{
  "best_metric": 2.2190773487091064,
  "best_model_checkpoint": "./model_tweets_2020_Q1_25/checkpoint-1952000",
  "epoch": 6.556194784547049,
  "eval_steps": 8000,
  "global_step": 2400000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "eval_loss": 2.5883126258850098,
      "eval_runtime": 332.1552,
      "eval_samples_per_second": 928.078,
      "eval_steps_per_second": 58.006,
      "step": 8000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.0726666666666665e-07,
      "loss": 2.7561,
      "step": 16000
    },
    {
      "epoch": 0.04,
      "eval_loss": 2.4957592487335205,
      "eval_runtime": 332.3107,
      "eval_samples_per_second": 927.644,
      "eval_steps_per_second": 57.979,
      "step": 16000
    },
    {
      "epoch": 0.07,
      "eval_loss": 2.437556266784668,
      "eval_runtime": 332.4781,
      "eval_samples_per_second": 927.177,
      "eval_steps_per_second": 57.95,
      "step": 24000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.0453333333333336e-07,
      "loss": 2.531,
      "step": 32000
    },
    {
      "epoch": 0.09,
      "eval_loss": 2.409024477005005,
      "eval_runtime": 332.2129,
      "eval_samples_per_second": 927.917,
      "eval_steps_per_second": 57.996,
      "step": 32000
    },
    {
      "epoch": 0.11,
      "eval_loss": 2.3791210651397705,
      "eval_runtime": 331.9964,
      "eval_samples_per_second": 928.522,
      "eval_steps_per_second": 58.034,
      "step": 40000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.018e-07,
      "loss": 2.4627,
      "step": 48000
    },
    {
      "epoch": 0.13,
      "eval_loss": 2.3666348457336426,
      "eval_runtime": 332.5374,
      "eval_samples_per_second": 927.011,
      "eval_steps_per_second": 57.939,
      "step": 48000
    },
    {
      "epoch": 0.15,
      "eval_loss": 2.3457391262054443,
      "eval_runtime": 333.7346,
      "eval_samples_per_second": 923.686,
      "eval_steps_per_second": 57.732,
      "step": 56000
    },
    {
      "epoch": 0.17,
      "learning_rate": 3.9906666666666667e-07,
      "loss": 2.4252,
      "step": 64000
    },
    {
      "epoch": 0.17,
      "eval_loss": 2.337965726852417,
      "eval_runtime": 331.9355,
      "eval_samples_per_second": 928.693,
      "eval_steps_per_second": 58.044,
      "step": 64000
    },
    {
      "epoch": 0.2,
      "eval_loss": 2.329805374145508,
      "eval_runtime": 333.3329,
      "eval_samples_per_second": 924.799,
      "eval_steps_per_second": 57.801,
      "step": 72000
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.963333333333333e-07,
      "loss": 2.4061,
      "step": 80000
    },
    {
      "epoch": 0.22,
      "eval_loss": 2.325295925140381,
      "eval_runtime": 334.0512,
      "eval_samples_per_second": 922.811,
      "eval_steps_per_second": 57.677,
      "step": 80000
    },
    {
      "epoch": 0.24,
      "eval_loss": 2.317714214324951,
      "eval_runtime": 334.2839,
      "eval_samples_per_second": 922.168,
      "eval_steps_per_second": 57.637,
      "step": 88000
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.936e-07,
      "loss": 2.395,
      "step": 96000
    },
    {
      "epoch": 0.26,
      "eval_loss": 2.3130738735198975,
      "eval_runtime": 334.6418,
      "eval_samples_per_second": 921.182,
      "eval_steps_per_second": 57.575,
      "step": 96000
    },
    {
      "epoch": 0.28,
      "eval_loss": 2.305846691131592,
      "eval_runtime": 333.8278,
      "eval_samples_per_second": 923.428,
      "eval_steps_per_second": 57.715,
      "step": 104000
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.908666666666667e-07,
      "loss": 2.3843,
      "step": 112000
    },
    {
      "epoch": 0.31,
      "eval_loss": 2.3009700775146484,
      "eval_runtime": 334.6239,
      "eval_samples_per_second": 921.231,
      "eval_steps_per_second": 57.578,
      "step": 112000
    },
    {
      "epoch": 0.33,
      "eval_loss": 2.2925422191619873,
      "eval_runtime": 335.4315,
      "eval_samples_per_second": 919.013,
      "eval_steps_per_second": 57.439,
      "step": 120000
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.8813333333333334e-07,
      "loss": 2.3738,
      "step": 128000
    },
    {
      "epoch": 0.35,
      "eval_loss": 2.2916228771209717,
      "eval_runtime": 334.3802,
      "eval_samples_per_second": 921.903,
      "eval_steps_per_second": 57.62,
      "step": 128000
    },
    {
      "epoch": 0.37,
      "eval_loss": 2.29465389251709,
      "eval_runtime": 333.899,
      "eval_samples_per_second": 923.231,
      "eval_steps_per_second": 57.703,
      "step": 136000
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.854e-07,
      "loss": 2.3686,
      "step": 144000
    },
    {
      "epoch": 0.39,
      "eval_loss": 2.2834742069244385,
      "eval_runtime": 334.2854,
      "eval_samples_per_second": 922.164,
      "eval_steps_per_second": 57.636,
      "step": 144000
    },
    {
      "epoch": 0.42,
      "eval_loss": 2.2863762378692627,
      "eval_runtime": 334.9469,
      "eval_samples_per_second": 920.343,
      "eval_steps_per_second": 57.523,
      "step": 152000
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.8266666666666665e-07,
      "loss": 2.3615,
      "step": 160000
    },
    {
      "epoch": 0.44,
      "eval_loss": 2.283402919769287,
      "eval_runtime": 340.7251,
      "eval_samples_per_second": 904.735,
      "eval_steps_per_second": 56.547,
      "step": 160000
    },
    {
      "epoch": 0.46,
      "eval_loss": 2.276758909225464,
      "eval_runtime": 337.1601,
      "eval_samples_per_second": 914.301,
      "eval_steps_per_second": 57.145,
      "step": 168000
    },
    {
      "epoch": 0.48,
      "learning_rate": 3.799333333333333e-07,
      "loss": 2.3515,
      "step": 176000
    },
    {
      "epoch": 0.48,
      "eval_loss": 2.280259847640991,
      "eval_runtime": 339.3606,
      "eval_samples_per_second": 908.373,
      "eval_steps_per_second": 56.774,
      "step": 176000
    },
    {
      "epoch": 0.5,
      "eval_loss": 2.280385732650757,
      "eval_runtime": 337.3298,
      "eval_samples_per_second": 913.841,
      "eval_steps_per_second": 57.116,
      "step": 184000
    },
    {
      "epoch": 0.52,
      "learning_rate": 3.772e-07,
      "loss": 2.3508,
      "step": 192000
    },
    {
      "epoch": 0.52,
      "eval_loss": 2.275428295135498,
      "eval_runtime": 335.956,
      "eval_samples_per_second": 917.578,
      "eval_steps_per_second": 57.35,
      "step": 192000
    },
    {
      "epoch": 0.55,
      "eval_loss": 2.2767016887664795,
      "eval_runtime": 335.4697,
      "eval_samples_per_second": 918.909,
      "eval_steps_per_second": 57.433,
      "step": 200000
    },
    {
      "epoch": 0.57,
      "learning_rate": 3.7446666666666667e-07,
      "loss": 2.35,
      "step": 208000
    },
    {
      "epoch": 0.57,
      "eval_loss": 2.2741713523864746,
      "eval_runtime": 335.6132,
      "eval_samples_per_second": 918.516,
      "eval_steps_per_second": 57.408,
      "step": 208000
    },
    {
      "epoch": 0.59,
      "eval_loss": 2.272230386734009,
      "eval_runtime": 334.8971,
      "eval_samples_per_second": 920.48,
      "eval_steps_per_second": 57.531,
      "step": 216000
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.7173333333333333e-07,
      "loss": 2.3385,
      "step": 224000
    },
    {
      "epoch": 0.61,
      "eval_loss": 2.266052484512329,
      "eval_runtime": 336.4044,
      "eval_samples_per_second": 916.356,
      "eval_steps_per_second": 57.273,
      "step": 224000
    },
    {
      "epoch": 0.63,
      "eval_loss": 2.2705941200256348,
      "eval_runtime": 333.5697,
      "eval_samples_per_second": 924.143,
      "eval_steps_per_second": 57.76,
      "step": 232000
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.69e-07,
      "loss": 2.3393,
      "step": 240000
    },
    {
      "epoch": 0.66,
      "eval_loss": 2.2633328437805176,
      "eval_runtime": 337.2689,
      "eval_samples_per_second": 914.007,
      "eval_steps_per_second": 57.127,
      "step": 240000
    },
    {
      "epoch": 0.68,
      "eval_loss": 2.26476788520813,
      "eval_runtime": 335.5887,
      "eval_samples_per_second": 918.583,
      "eval_steps_per_second": 57.413,
      "step": 248000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.6626666666666664e-07,
      "loss": 2.3392,
      "step": 256000
    },
    {
      "epoch": 0.7,
      "eval_loss": 2.2656445503234863,
      "eval_runtime": 336.8746,
      "eval_samples_per_second": 915.076,
      "eval_steps_per_second": 57.193,
      "step": 256000
    },
    {
      "epoch": 0.72,
      "eval_loss": 2.2660086154937744,
      "eval_runtime": 335.5863,
      "eval_samples_per_second": 918.589,
      "eval_steps_per_second": 57.413,
      "step": 264000
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.6353333333333335e-07,
      "loss": 2.3336,
      "step": 272000
    },
    {
      "epoch": 0.74,
      "eval_loss": 2.265676736831665,
      "eval_runtime": 336.9338,
      "eval_samples_per_second": 914.916,
      "eval_steps_per_second": 57.183,
      "step": 272000
    },
    {
      "epoch": 0.76,
      "eval_loss": 2.2604660987854004,
      "eval_runtime": 334.7066,
      "eval_samples_per_second": 921.004,
      "eval_steps_per_second": 57.564,
      "step": 280000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.608e-07,
      "loss": 2.3324,
      "step": 288000
    },
    {
      "epoch": 0.79,
      "eval_loss": 2.261540651321411,
      "eval_runtime": 336.5364,
      "eval_samples_per_second": 915.996,
      "eval_steps_per_second": 57.251,
      "step": 288000
    },
    {
      "epoch": 0.81,
      "eval_loss": 2.255148410797119,
      "eval_runtime": 334.2192,
      "eval_samples_per_second": 922.347,
      "eval_steps_per_second": 57.648,
      "step": 296000
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.5806666666666666e-07,
      "loss": 2.3312,
      "step": 304000
    },
    {
      "epoch": 0.83,
      "eval_loss": 2.2581000328063965,
      "eval_runtime": 334.0976,
      "eval_samples_per_second": 922.683,
      "eval_steps_per_second": 57.669,
      "step": 304000
    },
    {
      "epoch": 0.85,
      "eval_loss": 2.262571334838867,
      "eval_runtime": 333.6502,
      "eval_samples_per_second": 923.92,
      "eval_steps_per_second": 57.746,
      "step": 312000
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.553333333333333e-07,
      "loss": 2.3352,
      "step": 320000
    },
    {
      "epoch": 0.87,
      "eval_loss": 2.2576231956481934,
      "eval_runtime": 334.3477,
      "eval_samples_per_second": 921.992,
      "eval_steps_per_second": 57.626,
      "step": 320000
    },
    {
      "epoch": 0.9,
      "eval_loss": 2.2552905082702637,
      "eval_runtime": 334.1252,
      "eval_samples_per_second": 922.606,
      "eval_steps_per_second": 57.664,
      "step": 328000
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.5259999999999997e-07,
      "loss": 2.3287,
      "step": 336000
    },
    {
      "epoch": 0.92,
      "eval_loss": 2.259079694747925,
      "eval_runtime": 336.4319,
      "eval_samples_per_second": 916.281,
      "eval_steps_per_second": 57.269,
      "step": 336000
    },
    {
      "epoch": 0.94,
      "eval_loss": 2.2557525634765625,
      "eval_runtime": 334.0372,
      "eval_samples_per_second": 922.849,
      "eval_steps_per_second": 57.679,
      "step": 344000
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.498666666666667e-07,
      "loss": 2.321,
      "step": 352000
    },
    {
      "epoch": 0.96,
      "eval_loss": 2.260307788848877,
      "eval_runtime": 335.3355,
      "eval_samples_per_second": 919.276,
      "eval_steps_per_second": 57.456,
      "step": 352000
    },
    {
      "epoch": 0.98,
      "eval_loss": 2.256932258605957,
      "eval_runtime": 334.5306,
      "eval_samples_per_second": 921.488,
      "eval_steps_per_second": 57.594,
      "step": 360000
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.4713333333333333e-07,
      "loss": 2.3278,
      "step": 368000
    },
    {
      "epoch": 1.01,
      "eval_loss": 2.254404067993164,
      "eval_runtime": 335.0696,
      "eval_samples_per_second": 920.006,
      "eval_steps_per_second": 57.501,
      "step": 368000
    },
    {
      "epoch": 1.03,
      "eval_loss": 2.260352373123169,
      "eval_runtime": 334.4596,
      "eval_samples_per_second": 921.684,
      "eval_steps_per_second": 57.606,
      "step": 376000
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.444e-07,
      "loss": 2.319,
      "step": 384000
    },
    {
      "epoch": 1.05,
      "eval_loss": 2.25346040725708,
      "eval_runtime": 334.8356,
      "eval_samples_per_second": 920.649,
      "eval_steps_per_second": 57.542,
      "step": 384000
    },
    {
      "epoch": 1.07,
      "eval_loss": 2.241976499557495,
      "eval_runtime": 335.4873,
      "eval_samples_per_second": 918.86,
      "eval_steps_per_second": 57.43,
      "step": 392000
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.416666666666667e-07,
      "loss": 2.3151,
      "step": 400000
    },
    {
      "epoch": 1.09,
      "eval_loss": 2.2583224773406982,
      "eval_runtime": 335.4778,
      "eval_samples_per_second": 918.887,
      "eval_steps_per_second": 57.432,
      "step": 400000
    },
    {
      "epoch": 1.11,
      "eval_loss": 2.2534608840942383,
      "eval_runtime": 335.9955,
      "eval_samples_per_second": 917.471,
      "eval_steps_per_second": 57.343,
      "step": 408000
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.3893333333333335e-07,
      "loss": 2.3144,
      "step": 416000
    },
    {
      "epoch": 1.14,
      "eval_loss": 2.258167266845703,
      "eval_runtime": 335.4405,
      "eval_samples_per_second": 918.989,
      "eval_steps_per_second": 57.438,
      "step": 416000
    },
    {
      "epoch": 1.16,
      "eval_loss": 2.249640464782715,
      "eval_runtime": 334.4567,
      "eval_samples_per_second": 921.692,
      "eval_steps_per_second": 57.607,
      "step": 424000
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.3619999999999995e-07,
      "loss": 2.3191,
      "step": 432000
    },
    {
      "epoch": 1.18,
      "eval_loss": 2.2531871795654297,
      "eval_runtime": 335.2874,
      "eval_samples_per_second": 919.408,
      "eval_steps_per_second": 57.464,
      "step": 432000
    },
    {
      "epoch": 1.2,
      "eval_loss": 2.2515294551849365,
      "eval_runtime": 335.0984,
      "eval_samples_per_second": 919.927,
      "eval_steps_per_second": 57.497,
      "step": 440000
    },
    {
      "epoch": 1.22,
      "learning_rate": 3.3346666666666666e-07,
      "loss": 2.3168,
      "step": 448000
    },
    {
      "epoch": 1.22,
      "eval_loss": 2.2500967979431152,
      "eval_runtime": 336.248,
      "eval_samples_per_second": 916.782,
      "eval_steps_per_second": 57.3,
      "step": 448000
    },
    {
      "epoch": 1.25,
      "eval_loss": 2.245288848876953,
      "eval_runtime": 337.1188,
      "eval_samples_per_second": 914.414,
      "eval_steps_per_second": 57.152,
      "step": 456000
    },
    {
      "epoch": 1.27,
      "learning_rate": 3.307333333333333e-07,
      "loss": 2.3156,
      "step": 464000
    },
    {
      "epoch": 1.27,
      "eval_loss": 2.2444746494293213,
      "eval_runtime": 335.5009,
      "eval_samples_per_second": 918.823,
      "eval_steps_per_second": 57.428,
      "step": 464000
    },
    {
      "epoch": 1.29,
      "eval_loss": 2.2484536170959473,
      "eval_runtime": 334.7608,
      "eval_samples_per_second": 920.855,
      "eval_steps_per_second": 57.555,
      "step": 472000
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.28e-07,
      "loss": 2.3178,
      "step": 480000
    },
    {
      "epoch": 1.31,
      "eval_loss": 2.249347448348999,
      "eval_runtime": 335.2634,
      "eval_samples_per_second": 919.474,
      "eval_steps_per_second": 57.468,
      "step": 480000
    },
    {
      "epoch": 1.33,
      "eval_loss": 2.244321346282959,
      "eval_runtime": 335.8842,
      "eval_samples_per_second": 917.775,
      "eval_steps_per_second": 57.362,
      "step": 488000
    },
    {
      "epoch": 1.35,
      "learning_rate": 3.252666666666667e-07,
      "loss": 2.3113,
      "step": 496000
    },
    {
      "epoch": 1.35,
      "eval_loss": 2.2493255138397217,
      "eval_runtime": 336.0104,
      "eval_samples_per_second": 917.43,
      "eval_steps_per_second": 57.34,
      "step": 496000
    },
    {
      "epoch": 1.38,
      "eval_loss": 2.2493276596069336,
      "eval_runtime": 336.0104,
      "eval_samples_per_second": 917.43,
      "eval_steps_per_second": 57.34,
      "step": 504000
    },
    {
      "epoch": 1.4,
      "learning_rate": 3.2253333333333334e-07,
      "loss": 2.3116,
      "step": 512000
    },
    {
      "epoch": 1.4,
      "eval_loss": 2.2483131885528564,
      "eval_runtime": 335.3926,
      "eval_samples_per_second": 919.12,
      "eval_steps_per_second": 57.446,
      "step": 512000
    },
    {
      "epoch": 1.42,
      "eval_loss": 2.245854616165161,
      "eval_runtime": 335.4507,
      "eval_samples_per_second": 918.961,
      "eval_steps_per_second": 57.436,
      "step": 520000
    },
    {
      "epoch": 1.44,
      "learning_rate": 3.198e-07,
      "loss": 2.3166,
      "step": 528000
    },
    {
      "epoch": 1.44,
      "eval_loss": 2.248081922531128,
      "eval_runtime": 336.9072,
      "eval_samples_per_second": 914.988,
      "eval_steps_per_second": 57.188,
      "step": 528000
    },
    {
      "epoch": 1.46,
      "eval_loss": 2.254246711730957,
      "eval_runtime": 335.1811,
      "eval_samples_per_second": 919.7,
      "eval_steps_per_second": 57.482,
      "step": 536000
    },
    {
      "epoch": 1.49,
      "learning_rate": 3.1706666666666665e-07,
      "loss": 2.3158,
      "step": 544000
    },
    {
      "epoch": 1.49,
      "eval_loss": 2.244293451309204,
      "eval_runtime": 336.0363,
      "eval_samples_per_second": 917.359,
      "eval_steps_per_second": 57.336,
      "step": 544000
    },
    {
      "epoch": 1.51,
      "eval_loss": 2.240211009979248,
      "eval_runtime": 335.2536,
      "eval_samples_per_second": 919.501,
      "eval_steps_per_second": 57.47,
      "step": 552000
    },
    {
      "epoch": 1.53,
      "learning_rate": 3.1433333333333336e-07,
      "loss": 2.3148,
      "step": 560000
    },
    {
      "epoch": 1.53,
      "eval_loss": 2.2449421882629395,
      "eval_runtime": 335.5679,
      "eval_samples_per_second": 918.64,
      "eval_steps_per_second": 57.416,
      "step": 560000
    },
    {
      "epoch": 1.55,
      "eval_loss": 2.2415246963500977,
      "eval_runtime": 337.8499,
      "eval_samples_per_second": 912.435,
      "eval_steps_per_second": 57.028,
      "step": 568000
    },
    {
      "epoch": 1.57,
      "learning_rate": 3.116e-07,
      "loss": 2.3145,
      "step": 576000
    },
    {
      "epoch": 1.57,
      "eval_loss": 2.2471208572387695,
      "eval_runtime": 338.2882,
      "eval_samples_per_second": 911.253,
      "eval_steps_per_second": 56.954,
      "step": 576000
    },
    {
      "epoch": 1.6,
      "eval_loss": 2.2469303607940674,
      "eval_runtime": 338.2218,
      "eval_samples_per_second": 911.431,
      "eval_steps_per_second": 56.966,
      "step": 584000
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.0886666666666667e-07,
      "loss": 2.3119,
      "step": 592000
    },
    {
      "epoch": 1.62,
      "eval_loss": 2.244541645050049,
      "eval_runtime": 336.9016,
      "eval_samples_per_second": 915.003,
      "eval_steps_per_second": 57.189,
      "step": 592000
    },
    {
      "epoch": 1.64,
      "eval_loss": 2.248655080795288,
      "eval_runtime": 335.5748,
      "eval_samples_per_second": 918.621,
      "eval_steps_per_second": 57.415,
      "step": 600000
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.061333333333333e-07,
      "loss": 2.3045,
      "step": 608000
    },
    {
      "epoch": 1.66,
      "eval_loss": 2.2455668449401855,
      "eval_runtime": 335.9119,
      "eval_samples_per_second": 917.699,
      "eval_steps_per_second": 57.357,
      "step": 608000
    },
    {
      "epoch": 1.68,
      "eval_loss": 2.246622085571289,
      "eval_runtime": 335.9733,
      "eval_samples_per_second": 917.531,
      "eval_steps_per_second": 57.347,
      "step": 616000
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.034e-07,
      "loss": 2.3046,
      "step": 624000
    },
    {
      "epoch": 1.7,
      "eval_loss": 2.2357234954833984,
      "eval_runtime": 337.5898,
      "eval_samples_per_second": 913.138,
      "eval_steps_per_second": 57.072,
      "step": 624000
    },
    {
      "epoch": 1.73,
      "eval_loss": 2.244824171066284,
      "eval_runtime": 336.5413,
      "eval_samples_per_second": 915.983,
      "eval_steps_per_second": 57.25,
      "step": 632000
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.0066666666666663e-07,
      "loss": 2.3083,
      "step": 640000
    },
    {
      "epoch": 1.75,
      "eval_loss": 2.2381463050842285,
      "eval_runtime": 337.7594,
      "eval_samples_per_second": 912.679,
      "eval_steps_per_second": 57.044,
      "step": 640000
    },
    {
      "epoch": 1.77,
      "eval_loss": 2.24391770362854,
      "eval_runtime": 337.5805,
      "eval_samples_per_second": 913.163,
      "eval_steps_per_second": 57.074,
      "step": 648000
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.9793333333333334e-07,
      "loss": 2.3065,
      "step": 656000
    },
    {
      "epoch": 1.79,
      "eval_loss": 2.2402257919311523,
      "eval_runtime": 336.8532,
      "eval_samples_per_second": 915.135,
      "eval_steps_per_second": 57.197,
      "step": 656000
    },
    {
      "epoch": 1.81,
      "eval_loss": 2.2439496517181396,
      "eval_runtime": 337.4274,
      "eval_samples_per_second": 913.577,
      "eval_steps_per_second": 57.1,
      "step": 664000
    },
    {
      "epoch": 1.84,
      "learning_rate": 2.952e-07,
      "loss": 2.307,
      "step": 672000
    },
    {
      "epoch": 1.84,
      "eval_loss": 2.2409019470214844,
      "eval_runtime": 336.4075,
      "eval_samples_per_second": 916.347,
      "eval_steps_per_second": 57.273,
      "step": 672000
    },
    {
      "epoch": 1.86,
      "eval_loss": 2.2425954341888428,
      "eval_runtime": 336.4355,
      "eval_samples_per_second": 916.271,
      "eval_steps_per_second": 57.268,
      "step": 680000
    },
    {
      "epoch": 1.88,
      "learning_rate": 2.9246666666666665e-07,
      "loss": 2.3026,
      "step": 688000
    },
    {
      "epoch": 1.88,
      "eval_loss": 2.2386796474456787,
      "eval_runtime": 336.9441,
      "eval_samples_per_second": 914.888,
      "eval_steps_per_second": 57.182,
      "step": 688000
    },
    {
      "epoch": 1.9,
      "eval_loss": 2.235677719116211,
      "eval_runtime": 337.5327,
      "eval_samples_per_second": 913.292,
      "eval_steps_per_second": 57.082,
      "step": 696000
    },
    {
      "epoch": 1.92,
      "learning_rate": 2.897333333333333e-07,
      "loss": 2.2949,
      "step": 704000
    },
    {
      "epoch": 1.92,
      "eval_loss": 2.237877607345581,
      "eval_runtime": 336.876,
      "eval_samples_per_second": 915.073,
      "eval_steps_per_second": 57.193,
      "step": 704000
    },
    {
      "epoch": 1.95,
      "eval_loss": 2.2408130168914795,
      "eval_runtime": 338.2422,
      "eval_samples_per_second": 911.376,
      "eval_steps_per_second": 56.962,
      "step": 712000
    },
    {
      "epoch": 1.97,
      "learning_rate": 2.8699999999999996e-07,
      "loss": 2.2951,
      "step": 720000
    },
    {
      "epoch": 1.97,
      "eval_loss": 2.2431986331939697,
      "eval_runtime": 337.6174,
      "eval_samples_per_second": 913.063,
      "eval_steps_per_second": 57.068,
      "step": 720000
    },
    {
      "epoch": 1.99,
      "eval_loss": 2.244434118270874,
      "eval_runtime": 337.68,
      "eval_samples_per_second": 912.894,
      "eval_steps_per_second": 57.057,
      "step": 728000
    },
    {
      "epoch": 2.01,
      "learning_rate": 2.8426666666666667e-07,
      "loss": 2.3011,
      "step": 736000
    },
    {
      "epoch": 2.01,
      "eval_loss": 2.2381932735443115,
      "eval_runtime": 336.7238,
      "eval_samples_per_second": 915.486,
      "eval_steps_per_second": 57.219,
      "step": 736000
    },
    {
      "epoch": 2.03,
      "eval_loss": 2.2391436100006104,
      "eval_runtime": 337.1171,
      "eval_samples_per_second": 914.418,
      "eval_steps_per_second": 57.152,
      "step": 744000
    },
    {
      "epoch": 2.05,
      "learning_rate": 2.815333333333333e-07,
      "loss": 2.3017,
      "step": 752000
    },
    {
      "epoch": 2.05,
      "eval_loss": 2.236323833465576,
      "eval_runtime": 338.3716,
      "eval_samples_per_second": 911.028,
      "eval_steps_per_second": 56.94,
      "step": 752000
    },
    {
      "epoch": 2.08,
      "eval_loss": 2.2444024085998535,
      "eval_runtime": 337.1979,
      "eval_samples_per_second": 914.199,
      "eval_steps_per_second": 57.139,
      "step": 760000
    },
    {
      "epoch": 2.1,
      "learning_rate": 2.7880000000000003e-07,
      "loss": 2.2978,
      "step": 768000
    },
    {
      "epoch": 2.1,
      "eval_loss": 2.2370431423187256,
      "eval_runtime": 338.942,
      "eval_samples_per_second": 909.495,
      "eval_steps_per_second": 56.845,
      "step": 768000
    },
    {
      "epoch": 2.12,
      "eval_loss": 2.2350406646728516,
      "eval_runtime": 337.2566,
      "eval_samples_per_second": 914.04,
      "eval_steps_per_second": 57.129,
      "step": 776000
    },
    {
      "epoch": 2.14,
      "learning_rate": 2.7606666666666664e-07,
      "loss": 2.2961,
      "step": 784000
    },
    {
      "epoch": 2.14,
      "eval_loss": 2.234744071960449,
      "eval_runtime": 338.2021,
      "eval_samples_per_second": 911.485,
      "eval_steps_per_second": 56.969,
      "step": 784000
    },
    {
      "epoch": 2.16,
      "eval_loss": 2.238616704940796,
      "eval_runtime": 338.0434,
      "eval_samples_per_second": 911.913,
      "eval_steps_per_second": 56.996,
      "step": 792000
    },
    {
      "epoch": 2.19,
      "learning_rate": 2.733333333333333e-07,
      "loss": 2.2968,
      "step": 800000
    },
    {
      "epoch": 2.19,
      "eval_loss": 2.2322075366973877,
      "eval_runtime": 337.896,
      "eval_samples_per_second": 912.31,
      "eval_steps_per_second": 57.021,
      "step": 800000
    },
    {
      "epoch": 2.21,
      "eval_loss": 2.240255117416382,
      "eval_runtime": 338.7394,
      "eval_samples_per_second": 910.039,
      "eval_steps_per_second": 56.879,
      "step": 808000
    },
    {
      "epoch": 2.23,
      "learning_rate": 2.706e-07,
      "loss": 2.2962,
      "step": 816000
    },
    {
      "epoch": 2.23,
      "eval_loss": 2.2347311973571777,
      "eval_runtime": 339.8784,
      "eval_samples_per_second": 906.989,
      "eval_steps_per_second": 56.688,
      "step": 816000
    },
    {
      "epoch": 2.25,
      "eval_loss": 2.239776372909546,
      "eval_runtime": 338.4053,
      "eval_samples_per_second": 910.937,
      "eval_steps_per_second": 56.935,
      "step": 824000
    },
    {
      "epoch": 2.27,
      "learning_rate": 2.6786666666666666e-07,
      "loss": 2.2984,
      "step": 832000
    },
    {
      "epoch": 2.27,
      "eval_loss": 2.235778570175171,
      "eval_runtime": 338.4942,
      "eval_samples_per_second": 910.698,
      "eval_steps_per_second": 56.92,
      "step": 832000
    },
    {
      "epoch": 2.29,
      "eval_loss": 2.2412359714508057,
      "eval_runtime": 338.0942,
      "eval_samples_per_second": 911.775,
      "eval_steps_per_second": 56.987,
      "step": 840000
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.651333333333333e-07,
      "loss": 2.3029,
      "step": 848000
    },
    {
      "epoch": 2.32,
      "eval_loss": 2.238647937774658,
      "eval_runtime": 338.342,
      "eval_samples_per_second": 911.108,
      "eval_steps_per_second": 56.945,
      "step": 848000
    },
    {
      "epoch": 2.34,
      "eval_loss": 2.234633445739746,
      "eval_runtime": 339.1684,
      "eval_samples_per_second": 908.888,
      "eval_steps_per_second": 56.807,
      "step": 856000
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.624e-07,
      "loss": 2.2985,
      "step": 864000
    },
    {
      "epoch": 2.36,
      "eval_loss": 2.2323224544525146,
      "eval_runtime": 340.1121,
      "eval_samples_per_second": 906.366,
      "eval_steps_per_second": 56.649,
      "step": 864000
    },
    {
      "epoch": 2.38,
      "eval_loss": 2.2386910915374756,
      "eval_runtime": 338.6678,
      "eval_samples_per_second": 910.231,
      "eval_steps_per_second": 56.891,
      "step": 872000
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.596666666666667e-07,
      "loss": 2.2922,
      "step": 880000
    },
    {
      "epoch": 2.4,
      "eval_loss": 2.230320453643799,
      "eval_runtime": 338.2571,
      "eval_samples_per_second": 911.336,
      "eval_steps_per_second": 56.96,
      "step": 880000
    },
    {
      "epoch": 2.43,
      "eval_loss": 2.232644557952881,
      "eval_runtime": 338.4677,
      "eval_samples_per_second": 910.769,
      "eval_steps_per_second": 56.924,
      "step": 888000
    },
    {
      "epoch": 2.45,
      "learning_rate": 2.5693333333333333e-07,
      "loss": 2.2967,
      "step": 896000
    },
    {
      "epoch": 2.45,
      "eval_loss": 2.2422056198120117,
      "eval_runtime": 338.7421,
      "eval_samples_per_second": 910.031,
      "eval_steps_per_second": 56.878,
      "step": 896000
    },
    {
      "epoch": 2.47,
      "eval_loss": 2.235010862350464,
      "eval_runtime": 339.4694,
      "eval_samples_per_second": 908.082,
      "eval_steps_per_second": 56.756,
      "step": 904000
    },
    {
      "epoch": 2.49,
      "learning_rate": 2.542e-07,
      "loss": 2.2917,
      "step": 912000
    },
    {
      "epoch": 2.49,
      "eval_loss": 2.2299275398254395,
      "eval_runtime": 339.4601,
      "eval_samples_per_second": 908.107,
      "eval_steps_per_second": 56.758,
      "step": 912000
    },
    {
      "epoch": 2.51,
      "eval_loss": 2.2307727336883545,
      "eval_runtime": 339.5941,
      "eval_samples_per_second": 907.748,
      "eval_steps_per_second": 56.735,
      "step": 920000
    },
    {
      "epoch": 2.54,
      "learning_rate": 2.5146666666666664e-07,
      "loss": 2.2912,
      "step": 928000
    },
    {
      "epoch": 2.54,
      "eval_loss": 2.23453688621521,
      "eval_runtime": 339.4986,
      "eval_samples_per_second": 908.004,
      "eval_steps_per_second": 56.751,
      "step": 928000
    },
    {
      "epoch": 2.56,
      "eval_loss": 2.2263941764831543,
      "eval_runtime": 340.6179,
      "eval_samples_per_second": 905.02,
      "eval_steps_per_second": 56.565,
      "step": 936000
    },
    {
      "epoch": 2.58,
      "learning_rate": 2.4873333333333335e-07,
      "loss": 2.2887,
      "step": 944000
    },
    {
      "epoch": 2.58,
      "eval_loss": 2.236109972000122,
      "eval_runtime": 339.6326,
      "eval_samples_per_second": 907.646,
      "eval_steps_per_second": 56.729,
      "step": 944000
    },
    {
      "epoch": 2.6,
      "eval_loss": 2.2318520545959473,
      "eval_runtime": 342.6146,
      "eval_samples_per_second": 899.746,
      "eval_steps_per_second": 56.235,
      "step": 952000
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.46e-07,
      "loss": 2.2956,
      "step": 960000
    },
    {
      "epoch": 2.62,
      "eval_loss": 2.2339940071105957,
      "eval_runtime": 342.333,
      "eval_samples_per_second": 900.486,
      "eval_steps_per_second": 56.281,
      "step": 960000
    },
    {
      "epoch": 2.64,
      "eval_loss": 2.235605478286743,
      "eval_runtime": 342.7919,
      "eval_samples_per_second": 899.28,
      "eval_steps_per_second": 56.206,
      "step": 968000
    },
    {
      "epoch": 2.67,
      "learning_rate": 2.4326666666666666e-07,
      "loss": 2.2927,
      "step": 976000
    },
    {
      "epoch": 2.67,
      "eval_loss": 2.2365500926971436,
      "eval_runtime": 342.6264,
      "eval_samples_per_second": 899.715,
      "eval_steps_per_second": 56.233,
      "step": 976000
    },
    {
      "epoch": 2.69,
      "eval_loss": 2.2334821224212646,
      "eval_runtime": 342.0197,
      "eval_samples_per_second": 901.311,
      "eval_steps_per_second": 56.333,
      "step": 984000
    },
    {
      "epoch": 2.71,
      "learning_rate": 2.405333333333333e-07,
      "loss": 2.2872,
      "step": 992000
    },
    {
      "epoch": 2.71,
      "eval_loss": 2.232952833175659,
      "eval_runtime": 343.3686,
      "eval_samples_per_second": 897.77,
      "eval_steps_per_second": 56.112,
      "step": 992000
    },
    {
      "epoch": 2.73,
      "eval_loss": 2.225148916244507,
      "eval_runtime": 342.4862,
      "eval_samples_per_second": 900.083,
      "eval_steps_per_second": 56.256,
      "step": 1000000
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.3779999999999997e-07,
      "loss": 2.2936,
      "step": 1008000
    },
    {
      "epoch": 2.75,
      "eval_loss": 2.232741594314575,
      "eval_runtime": 342.7157,
      "eval_samples_per_second": 899.48,
      "eval_steps_per_second": 56.219,
      "step": 1008000
    },
    {
      "epoch": 2.78,
      "eval_loss": 2.2326343059539795,
      "eval_runtime": 344.8133,
      "eval_samples_per_second": 894.008,
      "eval_steps_per_second": 55.877,
      "step": 1016000
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.3506666666666668e-07,
      "loss": 2.2899,
      "step": 1024000
    },
    {
      "epoch": 2.8,
      "eval_loss": 2.2306628227233887,
      "eval_runtime": 344.8753,
      "eval_samples_per_second": 893.848,
      "eval_steps_per_second": 55.867,
      "step": 1024000
    },
    {
      "epoch": 2.82,
      "eval_loss": 2.2291181087493896,
      "eval_runtime": 343.9896,
      "eval_samples_per_second": 896.149,
      "eval_steps_per_second": 56.01,
      "step": 1032000
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.3233333333333334e-07,
      "loss": 2.2931,
      "step": 1040000
    },
    {
      "epoch": 2.84,
      "eval_loss": 2.228482723236084,
      "eval_runtime": 343.2535,
      "eval_samples_per_second": 898.071,
      "eval_steps_per_second": 56.131,
      "step": 1040000
    },
    {
      "epoch": 2.86,
      "eval_loss": 2.2326762676239014,
      "eval_runtime": 340.5017,
      "eval_samples_per_second": 905.329,
      "eval_steps_per_second": 56.584,
      "step": 1048000
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.2960000000000002e-07,
      "loss": 2.3042,
      "step": 1056000
    },
    {
      "epoch": 2.88,
      "eval_loss": 2.2366533279418945,
      "eval_runtime": 342.3036,
      "eval_samples_per_second": 900.563,
      "eval_steps_per_second": 56.286,
      "step": 1056000
    },
    {
      "epoch": 2.91,
      "eval_loss": 2.2344892024993896,
      "eval_runtime": 341.9516,
      "eval_samples_per_second": 901.49,
      "eval_steps_per_second": 56.344,
      "step": 1064000
    },
    {
      "epoch": 2.93,
      "learning_rate": 2.2686666666666667e-07,
      "loss": 2.2864,
      "step": 1072000
    },
    {
      "epoch": 2.93,
      "eval_loss": 2.2267308235168457,
      "eval_runtime": 341.6099,
      "eval_samples_per_second": 902.392,
      "eval_steps_per_second": 56.401,
      "step": 1072000
    },
    {
      "epoch": 2.95,
      "eval_loss": 2.2342631816864014,
      "eval_runtime": 343.6113,
      "eval_samples_per_second": 897.136,
      "eval_steps_per_second": 56.072,
      "step": 1080000
    },
    {
      "epoch": 2.97,
      "learning_rate": 2.2413333333333333e-07,
      "loss": 2.2933,
      "step": 1088000
    },
    {
      "epoch": 2.97,
      "eval_loss": 2.235445261001587,
      "eval_runtime": 342.1009,
      "eval_samples_per_second": 901.097,
      "eval_steps_per_second": 56.32,
      "step": 1088000
    },
    {
      "epoch": 2.99,
      "eval_loss": 2.226022243499756,
      "eval_runtime": 343.4712,
      "eval_samples_per_second": 897.502,
      "eval_steps_per_second": 56.095,
      "step": 1096000
    },
    {
      "epoch": 3.02,
      "learning_rate": 2.214e-07,
      "loss": 2.2909,
      "step": 1104000
    },
    {
      "epoch": 3.02,
      "eval_loss": 2.2340822219848633,
      "eval_runtime": 342.2127,
      "eval_samples_per_second": 900.802,
      "eval_steps_per_second": 56.301,
      "step": 1104000
    },
    {
      "epoch": 3.04,
      "eval_loss": 2.2265801429748535,
      "eval_runtime": 344.5846,
      "eval_samples_per_second": 894.602,
      "eval_steps_per_second": 55.914,
      "step": 1112000
    },
    {
      "epoch": 3.06,
      "learning_rate": 2.1866666666666667e-07,
      "loss": 2.2889,
      "step": 1120000
    },
    {
      "epoch": 3.06,
      "eval_loss": 2.225277900695801,
      "eval_runtime": 343.7811,
      "eval_samples_per_second": 896.693,
      "eval_steps_per_second": 56.044,
      "step": 1120000
    },
    {
      "epoch": 3.08,
      "eval_loss": 2.225517511367798,
      "eval_runtime": 344.8771,
      "eval_samples_per_second": 893.843,
      "eval_steps_per_second": 55.866,
      "step": 1128000
    },
    {
      "epoch": 3.1,
      "learning_rate": 2.1593333333333332e-07,
      "loss": 2.292,
      "step": 1136000
    },
    {
      "epoch": 3.1,
      "eval_loss": 2.219359874725342,
      "eval_runtime": 342.5656,
      "eval_samples_per_second": 899.874,
      "eval_steps_per_second": 56.243,
      "step": 1136000
    },
    {
      "epoch": 3.13,
      "eval_loss": 2.2318532466888428,
      "eval_runtime": 342.1989,
      "eval_samples_per_second": 900.839,
      "eval_steps_per_second": 56.304,
      "step": 1144000
    },
    {
      "epoch": 3.15,
      "learning_rate": 2.132e-07,
      "loss": 2.282,
      "step": 1152000
    },
    {
      "epoch": 3.15,
      "eval_loss": 2.2221035957336426,
      "eval_runtime": 342.4265,
      "eval_samples_per_second": 900.24,
      "eval_steps_per_second": 56.266,
      "step": 1152000
    },
    {
      "epoch": 3.17,
      "eval_loss": 2.2272608280181885,
      "eval_runtime": 341.618,
      "eval_samples_per_second": 902.371,
      "eval_steps_per_second": 56.399,
      "step": 1160000
    },
    {
      "epoch": 3.19,
      "learning_rate": 2.1046666666666666e-07,
      "loss": 2.2827,
      "step": 1168000
    },
    {
      "epoch": 3.19,
      "eval_loss": 2.2295727729797363,
      "eval_runtime": 344.8334,
      "eval_samples_per_second": 893.956,
      "eval_steps_per_second": 55.873,
      "step": 1168000
    },
    {
      "epoch": 3.21,
      "eval_loss": 2.2331955432891846,
      "eval_runtime": 343.112,
      "eval_samples_per_second": 898.441,
      "eval_steps_per_second": 56.154,
      "step": 1176000
    },
    {
      "epoch": 3.23,
      "learning_rate": 2.0773333333333334e-07,
      "loss": 2.2937,
      "step": 1184000
    },
    {
      "epoch": 3.23,
      "eval_loss": 2.230241298675537,
      "eval_runtime": 342.0014,
      "eval_samples_per_second": 901.359,
      "eval_steps_per_second": 56.336,
      "step": 1184000
    },
    {
      "epoch": 3.26,
      "eval_loss": 2.2262063026428223,
      "eval_runtime": 344.0166,
      "eval_samples_per_second": 896.079,
      "eval_steps_per_second": 56.006,
      "step": 1192000
    },
    {
      "epoch": 3.28,
      "learning_rate": 2.05e-07,
      "loss": 2.2845,
      "step": 1200000
    },
    {
      "epoch": 3.28,
      "eval_loss": 2.231752872467041,
      "eval_runtime": 342.9925,
      "eval_samples_per_second": 898.754,
      "eval_steps_per_second": 56.173,
      "step": 1200000
    },
    {
      "epoch": 3.3,
      "eval_loss": 2.229050636291504,
      "eval_runtime": 342.7011,
      "eval_samples_per_second": 899.519,
      "eval_steps_per_second": 56.221,
      "step": 1208000
    },
    {
      "epoch": 3.32,
      "learning_rate": 2.0226666666666668e-07,
      "loss": 2.284,
      "step": 1216000
    },
    {
      "epoch": 3.32,
      "eval_loss": 2.232661008834839,
      "eval_runtime": 343.2876,
      "eval_samples_per_second": 897.982,
      "eval_steps_per_second": 56.125,
      "step": 1216000
    },
    {
      "epoch": 3.34,
      "eval_loss": 2.230750560760498,
      "eval_runtime": 343.2951,
      "eval_samples_per_second": 897.962,
      "eval_steps_per_second": 56.124,
      "step": 1224000
    },
    {
      "epoch": 3.37,
      "learning_rate": 1.9953333333333333e-07,
      "loss": 2.2923,
      "step": 1232000
    },
    {
      "epoch": 3.37,
      "eval_loss": 2.226369857788086,
      "eval_runtime": 343.209,
      "eval_samples_per_second": 898.188,
      "eval_steps_per_second": 56.138,
      "step": 1232000
    },
    {
      "epoch": 3.39,
      "eval_loss": 2.2390074729919434,
      "eval_runtime": 342.5512,
      "eval_samples_per_second": 899.912,
      "eval_steps_per_second": 56.246,
      "step": 1240000
    },
    {
      "epoch": 3.41,
      "learning_rate": 1.968e-07,
      "loss": 2.2859,
      "step": 1248000
    },
    {
      "epoch": 3.41,
      "eval_loss": 2.2309505939483643,
      "eval_runtime": 343.4164,
      "eval_samples_per_second": 897.645,
      "eval_steps_per_second": 56.104,
      "step": 1248000
    },
    {
      "epoch": 3.43,
      "eval_loss": 2.22867751121521,
      "eval_runtime": 342.6067,
      "eval_samples_per_second": 899.766,
      "eval_steps_per_second": 56.236,
      "step": 1256000
    },
    {
      "epoch": 3.45,
      "learning_rate": 1.9406666666666667e-07,
      "loss": 2.2879,
      "step": 1264000
    },
    {
      "epoch": 3.45,
      "eval_loss": 2.228405714035034,
      "eval_runtime": 345.1005,
      "eval_samples_per_second": 893.264,
      "eval_steps_per_second": 55.83,
      "step": 1264000
    },
    {
      "epoch": 3.47,
      "eval_loss": 2.2228379249572754,
      "eval_runtime": 343.2335,
      "eval_samples_per_second": 898.123,
      "eval_steps_per_second": 56.134,
      "step": 1272000
    },
    {
      "epoch": 3.5,
      "learning_rate": 1.9133333333333333e-07,
      "loss": 2.292,
      "step": 1280000
    },
    {
      "epoch": 3.5,
      "eval_loss": 2.2295541763305664,
      "eval_runtime": 343.6906,
      "eval_samples_per_second": 896.929,
      "eval_steps_per_second": 56.059,
      "step": 1280000
    },
    {
      "epoch": 3.52,
      "eval_loss": 2.232851982116699,
      "eval_runtime": 343.5216,
      "eval_samples_per_second": 897.37,
      "eval_steps_per_second": 56.087,
      "step": 1288000
    },
    {
      "epoch": 3.54,
      "learning_rate": 1.886e-07,
      "loss": 2.2827,
      "step": 1296000
    },
    {
      "epoch": 3.54,
      "eval_loss": 2.226313352584839,
      "eval_runtime": 343.9653,
      "eval_samples_per_second": 896.213,
      "eval_steps_per_second": 56.014,
      "step": 1296000
    },
    {
      "epoch": 3.56,
      "eval_loss": 2.2323992252349854,
      "eval_runtime": 345.0144,
      "eval_samples_per_second": 893.487,
      "eval_steps_per_second": 55.844,
      "step": 1304000
    },
    {
      "epoch": 3.58,
      "learning_rate": 1.8586666666666666e-07,
      "loss": 2.2829,
      "step": 1312000
    },
    {
      "epoch": 3.58,
      "eval_loss": 2.2231664657592773,
      "eval_runtime": 343.513,
      "eval_samples_per_second": 897.393,
      "eval_steps_per_second": 56.088,
      "step": 1312000
    },
    {
      "epoch": 3.61,
      "eval_loss": 2.2273144721984863,
      "eval_runtime": 344.0809,
      "eval_samples_per_second": 895.911,
      "eval_steps_per_second": 55.996,
      "step": 1320000
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8313333333333332e-07,
      "loss": 2.2863,
      "step": 1328000
    },
    {
      "epoch": 3.63,
      "eval_loss": 2.2296173572540283,
      "eval_runtime": 344.7931,
      "eval_samples_per_second": 894.061,
      "eval_steps_per_second": 55.88,
      "step": 1328000
    },
    {
      "epoch": 3.65,
      "eval_loss": 2.2293524742126465,
      "eval_runtime": 343.7387,
      "eval_samples_per_second": 896.803,
      "eval_steps_per_second": 56.051,
      "step": 1336000
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.804e-07,
      "loss": 2.2796,
      "step": 1344000
    },
    {
      "epoch": 3.67,
      "eval_loss": 2.228300094604492,
      "eval_runtime": 345.0604,
      "eval_samples_per_second": 893.368,
      "eval_steps_per_second": 55.837,
      "step": 1344000
    },
    {
      "epoch": 3.69,
      "eval_loss": 2.2279834747314453,
      "eval_runtime": 343.4098,
      "eval_samples_per_second": 897.662,
      "eval_steps_per_second": 56.105,
      "step": 1352000
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.7766666666666666e-07,
      "loss": 2.2835,
      "step": 1360000
    },
    {
      "epoch": 3.72,
      "eval_loss": 2.226436138153076,
      "eval_runtime": 344.8134,
      "eval_samples_per_second": 894.008,
      "eval_steps_per_second": 55.877,
      "step": 1360000
    },
    {
      "epoch": 3.74,
      "eval_loss": 2.222442150115967,
      "eval_runtime": 344.7708,
      "eval_samples_per_second": 894.119,
      "eval_steps_per_second": 55.883,
      "step": 1368000
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.7493333333333334e-07,
      "loss": 2.2875,
      "step": 1376000
    },
    {
      "epoch": 3.76,
      "eval_loss": 2.2218754291534424,
      "eval_runtime": 346.0197,
      "eval_samples_per_second": 890.891,
      "eval_steps_per_second": 55.682,
      "step": 1376000
    },
    {
      "epoch": 3.78,
      "eval_loss": 2.224281072616577,
      "eval_runtime": 344.7226,
      "eval_samples_per_second": 894.244,
      "eval_steps_per_second": 55.891,
      "step": 1384000
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.722e-07,
      "loss": 2.2792,
      "step": 1392000
    },
    {
      "epoch": 3.8,
      "eval_loss": 2.232009172439575,
      "eval_runtime": 344.233,
      "eval_samples_per_second": 895.516,
      "eval_steps_per_second": 55.971,
      "step": 1392000
    },
    {
      "epoch": 3.82,
      "eval_loss": 2.227288246154785,
      "eval_runtime": 344.3453,
      "eval_samples_per_second": 895.223,
      "eval_steps_per_second": 55.953,
      "step": 1400000
    },
    {
      "epoch": 3.85,
      "learning_rate": 1.6946666666666668e-07,
      "loss": 2.2932,
      "step": 1408000
    },
    {
      "epoch": 3.85,
      "eval_loss": 2.225741386413574,
      "eval_runtime": 343.9791,
      "eval_samples_per_second": 896.177,
      "eval_steps_per_second": 56.012,
      "step": 1408000
    },
    {
      "epoch": 3.87,
      "eval_loss": 2.235980749130249,
      "eval_runtime": 344.5804,
      "eval_samples_per_second": 894.613,
      "eval_steps_per_second": 55.914,
      "step": 1416000
    },
    {
      "epoch": 3.89,
      "learning_rate": 1.6673333333333333e-07,
      "loss": 2.2899,
      "step": 1424000
    },
    {
      "epoch": 3.89,
      "eval_loss": 2.227717399597168,
      "eval_runtime": 345.0207,
      "eval_samples_per_second": 893.471,
      "eval_steps_per_second": 55.843,
      "step": 1424000
    },
    {
      "epoch": 3.91,
      "eval_loss": 2.227459192276001,
      "eval_runtime": 345.1331,
      "eval_samples_per_second": 893.18,
      "eval_steps_per_second": 55.825,
      "step": 1432000
    },
    {
      "epoch": 3.93,
      "learning_rate": 1.64e-07,
      "loss": 2.2859,
      "step": 1440000
    },
    {
      "epoch": 3.93,
      "eval_loss": 2.228656530380249,
      "eval_runtime": 345.5823,
      "eval_samples_per_second": 892.019,
      "eval_steps_per_second": 55.752,
      "step": 1440000
    },
    {
      "epoch": 3.96,
      "eval_loss": 2.2210566997528076,
      "eval_runtime": 345.2225,
      "eval_samples_per_second": 892.949,
      "eval_steps_per_second": 55.81,
      "step": 1448000
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.6126666666666667e-07,
      "loss": 2.2876,
      "step": 1456000
    },
    {
      "epoch": 3.98,
      "eval_loss": 2.2235565185546875,
      "eval_runtime": 345.2552,
      "eval_samples_per_second": 892.864,
      "eval_steps_per_second": 55.805,
      "step": 1456000
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.2287678718566895,
      "eval_runtime": 345.4865,
      "eval_samples_per_second": 892.266,
      "eval_steps_per_second": 55.768,
      "step": 1464000
    },
    {
      "epoch": 4.02,
      "learning_rate": 1.5853333333333332e-07,
      "loss": 2.2879,
      "step": 1472000
    },
    {
      "epoch": 4.02,
      "eval_loss": 2.2225778102874756,
      "eval_runtime": 346.081,
      "eval_samples_per_second": 890.734,
      "eval_steps_per_second": 55.672,
      "step": 1472000
    },
    {
      "epoch": 4.04,
      "eval_loss": 2.2241647243499756,
      "eval_runtime": 345.7366,
      "eval_samples_per_second": 891.621,
      "eval_steps_per_second": 55.727,
      "step": 1480000
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.558e-07,
      "loss": 2.282,
      "step": 1488000
    },
    {
      "epoch": 4.06,
      "eval_loss": 2.2286031246185303,
      "eval_runtime": 345.8095,
      "eval_samples_per_second": 891.433,
      "eval_steps_per_second": 55.716,
      "step": 1488000
    },
    {
      "epoch": 4.09,
      "eval_loss": 2.220984697341919,
      "eval_runtime": 346.175,
      "eval_samples_per_second": 890.492,
      "eval_steps_per_second": 55.657,
      "step": 1496000
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.5306666666666666e-07,
      "loss": 2.2828,
      "step": 1504000
    },
    {
      "epoch": 4.11,
      "eval_loss": 2.2303643226623535,
      "eval_runtime": 345.9183,
      "eval_samples_per_second": 891.153,
      "eval_steps_per_second": 55.698,
      "step": 1504000
    },
    {
      "epoch": 4.13,
      "eval_loss": 2.2310214042663574,
      "eval_runtime": 345.0795,
      "eval_samples_per_second": 893.319,
      "eval_steps_per_second": 55.834,
      "step": 1512000
    },
    {
      "epoch": 4.15,
      "learning_rate": 1.5033333333333332e-07,
      "loss": 2.2765,
      "step": 1520000
    },
    {
      "epoch": 4.15,
      "eval_loss": 2.229534387588501,
      "eval_runtime": 346.2314,
      "eval_samples_per_second": 890.347,
      "eval_steps_per_second": 55.648,
      "step": 1520000
    },
    {
      "epoch": 4.17,
      "eval_loss": 2.2276086807250977,
      "eval_runtime": 345.5194,
      "eval_samples_per_second": 892.181,
      "eval_steps_per_second": 55.762,
      "step": 1528000
    },
    {
      "epoch": 4.2,
      "learning_rate": 1.476e-07,
      "loss": 2.2839,
      "step": 1536000
    },
    {
      "epoch": 4.2,
      "eval_loss": 2.226030111312866,
      "eval_runtime": 345.3689,
      "eval_samples_per_second": 892.57,
      "eval_steps_per_second": 55.787,
      "step": 1536000
    },
    {
      "epoch": 4.22,
      "eval_loss": 2.225531578063965,
      "eval_runtime": 346.1485,
      "eval_samples_per_second": 890.56,
      "eval_steps_per_second": 55.661,
      "step": 1544000
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.4486666666666665e-07,
      "loss": 2.2845,
      "step": 1552000
    },
    {
      "epoch": 4.24,
      "eval_loss": 2.2199981212615967,
      "eval_runtime": 345.8554,
      "eval_samples_per_second": 891.315,
      "eval_steps_per_second": 55.708,
      "step": 1552000
    },
    {
      "epoch": 4.26,
      "eval_loss": 2.222754955291748,
      "eval_runtime": 346.6039,
      "eval_samples_per_second": 889.39,
      "eval_steps_per_second": 55.588,
      "step": 1560000
    },
    {
      "epoch": 4.28,
      "learning_rate": 1.4213333333333334e-07,
      "loss": 2.2816,
      "step": 1568000
    },
    {
      "epoch": 4.28,
      "eval_loss": 2.2322280406951904,
      "eval_runtime": 346.0535,
      "eval_samples_per_second": 890.804,
      "eval_steps_per_second": 55.676,
      "step": 1568000
    },
    {
      "epoch": 4.31,
      "eval_loss": 2.225015163421631,
      "eval_runtime": 345.7846,
      "eval_samples_per_second": 891.497,
      "eval_steps_per_second": 55.72,
      "step": 1576000
    },
    {
      "epoch": 4.33,
      "learning_rate": 1.3940000000000002e-07,
      "loss": 2.2965,
      "step": 1584000
    },
    {
      "epoch": 4.33,
      "eval_loss": 2.2242190837860107,
      "eval_runtime": 348.9458,
      "eval_samples_per_second": 883.421,
      "eval_steps_per_second": 55.215,
      "step": 1584000
    },
    {
      "epoch": 4.35,
      "eval_loss": 2.22951340675354,
      "eval_runtime": 346.387,
      "eval_samples_per_second": 889.947,
      "eval_steps_per_second": 55.623,
      "step": 1592000
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.3666666666666665e-07,
      "loss": 2.2806,
      "step": 1600000
    },
    {
      "epoch": 4.37,
      "eval_loss": 2.219784736633301,
      "eval_runtime": 346.2214,
      "eval_samples_per_second": 890.372,
      "eval_steps_per_second": 55.649,
      "step": 1600000
    },
    {
      "epoch": 4.39,
      "eval_loss": 2.230062961578369,
      "eval_runtime": 346.9441,
      "eval_samples_per_second": 888.518,
      "eval_steps_per_second": 55.533,
      "step": 1608000
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.3393333333333333e-07,
      "loss": 2.2868,
      "step": 1616000
    },
    {
      "epoch": 4.41,
      "eval_loss": 2.2308502197265625,
      "eval_runtime": 348.3618,
      "eval_samples_per_second": 884.902,
      "eval_steps_per_second": 55.307,
      "step": 1616000
    },
    {
      "epoch": 4.44,
      "eval_loss": 2.226969003677368,
      "eval_runtime": 347.2742,
      "eval_samples_per_second": 887.673,
      "eval_steps_per_second": 55.481,
      "step": 1624000
    },
    {
      "epoch": 4.46,
      "learning_rate": 1.312e-07,
      "loss": 2.2907,
      "step": 1632000
    },
    {
      "epoch": 4.46,
      "eval_loss": 2.2290947437286377,
      "eval_runtime": 346.9393,
      "eval_samples_per_second": 888.53,
      "eval_steps_per_second": 55.534,
      "step": 1632000
    },
    {
      "epoch": 4.48,
      "eval_loss": 2.226861000061035,
      "eval_runtime": 347.5547,
      "eval_samples_per_second": 886.957,
      "eval_steps_per_second": 55.436,
      "step": 1640000
    },
    {
      "epoch": 4.5,
      "learning_rate": 1.2846666666666667e-07,
      "loss": 2.2809,
      "step": 1648000
    },
    {
      "epoch": 4.5,
      "eval_loss": 2.2260689735412598,
      "eval_runtime": 347.4948,
      "eval_samples_per_second": 887.11,
      "eval_steps_per_second": 55.445,
      "step": 1648000
    },
    {
      "epoch": 4.52,
      "eval_loss": 2.231820583343506,
      "eval_runtime": 348.114,
      "eval_samples_per_second": 885.532,
      "eval_steps_per_second": 55.347,
      "step": 1656000
    },
    {
      "epoch": 4.55,
      "learning_rate": 1.2573333333333332e-07,
      "loss": 2.2876,
      "step": 1664000
    },
    {
      "epoch": 4.55,
      "eval_loss": 2.22523832321167,
      "eval_runtime": 347.061,
      "eval_samples_per_second": 888.219,
      "eval_steps_per_second": 55.515,
      "step": 1664000
    },
    {
      "epoch": 4.57,
      "eval_loss": 2.2248425483703613,
      "eval_runtime": 347.102,
      "eval_samples_per_second": 888.114,
      "eval_steps_per_second": 55.508,
      "step": 1672000
    },
    {
      "epoch": 4.59,
      "learning_rate": 1.23e-07,
      "loss": 2.2844,
      "step": 1680000
    },
    {
      "epoch": 4.59,
      "eval_loss": 2.222309112548828,
      "eval_runtime": 347.6799,
      "eval_samples_per_second": 886.637,
      "eval_steps_per_second": 55.416,
      "step": 1680000
    },
    {
      "epoch": 4.61,
      "eval_loss": 2.2250306606292725,
      "eval_runtime": 348.9099,
      "eval_samples_per_second": 883.512,
      "eval_steps_per_second": 55.221,
      "step": 1688000
    },
    {
      "epoch": 4.63,
      "learning_rate": 1.2026666666666666e-07,
      "loss": 2.2841,
      "step": 1696000
    },
    {
      "epoch": 4.63,
      "eval_loss": 2.227815866470337,
      "eval_runtime": 347.3551,
      "eval_samples_per_second": 887.466,
      "eval_steps_per_second": 55.468,
      "step": 1696000
    },
    {
      "epoch": 4.65,
      "eval_loss": 2.222553014755249,
      "eval_runtime": 347.2892,
      "eval_samples_per_second": 887.635,
      "eval_steps_per_second": 55.478,
      "step": 1704000
    },
    {
      "epoch": 4.68,
      "learning_rate": 1.1753333333333334e-07,
      "loss": 2.2851,
      "step": 1712000
    },
    {
      "epoch": 4.68,
      "eval_loss": 2.2273736000061035,
      "eval_runtime": 347.2652,
      "eval_samples_per_second": 887.696,
      "eval_steps_per_second": 55.482,
      "step": 1712000
    },
    {
      "epoch": 4.7,
      "eval_loss": 2.2246508598327637,
      "eval_runtime": 348.4684,
      "eval_samples_per_second": 884.631,
      "eval_steps_per_second": 55.291,
      "step": 1720000
    },
    {
      "epoch": 4.72,
      "learning_rate": 1.1480000000000001e-07,
      "loss": 2.2863,
      "step": 1728000
    },
    {
      "epoch": 4.72,
      "eval_loss": 2.223870277404785,
      "eval_runtime": 347.9398,
      "eval_samples_per_second": 885.975,
      "eval_steps_per_second": 55.375,
      "step": 1728000
    },
    {
      "epoch": 4.74,
      "eval_loss": 2.2227413654327393,
      "eval_runtime": 350.0067,
      "eval_samples_per_second": 880.743,
      "eval_steps_per_second": 55.048,
      "step": 1736000
    },
    {
      "epoch": 4.76,
      "learning_rate": 1.1206666666666666e-07,
      "loss": 2.2788,
      "step": 1744000
    },
    {
      "epoch": 4.76,
      "eval_loss": 2.223409414291382,
      "eval_runtime": 350.9466,
      "eval_samples_per_second": 878.384,
      "eval_steps_per_second": 54.9,
      "step": 1744000
    },
    {
      "epoch": 4.79,
      "eval_loss": 2.2293310165405273,
      "eval_runtime": 350.7999,
      "eval_samples_per_second": 878.752,
      "eval_steps_per_second": 54.923,
      "step": 1752000
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.0933333333333333e-07,
      "loss": 2.2849,
      "step": 1760000
    },
    {
      "epoch": 4.81,
      "eval_loss": 2.2198665142059326,
      "eval_runtime": 350.5794,
      "eval_samples_per_second": 879.304,
      "eval_steps_per_second": 54.958,
      "step": 1760000
    },
    {
      "epoch": 4.83,
      "eval_loss": 2.2308552265167236,
      "eval_runtime": 351.9489,
      "eval_samples_per_second": 875.883,
      "eval_steps_per_second": 54.744,
      "step": 1768000
    },
    {
      "epoch": 4.85,
      "learning_rate": 1.066e-07,
      "loss": 2.2826,
      "step": 1776000
    },
    {
      "epoch": 4.85,
      "eval_loss": 2.223459243774414,
      "eval_runtime": 352.3884,
      "eval_samples_per_second": 874.79,
      "eval_steps_per_second": 54.675,
      "step": 1776000
    },
    {
      "epoch": 4.87,
      "eval_loss": 2.2291903495788574,
      "eval_runtime": 350.0779,
      "eval_samples_per_second": 880.564,
      "eval_steps_per_second": 55.036,
      "step": 1784000
    },
    {
      "epoch": 4.9,
      "learning_rate": 1.0386666666666667e-07,
      "loss": 2.2809,
      "step": 1792000
    },
    {
      "epoch": 4.9,
      "eval_loss": 2.224785089492798,
      "eval_runtime": 351.7673,
      "eval_samples_per_second": 876.335,
      "eval_steps_per_second": 54.772,
      "step": 1792000
    },
    {
      "epoch": 4.92,
      "eval_loss": 2.218683958053589,
      "eval_runtime": 353.6242,
      "eval_samples_per_second": 871.733,
      "eval_steps_per_second": 54.484,
      "step": 1800000
    },
    {
      "epoch": 4.94,
      "learning_rate": 1.0113333333333334e-07,
      "loss": 2.2865,
      "step": 1808000
    },
    {
      "epoch": 4.94,
      "eval_loss": 2.233116388320923,
      "eval_runtime": 351.3999,
      "eval_samples_per_second": 877.251,
      "eval_steps_per_second": 54.829,
      "step": 1808000
    },
    {
      "epoch": 4.96,
      "eval_loss": 2.2243831157684326,
      "eval_runtime": 349.1059,
      "eval_samples_per_second": 883.016,
      "eval_steps_per_second": 55.19,
      "step": 1816000
    },
    {
      "epoch": 4.98,
      "learning_rate": 9.84e-08,
      "loss": 2.2773,
      "step": 1824000
    },
    {
      "epoch": 4.98,
      "eval_loss": 2.2246005535125732,
      "eval_runtime": 349.0697,
      "eval_samples_per_second": 883.107,
      "eval_steps_per_second": 55.195,
      "step": 1824000
    },
    {
      "epoch": 5.0,
      "eval_loss": 2.2314696311950684,
      "eval_runtime": 349.3246,
      "eval_samples_per_second": 882.463,
      "eval_steps_per_second": 55.155,
      "step": 1832000
    },
    {
      "epoch": 5.03,
      "learning_rate": 9.566666666666666e-08,
      "loss": 2.2738,
      "step": 1840000
    },
    {
      "epoch": 5.03,
      "eval_loss": 2.231853485107422,
      "eval_runtime": 349.0971,
      "eval_samples_per_second": 883.038,
      "eval_steps_per_second": 55.191,
      "step": 1840000
    },
    {
      "epoch": 5.05,
      "eval_loss": 2.225752115249634,
      "eval_runtime": 350.4379,
      "eval_samples_per_second": 879.659,
      "eval_steps_per_second": 54.98,
      "step": 1848000
    },
    {
      "epoch": 5.07,
      "learning_rate": 9.293333333333333e-08,
      "loss": 2.2806,
      "step": 1856000
    },
    {
      "epoch": 5.07,
      "eval_loss": 2.2240936756134033,
      "eval_runtime": 350.1455,
      "eval_samples_per_second": 880.394,
      "eval_steps_per_second": 55.026,
      "step": 1856000
    },
    {
      "epoch": 5.09,
      "eval_loss": 2.2228317260742188,
      "eval_runtime": 352.1603,
      "eval_samples_per_second": 875.357,
      "eval_steps_per_second": 54.711,
      "step": 1864000
    },
    {
      "epoch": 5.11,
      "learning_rate": 9.02e-08,
      "loss": 2.2822,
      "step": 1872000
    },
    {
      "epoch": 5.11,
      "eval_loss": 2.2218220233917236,
      "eval_runtime": 352.822,
      "eval_samples_per_second": 873.715,
      "eval_steps_per_second": 54.608,
      "step": 1872000
    },
    {
      "epoch": 5.14,
      "eval_loss": 2.227595329284668,
      "eval_runtime": 350.713,
      "eval_samples_per_second": 878.969,
      "eval_steps_per_second": 54.937,
      "step": 1880000
    },
    {
      "epoch": 5.16,
      "learning_rate": 8.746666666666667e-08,
      "loss": 2.2866,
      "step": 1888000
    },
    {
      "epoch": 5.16,
      "eval_loss": 2.2233176231384277,
      "eval_runtime": 351.7914,
      "eval_samples_per_second": 876.275,
      "eval_steps_per_second": 54.768,
      "step": 1888000
    },
    {
      "epoch": 5.18,
      "eval_loss": 2.2265591621398926,
      "eval_runtime": 350.4277,
      "eval_samples_per_second": 879.685,
      "eval_steps_per_second": 54.981,
      "step": 1896000
    },
    {
      "epoch": 5.2,
      "learning_rate": 8.473333333333334e-08,
      "loss": 2.2831,
      "step": 1904000
    },
    {
      "epoch": 5.2,
      "eval_loss": 2.2230618000030518,
      "eval_runtime": 353.0183,
      "eval_samples_per_second": 873.229,
      "eval_steps_per_second": 54.578,
      "step": 1904000
    },
    {
      "epoch": 5.22,
      "eval_loss": 2.224078416824341,
      "eval_runtime": 349.9623,
      "eval_samples_per_second": 880.855,
      "eval_steps_per_second": 55.055,
      "step": 1912000
    },
    {
      "epoch": 5.24,
      "learning_rate": 8.2e-08,
      "loss": 2.2875,
      "step": 1920000
    },
    {
      "epoch": 5.24,
      "eval_loss": 2.226329803466797,
      "eval_runtime": 352.5996,
      "eval_samples_per_second": 874.266,
      "eval_steps_per_second": 54.643,
      "step": 1920000
    },
    {
      "epoch": 5.27,
      "eval_loss": 2.22342586517334,
      "eval_runtime": 350.4146,
      "eval_samples_per_second": 879.718,
      "eval_steps_per_second": 54.983,
      "step": 1928000
    },
    {
      "epoch": 5.29,
      "learning_rate": 7.926666666666666e-08,
      "loss": 2.2802,
      "step": 1936000
    },
    {
      "epoch": 5.29,
      "eval_loss": 2.2269349098205566,
      "eval_runtime": 352.2475,
      "eval_samples_per_second": 875.14,
      "eval_steps_per_second": 54.697,
      "step": 1936000
    },
    {
      "epoch": 5.31,
      "eval_loss": 2.2252981662750244,
      "eval_runtime": 351.5457,
      "eval_samples_per_second": 876.887,
      "eval_steps_per_second": 54.807,
      "step": 1944000
    },
    {
      "epoch": 5.33,
      "learning_rate": 7.653333333333333e-08,
      "loss": 2.2905,
      "step": 1952000
    },
    {
      "epoch": 5.33,
      "eval_loss": 2.2190773487091064,
      "eval_runtime": 351.1477,
      "eval_samples_per_second": 877.881,
      "eval_steps_per_second": 54.869,
      "step": 1952000
    },
    {
      "epoch": 5.35,
      "eval_loss": 2.2216830253601074,
      "eval_runtime": 350.8504,
      "eval_samples_per_second": 878.625,
      "eval_steps_per_second": 54.915,
      "step": 1960000
    },
    {
      "epoch": 5.38,
      "learning_rate": 7.38e-08,
      "loss": 2.282,
      "step": 1968000
    },
    {
      "epoch": 5.38,
      "eval_loss": 2.221177577972412,
      "eval_runtime": 352.0571,
      "eval_samples_per_second": 875.614,
      "eval_steps_per_second": 54.727,
      "step": 1968000
    },
    {
      "epoch": 5.4,
      "eval_loss": 2.221277952194214,
      "eval_runtime": 350.8021,
      "eval_samples_per_second": 878.746,
      "eval_steps_per_second": 54.923,
      "step": 1976000
    },
    {
      "epoch": 5.42,
      "learning_rate": 7.106666666666667e-08,
      "loss": 2.2798,
      "step": 1984000
    },
    {
      "epoch": 5.42,
      "eval_loss": 2.2217955589294434,
      "eval_runtime": 352.0687,
      "eval_samples_per_second": 875.585,
      "eval_steps_per_second": 54.725,
      "step": 1984000
    },
    {
      "epoch": 5.44,
      "eval_loss": 2.222245216369629,
      "eval_runtime": 351.0214,
      "eval_samples_per_second": 878.197,
      "eval_steps_per_second": 54.888,
      "step": 1992000
    },
    {
      "epoch": 5.46,
      "learning_rate": 6.833333333333332e-08,
      "loss": 2.2864,
      "step": 2000000
    },
    {
      "epoch": 5.46,
      "eval_loss": 2.2212188243865967,
      "eval_runtime": 351.5119,
      "eval_samples_per_second": 876.972,
      "eval_steps_per_second": 54.812,
      "step": 2000000
    },
    {
      "epoch": 5.49,
      "eval_loss": 2.228152275085449,
      "eval_runtime": 351.2144,
      "eval_samples_per_second": 877.715,
      "eval_steps_per_second": 54.858,
      "step": 2008000
    },
    {
      "epoch": 5.51,
      "learning_rate": 6.56e-08,
      "loss": 2.2867,
      "step": 2016000
    },
    {
      "epoch": 5.51,
      "eval_loss": 2.2304341793060303,
      "eval_runtime": 352.0733,
      "eval_samples_per_second": 875.573,
      "eval_steps_per_second": 54.724,
      "step": 2016000
    },
    {
      "epoch": 5.53,
      "eval_loss": 2.2221643924713135,
      "eval_runtime": 353.329,
      "eval_samples_per_second": 872.462,
      "eval_steps_per_second": 54.53,
      "step": 2024000
    },
    {
      "epoch": 5.55,
      "learning_rate": 6.286666666666666e-08,
      "loss": 2.2834,
      "step": 2032000
    },
    {
      "epoch": 5.55,
      "eval_loss": 2.2284741401672363,
      "eval_runtime": 352.1812,
      "eval_samples_per_second": 875.305,
      "eval_steps_per_second": 54.708,
      "step": 2032000
    },
    {
      "epoch": 5.57,
      "eval_loss": 2.222963571548462,
      "eval_runtime": 351.3628,
      "eval_samples_per_second": 877.344,
      "eval_steps_per_second": 54.835,
      "step": 2040000
    },
    {
      "epoch": 5.59,
      "learning_rate": 6.013333333333333e-08,
      "loss": 2.2851,
      "step": 2048000
    },
    {
      "epoch": 5.59,
      "eval_loss": 2.223684072494507,
      "eval_runtime": 351.5337,
      "eval_samples_per_second": 876.917,
      "eval_steps_per_second": 54.808,
      "step": 2048000
    },
    {
      "epoch": 5.62,
      "eval_loss": 2.228254795074463,
      "eval_runtime": 352.7325,
      "eval_samples_per_second": 873.937,
      "eval_steps_per_second": 54.622,
      "step": 2056000
    },
    {
      "epoch": 5.64,
      "learning_rate": 5.7400000000000004e-08,
      "loss": 2.2774,
      "step": 2064000
    },
    {
      "epoch": 5.64,
      "eval_loss": 2.2232439517974854,
      "eval_runtime": 352.1455,
      "eval_samples_per_second": 875.394,
      "eval_steps_per_second": 54.713,
      "step": 2064000
    },
    {
      "epoch": 5.66,
      "eval_loss": 2.2282047271728516,
      "eval_runtime": 352.0221,
      "eval_samples_per_second": 875.701,
      "eval_steps_per_second": 54.732,
      "step": 2072000
    },
    {
      "epoch": 5.68,
      "learning_rate": 5.4666666666666666e-08,
      "loss": 2.277,
      "step": 2080000
    },
    {
      "epoch": 5.68,
      "eval_loss": 2.2271482944488525,
      "eval_runtime": 351.8672,
      "eval_samples_per_second": 876.086,
      "eval_steps_per_second": 54.756,
      "step": 2080000
    },
    {
      "epoch": 5.7,
      "eval_loss": 2.2255890369415283,
      "eval_runtime": 351.9475,
      "eval_samples_per_second": 875.886,
      "eval_steps_per_second": 54.744,
      "step": 2088000
    },
    {
      "epoch": 5.73,
      "learning_rate": 5.1933333333333335e-08,
      "loss": 2.2868,
      "step": 2096000
    },
    {
      "epoch": 5.73,
      "eval_loss": 2.2252378463745117,
      "eval_runtime": 352.5562,
      "eval_samples_per_second": 874.374,
      "eval_steps_per_second": 54.649,
      "step": 2096000
    },
    {
      "epoch": 5.75,
      "eval_loss": 2.228463888168335,
      "eval_runtime": 352.4037,
      "eval_samples_per_second": 874.753,
      "eval_steps_per_second": 54.673,
      "step": 2104000
    },
    {
      "epoch": 5.77,
      "learning_rate": 4.92e-08,
      "loss": 2.2727,
      "step": 2112000
    },
    {
      "epoch": 5.77,
      "eval_loss": 2.2250723838806152,
      "eval_runtime": 351.7999,
      "eval_samples_per_second": 876.254,
      "eval_steps_per_second": 54.767,
      "step": 2112000
    },
    {
      "epoch": 5.79,
      "eval_loss": 2.2239432334899902,
      "eval_runtime": 352.3889,
      "eval_samples_per_second": 874.789,
      "eval_steps_per_second": 54.675,
      "step": 2120000
    },
    {
      "epoch": 5.81,
      "learning_rate": 4.6466666666666666e-08,
      "loss": 2.2803,
      "step": 2128000
    },
    {
      "epoch": 5.81,
      "eval_loss": 2.228705883026123,
      "eval_runtime": 352.9086,
      "eval_samples_per_second": 873.501,
      "eval_steps_per_second": 54.595,
      "step": 2128000
    },
    {
      "epoch": 5.84,
      "eval_loss": 2.227353096008301,
      "eval_runtime": 353.6833,
      "eval_samples_per_second": 871.588,
      "eval_steps_per_second": 54.475,
      "step": 2136000
    },
    {
      "epoch": 5.86,
      "learning_rate": 4.3733333333333335e-08,
      "loss": 2.2785,
      "step": 2144000
    },
    {
      "epoch": 5.86,
      "eval_loss": 2.2227251529693604,
      "eval_runtime": 353.3218,
      "eval_samples_per_second": 872.479,
      "eval_steps_per_second": 54.531,
      "step": 2144000
    },
    {
      "epoch": 5.88,
      "eval_loss": 2.226724863052368,
      "eval_runtime": 352.6583,
      "eval_samples_per_second": 874.121,
      "eval_steps_per_second": 54.634,
      "step": 2152000
    },
    {
      "epoch": 5.9,
      "learning_rate": 4.1e-08,
      "loss": 2.2829,
      "step": 2160000
    },
    {
      "epoch": 5.9,
      "eval_loss": 2.225097894668579,
      "eval_runtime": 352.6716,
      "eval_samples_per_second": 874.088,
      "eval_steps_per_second": 54.632,
      "step": 2160000
    },
    {
      "epoch": 5.92,
      "eval_loss": 2.222792387008667,
      "eval_runtime": 353.112,
      "eval_samples_per_second": 872.998,
      "eval_steps_per_second": 54.563,
      "step": 2168000
    },
    {
      "epoch": 5.94,
      "learning_rate": 3.8266666666666665e-08,
      "loss": 2.2816,
      "step": 2176000
    },
    {
      "epoch": 5.94,
      "eval_loss": 2.22352933883667,
      "eval_runtime": 353.4605,
      "eval_samples_per_second": 872.137,
      "eval_steps_per_second": 54.51,
      "step": 2176000
    },
    {
      "epoch": 5.97,
      "eval_loss": 2.22891902923584,
      "eval_runtime": 352.6495,
      "eval_samples_per_second": 874.143,
      "eval_steps_per_second": 54.635,
      "step": 2184000
    },
    {
      "epoch": 5.99,
      "learning_rate": 3.5533333333333334e-08,
      "loss": 2.283,
      "step": 2192000
    },
    {
      "epoch": 5.99,
      "eval_loss": 2.2238047122955322,
      "eval_runtime": 353.765,
      "eval_samples_per_second": 871.386,
      "eval_steps_per_second": 54.463,
      "step": 2192000
    },
    {
      "epoch": 6.01,
      "eval_loss": 2.224536657333374,
      "eval_runtime": 353.3721,
      "eval_samples_per_second": 872.355,
      "eval_steps_per_second": 54.523,
      "step": 2200000
    },
    {
      "epoch": 6.03,
      "learning_rate": 3.28e-08,
      "loss": 2.2761,
      "step": 2208000
    },
    {
      "epoch": 6.03,
      "eval_loss": 2.2296528816223145,
      "eval_runtime": 353.2693,
      "eval_samples_per_second": 872.609,
      "eval_steps_per_second": 54.539,
      "step": 2208000
    },
    {
      "epoch": 6.05,
      "eval_loss": 2.230041265487671,
      "eval_runtime": 355.1155,
      "eval_samples_per_second": 868.073,
      "eval_steps_per_second": 54.256,
      "step": 2216000
    },
    {
      "epoch": 6.08,
      "learning_rate": 3.0066666666666665e-08,
      "loss": 2.2823,
      "step": 2224000
    },
    {
      "epoch": 6.08,
      "eval_loss": 2.22680401802063,
      "eval_runtime": 364.0284,
      "eval_samples_per_second": 846.818,
      "eval_steps_per_second": 52.927,
      "step": 2224000
    },
    {
      "epoch": 6.1,
      "eval_loss": 2.2252140045166016,
      "eval_runtime": 353.3986,
      "eval_samples_per_second": 872.29,
      "eval_steps_per_second": 54.519,
      "step": 2232000
    },
    {
      "epoch": 6.12,
      "learning_rate": 2.7333333333333333e-08,
      "loss": 2.2715,
      "step": 2240000
    },
    {
      "epoch": 6.12,
      "eval_loss": 2.2239723205566406,
      "eval_runtime": 353.6515,
      "eval_samples_per_second": 871.666,
      "eval_steps_per_second": 54.48,
      "step": 2240000
    },
    {
      "epoch": 6.14,
      "eval_loss": 2.2233335971832275,
      "eval_runtime": 353.2896,
      "eval_samples_per_second": 872.559,
      "eval_steps_per_second": 54.536,
      "step": 2248000
    },
    {
      "epoch": 6.16,
      "learning_rate": 2.46e-08,
      "loss": 2.2809,
      "step": 2256000
    },
    {
      "epoch": 6.16,
      "eval_loss": 2.223785161972046,
      "eval_runtime": 354.1994,
      "eval_samples_per_second": 870.318,
      "eval_steps_per_second": 54.396,
      "step": 2256000
    },
    {
      "epoch": 6.18,
      "eval_loss": 2.220431089401245,
      "eval_runtime": 353.1693,
      "eval_samples_per_second": 872.856,
      "eval_steps_per_second": 54.555,
      "step": 2264000
    },
    {
      "epoch": 6.21,
      "learning_rate": 2.1866666666666667e-08,
      "loss": 2.2823,
      "step": 2272000
    },
    {
      "epoch": 6.21,
      "eval_loss": 2.2218103408813477,
      "eval_runtime": 354.0959,
      "eval_samples_per_second": 870.572,
      "eval_steps_per_second": 54.412,
      "step": 2272000
    },
    {
      "epoch": 6.23,
      "eval_loss": 2.2294769287109375,
      "eval_runtime": 353.4329,
      "eval_samples_per_second": 872.205,
      "eval_steps_per_second": 54.514,
      "step": 2280000
    },
    {
      "epoch": 6.25,
      "learning_rate": 1.9133333333333333e-08,
      "loss": 2.2848,
      "step": 2288000
    },
    {
      "epoch": 6.25,
      "eval_loss": 2.2298202514648438,
      "eval_runtime": 353.5589,
      "eval_samples_per_second": 871.894,
      "eval_steps_per_second": 54.494,
      "step": 2288000
    },
    {
      "epoch": 6.27,
      "eval_loss": 2.2298853397369385,
      "eval_runtime": 354.6654,
      "eval_samples_per_second": 869.174,
      "eval_steps_per_second": 54.324,
      "step": 2296000
    },
    {
      "epoch": 6.29,
      "learning_rate": 1.64e-08,
      "loss": 2.2847,
      "step": 2304000
    },
    {
      "epoch": 6.29,
      "eval_loss": 2.224604606628418,
      "eval_runtime": 354.3384,
      "eval_samples_per_second": 869.976,
      "eval_steps_per_second": 54.375,
      "step": 2304000
    },
    {
      "epoch": 6.32,
      "eval_loss": 2.222991704940796,
      "eval_runtime": 353.8486,
      "eval_samples_per_second": 871.181,
      "eval_steps_per_second": 54.45,
      "step": 2312000
    },
    {
      "epoch": 6.34,
      "learning_rate": 1.3666666666666667e-08,
      "loss": 2.2783,
      "step": 2320000
    },
    {
      "epoch": 6.34,
      "eval_loss": 2.2260053157806396,
      "eval_runtime": 354.5313,
      "eval_samples_per_second": 869.503,
      "eval_steps_per_second": 54.345,
      "step": 2320000
    },
    {
      "epoch": 6.36,
      "eval_loss": 2.217644453048706,
      "eval_runtime": 355.0802,
      "eval_samples_per_second": 868.159,
      "eval_steps_per_second": 54.261,
      "step": 2328000
    },
    {
      "epoch": 6.38,
      "learning_rate": 1.0933333333333334e-08,
      "loss": 2.2791,
      "step": 2336000
    },
    {
      "epoch": 6.38,
      "eval_loss": 2.221074342727661,
      "eval_runtime": 353.9231,
      "eval_samples_per_second": 870.997,
      "eval_steps_per_second": 54.438,
      "step": 2336000
    },
    {
      "epoch": 6.4,
      "eval_loss": 2.2261996269226074,
      "eval_runtime": 355.1404,
      "eval_samples_per_second": 868.012,
      "eval_steps_per_second": 54.252,
      "step": 2344000
    },
    {
      "epoch": 6.43,
      "learning_rate": 8.2e-09,
      "loss": 2.2797,
      "step": 2352000
    },
    {
      "epoch": 6.43,
      "eval_loss": 2.2293312549591064,
      "eval_runtime": 354.4371,
      "eval_samples_per_second": 869.734,
      "eval_steps_per_second": 54.359,
      "step": 2352000
    },
    {
      "epoch": 6.45,
      "eval_loss": 2.221876859664917,
      "eval_runtime": 356.764,
      "eval_samples_per_second": 864.061,
      "eval_steps_per_second": 54.005,
      "step": 2360000
    },
    {
      "epoch": 6.47,
      "learning_rate": 5.466666666666667e-09,
      "loss": 2.2784,
      "step": 2368000
    },
    {
      "epoch": 6.47,
      "eval_loss": 2.2249085903167725,
      "eval_runtime": 354.8521,
      "eval_samples_per_second": 868.717,
      "eval_steps_per_second": 54.296,
      "step": 2368000
    },
    {
      "epoch": 6.49,
      "eval_loss": 2.2216453552246094,
      "eval_runtime": 355.228,
      "eval_samples_per_second": 867.798,
      "eval_steps_per_second": 54.238,
      "step": 2376000
    },
    {
      "epoch": 6.51,
      "learning_rate": 2.7333333333333334e-09,
      "loss": 2.271,
      "step": 2384000
    },
    {
      "epoch": 6.51,
      "eval_loss": 2.225550413131714,
      "eval_runtime": 355.1559,
      "eval_samples_per_second": 867.974,
      "eval_steps_per_second": 54.249,
      "step": 2384000
    },
    {
      "epoch": 6.53,
      "eval_loss": 2.2295968532562256,
      "eval_runtime": 355.7601,
      "eval_samples_per_second": 866.5,
      "eval_steps_per_second": 54.157,
      "step": 2392000
    },
    {
      "epoch": 6.56,
      "learning_rate": 0.0,
      "loss": 2.2787,
      "step": 2400000
    },
    {
      "epoch": 6.56,
      "eval_loss": 2.2274532318115234,
      "eval_runtime": 356.8124,
      "eval_samples_per_second": 863.944,
      "eval_steps_per_second": 53.998,
      "step": 2400000
    },
    {
      "epoch": 6.56,
      "step": 2400000,
      "total_flos": 7.335203818962209e+17,
      "train_loss": 2.304743234049479,
      "train_runtime": 256347.6755,
      "train_samples_per_second": 149.797,
      "train_steps_per_second": 9.362
    }
  ],
  "logging_steps": 16000,
  "max_steps": 2400000,
  "num_train_epochs": 7,
  "save_steps": 32000,
  "total_flos": 7.335203818962209e+17,
  "trial_name": null,
  "trial_params": null
}