{
  "best_metric": 3.276972532272339,
  "best_model_checkpoint": "bert_12_layer_model_v3_complete_training_48/checkpoint-210000",
  "epoch": 1.7205017286863622,
  "eval_steps": 10000,
  "global_step": 210001,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5.000000000000001e-07,
      "loss": 10.0329,
      "step": 500
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 9.2587,
      "step": 1000
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.5e-06,
      "loss": 8.8486,
      "step": 1500
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 8.3677,
      "step": 2000
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.5e-06,
      "loss": 7.88,
      "step": 2500
    },
    {
      "epoch": 0.02,
      "learning_rate": 3e-06,
      "loss": 7.4747,
      "step": 3000
    },
    {
      "epoch": 0.03,
      "learning_rate": 3.5e-06,
      "loss": 7.204,
      "step": 3500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.000000000000001e-06,
      "loss": 7.0561,
      "step": 4000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.5e-06,
      "loss": 6.9694,
      "step": 4500
    },
    {
      "epoch": 0.04,
      "learning_rate": 5e-06,
      "loss": 6.8997,
      "step": 5000
    },
    {
      "epoch": 0.05,
      "learning_rate": 5.500000000000001e-06,
      "loss": 6.8455,
      "step": 5500
    },
    {
      "epoch": 0.05,
      "learning_rate": 6e-06,
      "loss": 6.7998,
      "step": 6000
    },
    {
      "epoch": 0.05,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 6.7626,
      "step": 6500
    },
    {
      "epoch": 0.06,
      "learning_rate": 7e-06,
      "loss": 6.7341,
      "step": 7000
    },
    {
      "epoch": 0.06,
      "learning_rate": 7.500000000000001e-06,
      "loss": 6.7025,
      "step": 7500
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.000000000000001e-06,
      "loss": 6.6692,
      "step": 8000
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.5e-06,
      "loss": 6.6379,
      "step": 8500
    },
    {
      "epoch": 0.07,
      "learning_rate": 9e-06,
      "loss": 6.6169,
      "step": 9000
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.5e-06,
      "loss": 6.5932,
      "step": 9500
    },
    {
      "epoch": 0.08,
      "learning_rate": 1e-05,
      "loss": 6.5785,
      "step": 10000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.1281588728537004,
      "eval_loss": 6.536772727966309,
      "eval_runtime": 1720.6981,
      "eval_samples_per_second": 179.204,
      "eval_steps_per_second": 3.734,
      "step": 10000
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.991670692498626e-06,
      "loss": 6.5603,
      "step": 10500
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.983341384997253e-06,
      "loss": 6.5317,
      "step": 11000
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.975012077495878e-06,
      "loss": 6.5131,
      "step": 11500
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.966682769994504e-06,
      "loss": 6.5036,
      "step": 12000
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.95835346249313e-06,
      "loss": 6.4902,
      "step": 12500
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.950024154991756e-06,
      "loss": 6.4707,
      "step": 13000
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.941694847490381e-06,
      "loss": 6.4644,
      "step": 13500
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.933365539989006e-06,
      "loss": 6.4583,
      "step": 14000
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.925036232487631e-06,
      "loss": 6.4372,
      "step": 14500
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.916706924986258e-06,
      "loss": 6.4306,
      "step": 15000
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.908377617484883e-06,
      "loss": 6.4226,
      "step": 15500
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.90004830998351e-06,
      "loss": 6.4187,
      "step": 16000
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.891719002482134e-06,
      "loss": 6.4088,
      "step": 16500
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.883389694980761e-06,
      "loss": 6.3977,
      "step": 17000
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.875060387479386e-06,
      "loss": 6.3956,
      "step": 17500
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.866731079978011e-06,
      "loss": 6.3832,
      "step": 18000
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.858401772476636e-06,
      "loss": 6.3764,
      "step": 18500
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.850072464975263e-06,
      "loss": 6.3653,
      "step": 19000
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.84174315747389e-06,
      "loss": 6.3657,
      "step": 19500
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.833413849972514e-06,
      "loss": 6.3475,
      "step": 20000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.1409108328280017,
      "eval_loss": 6.334103584289551,
      "eval_runtime": 1722.0793,
      "eval_samples_per_second": 179.06,
      "eval_steps_per_second": 3.731,
      "step": 20000
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.82508454247114e-06,
      "loss": 6.348,
      "step": 20500
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.816755234969766e-06,
      "loss": 6.3477,
      "step": 21000
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.808425927468391e-06,
      "loss": 6.3332,
      "step": 21500
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.800096619967016e-06,
      "loss": 6.3293,
      "step": 22000
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.791767312465643e-06,
      "loss": 6.3264,
      "step": 22500
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.783438004964268e-06,
      "loss": 6.3268,
      "step": 23000
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.775108697462895e-06,
      "loss": 6.3126,
      "step": 23500
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.76677938996152e-06,
      "loss": 6.3155,
      "step": 24000
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.758450082460145e-06,
      "loss": 6.3013,
      "step": 24500
    },
    {
      "epoch": 0.2,
      "learning_rate": 9.750120774958771e-06,
      "loss": 6.3143,
      "step": 25000
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.741791467457396e-06,
      "loss": 6.3005,
      "step": 25500
    },
    {
      "epoch": 0.21,
      "learning_rate": 9.733462159956021e-06,
      "loss": 6.2994,
      "step": 26000
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.725132852454648e-06,
      "loss": 6.2936,
      "step": 26500
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.716803544953275e-06,
      "loss": 6.2832,
      "step": 27000
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.7084742374519e-06,
      "loss": 6.2783,
      "step": 27500
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.700144929950525e-06,
      "loss": 6.2787,
      "step": 28000
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.69181562244915e-06,
      "loss": 6.281,
      "step": 28500
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.683486314947776e-06,
      "loss": 6.2738,
      "step": 29000
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.675157007446401e-06,
      "loss": 6.271,
      "step": 29500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.666827699945028e-06,
      "loss": 6.2664,
      "step": 30000
    },
    {
      "epoch": 0.25,
      "eval_accuracy": 0.14461693539515286,
      "eval_loss": 6.252469539642334,
      "eval_runtime": 1724.7009,
      "eval_samples_per_second": 178.788,
      "eval_steps_per_second": 3.725,
      "step": 30000
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.658498392443653e-06,
      "loss": 6.2588,
      "step": 30500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.65016908494228e-06,
      "loss": 6.2593,
      "step": 31000
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.641839777440905e-06,
      "loss": 6.2593,
      "step": 31500
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.63351046993953e-06,
      "loss": 6.2613,
      "step": 32000
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.625181162438155e-06,
      "loss": 6.2533,
      "step": 32500
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.616851854936781e-06,
      "loss": 6.2452,
      "step": 33000
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.608522547435406e-06,
      "loss": 6.2417,
      "step": 33500
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.600193239934033e-06,
      "loss": 6.2437,
      "step": 34000
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.591863932432658e-06,
      "loss": 6.2389,
      "step": 34500
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.583534624931285e-06,
      "loss": 6.2385,
      "step": 35000
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.57520531742991e-06,
      "loss": 6.2453,
      "step": 35500
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.566876009928535e-06,
      "loss": 6.2325,
      "step": 36000
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.55854670242716e-06,
      "loss": 6.2267,
      "step": 36500
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.550217394925787e-06,
      "loss": 6.2266,
      "step": 37000
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.541888087424413e-06,
      "loss": 6.2342,
      "step": 37500
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.533558779923038e-06,
      "loss": 6.2164,
      "step": 38000
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.525229472421663e-06,
      "loss": 6.2245,
      "step": 38500
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.51690016492029e-06,
      "loss": 6.2191,
      "step": 39000
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.508570857418915e-06,
      "loss": 6.2151,
      "step": 39500
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.50024154991754e-06,
      "loss": 6.2198,
      "step": 40000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.1473662192399404,
      "eval_loss": 6.200199604034424,
      "eval_runtime": 1724.9579,
      "eval_samples_per_second": 178.761,
      "eval_steps_per_second": 3.725,
      "step": 40000
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.491912242416167e-06,
      "loss": 6.2087,
      "step": 40500
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.483582934914792e-06,
      "loss": 6.2109,
      "step": 41000
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.475253627413418e-06,
      "loss": 6.2109,
      "step": 41500
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.466924319912043e-06,
      "loss": 6.2089,
      "step": 42000
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.458595012410668e-06,
      "loss": 6.2088,
      "step": 42500
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.450265704909295e-06,
      "loss": 6.1999,
      "step": 43000
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.44193639740792e-06,
      "loss": 6.2022,
      "step": 43500
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.433607089906545e-06,
      "loss": 6.2002,
      "step": 44000
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.425277782405172e-06,
      "loss": 6.1945,
      "step": 44500
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.416948474903798e-06,
      "loss": 6.1907,
      "step": 45000
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.408619167402423e-06,
      "loss": 6.1858,
      "step": 45500
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.400289859901048e-06,
      "loss": 6.1872,
      "step": 46000
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.391960552399673e-06,
      "loss": 6.1799,
      "step": 46500
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.3836312448983e-06,
      "loss": 6.1826,
      "step": 47000
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.375301937396925e-06,
      "loss": 6.181,
      "step": 47500
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.366972629895552e-06,
      "loss": 6.1716,
      "step": 48000
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.358643322394177e-06,
      "loss": 6.1592,
      "step": 48500
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.350314014892804e-06,
      "loss": 6.1546,
      "step": 49000
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.341984707391429e-06,
      "loss": 6.1542,
      "step": 49500
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.333655399890054e-06,
      "loss": 6.1403,
      "step": 50000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.14970864377614984,
      "eval_loss": 6.124306678771973,
      "eval_runtime": 1727.7237,
      "eval_samples_per_second": 178.475,
      "eval_steps_per_second": 3.719,
      "step": 50000
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.325326092388679e-06,
      "loss": 6.1416,
      "step": 50500
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.316996784887305e-06,
      "loss": 6.1331,
      "step": 51000
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.30866747738593e-06,
      "loss": 6.1328,
      "step": 51500
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.300338169884557e-06,
      "loss": 6.122,
      "step": 52000
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.292008862383182e-06,
      "loss": 6.1199,
      "step": 52500
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.283679554881809e-06,
      "loss": 6.1099,
      "step": 53000
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.275350247380434e-06,
      "loss": 6.1117,
      "step": 53500
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.267020939879059e-06,
      "loss": 6.1092,
      "step": 54000
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.258691632377685e-06,
      "loss": 6.0971,
      "step": 54500
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.25036232487631e-06,
      "loss": 6.1019,
      "step": 55000
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.242033017374937e-06,
      "loss": 6.1,
      "step": 55500
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.233703709873562e-06,
      "loss": 6.0994,
      "step": 56000
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.225374402372189e-06,
      "loss": 6.0944,
      "step": 56500
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.217045094870814e-06,
      "loss": 6.0903,
      "step": 57000
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.208715787369439e-06,
      "loss": 6.0891,
      "step": 57500
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.200386479868064e-06,
      "loss": 6.0772,
      "step": 58000
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.19205717236669e-06,
      "loss": 6.0875,
      "step": 58500
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.183727864865315e-06,
      "loss": 6.0704,
      "step": 59000
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.175398557363942e-06,
      "loss": 6.0731,
      "step": 59500
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.167069249862567e-06,
      "loss": 6.0674,
      "step": 60000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.15200529744836427,
      "eval_loss": 6.038893699645996,
      "eval_runtime": 1726.8486,
      "eval_samples_per_second": 178.566,
      "eval_steps_per_second": 3.721,
      "step": 60000
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.158739942361194e-06,
      "loss": 6.0683,
      "step": 60500
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.150410634859819e-06,
      "loss": 6.0596,
      "step": 61000
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.142081327358444e-06,
      "loss": 6.0569,
      "step": 61500
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.133752019857069e-06,
      "loss": 6.0533,
      "step": 62000
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.125422712355696e-06,
      "loss": 6.0434,
      "step": 62500
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.11709340485432e-06,
      "loss": 6.0474,
      "step": 63000
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.108764097352947e-06,
      "loss": 6.0405,
      "step": 63500
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.100434789851572e-06,
      "loss": 6.0249,
      "step": 64000
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.092105482350199e-06,
      "loss": 6.0271,
      "step": 64500
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.083776174848824e-06,
      "loss": 6.0265,
      "step": 65000
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.075446867347449e-06,
      "loss": 6.0222,
      "step": 65500
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.067117559846076e-06,
      "loss": 6.0264,
      "step": 66000
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.0587882523447e-06,
      "loss": 6.0097,
      "step": 66500
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.050458944843327e-06,
      "loss": 6.0166,
      "step": 67000
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.042129637341952e-06,
      "loss": 6.0065,
      "step": 67500
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.033800329840577e-06,
      "loss": 5.9989,
      "step": 68000
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.025471022339204e-06,
      "loss": 6.0003,
      "step": 68500
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.017141714837829e-06,
      "loss": 5.9957,
      "step": 69000
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.008812407336454e-06,
      "loss": 5.9931,
      "step": 69500
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.00048309983508e-06,
      "loss": 5.9921,
      "step": 70000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.15476677622749932,
      "eval_loss": 5.958353519439697,
      "eval_runtime": 1726.9067,
      "eval_samples_per_second": 178.56,
      "eval_steps_per_second": 3.721,
      "step": 70000
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.992153792333706e-06,
      "loss": 5.9871,
      "step": 70500
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.983824484832332e-06,
      "loss": 5.9847,
      "step": 71000
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.975495177330957e-06,
      "loss": 5.9822,
      "step": 71500
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.967165869829582e-06,
      "loss": 5.9836,
      "step": 72000
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.958836562328209e-06,
      "loss": 5.9741,
      "step": 72500
    },
    {
      "epoch": 0.6,
      "learning_rate": 8.950507254826834e-06,
      "loss": 5.9702,
      "step": 73000
    },
    {
      "epoch": 0.6,
      "learning_rate": 8.942177947325459e-06,
      "loss": 5.968,
      "step": 73500
    },
    {
      "epoch": 0.61,
      "learning_rate": 8.933848639824086e-06,
      "loss": 5.9635,
      "step": 74000
    },
    {
      "epoch": 0.61,
      "learning_rate": 8.925519332322713e-06,
      "loss": 5.9605,
      "step": 74500
    },
    {
      "epoch": 0.61,
      "learning_rate": 8.917190024821338e-06,
      "loss": 5.964,
      "step": 75000
    },
    {
      "epoch": 0.62,
      "learning_rate": 8.908860717319963e-06,
      "loss": 5.9591,
      "step": 75500
    },
    {
      "epoch": 0.62,
      "learning_rate": 8.900531409818588e-06,
      "loss": 5.9585,
      "step": 76000
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.892202102317214e-06,
      "loss": 5.9523,
      "step": 76500
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.88387279481584e-06,
      "loss": 5.9508,
      "step": 77000
    },
    {
      "epoch": 0.63,
      "learning_rate": 8.875543487314466e-06,
      "loss": 5.9423,
      "step": 77500
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.867214179813091e-06,
      "loss": 5.9395,
      "step": 78000
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.858884872311718e-06,
      "loss": 5.9401,
      "step": 78500
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.850555564810343e-06,
      "loss": 5.9308,
      "step": 79000
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.842226257308968e-06,
      "loss": 5.9314,
      "step": 79500
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.833896949807593e-06,
      "loss": 5.9337,
      "step": 80000
    },
    {
      "epoch": 0.66,
      "eval_accuracy": 0.15771381455283048,
      "eval_loss": 5.892898082733154,
      "eval_runtime": 1725.7395,
      "eval_samples_per_second": 178.681,
      "eval_steps_per_second": 3.723,
      "step": 80000
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.82556764230622e-06,
      "loss": 5.9274,
      "step": 80500
    },
    {
      "epoch": 0.66,
      "learning_rate": 8.817238334804844e-06,
      "loss": 5.9209,
      "step": 81000
    },
    {
      "epoch": 0.67,
      "learning_rate": 8.808909027303471e-06,
      "loss": 5.9167,
      "step": 81500
    },
    {
      "epoch": 0.67,
      "learning_rate": 8.800579719802096e-06,
      "loss": 5.9163,
      "step": 82000
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.792250412300723e-06,
      "loss": 5.9106,
      "step": 82500
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.783921104799348e-06,
      "loss": 5.9035,
      "step": 83000
    },
    {
      "epoch": 0.68,
      "learning_rate": 8.775591797297973e-06,
      "loss": 5.9026,
      "step": 83500
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.767262489796598e-06,
      "loss": 5.897,
      "step": 84000
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.758933182295224e-06,
      "loss": 5.8837,
      "step": 84500
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.750603874793851e-06,
      "loss": 5.8863,
      "step": 85000
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.742274567292476e-06,
      "loss": 5.8799,
      "step": 85500
    },
    {
      "epoch": 0.7,
      "learning_rate": 8.733945259791101e-06,
      "loss": 5.8787,
      "step": 86000
    },
    {
      "epoch": 0.71,
      "learning_rate": 8.725615952289728e-06,
      "loss": 5.8655,
      "step": 86500
    },
    {
      "epoch": 0.71,
      "learning_rate": 8.717286644788353e-06,
      "loss": 5.8605,
      "step": 87000
    },
    {
      "epoch": 0.72,
      "learning_rate": 8.708957337286978e-06,
      "loss": 5.864,
      "step": 87500
    },
    {
      "epoch": 0.72,
      "learning_rate": 8.700628029785605e-06,
      "loss": 5.8508,
      "step": 88000
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.69229872228423e-06,
      "loss": 5.8478,
      "step": 88500
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.683969414782856e-06,
      "loss": 5.8422,
      "step": 89000
    },
    {
      "epoch": 0.73,
      "learning_rate": 8.675640107281481e-06,
      "loss": 5.8289,
      "step": 89500
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.667310799780106e-06,
      "loss": 5.8269,
      "step": 90000
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 0.1685354441535283,
      "eval_loss": 5.765416622161865,
      "eval_runtime": 1728.0921,
      "eval_samples_per_second": 178.437,
      "eval_steps_per_second": 3.718,
      "step": 90000
    },
    {
      "epoch": 0.74,
      "learning_rate": 8.658981492278733e-06,
      "loss": 5.8188,
      "step": 90500
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.650652184777358e-06,
      "loss": 5.8086,
      "step": 91000
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.642322877275983e-06,
      "loss": 5.7977,
      "step": 91500
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.63399356977461e-06,
      "loss": 5.796,
      "step": 92000
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.625664262273236e-06,
      "loss": 5.7847,
      "step": 92500
    },
    {
      "epoch": 0.76,
      "learning_rate": 8.617334954771861e-06,
      "loss": 5.7728,
      "step": 93000
    },
    {
      "epoch": 0.77,
      "learning_rate": 8.609005647270486e-06,
      "loss": 5.7654,
      "step": 93500
    },
    {
      "epoch": 0.77,
      "learning_rate": 8.600676339769111e-06,
      "loss": 5.754,
      "step": 94000
    },
    {
      "epoch": 0.77,
      "learning_rate": 8.592347032267738e-06,
      "loss": 5.7407,
      "step": 94500
    },
    {
      "epoch": 0.78,
      "learning_rate": 8.584017724766363e-06,
      "loss": 5.7215,
      "step": 95000
    },
    {
      "epoch": 0.78,
      "learning_rate": 8.57568841726499e-06,
      "loss": 5.7266,
      "step": 95500
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.567359109763615e-06,
      "loss": 5.7056,
      "step": 96000
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.559029802262241e-06,
      "loss": 5.686,
      "step": 96500
    },
    {
      "epoch": 0.79,
      "learning_rate": 8.550700494760866e-06,
      "loss": 5.6728,
      "step": 97000
    },
    {
      "epoch": 0.8,
      "learning_rate": 8.542371187259491e-06,
      "loss": 5.6472,
      "step": 97500
    },
    {
      "epoch": 0.8,
      "learning_rate": 8.534041879758118e-06,
      "loss": 5.6234,
      "step": 98000
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.525712572256743e-06,
      "loss": 5.6006,
      "step": 98500
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.517383264755368e-06,
      "loss": 5.5717,
      "step": 99000
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.509053957253995e-06,
      "loss": 5.5411,
      "step": 99500
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.500724649752622e-06,
      "loss": 5.5053,
      "step": 100000
    },
    {
      "epoch": 0.82,
      "eval_accuracy": 0.2274401366453315,
      "eval_loss": 5.321637153625488,
      "eval_runtime": 1731.0741,
      "eval_samples_per_second": 178.13,
      "eval_steps_per_second": 3.712,
      "step": 100000
    },
    {
      "epoch": 0.82,
      "learning_rate": 8.492395342251247e-06,
      "loss": 5.4662,
      "step": 100500
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.484066034749872e-06,
      "loss": 5.4259,
      "step": 101000
    },
    {
      "epoch": 0.83,
      "learning_rate": 8.475736727248497e-06,
      "loss": 5.3922,
      "step": 101500
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.467407419747123e-06,
      "loss": 5.3559,
      "step": 102000
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.459078112245748e-06,
      "loss": 5.3287,
      "step": 102500
    },
    {
      "epoch": 0.84,
      "learning_rate": 8.450748804744375e-06,
      "loss": 5.2922,
      "step": 103000
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.442419497243e-06,
      "loss": 5.2612,
      "step": 103500
    },
    {
      "epoch": 0.85,
      "learning_rate": 8.434090189741627e-06,
      "loss": 5.2302,
      "step": 104000
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.425760882240252e-06,
      "loss": 5.2009,
      "step": 104500
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.417431574738877e-06,
      "loss": 5.1624,
      "step": 105000
    },
    {
      "epoch": 0.86,
      "learning_rate": 8.409102267237502e-06,
      "loss": 5.1332,
      "step": 105500
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.400772959736128e-06,
      "loss": 5.1129,
      "step": 106000
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.392443652234753e-06,
      "loss": 5.0846,
      "step": 106500
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.38411434473338e-06,
      "loss": 5.0591,
      "step": 107000
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.375785037232005e-06,
      "loss": 5.0155,
      "step": 107500
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.367455729730632e-06,
      "loss": 4.9745,
      "step": 108000
    },
    {
      "epoch": 0.89,
      "learning_rate": 8.359126422229257e-06,
      "loss": 4.9379,
      "step": 108500
    },
    {
      "epoch": 0.89,
      "learning_rate": 8.350797114727882e-06,
      "loss": 4.8971,
      "step": 109000
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.342467807226507e-06,
      "loss": 4.8627,
      "step": 109500
    },
    {
      "epoch": 0.9,
      "learning_rate": 8.334138499725133e-06,
      "loss": 4.8237,
      "step": 110000
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.3053704435038631,
      "eval_loss": 4.63502836227417,
      "eval_runtime": 1724.1947,
      "eval_samples_per_second": 178.841,
      "eval_steps_per_second": 3.726,
      "step": 110000
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.32580919222376e-06,
      "loss": 4.7891,
      "step": 110500
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.317479884722385e-06,
      "loss": 4.7584,
      "step": 111000
    },
    {
      "epoch": 0.91,
      "learning_rate": 8.30915057722101e-06,
      "loss": 4.736,
      "step": 111500
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.300821269719637e-06,
      "loss": 4.7001,
      "step": 112000
    },
    {
      "epoch": 0.92,
      "learning_rate": 8.292491962218262e-06,
      "loss": 4.6641,
      "step": 112500
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.284162654716887e-06,
      "loss": 4.6343,
      "step": 113000
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.275833347215514e-06,
      "loss": 4.6078,
      "step": 113500
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.267504039714139e-06,
      "loss": 4.5865,
      "step": 114000
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.259174732212765e-06,
      "loss": 4.5528,
      "step": 114500
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.25084542471139e-06,
      "loss": 4.5327,
      "step": 115000
    },
    {
      "epoch": 0.95,
      "learning_rate": 8.242516117210015e-06,
      "loss": 4.51,
      "step": 115500
    },
    {
      "epoch": 0.95,
      "learning_rate": 8.234186809708642e-06,
      "loss": 4.4905,
      "step": 116000
    },
    {
      "epoch": 0.95,
      "learning_rate": 8.225857502207267e-06,
      "loss": 4.4668,
      "step": 116500
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.217528194705892e-06,
      "loss": 4.4387,
      "step": 117000
    },
    {
      "epoch": 0.96,
      "learning_rate": 8.209198887204519e-06,
      "loss": 4.4218,
      "step": 117500
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.200869579703145e-06,
      "loss": 4.3992,
      "step": 118000
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.19254027220177e-06,
      "loss": 4.3818,
      "step": 118500
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.184210964700395e-06,
      "loss": 4.3741,
      "step": 119000
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.17588165719902e-06,
      "loss": 4.3455,
      "step": 119500
    },
    {
      "epoch": 0.98,
      "learning_rate": 8.167552349697647e-06,
      "loss": 4.3302,
      "step": 120000
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.34964174989734176,
      "eval_loss": 4.172255039215088,
      "eval_runtime": 1724.4645,
      "eval_samples_per_second": 178.813,
      "eval_steps_per_second": 3.726,
      "step": 120000
    },
    {
      "epoch": 0.99,
      "learning_rate": 8.159223042196272e-06,
      "loss": 4.3019,
      "step": 120500
    },
    {
      "epoch": 0.99,
      "learning_rate": 8.150893734694899e-06,
      "loss": 4.3005,
      "step": 121000
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.142564427193524e-06,
      "loss": 4.2779,
      "step": 121500
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.13423511969215e-06,
      "loss": 4.2619,
      "step": 122000
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.125905812190775e-06,
      "loss": 4.247,
      "step": 122500
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.1175765046894e-06,
      "loss": 4.2294,
      "step": 123000
    },
    {
      "epoch": 1.01,
      "learning_rate": 8.109247197188025e-06,
      "loss": 4.212,
      "step": 123500
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.100917889686652e-06,
      "loss": 4.2155,
      "step": 124000
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.092588582185277e-06,
      "loss": 4.1914,
      "step": 124500
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.084259274683904e-06,
      "loss": 4.1819,
      "step": 125000
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.075929967182529e-06,
      "loss": 4.1672,
      "step": 125500
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.067600659681156e-06,
      "loss": 4.1501,
      "step": 126000
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.05927135217978e-06,
      "loss": 4.1457,
      "step": 126500
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.050942044678406e-06,
      "loss": 4.1164,
      "step": 127000
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.04261273717703e-06,
      "loss": 4.117,
      "step": 127500
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.034283429675657e-06,
      "loss": 4.1062,
      "step": 128000
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.025954122174284e-06,
      "loss": 4.0928,
      "step": 128500
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.017624814672909e-06,
      "loss": 4.0904,
      "step": 129000
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.009295507171534e-06,
      "loss": 4.0701,
      "step": 129500
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.00096619967016e-06,
      "loss": 4.0721,
      "step": 130000
    },
    {
      "epoch": 1.07,
      "eval_accuracy": 0.3740549523966965,
      "eval_loss": 3.9373106956481934,
      "eval_runtime": 1726.1642,
      "eval_samples_per_second": 178.637,
      "eval_steps_per_second": 3.722,
      "step": 130000
    },
    {
      "epoch": 1.07,
      "learning_rate": 7.992636892168786e-06,
      "loss": 4.0592,
      "step": 130500
    },
    {
      "epoch": 1.07,
      "learning_rate": 7.98430758466741e-06,
      "loss": 4.0396,
      "step": 131000
    },
    {
      "epoch": 1.08,
      "learning_rate": 7.975978277166037e-06,
      "loss": 4.0345,
      "step": 131500
    },
    {
      "epoch": 1.08,
      "learning_rate": 7.967648969664662e-06,
      "loss": 4.0213,
      "step": 132000
    },
    {
      "epoch": 1.09,
      "learning_rate": 7.959319662163289e-06,
      "loss": 4.0221,
      "step": 132500
    },
    {
      "epoch": 1.09,
      "learning_rate": 7.950990354661914e-06,
      "loss": 4.012,
      "step": 133000
    },
    {
      "epoch": 1.09,
      "learning_rate": 7.942661047160539e-06,
      "loss": 3.9936,
      "step": 133500
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.934331739659166e-06,
      "loss": 3.9952,
      "step": 134000
    },
    {
      "epoch": 1.1,
      "learning_rate": 7.92600243215779e-06,
      "loss": 3.9885,
      "step": 134500
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.917673124656416e-06,
      "loss": 3.9682,
      "step": 135000
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.909343817155042e-06,
      "loss": 3.968,
      "step": 135500
    },
    {
      "epoch": 1.11,
      "learning_rate": 7.901014509653669e-06,
      "loss": 3.9526,
      "step": 136000
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.892685202152294e-06,
      "loss": 3.9439,
      "step": 136500
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.884355894650919e-06,
      "loss": 3.9409,
      "step": 137000
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.876026587149544e-06,
      "loss": 3.932,
      "step": 137500
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.86769727964817e-06,
      "loss": 3.9288,
      "step": 138000
    },
    {
      "epoch": 1.13,
      "learning_rate": 7.859367972146796e-06,
      "loss": 3.9232,
      "step": 138500
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.851038664645423e-06,
      "loss": 3.9049,
      "step": 139000
    },
    {
      "epoch": 1.14,
      "learning_rate": 7.842709357144048e-06,
      "loss": 3.9075,
      "step": 139500
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.834380049642674e-06,
      "loss": 3.8896,
      "step": 140000
    },
    {
      "epoch": 1.15,
      "eval_accuracy": 0.3903213950637094,
      "eval_loss": 3.776296377182007,
      "eval_runtime": 1723.9466,
      "eval_samples_per_second": 178.866,
      "eval_steps_per_second": 3.727,
      "step": 140000
    },
    {
      "epoch": 1.15,
      "learning_rate": 7.8260507421413e-06,
      "loss": 3.8844,
      "step": 140500
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.817721434639924e-06,
      "loss": 3.8857,
      "step": 141000
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.809392127138551e-06,
      "loss": 3.8744,
      "step": 141500
    },
    {
      "epoch": 1.16,
      "learning_rate": 7.801062819637176e-06,
      "loss": 3.8706,
      "step": 142000
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.792733512135801e-06,
      "loss": 3.8717,
      "step": 142500
    },
    {
      "epoch": 1.17,
      "learning_rate": 7.784404204634428e-06,
      "loss": 3.8542,
      "step": 143000
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.776074897133054e-06,
      "loss": 3.8553,
      "step": 143500
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.76774558963168e-06,
      "loss": 3.8442,
      "step": 144000
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.759416282130304e-06,
      "loss": 3.8403,
      "step": 144500
    },
    {
      "epoch": 1.19,
      "learning_rate": 7.75108697462893e-06,
      "loss": 3.8396,
      "step": 145000
    },
    {
      "epoch": 1.19,
      "learning_rate": 7.742757667127556e-06,
      "loss": 3.8303,
      "step": 145500
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.734428359626181e-06,
      "loss": 3.8179,
      "step": 146000
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.726099052124808e-06,
      "loss": 3.8146,
      "step": 146500
    },
    {
      "epoch": 1.2,
      "learning_rate": 7.717769744623433e-06,
      "loss": 3.8204,
      "step": 147000
    },
    {
      "epoch": 1.21,
      "learning_rate": 7.70944043712206e-06,
      "loss": 3.8103,
      "step": 147500
    },
    {
      "epoch": 1.21,
      "learning_rate": 7.701111129620684e-06,
      "loss": 3.7987,
      "step": 148000
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.69278182211931e-06,
      "loss": 3.7987,
      "step": 148500
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.684452514617934e-06,
      "loss": 3.7856,
      "step": 149000
    },
    {
      "epoch": 1.22,
      "learning_rate": 7.676123207116561e-06,
      "loss": 3.7868,
      "step": 149500
    },
    {
      "epoch": 1.23,
      "learning_rate": 7.667793899615186e-06,
      "loss": 3.7842,
      "step": 150000
    },
    {
      "epoch": 1.23,
      "eval_accuracy": 0.40179710200093327,
      "eval_loss": 3.6628007888793945,
      "eval_runtime": 1725.8659,
      "eval_samples_per_second": 178.667,
      "eval_steps_per_second": 3.723,
      "step": 150000
    },
    {
      "epoch": 1.23,
      "learning_rate": 7.659464592113813e-06,
      "loss": 3.7749,
      "step": 150500
    },
    {
      "epoch": 1.24,
      "learning_rate": 7.651135284612438e-06,
      "loss": 3.7717,
      "step": 151000
    },
    {
      "epoch": 1.24,
      "learning_rate": 7.642805977111065e-06,
      "loss": 3.7711,
      "step": 151500
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.63447666960969e-06,
      "loss": 3.7649,
      "step": 152000
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.6261473621083145e-06,
      "loss": 3.7623,
      "step": 152500
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.61781805460694e-06,
      "loss": 3.7533,
      "step": 153000
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.609488747105567e-06,
      "loss": 3.7552,
      "step": 153500
    },
    {
      "epoch": 1.26,
      "learning_rate": 7.601159439604192e-06,
      "loss": 3.7439,
      "step": 154000
    },
    {
      "epoch": 1.27,
      "learning_rate": 7.592830132102817e-06,
      "loss": 3.738,
      "step": 154500
    },
    {
      "epoch": 1.27,
      "learning_rate": 7.584500824601443e-06,
      "loss": 3.7332,
      "step": 155000
    },
    {
      "epoch": 1.27,
      "learning_rate": 7.57617151710007e-06,
      "loss": 3.7265,
      "step": 155500
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.567842209598695e-06,
      "loss": 3.7252,
      "step": 156000
    },
    {
      "epoch": 1.28,
      "learning_rate": 7.5595129020973205e-06,
      "loss": 3.7141,
      "step": 156500
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.5511835945959455e-06,
      "loss": 3.7048,
      "step": 157000
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.542854287094572e-06,
      "loss": 3.7097,
      "step": 157500
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.534524979593197e-06,
      "loss": 3.709,
      "step": 158000
    },
    {
      "epoch": 1.3,
      "learning_rate": 7.526195672091823e-06,
      "loss": 3.7049,
      "step": 158500
    },
    {
      "epoch": 1.3,
      "learning_rate": 7.517866364590448e-06,
      "loss": 3.6925,
      "step": 159000
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.509537057089075e-06,
      "loss": 3.6936,
      "step": 159500
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.5012077495877e-06,
      "loss": 3.6871,
      "step": 160000
    },
    {
      "epoch": 1.31,
      "eval_accuracy": 0.41109455601253003,
      "eval_loss": 3.5734403133392334,
      "eval_runtime": 1726.4499,
      "eval_samples_per_second": 178.607,
      "eval_steps_per_second": 3.722,
      "step": 160000
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.492878442086326e-06,
      "loss": 3.6833,
      "step": 160500
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.484549134584951e-06,
      "loss": 3.6789,
      "step": 161000
    },
    {
      "epoch": 1.32,
      "learning_rate": 7.476219827083577e-06,
      "loss": 3.6832,
      "step": 161500
    },
    {
      "epoch": 1.33,
      "learning_rate": 7.467890519582202e-06,
      "loss": 3.6663,
      "step": 162000
    },
    {
      "epoch": 1.33,
      "learning_rate": 7.459561212080828e-06,
      "loss": 3.6736,
      "step": 162500
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.451231904579453e-06,
      "loss": 3.6681,
      "step": 163000
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.44290259707808e-06,
      "loss": 3.6638,
      "step": 163500
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.434573289576706e-06,
      "loss": 3.6573,
      "step": 164000
    },
    {
      "epoch": 1.35,
      "learning_rate": 7.426243982075331e-06,
      "loss": 3.6478,
      "step": 164500
    },
    {
      "epoch": 1.35,
      "learning_rate": 7.417914674573956e-06,
      "loss": 3.6517,
      "step": 165000
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.409585367072582e-06,
      "loss": 3.6423,
      "step": 165500
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.401256059571208e-06,
      "loss": 3.6381,
      "step": 166000
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.392926752069833e-06,
      "loss": 3.6294,
      "step": 166500
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.384597444568459e-06,
      "loss": 3.6372,
      "step": 167000
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.376268137067085e-06,
      "loss": 3.632,
      "step": 167500
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.367938829565711e-06,
      "loss": 3.6243,
      "step": 168000
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.359609522064336e-06,
      "loss": 3.6227,
      "step": 168500
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.351280214562962e-06,
      "loss": 3.6154,
      "step": 169000
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.3429509070615875e-06,
      "loss": 3.613,
      "step": 169500
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.334621599560213e-06,
      "loss": 3.6099,
      "step": 170000
    },
    {
      "epoch": 1.39,
      "eval_accuracy": 0.4187936750226971,
      "eval_loss": 3.4986298084259033,
      "eval_runtime": 1725.7302,
      "eval_samples_per_second": 178.681,
      "eval_steps_per_second": 3.723,
      "step": 170000
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.326292292058838e-06,
      "loss": 3.6011,
      "step": 170500
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.317962984557464e-06,
      "loss": 3.601,
      "step": 171000
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.309633677056091e-06,
      "loss": 3.603,
      "step": 171500
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.301304369554716e-06,
      "loss": 3.5999,
      "step": 172000
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.292975062053341e-06,
      "loss": 3.5966,
      "step": 172500
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.284645754551967e-06,
      "loss": 3.5932,
      "step": 173000
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.2763164470505934e-06,
      "loss": 3.5929,
      "step": 173500
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.2679871395492184e-06,
      "loss": 3.5886,
      "step": 174000
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.259657832047844e-06,
      "loss": 3.5809,
      "step": 174500
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.251328524546469e-06,
      "loss": 3.5818,
      "step": 175000
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.242999217045096e-06,
      "loss": 3.573,
      "step": 175500
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.234669909543721e-06,
      "loss": 3.577,
      "step": 176000
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.226340602042347e-06,
      "loss": 3.5687,
      "step": 176500
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.218011294540972e-06,
      "loss": 3.5677,
      "step": 177000
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.2096819870395985e-06,
      "loss": 3.5615,
      "step": 177500
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.2013526795382235e-06,
      "loss": 3.5601,
      "step": 178000
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.193023372036849e-06,
      "loss": 3.555,
      "step": 178500
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.184694064535474e-06,
      "loss": 3.5514,
      "step": 179000
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.176364757034101e-06,
      "loss": 3.5523,
      "step": 179500
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.168035449532726e-06,
      "loss": 3.5585,
      "step": 180000
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.4258632688973274,
      "eval_loss": 3.4324440956115723,
      "eval_runtime": 1723.9339,
      "eval_samples_per_second": 178.868,
      "eval_steps_per_second": 3.727,
      "step": 180000
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.159706142031352e-06,
      "loss": 3.5419,
      "step": 180500
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.151376834529977e-06,
      "loss": 3.5335,
      "step": 181000
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.143047527028604e-06,
      "loss": 3.5335,
      "step": 181500
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.1347182195272295e-06,
      "loss": 3.5402,
      "step": 182000
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.1263889120258545e-06,
      "loss": 3.5347,
      "step": 182500
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.118059604524481e-06,
      "loss": 3.525,
      "step": 183000
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.109730297023106e-06,
      "loss": 3.5263,
      "step": 183500
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.101400989521732e-06,
      "loss": 3.5225,
      "step": 184000
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.093071682020357e-06,
      "loss": 3.5255,
      "step": 184500
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.084742374518984e-06,
      "loss": 3.5103,
      "step": 185000
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.076413067017609e-06,
      "loss": 3.5188,
      "step": 185500
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.068083759516235e-06,
      "loss": 3.5088,
      "step": 186000
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.05975445201486e-06,
      "loss": 3.4969,
      "step": 186500
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.051425144513486e-06,
      "loss": 3.5042,
      "step": 187000
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.043095837012111e-06,
      "loss": 3.5034,
      "step": 187500
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.034766529510737e-06,
      "loss": 3.4992,
      "step": 188000
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.026437222009362e-06,
      "loss": 3.4957,
      "step": 188500
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.018107914507989e-06,
      "loss": 3.4922,
      "step": 189000
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.009778607006615e-06,
      "loss": 3.4931,
      "step": 189500
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.00144929950524e-06,
      "loss": 3.4922,
      "step": 190000
    },
    {
      "epoch": 1.56,
      "eval_accuracy": 0.431615353474136,
      "eval_loss": 3.3766753673553467,
      "eval_runtime": 1724.6809,
      "eval_samples_per_second": 178.79,
      "eval_steps_per_second": 3.725,
      "step": 190000
    },
    {
      "epoch": 1.56,
      "learning_rate": 6.993119992003865e-06,
      "loss": 3.488,
      "step": 190500
    },
    {
      "epoch": 1.56,
      "learning_rate": 6.984790684502491e-06,
      "loss": 3.4853,
      "step": 191000
    },
    {
      "epoch": 1.57,
      "learning_rate": 6.976461377001117e-06,
      "loss": 3.4766,
      "step": 191500
    },
    {
      "epoch": 1.57,
      "learning_rate": 6.968132069499742e-06,
      "loss": 3.4742,
      "step": 192000
    },
    {
      "epoch": 1.58,
      "learning_rate": 6.959802761998368e-06,
      "loss": 3.4747,
      "step": 192500
    },
    {
      "epoch": 1.58,
      "learning_rate": 6.951473454496994e-06,
      "loss": 3.4748,
      "step": 193000
    },
    {
      "epoch": 1.59,
      "learning_rate": 6.94314414699562e-06,
      "loss": 3.4671,
      "step": 193500
    },
    {
      "epoch": 1.59,
      "learning_rate": 6.934814839494245e-06,
      "loss": 3.4656,
      "step": 194000
    },
    {
      "epoch": 1.59,
      "learning_rate": 6.926485531992871e-06,
      "loss": 3.4689,
      "step": 194500
    },
    {
      "epoch": 1.6,
      "learning_rate": 6.9181562244914965e-06,
      "loss": 3.4581,
      "step": 195000
    },
    {
      "epoch": 1.6,
      "learning_rate": 6.909826916990122e-06,
      "loss": 3.4493,
      "step": 195500
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.901497609488747e-06,
      "loss": 3.4517,
      "step": 196000
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.893168301987373e-06,
      "loss": 3.4527,
      "step": 196500
    },
    {
      "epoch": 1.61,
      "learning_rate": 6.884838994485999e-06,
      "loss": 3.455,
      "step": 197000
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.876509686984625e-06,
      "loss": 3.4463,
      "step": 197500
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.86818037948325e-06,
      "loss": 3.4449,
      "step": 198000
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.859851071981876e-06,
      "loss": 3.4384,
      "step": 198500
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.8515217644805024e-06,
      "loss": 3.4395,
      "step": 199000
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.8431924569791274e-06,
      "loss": 3.4325,
      "step": 199500
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.8348631494777525e-06,
      "loss": 3.4342,
      "step": 200000
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 0.43735020422872606,
      "eval_loss": 3.3234128952026367,
      "eval_runtime": 1731.0697,
      "eval_samples_per_second": 178.13,
      "eval_steps_per_second": 3.712,
      "step": 200000
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.826533841976378e-06,
      "loss": 3.4376,
      "step": 200500
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.818204534475005e-06,
      "loss": 3.4306,
      "step": 201000
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.80987522697363e-06,
      "loss": 3.425,
      "step": 201500
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.801545919472256e-06,
      "loss": 3.4245,
      "step": 202000
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.793216611970881e-06,
      "loss": 3.4192,
      "step": 202500
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.7848873044695075e-06,
      "loss": 3.4209,
      "step": 203000
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.7765579969681325e-06,
      "loss": 3.4162,
      "step": 203500
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.768228689466758e-06,
      "loss": 3.4149,
      "step": 204000
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.759899381965383e-06,
      "loss": 3.4099,
      "step": 204500
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.75157007446401e-06,
      "loss": 3.4152,
      "step": 205000
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.743240766962635e-06,
      "loss": 3.4053,
      "step": 205500
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.734911459461261e-06,
      "loss": 3.4121,
      "step": 206000
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.726582151959886e-06,
      "loss": 3.4016,
      "step": 206500
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.718252844458513e-06,
      "loss": 3.4082,
      "step": 207000
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.709923536957138e-06,
      "loss": 3.3972,
      "step": 207500
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.7015942294557635e-06,
      "loss": 3.3905,
      "step": 208000
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.6932649219543885e-06,
      "loss": 3.3943,
      "step": 208500
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.684935614453015e-06,
      "loss": 3.3929,
      "step": 209000
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.676606306951641e-06,
      "loss": 3.3955,
      "step": 209500
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.668276999450266e-06,
      "loss": 3.3799,
      "step": 210000
    },
    {
      "epoch": 1.72,
      "eval_accuracy": 0.4429012463715165,
      "eval_loss": 3.276972532272339,
      "eval_runtime": 1731.0997,
      "eval_samples_per_second": 178.127,
      "eval_steps_per_second": 3.712,
      "step": 210000
    },
    {
      "epoch": 1.72,
      "step": 210001,
      "total_flos": 2.91571589780498e+18,
      "train_loss": 5.009029721765974,
      "train_runtime": 198560.4752,
      "train_samples_per_second": 147.531,
      "train_steps_per_second": 3.074
    }
  ],
  "logging_steps": 500,
  "max_steps": 610290,
  "num_train_epochs": 5,
  "save_steps": 10000,
  "total_flos": 2.91571589780498e+18,
  "trial_name": null,
  "trial_params": null
}