
      "eval_runtime": 111.3995,
      "eval_samples_per_second": 32.244,
      "eval_steps_per_second": 2.02,
      "step": 16000
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.5675227642059326,
      "learning_rate": 4.17515e-05,
      "loss": 6.5233,
      "step": 16500
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5619395971298218,
      "learning_rate": 4.15015e-05,
      "loss": 6.5051,
      "step": 17000
    },
    {
      "epoch": 0.17,
      "eval_accuracy": 0.035478945282990115,
      "eval_loss": 6.467820644378662,
      "eval_runtime": 111.7569,
      "eval_samples_per_second": 32.141,
      "eval_steps_per_second": 2.013,
      "step": 17000
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.4741845726966858,
      "learning_rate": 4.1252000000000004e-05,
      "loss": 6.491,
      "step": 17500
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5824158191680908,
      "learning_rate": 4.1002000000000005e-05,
      "loss": 6.4391,
      "step": 18000
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.03599301260322167,
      "eval_loss": 6.432408332824707,
      "eval_runtime": 113.2507,
      "eval_samples_per_second": 31.717,
      "eval_steps_per_second": 1.987,
      "step": 18000
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.740391194820404,
      "learning_rate": 4.0752e-05,
      "loss": 6.4602,
      "step": 18500
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.55320805311203,
      "learning_rate": 4.0502e-05,
      "loss": 6.4242,
      "step": 19000
    },
    {
      "epoch": 0.19,
      "eval_accuracy": 0.03552126262989112,
      "eval_loss": 6.40146541595459,
      "eval_runtime": 111.8361,
      "eval_samples_per_second": 32.118,
      "eval_steps_per_second": 2.012,
      "step": 19000
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.7065308094024658,
      "learning_rate": 4.0252e-05,
      "loss": 6.4047,
      "step": 19500
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6466010212898254,
      "learning_rate": 4.0003e-05,
      "loss": 6.3889,
      "step": 20000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.03732308355485308,
      "eval_loss": 6.355311393737793,
      "eval_runtime": 112.8526,
      "eval_samples_per_second": 31.829,
      "eval_steps_per_second": 1.994,
      "step": 20000
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.5905252695083618,
      "learning_rate": 3.9753000000000004e-05,
      "loss": 6.3693,
      "step": 20500
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5333964824676514,
      "learning_rate": 3.9503000000000004e-05,
      "loss": 6.3631,
      "step": 21000
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.036739757297089004,
      "eval_loss": 6.328461647033691,
      "eval_runtime": 112.8636,
      "eval_samples_per_second": 31.826,
      "eval_steps_per_second": 1.994,
      "step": 21000
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.5893262028694153,
      "learning_rate": 3.9253e-05,
      "loss": 6.3469,
      "step": 21500
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.596720278263092,
      "learning_rate": 3.9003e-05,
      "loss": 6.3296,
      "step": 22000
    },
    {
      "epoch": 0.22,
      "eval_accuracy": 0.03686983891650175,
      "eval_loss": 6.3014631271362305,
      "eval_runtime": 111.598,
      "eval_samples_per_second": 32.187,
      "eval_steps_per_second": 2.016,
      "step": 22000
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.49506717920303345,
      "learning_rate": 3.8753e-05,
      "loss": 6.3161,
      "step": 22500
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6392377614974976,
      "learning_rate": 3.8503e-05,
      "loss": 6.3081,
      "step": 23000
    },
    {
      "epoch": 0.23,
      "eval_accuracy": 0.03635019278204852,
      "eval_loss": 6.269931316375732,
      "eval_runtime": 112.6212,
      "eval_samples_per_second": 31.895,
      "eval_steps_per_second": 1.998,
      "step": 23000
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.5753453969955444,
      "learning_rate": 3.8253e-05,
      "loss": 6.2806,
      "step": 23500
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5371689796447754,
      "learning_rate": 3.80035e-05,
      "loss": 6.2784,
      "step": 24000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.036991620348901764,
      "eval_loss": 6.2454352378845215,
      "eval_runtime": 115.1822,
      "eval_samples_per_second": 31.185,
      "eval_steps_per_second": 1.953,
      "step": 24000
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.5090768337249756,
      "learning_rate": 3.7753500000000004e-05,
      "loss": 6.2632,
      "step": 24500
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5666926503181458,
      "learning_rate": 3.7503500000000004e-05,
      "loss": 6.2589,
      "step": 25000
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.03744921918371879,
      "eval_loss": 6.216660499572754,
      "eval_runtime": 112.1843,
      "eval_samples_per_second": 32.019,
      "eval_steps_per_second": 2.006,
      "step": 25000
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.5514094233512878,
      "learning_rate": 3.7253500000000005e-05,
      "loss": 6.2441,
      "step": 25500
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5683887600898743,
      "learning_rate": 3.7004e-05,
      "loss": 6.2371,
      "step": 26000
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.03695529002214109,
      "eval_loss": 6.189018726348877,
      "eval_runtime": 110.9014,
      "eval_samples_per_second": 32.389,
      "eval_steps_per_second": 2.029,
      "step": 26000
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.6400378346443176,
      "learning_rate": 3.6754e-05,
      "loss": 6.2257,
      "step": 26500
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5867771506309509,
      "learning_rate": 3.6504e-05,
      "loss": 6.1978,
      "step": 27000
    },
    {
      "epoch": 0.27,
      "eval_accuracy": 0.0376371299749416,
      "eval_loss": 6.165986061096191,
      "eval_runtime": 112.4076,
      "eval_samples_per_second": 31.955,
      "eval_steps_per_second": 2.002,
      "step": 27000
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.7539021372795105,
      "learning_rate": 3.62545e-05,
      "loss": 6.2078,
      "step": 27500
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.5702797174453735,
      "learning_rate": 3.6004500000000004e-05,
      "loss": 6.1895,
      "step": 28000
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.03748677412823544,
      "eval_loss": 6.1377997398376465,
      "eval_runtime": 111.0948,
      "eval_samples_per_second": 32.333,
      "eval_steps_per_second": 2.025,
      "step": 28000
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.5679642558097839,
      "learning_rate": 3.5754500000000005e-05,
      "loss": 6.1921,
      "step": 28500
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6666756868362427,
      "learning_rate": 3.55045e-05,
      "loss": 6.1636,
      "step": 29000
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.03662328254163156,
      "eval_loss": 6.121272087097168,
      "eval_runtime": 111.2726,
      "eval_samples_per_second": 32.281,
      "eval_steps_per_second": 2.022,
      "step": 29000
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.5925132632255554,
      "learning_rate": 3.52545e-05,
      "loss": 6.1639,
      "step": 29500
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5706653594970703,
      "learning_rate": 3.50045e-05,
      "loss": 6.1262,
      "step": 30000
    },
    {
      "epoch": 0.3,
      "eval_accuracy": 0.03699611061400702,
      "eval_loss": 6.096695899963379,
      "eval_runtime": 112.2709,
      "eval_samples_per_second": 31.994,
      "eval_steps_per_second": 2.004,
      "step": 30000
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.6464671492576599,
      "learning_rate": 3.47545e-05,
      "loss": 6.1365,
      "step": 30500
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.5408177375793457,
      "learning_rate": 3.4505e-05,
      "loss": 6.1345,
      "step": 31000
    },
    {
      "epoch": 0.31,
      "eval_accuracy": 0.03612323029127397,
      "eval_loss": 6.074455738067627,
      "eval_runtime": 113.9504,
      "eval_samples_per_second": 31.522,
      "eval_steps_per_second": 1.975,
      "step": 31000
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.5881854891777039,
      "learning_rate": 3.4255e-05,
      "loss": 6.1113,
      "step": 31500
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.6097517609596252,
      "learning_rate": 3.4005000000000004e-05,
      "loss": 6.1096,
      "step": 32000
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.03600621126125832,
      "eval_loss": 6.055612087249756,
      "eval_runtime": 111.7549,
      "eval_samples_per_second": 32.142,
      "eval_steps_per_second": 2.013,
      "step": 32000
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.6643844246864319,
      "learning_rate": 3.3755000000000005e-05,
      "loss": 6.1049,
      "step": 32500
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.6951597929000854,
      "learning_rate": 3.3505000000000005e-05,
      "loss": 6.0794,
      "step": 33000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.03573761176678053,
      "eval_loss": 6.041288375854492,
      "eval_runtime": 112.9798,
      "eval_samples_per_second": 31.793,
      "eval_steps_per_second": 1.992,
      "step": 33000
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.8083083033561707,
      "learning_rate": 3.3255000000000006e-05,
      "loss": 6.1019,
      "step": 33500
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.6118751764297485,
      "learning_rate": 3.3005e-05,
      "loss": 6.0643,
      "step": 34000
    },
    {
      "epoch": 0.34,
      "eval_accuracy": 0.03625793824443153,
      "eval_loss": 6.013612270355225,
      "eval_runtime": 111.1363,
      "eval_samples_per_second": 32.321,
      "eval_steps_per_second": 2.025,
      "step": 34000
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.6557193398475647,
      "learning_rate": 3.27555e-05,
      "loss": 6.0551,
      "step": 34500
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.5440818667411804,
      "learning_rate": 3.25055e-05,
      "loss": 6.057,
      "step": 35000
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.03619099247377141,
      "eval_loss": 5.996499538421631,
      "eval_runtime": 114.4917,
      "eval_samples_per_second": 31.373,
      "eval_steps_per_second": 1.965,
      "step": 35000
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.5865129828453064,
      "learning_rate": 3.2255499999999996e-05,
      "loss": 6.0495,
      "step": 35500
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.598119854927063,
      "learning_rate": 3.20055e-05,
      "loss": 6.0337,
      "step": 36000
    },
    {
      "epoch": 0.36,
      "eval_accuracy": 0.03535702778195055,
      "eval_loss": 5.980613708496094,
      "eval_runtime": 112.638,
      "eval_samples_per_second": 31.89,
      "eval_steps_per_second": 1.998,
      "step": 36000
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.5838174819946289,
      "learning_rate": 3.1756000000000006e-05,
      "loss": 6.0167,
      "step": 36500
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.7071039080619812,
      "learning_rate": 3.1506e-05,
      "loss": 6.0217,
      "step": 37000
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.03629073078656382,
      "eval_loss": 5.958355903625488,
      "eval_runtime": 112.2627,
      "eval_samples_per_second": 31.996,
      "eval_steps_per_second": 2.004,
      "step": 37000
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.606131374835968,
      "learning_rate": 3.1256e-05,
      "loss": 6.0086,
      "step": 37500
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.6173387169837952,
      "learning_rate": 3.1006e-05,
      "loss": 6.0045,
      "step": 38000
    },
    {
      "epoch": 0.38,
      "eval_accuracy": 0.035918719126025685,
      "eval_loss": 5.952617168426514,
      "eval_runtime": 116.0654,
      "eval_samples_per_second": 30.948,
      "eval_steps_per_second": 1.939,
      "step": 38000
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.6699332594871521,
      "learning_rate": 3.0756e-05,
      "loss": 5.9996,
      "step": 38500
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.6398167610168457,
      "learning_rate": 3.0506000000000002e-05,
      "loss": 5.9896,
      "step": 39000
    },
    {
      "epoch": 0.39,
      "eval_accuracy": 0.0355204462180538,
      "eval_loss": 5.9288482666015625,
      "eval_runtime": 110.5071,
      "eval_samples_per_second": 32.505,
      "eval_steps_per_second": 2.036,
      "step": 39000
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.7196072936058044,
      "learning_rate": 3.0256e-05,
      "loss": 5.9879,
      "step": 39500
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6065074801445007,
      "learning_rate": 3.0006e-05,
      "loss": 5.9711,
      "step": 40000
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.03516381031378517,
      "eval_loss": 5.915222644805908,
      "eval_runtime": 110.7645,
      "eval_samples_per_second": 32.429,
      "eval_steps_per_second": 2.031,
      "step": 40000
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.6913782954216003,
      "learning_rate": 2.9757e-05,
      "loss": 5.9671,
      "step": 40500
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.6546822190284729,
      "learning_rate": 2.9507e-05,
      "loss": 5.9629,
      "step": 41000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.03489303372107453,
      "eval_loss": 5.896161079406738,
      "eval_runtime": 111.3465,
      "eval_samples_per_second": 32.26,
      "eval_steps_per_second": 2.021,
      "step": 41000
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.6649766564369202,
      "learning_rate": 2.9257e-05,
      "loss": 5.9631,
      "step": 41500
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.7166668176651001,
      "learning_rate": 2.9007000000000002e-05,
      "loss": 5.9465,
      "step": 42000
    },
    {
      "epoch": 0.42,
      "eval_accuracy": 0.035887695476207584,
      "eval_loss": 5.882122993469238,
      "eval_runtime": 110.5783,
      "eval_samples_per_second": 32.484,
      "eval_steps_per_second": 2.035,
      "step": 42000
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.6148796081542969,
      "learning_rate": 2.8757e-05,
      "loss": 5.9383,
      "step": 42500
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.712359607219696,
      "learning_rate": 2.8507e-05,
      "loss": 5.9463,
      "step": 43000
    },
    {
      "epoch": 0.43,
      "eval_accuracy": 0.034527145149316284,
      "eval_loss": 5.869154930114746,
      "eval_runtime": 113.9173,
      "eval_samples_per_second": 31.532,
      "eval_steps_per_second": 1.975,
      "step": 43000
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.6819197535514832,
      "learning_rate": 2.8257e-05,
      "loss": 5.9363,
      "step": 43500
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7428186535835266,
      "learning_rate": 2.8007e-05,
      "loss": 5.9317,
      "step": 44000
    },
    {
      "epoch": 0.44,
      "eval_accuracy": 0.034260450615792234,
      "eval_loss": 5.869858741760254,
      "eval_runtime": 111.9079,
      "eval_samples_per_second": 32.098,
      "eval_steps_per_second": 2.011,
      "step": 44000
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.6326854825019836,
      "learning_rate": 2.7757500000000003e-05,
      "loss": 5.9114,
      "step": 44500
    },
    {
      "epoch": 1.00345,
      "grad_norm": 0.6722562313079834,
      "learning_rate": 2.7507500000000004e-05,
      "loss": 5.9097,
      "step": 45000
    },
    {
      "epoch": 1.00345,
      "eval_accuracy": 0.034564291887914274,
      "eval_loss": 5.8482561111450195,
      "eval_runtime": 111.3739,
      "eval_samples_per_second": 32.252,
      "eval_steps_per_second": 2.02,
      "step": 45000
    },
    {
      "epoch": 1.00845,
      "grad_norm": 0.6160574555397034,
      "learning_rate": 2.72575e-05,
      "loss": 5.8948,
      "step": 45500
    },
    {
      "epoch": 1.01345,
      "grad_norm": 0.6146988868713379,
      "learning_rate": 2.7007500000000002e-05,
      "loss": 5.9107,
      "step": 46000
    },
    {
      "epoch": 1.01345,
      "eval_accuracy": 0.03477846392657083,
      "eval_loss": 5.8351731300354,
      "eval_runtime": 112.8785,
      "eval_samples_per_second": 31.822,
      "eval_steps_per_second": 1.993,
      "step": 46000
    },
    {
      "epoch": 1.01845,
      "grad_norm": 0.6342368721961975,
      "learning_rate": 2.6758e-05,
      "loss": 5.893,
      "step": 46500
    },
    {
      "epoch": 1.02345,
      "grad_norm": 0.7240073680877686,
      "learning_rate": 2.6508e-05,
      "loss": 5.8838,
      "step": 47000
    },
    {
      "epoch": 1.02345,
      "eval_accuracy": 0.034256504625245196,
      "eval_loss": 5.818812847137451,
      "eval_runtime": 112.1297,
      "eval_samples_per_second": 32.034,
      "eval_steps_per_second": 2.007,
      "step": 47000
    },
    {
      "epoch": 1.02845,
      "grad_norm": 0.6988590955734253,
      "learning_rate": 2.6257999999999998e-05,
      "loss": 5.8907,
      "step": 47500
    },
    {
      "epoch": 1.03345,
      "grad_norm": 0.6514461636543274,
      "learning_rate": 2.6008e-05,
      "loss": 5.887,
      "step": 48000
    },
    {
      "epoch": 1.03345,
      "eval_accuracy": 0.03395511258863511,
      "eval_loss": 5.808635234832764,
      "eval_runtime": 112.3746,
      "eval_samples_per_second": 31.965,
      "eval_steps_per_second": 2.002,
      "step": 48000
    },
    {
      "epoch": 1.03845,
      "grad_norm": 0.6310043931007385,
      "learning_rate": 2.5758e-05,
      "loss": 5.873,
      "step": 48500
    },
    {
      "epoch": 1.04345,
      "grad_norm": 0.7601247429847717,
      "learning_rate": 2.5507999999999997e-05,
      "loss": 5.8563,
      "step": 49000
    },
    {
      "epoch": 1.04345,
      "eval_accuracy": 0.03384244775508516,
      "eval_loss": 5.797094821929932,
      "eval_runtime": 112.1133,
      "eval_samples_per_second": 32.039,
      "eval_steps_per_second": 2.007,
      "step": 49000
    },
    {
      "epoch": 1.04845,
      "grad_norm": 0.6604752540588379,
      "learning_rate": 2.5258500000000002e-05,
      "loss": 5.8583,
      "step": 49500
    },
    {
      "epoch": 1.05345,
      "grad_norm": 0.681904673576355,
      "learning_rate": 2.5008500000000003e-05,
      "loss": 5.8576,
      "step": 50000
    },
    {
      "epoch": 1.05345,
      "eval_accuracy": 0.03392898740984092,
      "eval_loss": 5.796751022338867,
      "eval_runtime": 113.244,
      "eval_samples_per_second": 31.719,
      "eval_steps_per_second": 1.987,
      "step": 50000
    },
    {
      "epoch": 1.0584500000000001,
      "grad_norm": 0.6337283849716187,
      "learning_rate": 2.47585e-05,
      "loss": 5.8495,
      "step": 50500
    },
    {
      "epoch": 1.06345,
      "grad_norm": 0.6972376108169556,
      "learning_rate": 2.45085e-05,
      "loss": 5.8567,
      "step": 51000
    },
    {
      "epoch": 1.06345,
      "eval_accuracy": 0.034259225998036255,
      "eval_loss": 5.779718399047852,
      "eval_runtime": 110.7791,
      "eval_samples_per_second": 32.425,
      "eval_steps_per_second": 2.031,
      "step": 51000
    },
    {
      "epoch": 1.06845,
      "grad_norm": 0.6530519127845764,
      "learning_rate": 2.42585e-05,
      "loss": 5.8519,
      "step": 51500
    },
    {
      "epoch": 1.07345,
      "grad_norm": 0.6328603625297546,
      "learning_rate": 2.4009e-05,
      "loss": 5.841,
      "step": 52000
    },
    {
      "epoch": 1.07345,
      "eval_accuracy": 0.033661476464479555,
      "eval_loss": 5.767716407775879,
      "eval_runtime": 114.652,
      "eval_samples_per_second": 31.33,
      "eval_steps_per_second": 1.962,
      "step": 52000
    },
    {
      "epoch": 1.07845,
      "grad_norm": 0.6603217124938965,
      "learning_rate": 2.3759e-05,
      "loss": 5.8284,
      "step": 52500
    },
    {
      "epoch": 1.08345,
      "grad_norm": 0.6852928400039673,
      "learning_rate": 2.3509e-05,
      "loss": 5.8192,
      "step": 53000
    },
    {
      "epoch": 1.08345,
      "eval_accuracy": 0.03317965741182208,
      "eval_loss": 5.76131534576416,
      "eval_runtime": 111.561,
      "eval_samples_per_second": 32.198,
      "eval_steps_per_second": 2.017,
      "step": 53000
    },
    {
      "epoch": 1.08845,
      "grad_norm": 0.6923466324806213,
      "learning_rate": 2.3259e-05,
      "loss": 5.8147,
      "step": 53500
    },
    {
      "epoch": 1.09345,
      "grad_norm": 0.7163240313529968,
      "learning_rate": 2.3009e-05,
      "loss": 5.8214,
      "step": 54000
    },
    {
      "epoch": 1.09345,
      "eval_accuracy": 0.03383822962725901,
      "eval_loss": 5.748566150665283,
      "eval_runtime": 113.5245,
      "eval_samples_per_second": 31.641,
      "eval_steps_per_second": 1.982,
      "step": 54000
    },
    {
      "epoch": 1.09845,
      "grad_norm": 0.7053197026252747,
      "learning_rate": 2.27595e-05,
      "loss": 5.8176,
      "step": 54500
    },
    {
      "epoch": 1.10345,
      "grad_norm": 0.7318024039268494,
      "learning_rate": 2.25095e-05,
      "loss": 5.8166,
      "step": 55000
    },
    {
      "epoch": 1.10345,
      "eval_accuracy": 0.033826935930176105,
      "eval_loss": 5.74090576171875,
      "eval_runtime": 112.6548,
      "eval_samples_per_second": 31.885,
      "eval_steps_per_second": 1.997,
      "step": 55000
    },
    {
      "epoch": 1.10845,
      "grad_norm": 0.727443516254425,
      "learning_rate": 2.22595e-05,
      "loss": 5.8054,
      "step": 55500
    },
    {
      "epoch": 1.11345,
      "grad_norm": 0.6675511598587036,
      "learning_rate": 2.2009500000000003e-05,
      "loss": 5.806,
      "step": 56000
    },
    {
      "epoch": 1.11345,
      "eval_accuracy": 0.03327177588079952,
      "eval_loss": 5.734244346618652,
      "eval_runtime": 114.861,
      "eval_samples_per_second": 31.273,
      "eval_steps_per_second": 1.959,
      "step": 56000
    },
    {
      "epoch": 1.11845,
      "grad_norm": 0.7548192143440247,
      "learning_rate": 2.17595e-05,
      "loss": 5.7987,
      "step": 56500
    },
    {
      "epoch": 1.12345,
      "grad_norm": 0.6868897676467896,
      "learning_rate": 2.15095e-05,
      "loss": 5.7961,
      "step": 57000
    },
    {
      "epoch": 1.12345,
      "eval_accuracy": 0.033481321585711266,
      "eval_loss": 5.723623752593994,
      "eval_runtime": 113.4262,
      "eval_samples_per_second": 31.668,
      "eval_steps_per_second": 1.984,
      "step": 57000
    },
    {
      "epoch": 1.12845,
      "grad_norm": 0.6014879941940308,
      "learning_rate": 2.1259500000000002e-05,
      "loss": 5.7892,
      "step": 57500
    },
    {
      "epoch": 1.13345,
      "grad_norm": 0.6603185534477234,
      "learning_rate": 2.101e-05,
      "loss": 5.7847,
      "step": 58000
    },
    {
      "epoch": 1.13345,
      "eval_accuracy": 0.033258985428681526,
      "eval_loss": 5.716407775878906,
      "eval_runtime": 112.5491,
      "eval_samples_per_second": 31.915,
      "eval_steps_per_second": 1.999,
      "step": 58000
    },
    {
      "epoch": 1.13845,
      "grad_norm": 0.7514461278915405,
      "learning_rate": 2.076e-05,
      "loss": 5.7785,
      "step": 58500
    },
    {
      "epoch": 1.14345,
      "grad_norm": 0.9591528177261353,
      "learning_rate": 2.0510000000000002e-05,
      "loss": 5.787,
      "step": 59000
    },
    {
      "epoch": 1.14345,
      "eval_accuracy": 0.03300685023958966,
      "eval_loss": 5.7095818519592285,
      "eval_runtime": 112.4202,
      "eval_samples_per_second": 31.952,
      "eval_steps_per_second": 2.001,
      "step": 59000
    },
    {
      "epoch": 1.14845,
      "grad_norm": 0.6442362666130066,
      "learning_rate": 2.0260000000000003e-05,
      "loss": 5.7765,
      "step": 59500
    },
    {
      "epoch": 1.15345,
      "grad_norm": 0.7159613370895386,
      "learning_rate": 2.001e-05,
      "loss": 5.7711,
      "step": 60000
    },
    {
      "epoch": 1.15345,
      "eval_accuracy": 0.03279621598556148,
      "eval_loss": 5.703481674194336,
      "eval_runtime": 112.2964,
      "eval_samples_per_second": 31.987,
      "eval_steps_per_second": 2.004,
      "step": 60000
    },
    {
      "epoch": 1.15845,
      "grad_norm": 0.6875385642051697,
      "learning_rate": 1.976e-05,
      "loss": 5.7593,
      "step": 60500
    },
    {
      "epoch": 1.16345,
      "grad_norm": 0.6751101016998291,
      "learning_rate": 1.951e-05,
      "loss": 5.7699,
      "step": 61000
    },
    {
      "epoch": 1.16345,
      "eval_accuracy": 0.03306223017588777,
      "eval_loss": 5.6887898445129395,
      "eval_runtime": 113.7541,
      "eval_samples_per_second": 31.577,
      "eval_steps_per_second": 1.978,
      "step": 61000
    },
    {
      "epoch": 1.16845,
      "grad_norm": 0.7514244914054871,
      "learning_rate": 1.92605e-05,
      "loss": 5.7672,
      "step": 61500
    },
    {
      "epoch": 1.1734499999999999,
      "grad_norm": 0.697762131690979,
      "learning_rate": 1.90105e-05,
      "loss": 5.763,
      "step": 62000
    },
    {
      "epoch": 1.1734499999999999,
      "eval_accuracy": 0.03338308002795394,
      "eval_loss": 5.687499046325684,
      "eval_runtime": 111.5862,
      "eval_samples_per_second": 32.19,
      "eval_steps_per_second": 2.016,
      "step": 62000
    },
    {
      "epoch": 1.17845,
      "grad_norm": 0.6761494874954224,
      "learning_rate": 1.87605e-05,
      "loss": 5.7669,
      "step": 62500
    },
    {
      "epoch": 1.1834500000000001,
      "grad_norm": 0.6467716693878174,
      "learning_rate": 1.8510500000000002e-05,
      "loss": 5.7434,
      "step": 63000
    },
    {
      "epoch": 1.1834500000000001,
      "eval_accuracy": 0.03302726053552262,
      "eval_loss": 5.680927276611328,
      "eval_runtime": 112.7327,
      "eval_samples_per_second": 31.863,
      "eval_steps_per_second": 1.996,
      "step": 63000
    },
    {
      "epoch": 1.18845,
      "grad_norm": 0.7104445695877075,
      "learning_rate": 1.82605e-05,
      "loss": 5.7506,
      "step": 63500
    },
    {
      "epoch": 1.19345,
      "grad_norm": 0.6914287805557251,
      "learning_rate": 1.8011e-05,
      "loss": 5.7477,
      "step": 64000
    },
    {
      "epoch": 1.19345,
      "eval_accuracy": 0.032914323564693565,
      "eval_loss": 5.668553829193115,
      "eval_runtime": 113.1627,
      "eval_samples_per_second": 31.742,
      "eval_steps_per_second": 1.988,
      "step": 64000
    },
    {
      "epoch": 1.19845,
      "grad_norm": 0.7188768982887268,
      "learning_rate": 1.7760999999999998e-05,
      "loss": 5.7329,
      "step": 64500
    },
    {
      "epoch": 1.20345,
      "grad_norm": 0.6479863524436951,
      "learning_rate": 1.7511e-05,
      "loss": 5.7409,
      "step": 65000
    },
    {
      "epoch": 1.20345,
      "eval_accuracy": 0.03304100346811748,
      "eval_loss": 5.662350654602051,
      "eval_runtime": 110.8399,
      "eval_samples_per_second": 32.407,
      "eval_steps_per_second": 2.03,
      "step": 65000
    },
    {
      "epoch": 1.20845,
      "grad_norm": 0.678534209728241,
      "learning_rate": 1.7261000000000003e-05,
      "loss": 5.7384,
      "step": 65500
    },
    {
      "epoch": 1.21345,
      "grad_norm": 0.8188093900680542,
      "learning_rate": 1.7011e-05,
      "loss": 5.737,
      "step": 66000
    },
    {
      "epoch": 1.21345,
      "eval_accuracy": 0.03385346931488896,
      "eval_loss": 5.675750732421875,
      "eval_runtime": 111.9995,
      "eval_samples_per_second": 32.072,
      "eval_steps_per_second": 2.009,
      "step": 66000
    },
    {
      "epoch": 1.21845,
      "grad_norm": 0.7928422689437866,
      "learning_rate": 1.6761e-05,
      "loss": 5.7299,
      "step": 66500
    },
    {
      "epoch": 1.22345,
      "grad_norm": 0.7139099836349487,
      "learning_rate": 1.6511e-05,
      "loss": 5.729,
      "step": 67000
    },
    {
      "epoch": 1.22345,
      "eval_accuracy": 0.032598372183651296,
      "eval_loss": 5.654570579528809,
      "eval_runtime": 111.7834,
      "eval_samples_per_second": 32.134,
      "eval_steps_per_second": 2.013,
      "step": 67000
    },
    {
      "epoch": 1.22845,
      "grad_norm": 0.6944179534912109,
      "learning_rate": 1.6261000000000002e-05,
      "loss": 5.7183,
      "step": 67500
    },
    {
      "epoch": 1.23345,
      "grad_norm": 0.7094106078147888,
      "learning_rate": 1.60115e-05,
      "loss": 5.7232,
      "step": 68000
    },
    {
      "epoch": 1.23345,
      "eval_accuracy": 0.032930243595521276,
      "eval_loss": 5.646746635437012,
      "eval_runtime": 117.6522,
      "eval_samples_per_second": 30.531,
      "eval_steps_per_second": 1.912,
      "step": 68000
    },
    {
      "epoch": 1.23845,
      "grad_norm": 0.6927877068519592,
      "learning_rate": 1.57615e-05,
      "loss": 5.7162,
      "step": 68500
    },
    {
      "epoch": 1.24345,
      "grad_norm": 0.7557797431945801,
      "learning_rate": 1.5511500000000002e-05,
      "loss": 5.7127,
      "step": 69000
    },
    {
      "epoch": 1.24345,
      "eval_accuracy": 0.032887245905422496,
      "eval_loss": 5.644942283630371,
      "eval_runtime": 111.9326,
      "eval_samples_per_second": 32.091,
      "eval_steps_per_second": 2.01,
      "step": 69000
    },
    {
      "epoch": 1.24845,
      "grad_norm": 0.673581063747406,
      "learning_rate": 1.52615e-05,
      "loss": 5.7094,
      "step": 69500
    },
    {
      "epoch": 1.25345,
      "grad_norm": 0.6678490042686462,
      "learning_rate": 1.50115e-05,
      "loss": 5.7187,
      "step": 70000
    },
    {
      "epoch": 1.25345,
      "eval_accuracy": 0.03288316384623591,
      "eval_loss": 5.635218143463135,
      "eval_runtime": 116.4191,
      "eval_samples_per_second": 30.854,
      "eval_steps_per_second": 1.933,
      "step": 70000
    },
    {
      "epoch": 1.25845,
      "grad_norm": 0.7218269109725952,
      "learning_rate": 1.4761500000000001e-05,
      "loss": 5.7138,
      "step": 70500
    },
    {
      "epoch": 1.26345,
      "grad_norm": 0.7058696150779724,
      "learning_rate": 1.4512000000000001e-05,
      "loss": 5.717,
      "step": 71000
    },
    {
      "epoch": 1.26345,
      "eval_accuracy": 0.03260463134107074,
      "eval_loss": 5.626367568969727,
      "eval_runtime": 111.4646,
      "eval_samples_per_second": 32.225,
      "eval_steps_per_second": 2.019,
      "step": 71000
    },
    {
      "epoch": 1.26845,
      "grad_norm": 0.7154064774513245,
      "learning_rate": 1.4262e-05,
      "loss": 5.6955,
      "step": 71500
    },
    {
      "epoch": 1.27345,
      "grad_norm": 0.6907570362091064,
      "learning_rate": 1.4012e-05,
      "loss": 5.714,
      "step": 72000
    },
    {
      "epoch": 1.27345,
      "eval_accuracy": 0.03297637086432977,
      "eval_loss": 5.6219401359558105,
      "eval_runtime": 113.1231,
      "eval_samples_per_second": 31.753,
      "eval_steps_per_second": 1.989,
      "step": 72000
    },
    {
      "epoch": 1.2784499999999999,
      "grad_norm": 0.708003580570221,
      "learning_rate": 1.3762e-05,
      "loss": 5.6993,
      "step": 72500
    },
    {
      "epoch": 1.28345,
      "grad_norm": 0.8350797295570374,
      "learning_rate": 1.35125e-05,
      "loss": 5.7079,
      "step": 73000
    },
    {
      "epoch": 1.28345,
      "eval_accuracy": 0.032977323344806644,
      "eval_loss": 5.616916656494141,
      "eval_runtime": 112.8971,
      "eval_samples_per_second": 31.817,
      "eval_steps_per_second": 1.993,
      "step": 73000
    },
    {
      "epoch": 1.28845,
      "grad_norm": 0.6496825218200684,
      "learning_rate": 1.32625e-05,
      "loss": 5.7047,
      "step": 73500
    },
    {
      "epoch": 1.29345,
      "grad_norm": 0.7288230657577515,
      "learning_rate": 1.30125e-05,
      "loss": 5.7034,
      "step": 74000
    },
    {
      "epoch": 1.29345,
      "eval_accuracy": 0.032624225225166385,
      "eval_loss": 5.613090991973877,
      "eval_runtime": 112.7724,
      "eval_samples_per_second": 31.852,
      "eval_steps_per_second": 1.995,
      "step": 74000
    },
    {
      "epoch": 1.2984499999999999,
      "grad_norm": 0.6955094337463379,
      "learning_rate": 1.27625e-05,
      "loss": 5.6884,
      "step": 74500
    },
    {
      "epoch": 1.30345,
      "grad_norm": 0.6996705532073975,
      "learning_rate": 1.25125e-05,
      "loss": 5.6768,
      "step": 75000
    },
    {
      "epoch": 1.30345,
      "eval_accuracy": 0.03247931212404235,
      "eval_loss": 5.61249303817749,
      "eval_runtime": 113.0871,
      "eval_samples_per_second": 31.763,
      "eval_steps_per_second": 1.99,
      "step": 75000
    },
    {
      "epoch": 1.3084500000000001,
      "grad_norm": 0.7164750695228577,
      "learning_rate": 1.22625e-05,
      "loss": 5.685,
      "step": 75500
    },
    {
      "epoch": 1.31345,
      "grad_norm": 0.6893213391304016,
      "learning_rate": 1.20125e-05,
      "loss": 5.6955,
      "step": 76000
    },
    {
      "epoch": 1.31345,
      "eval_accuracy": 0.0328188033797273,
      "eval_loss": 5.6074957847595215,
      "eval_runtime": 114.2316,
      "eval_samples_per_second": 31.445,
      "eval_steps_per_second": 1.97,
      "step": 76000
    },
    {
      "epoch": 1.31845,
      "grad_norm": 0.7451682090759277,
      "learning_rate": 1.17625e-05,
      "loss": 5.6853,
      "step": 76500
    },
    {
      "epoch": 1.32345,
      "grad_norm": 0.7856729626655579,
      "learning_rate": 1.15125e-05,
      "loss": 5.6947,
      "step": 77000
    },
    {
      "epoch": 1.32345,
      "eval_accuracy": 0.032477407163088605,
      "eval_loss": 5.601708889007568,
      "eval_runtime": 111.6415,
      "eval_samples_per_second": 32.174,
      "eval_steps_per_second": 2.015,
      "step": 77000
    },
    {
      "epoch": 1.3284500000000001,
      "grad_norm": 0.7268499135971069,
      "learning_rate": 1.1262500000000001e-05,
      "loss": 5.6676,
      "step": 77500
    },
    {
      "epoch": 1.33345,
      "grad_norm": 0.6918110847473145,
      "learning_rate": 1.1013000000000001e-05,
      "loss": 5.7056,
      "step": 78000
    },
    {
      "epoch": 1.33345,
      "eval_accuracy": 0.032296980147041215,
      "eval_loss": 5.595623970031738,
      "eval_runtime": 110.9779,
      "eval_samples_per_second": 32.367,
      "eval_steps_per_second": 2.027,
      "step": 78000
    },
    {
      "epoch": 1.33845,
      "grad_norm": 0.7075666189193726,
      "learning_rate": 1.0763e-05,
      "loss": 5.6793,
      "step": 78500
    },
    {
      "epoch": 1.34345,
      "grad_norm": 0.7042005658149719,
      "learning_rate": 1.0513e-05,
      "loss": 5.6636,
      "step": 79000
    },
    {
      "epoch": 1.34345,
      "eval_accuracy": 0.03255183670892414,
      "eval_loss": 5.592087268829346,
      "eval_runtime": 111.6356,
      "eval_samples_per_second": 32.176,
      "eval_steps_per_second": 2.015,
      "step": 79000
    },
    {
      "epoch": 1.34845,
      "grad_norm": 0.7470083832740784,
      "learning_rate": 1.0263e-05,
      "loss": 5.6727,
      "step": 79500
    },
    {
      "epoch": 1.35345,
      "grad_norm": 0.7401617169380188,
      "learning_rate": 1.00135e-05,
      "loss": 5.6723,
      "step": 80000
    },
    {
      "epoch": 1.35345,
      "eval_accuracy": 0.03255442201307565,
      "eval_loss": 5.588088512420654,
      "eval_runtime": 112.4329,
      "eval_samples_per_second": 31.948,
      "eval_steps_per_second": 2.001,
      "step": 80000
    },
    {
      "epoch": 1.35845,
      "grad_norm": 0.7123810648918152,
      "learning_rate": 9.7635e-06,
      "loss": 5.6695,
      "step": 80500
    },
    {
      "epoch": 1.36345,
      "grad_norm": 0.7234753966331482,
      "learning_rate": 9.5135e-06,
      "loss": 5.659,
      "step": 81000
    },
    {
      "epoch": 1.36345,
      "eval_accuracy": 0.03243753905169955,
      "eval_loss": 5.5822529792785645,
      "eval_runtime": 115.5372,
      "eval_samples_per_second": 31.09,
      "eval_steps_per_second": 1.947,
      "step": 81000
    },
    {
      "epoch": 1.36845,
      "grad_norm": 0.7145719528198242,
      "learning_rate": 9.2635e-06,
      "loss": 5.6632,
      "step": 81500
    },
    {
      "epoch": 1.37345,
      "grad_norm": 0.7814493179321289,
      "learning_rate": 9.013500000000001e-06,
      "loss": 5.6729,
      "step": 82000
    },
    {
      "epoch": 1.37345,
      "eval_accuracy": 0.032616741449990966,
      "eval_loss": 5.579476833343506,
      "eval_runtime": 111.3925,
      "eval_samples_per_second": 32.246,
      "eval_steps_per_second": 2.02,
      "step": 82000
    },
    {
      "epoch": 1.37845,
      "grad_norm": 0.7196946144104004,
      "learning_rate": 8.7635e-06,
      "loss": 5.6638,
      "step": 82500
    },
    {
      "epoch": 1.38345,
      "grad_norm": 0.7334076762199402,
      "learning_rate": 8.514e-06,
      "loss": 5.6595,
      "step": 83000
    },
    {
      "epoch": 1.38345,
      "eval_accuracy": 0.03224581833856925,
      "eval_loss": 5.579442977905273,
      "eval_runtime": 111.9776,
      "eval_samples_per_second": 32.078,
      "eval_steps_per_second": 2.009,
      "step": 83000
    },
    {
      "epoch": 1.38845,
      "grad_norm": 0.797461748123169,
      "learning_rate": 8.264e-06,
      "loss": 5.66,
      "step": 83500
    },
    {
      "epoch": 1.39345,
      "grad_norm": 0.7179501056671143,
      "learning_rate": 8.014e-06,
      "loss": 5.6565,
      "step": 84000
    },
    {
      "epoch": 1.39345,
      "eval_accuracy": 0.032768049777174,
      "eval_loss": 5.575778961181641,
      "eval_runtime": 113.2284,
      "eval_samples_per_second": 31.723,
      "eval_steps_per_second": 1.987,
      "step": 84000
    },
    {
      "epoch": 1.39845,
      "grad_norm": 0.7444576025009155,
      "learning_rate": 7.764e-06,
      "loss": 5.6539,
      "step": 84500
    },
    {
      "epoch": 1.4034499999999999,
      "grad_norm": 0.681348979473114,
      "learning_rate": 7.514500000000001e-06,
      "loss": 5.6649,
      "step": 85000
    },
    {
      "epoch": 1.4034499999999999,
      "eval_accuracy": 0.03250598157739475,
      "eval_loss": 5.571649074554443,
      "eval_runtime": 110.7941,
      "eval_samples_per_second": 32.42,
      "eval_steps_per_second": 2.031,
      "step": 85000
    },
    {
      "epoch": 1.40845,
      "grad_norm": 0.6666921973228455,
      "learning_rate": 7.2645000000000005e-06,
      "loss": 5.6487,
      "step": 85500
    },
    {
      "epoch": 1.41345,
      "grad_norm": 0.7330692410469055,
      "learning_rate": 7.0145e-06,
      "loss": 5.6561,
      "step": 86000
    },
    {
      "epoch": 1.41345,
      "eval_accuracy": 0.032128118965355834,
      "eval_loss": 5.5695481300354,
      "eval_runtime": 113.9227,
      "eval_samples_per_second": 31.53,
      "eval_steps_per_second": 1.975,
      "step": 86000
    },
    {
      "epoch": 1.41845,
      "grad_norm": 0.7140607237815857,
      "learning_rate": 6.7645e-06,
      "loss": 5.6569,
      "step": 86500
    },
    {
      "epoch": 1.4234499999999999,
      "grad_norm": 0.695405125617981,
      "learning_rate": 6.5145e-06,
      "loss": 5.6405,
      "step": 87000
    },
    {
      "epoch": 1.4234499999999999,
      "eval_accuracy": 0.03226895000729328,
      "eval_loss": 5.565379619598389,
      "eval_runtime": 110.9885,
      "eval_samples_per_second": 32.364,
      "eval_steps_per_second": 2.027,
      "step": 87000
    },
    {
      "epoch": 1.42845,
      "grad_norm": 0.748406708240509,
      "learning_rate": 6.265e-06,
      "loss": 5.661,
      "step": 87500
    },
    {
      "epoch": 1.4334500000000001,
      "grad_norm": 0.7667502760887146,
      "learning_rate": 6.015000000000001e-06,
      "loss": 5.6482,
      "step": 88000
    },
    {
      "epoch": 1.4334500000000001,
      "eval_accuracy": 0.03212390083752969,
      "eval_loss": 5.562798023223877,
      "eval_runtime": 113.7863,
      "eval_samples_per_second": 31.568,
      "eval_steps_per_second": 1.977,
      "step": 88000
    },
    {
      "epoch": 1.43845,
      "grad_norm": 0.7025783658027649,
      "learning_rate": 5.765e-06,
      "loss": 5.6537,
      "step": 88500
    },
    {
      "epoch": 1.44345,
      "grad_norm": 0.7218544483184814,
      "learning_rate": 5.515e-06,
      "loss": 5.6425,
      "step": 89000
    },
    {
      "epoch": 1.44345,
      "eval_accuracy": 0.03228555038131876,
      "eval_loss": 5.562201023101807,
      "eval_runtime": 110.7556,
      "eval_samples_per_second": 32.432,
      "eval_steps_per_second": 2.032,
      "step": 89000
    },
    {
      "epoch": 2.0019,
      "grad_norm": 0.7732229232788086,
      "learning_rate": 5.265e-06,
      "loss": 5.6439,
      "step": 89500
    },
    {
      "epoch": 2.0069,
      "grad_norm": 0.7334641814231873,
      "learning_rate": 5.015e-06,
      "loss": 5.6379,
      "step": 90000
    },
    {
      "epoch": 2.0069,
      "eval_accuracy": 0.032302831098542,
      "eval_loss": 5.558170318603516,
      "eval_runtime": 110.8037,
      "eval_samples_per_second": 32.418,
      "eval_steps_per_second": 2.031,
      "step": 90000
    },
    {
      "epoch": 2.0119,
      "grad_norm": 0.7253163456916809,
      "learning_rate": 4.765e-06,
      "loss": 5.6416,
      "step": 90500
    },
    {
      "epoch": 2.0169,
      "grad_norm": 0.6948328614234924,
      "learning_rate": 4.515000000000001e-06,
      "loss": 5.6357,
      "step": 91000
    },
    {
      "epoch": 2.0169,
      "eval_accuracy": 0.03219384011825997,
      "eval_loss": 5.557282447814941,
      "eval_runtime": 112.1174,
      "eval_samples_per_second": 32.038,
      "eval_steps_per_second": 2.007,
      "step": 91000
    },
    {
      "epoch": 2.0219,
      "grad_norm": 0.8627682328224182,
      "learning_rate": 4.2655e-06,
      "loss": 5.6417,
      "step": 91500
    },
    {
      "epoch": 2.0269,
      "grad_norm": 0.6853375434875488,
      "learning_rate": 4.015500000000001e-06,
      "loss": 5.6381,
      "step": 92000
    },
    {
      "epoch": 2.0269,
      "eval_accuracy": 0.032045117095228455,
      "eval_loss": 5.556839942932129,
      "eval_runtime": 112.6404,
      "eval_samples_per_second": 31.889,
      "eval_steps_per_second": 1.998,
      "step": 92000
    },
    {
      "epoch": 2.0319,
      "grad_norm": 0.7689598798751831,
      "learning_rate": 3.7655e-06,
      "loss": 5.6349,
      "step": 92500
    },
    {
      "epoch": 2.0369,
      "grad_norm": 0.714249849319458,
      "learning_rate": 3.516e-06,
      "loss": 5.6427,
      "step": 93000
    },
    {
      "epoch": 2.0369,
      "eval_accuracy": 0.032379982017168595,
      "eval_loss": 5.5526018142700195,
      "eval_runtime": 114.2105,
      "eval_samples_per_second": 31.451,
      "eval_steps_per_second": 1.97,
      "step": 93000
    },
    {
      "epoch": 2.0419,
      "grad_norm": 0.7157047390937805,
      "learning_rate": 3.266e-06,
      "loss": 5.6238,
      "step": 93500
    },
    {
      "epoch": 2.0469,
      "grad_norm": 0.702506959438324,
      "learning_rate": 3.016e-06,
      "loss": 5.6364,
      "step": 94000
    },
    {
      "epoch": 2.0469,
      "eval_accuracy": 0.03230364751037931,
      "eval_loss": 5.55258321762085,
      "eval_runtime": 113.8093,
      "eval_samples_per_second": 31.562,
      "eval_steps_per_second": 1.977,
      "step": 94000
    },
    {
      "epoch": 2.0519,
      "grad_norm": 0.7739561200141907,
      "learning_rate": 2.7660000000000003e-06,
      "loss": 5.6289,
      "step": 94500
    },
    {
      "epoch": 2.0569,
      "grad_norm": 0.700249195098877,
      "learning_rate": 2.516e-06,
      "loss": 5.626,
      "step": 95000
    },
    {
      "epoch": 2.0569,
      "eval_accuracy": 0.032115600650516954,
      "eval_loss": 5.550052642822266,
      "eval_runtime": 112.9135,
      "eval_samples_per_second": 31.812,
      "eval_steps_per_second": 1.993,
      "step": 95000
    },
    {
      "epoch": 2.0619,
      "grad_norm": 0.694694995880127,
      "learning_rate": 2.266e-06,
      "loss": 5.6414,
      "step": 95500
    },
    {
      "epoch": 2.0669,
      "grad_norm": 0.8426607251167297,
      "learning_rate": 2.0165e-06,
      "loss": 5.636,
      "step": 96000
    },
    {
      "epoch": 2.0669,
      "eval_accuracy": 0.032368688320085694,
      "eval_loss": 5.549187183380127,
      "eval_runtime": 111.6948,
      "eval_samples_per_second": 32.159,
      "eval_steps_per_second": 2.014,
      "step": 96000
    },
    {
      "epoch": 2.0719,
      "grad_norm": 0.7448583841323853,
      "learning_rate": 1.7665000000000002e-06,
      "loss": 5.633,
      "step": 96500
    },
    {
      "epoch": 2.0769,
      "grad_norm": 0.7047140002250671,
      "learning_rate": 1.5165e-06,
      "loss": 5.632,
      "step": 97000
    },
    {
      "epoch": 2.0769,
      "eval_accuracy": 0.032333582611081,
      "eval_loss": 5.548858165740967,
      "eval_runtime": 111.3384,
      "eval_samples_per_second": 32.262,
      "eval_steps_per_second": 2.021,
      "step": 97000
    },
    {
      "epoch": 2.0819,
      "grad_norm": 0.7185714244842529,
      "learning_rate": 1.2665e-06,
      "loss": 5.6218,
      "step": 97500
    },
    {
      "epoch": 2.0869,
      "grad_norm": 0.906790018081665,
      "learning_rate": 1.0165000000000001e-06,
      "loss": 5.6133,
      "step": 98000
    },
    {
      "epoch": 2.0869,
      "eval_accuracy": 0.03233031696373172,
      "eval_loss": 5.547926425933838,
      "eval_runtime": 112.1516,
      "eval_samples_per_second": 32.028,
      "eval_steps_per_second": 2.006,
      "step": 98000
    },
    {
      "epoch": 2.0919,
      "grad_norm": 0.7219076752662659,
      "learning_rate": 7.67e-07,
      "loss": 5.6305,
      "step": 98500
    },
    {
      "epoch": 2.0969,
      "grad_norm": 0.7102829217910767,
      "learning_rate": 5.17e-07,
      "loss": 5.6291,
      "step": 99000
    },
    {
      "epoch": 2.0969,
      "eval_accuracy": 0.032251125015511826,
      "eval_loss": 5.54772424697876,
      "eval_runtime": 113.02,
      "eval_samples_per_second": 31.782,
      "eval_steps_per_second": 1.991,
      "step": 99000
    },
    {
      "epoch": 2.1019,
      "grad_norm": 0.7246349453926086,
      "learning_rate": 2.67e-07,
      "loss": 5.6351,
      "step": 99500
    },
    {
      "epoch": 2.1069,
      "grad_norm": 0.6908907294273376,
      "learning_rate": 1.7000000000000003e-08,
      "loss": 5.6271,
      "step": 100000
    },
    {
      "epoch": 2.1069,
      "eval_accuracy": 0.0322300343763811,
      "eval_loss": 5.546974182128906,
      "eval_runtime": 114.2737,
      "eval_samples_per_second": 31.433,
      "eval_steps_per_second": 1.969,
      "step": 100000
    },
    {
      "epoch": 2.1069,
      "step": 100000,
      "total_flos": 9.182034338135409e+17,
      "train_loss": 0.0,
      "train_runtime": 1033.1844,
      "train_samples_per_second": 3097.221,
      "train_steps_per_second": 193.576
    }
  ],
  "logging_steps": 500,
  "max_steps": 200000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.182034338135409e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}