{
  "best_metric": 2.4417428970336914,
  "best_model_checkpoint": "./model_tweets_2020_Q1_50/checkpoint-1888000",
  "epoch": 9.834292176820574,
  "eval_steps": 8000,
  "global_step": 2400000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "eval_loss": 2.89373779296875,
      "eval_runtime": 220.7068,
      "eval_samples_per_second": 931.145,
      "eval_steps_per_second": 58.199,
      "step": 8000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.0726666666666665e-07,
      "loss": 3.073,
      "step": 16000
    },
    {
      "epoch": 0.07,
      "eval_loss": 2.76598858833313,
      "eval_runtime": 221.0774,
      "eval_samples_per_second": 929.584,
      "eval_steps_per_second": 58.102,
      "step": 16000
    },
    {
      "epoch": 0.1,
      "eval_loss": 2.7232513427734375,
      "eval_runtime": 221.6288,
      "eval_samples_per_second": 927.271,
      "eval_steps_per_second": 57.957,
      "step": 24000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.0453333333333336e-07,
      "loss": 2.8244,
      "step": 32000
    },
    {
      "epoch": 0.13,
      "eval_loss": 2.687758207321167,
      "eval_runtime": 220.1262,
      "eval_samples_per_second": 933.601,
      "eval_steps_per_second": 58.353,
      "step": 32000
    },
    {
      "epoch": 0.16,
      "eval_loss": 2.6519503593444824,
      "eval_runtime": 220.495,
      "eval_samples_per_second": 932.039,
      "eval_steps_per_second": 58.255,
      "step": 40000
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.018e-07,
      "loss": 2.7542,
      "step": 48000
    },
    {
      "epoch": 0.2,
      "eval_loss": 2.63004469871521,
      "eval_runtime": 220.2803,
      "eval_samples_per_second": 932.948,
      "eval_steps_per_second": 58.312,
      "step": 48000
    },
    {
      "epoch": 0.23,
      "eval_loss": 2.613522529602051,
      "eval_runtime": 221.2317,
      "eval_samples_per_second": 928.936,
      "eval_steps_per_second": 58.061,
      "step": 56000
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.9906666666666667e-07,
      "loss": 2.7083,
      "step": 64000
    },
    {
      "epoch": 0.26,
      "eval_loss": 2.6067709922790527,
      "eval_runtime": 220.3177,
      "eval_samples_per_second": 932.789,
      "eval_steps_per_second": 58.302,
      "step": 64000
    },
    {
      "epoch": 0.3,
      "eval_loss": 2.5854294300079346,
      "eval_runtime": 220.8061,
      "eval_samples_per_second": 930.726,
      "eval_steps_per_second": 58.173,
      "step": 72000
    },
    {
      "epoch": 0.33,
      "learning_rate": 3.963333333333333e-07,
      "loss": 2.6752,
      "step": 80000
    },
    {
      "epoch": 0.33,
      "eval_loss": 2.575528860092163,
      "eval_runtime": 221.8521,
      "eval_samples_per_second": 926.338,
      "eval_steps_per_second": 57.899,
      "step": 80000
    },
    {
      "epoch": 0.36,
      "eval_loss": 2.5720720291137695,
      "eval_runtime": 221.4472,
      "eval_samples_per_second": 928.032,
      "eval_steps_per_second": 58.005,
      "step": 88000
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.936e-07,
      "loss": 2.6657,
      "step": 96000
    },
    {
      "epoch": 0.39,
      "eval_loss": 2.5709290504455566,
      "eval_runtime": 220.9006,
      "eval_samples_per_second": 930.328,
      "eval_steps_per_second": 58.148,
      "step": 96000
    },
    {
      "epoch": 0.43,
      "eval_loss": 2.5656096935272217,
      "eval_runtime": 220.5433,
      "eval_samples_per_second": 931.835,
      "eval_steps_per_second": 58.243,
      "step": 104000
    },
    {
      "epoch": 0.46,
      "learning_rate": 3.908666666666667e-07,
      "loss": 2.6534,
      "step": 112000
    },
    {
      "epoch": 0.46,
      "eval_loss": 2.5558407306671143,
      "eval_runtime": 221.6371,
      "eval_samples_per_second": 927.236,
      "eval_steps_per_second": 57.955,
      "step": 112000
    },
    {
      "epoch": 0.49,
      "eval_loss": 2.5495829582214355,
      "eval_runtime": 220.7733,
      "eval_samples_per_second": 930.864,
      "eval_steps_per_second": 58.182,
      "step": 120000
    },
    {
      "epoch": 0.52,
      "learning_rate": 3.8813333333333334e-07,
      "loss": 2.646,
      "step": 128000
    },
    {
      "epoch": 0.52,
      "eval_loss": 2.547106981277466,
      "eval_runtime": 221.6448,
      "eval_samples_per_second": 927.204,
      "eval_steps_per_second": 57.953,
      "step": 128000
    },
    {
      "epoch": 0.56,
      "eval_loss": 2.5408244132995605,
      "eval_runtime": 221.6302,
      "eval_samples_per_second": 927.265,
      "eval_steps_per_second": 57.957,
      "step": 136000
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.854e-07,
      "loss": 2.625,
      "step": 144000
    },
    {
      "epoch": 0.59,
      "eval_loss": 2.531517744064331,
      "eval_runtime": 223.6683,
      "eval_samples_per_second": 918.816,
      "eval_steps_per_second": 57.429,
      "step": 144000
    },
    {
      "epoch": 0.62,
      "eval_loss": 2.5364675521850586,
      "eval_runtime": 224.1465,
      "eval_samples_per_second": 916.856,
      "eval_steps_per_second": 57.306,
      "step": 152000
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.8266666666666665e-07,
      "loss": 2.6222,
      "step": 160000
    },
    {
      "epoch": 0.66,
      "eval_loss": 2.5372273921966553,
      "eval_runtime": 221.7325,
      "eval_samples_per_second": 926.837,
      "eval_steps_per_second": 57.93,
      "step": 160000
    },
    {
      "epoch": 0.69,
      "eval_loss": 2.534186363220215,
      "eval_runtime": 222.3525,
      "eval_samples_per_second": 924.253,
      "eval_steps_per_second": 57.769,
      "step": 168000
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.799333333333333e-07,
      "loss": 2.6256,
      "step": 176000
    },
    {
      "epoch": 0.72,
      "eval_loss": 2.5308265686035156,
      "eval_runtime": 221.538,
      "eval_samples_per_second": 927.651,
      "eval_steps_per_second": 57.981,
      "step": 176000
    },
    {
      "epoch": 0.75,
      "eval_loss": 2.5311617851257324,
      "eval_runtime": 224.2919,
      "eval_samples_per_second": 916.261,
      "eval_steps_per_second": 57.269,
      "step": 184000
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.772e-07,
      "loss": 2.6074,
      "step": 192000
    },
    {
      "epoch": 0.79,
      "eval_loss": 2.522848129272461,
      "eval_runtime": 224.4507,
      "eval_samples_per_second": 915.613,
      "eval_steps_per_second": 57.229,
      "step": 192000
    },
    {
      "epoch": 0.82,
      "eval_loss": 2.529161214828491,
      "eval_runtime": 222.7477,
      "eval_samples_per_second": 922.613,
      "eval_steps_per_second": 57.666,
      "step": 200000
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.7446666666666667e-07,
      "loss": 2.6071,
      "step": 208000
    },
    {
      "epoch": 0.85,
      "eval_loss": 2.5295047760009766,
      "eval_runtime": 223.0891,
      "eval_samples_per_second": 921.201,
      "eval_steps_per_second": 57.578,
      "step": 208000
    },
    {
      "epoch": 0.89,
      "eval_loss": 2.523491621017456,
      "eval_runtime": 221.4007,
      "eval_samples_per_second": 928.227,
      "eval_steps_per_second": 58.017,
      "step": 216000
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.7173333333333333e-07,
      "loss": 2.5955,
      "step": 224000
    },
    {
      "epoch": 0.92,
      "eval_loss": 2.5219199657440186,
      "eval_runtime": 221.3605,
      "eval_samples_per_second": 928.395,
      "eval_steps_per_second": 58.028,
      "step": 224000
    },
    {
      "epoch": 0.95,
      "eval_loss": 2.5190882682800293,
      "eval_runtime": 221.1449,
      "eval_samples_per_second": 929.3,
      "eval_steps_per_second": 58.084,
      "step": 232000
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.69e-07,
      "loss": 2.6036,
      "step": 240000
    },
    {
      "epoch": 0.98,
      "eval_loss": 2.517120361328125,
      "eval_runtime": 220.9198,
      "eval_samples_per_second": 930.247,
      "eval_steps_per_second": 58.143,
      "step": 240000
    },
    {
      "epoch": 1.02,
      "eval_loss": 2.5102434158325195,
      "eval_runtime": 221.7647,
      "eval_samples_per_second": 926.703,
      "eval_steps_per_second": 57.922,
      "step": 248000
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.6626666666666664e-07,
      "loss": 2.6046,
      "step": 256000
    },
    {
      "epoch": 1.05,
      "eval_loss": 2.5070137977600098,
      "eval_runtime": 221.6584,
      "eval_samples_per_second": 927.147,
      "eval_steps_per_second": 57.95,
      "step": 256000
    },
    {
      "epoch": 1.08,
      "eval_loss": 2.5109376907348633,
      "eval_runtime": 221.3382,
      "eval_samples_per_second": 928.489,
      "eval_steps_per_second": 58.033,
      "step": 264000
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.6353333333333335e-07,
      "loss": 2.5892,
      "step": 272000
    },
    {
      "epoch": 1.11,
      "eval_loss": 2.5104565620422363,
      "eval_runtime": 222.1683,
      "eval_samples_per_second": 925.019,
      "eval_steps_per_second": 57.817,
      "step": 272000
    },
    {
      "epoch": 1.15,
      "eval_loss": 2.508704423904419,
      "eval_runtime": 222.9629,
      "eval_samples_per_second": 921.723,
      "eval_steps_per_second": 57.61,
      "step": 280000
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.608e-07,
      "loss": 2.5929,
      "step": 288000
    },
    {
      "epoch": 1.18,
      "eval_loss": 2.509392738342285,
      "eval_runtime": 223.3494,
      "eval_samples_per_second": 920.128,
      "eval_steps_per_second": 57.511,
      "step": 288000
    },
    {
      "epoch": 1.21,
      "eval_loss": 2.508585214614868,
      "eval_runtime": 222.7314,
      "eval_samples_per_second": 922.681,
      "eval_steps_per_second": 57.67,
      "step": 296000
    },
    {
      "epoch": 1.25,
      "learning_rate": 3.5806666666666666e-07,
      "loss": 2.5857,
      "step": 304000
    },
    {
      "epoch": 1.25,
      "eval_loss": 2.4991345405578613,
      "eval_runtime": 223.3332,
      "eval_samples_per_second": 920.195,
      "eval_steps_per_second": 57.515,
      "step": 304000
    },
    {
      "epoch": 1.28,
      "eval_loss": 2.508927822113037,
      "eval_runtime": 224.1404,
      "eval_samples_per_second": 916.881,
      "eval_steps_per_second": 57.308,
      "step": 312000
    },
    {
      "epoch": 1.31,
      "learning_rate": 3.553333333333333e-07,
      "loss": 2.5828,
      "step": 320000
    },
    {
      "epoch": 1.31,
      "eval_loss": 2.501734972000122,
      "eval_runtime": 223.1146,
      "eval_samples_per_second": 921.096,
      "eval_steps_per_second": 57.571,
      "step": 320000
    },
    {
      "epoch": 1.34,
      "eval_loss": 2.503918409347534,
      "eval_runtime": 223.3327,
      "eval_samples_per_second": 920.196,
      "eval_steps_per_second": 57.515,
      "step": 328000
    },
    {
      "epoch": 1.38,
      "learning_rate": 3.5259999999999997e-07,
      "loss": 2.5812,
      "step": 336000
    },
    {
      "epoch": 1.38,
      "eval_loss": 2.5064587593078613,
      "eval_runtime": 224.1587,
      "eval_samples_per_second": 916.806,
      "eval_steps_per_second": 57.303,
      "step": 336000
    },
    {
      "epoch": 1.41,
      "eval_loss": 2.508263111114502,
      "eval_runtime": 222.503,
      "eval_samples_per_second": 923.628,
      "eval_steps_per_second": 57.73,
      "step": 344000
    },
    {
      "epoch": 1.44,
      "learning_rate": 3.498666666666667e-07,
      "loss": 2.5775,
      "step": 352000
    },
    {
      "epoch": 1.44,
      "eval_loss": 2.509936571121216,
      "eval_runtime": 223.039,
      "eval_samples_per_second": 921.408,
      "eval_steps_per_second": 57.591,
      "step": 352000
    },
    {
      "epoch": 1.48,
      "eval_loss": 2.5078811645507812,
      "eval_runtime": 221.7646,
      "eval_samples_per_second": 926.703,
      "eval_steps_per_second": 57.922,
      "step": 360000
    },
    {
      "epoch": 1.51,
      "learning_rate": 3.4713333333333333e-07,
      "loss": 2.5711,
      "step": 368000
    },
    {
      "epoch": 1.51,
      "eval_loss": 2.4922046661376953,
      "eval_runtime": 223.0544,
      "eval_samples_per_second": 921.345,
      "eval_steps_per_second": 57.587,
      "step": 368000
    },
    {
      "epoch": 1.54,
      "eval_loss": 2.5012030601501465,
      "eval_runtime": 222.0392,
      "eval_samples_per_second": 925.557,
      "eval_steps_per_second": 57.85,
      "step": 376000
    },
    {
      "epoch": 1.57,
      "learning_rate": 3.444e-07,
      "loss": 2.5797,
      "step": 384000
    },
    {
      "epoch": 1.57,
      "eval_loss": 2.49989914894104,
      "eval_runtime": 223.8829,
      "eval_samples_per_second": 917.935,
      "eval_steps_per_second": 57.374,
      "step": 384000
    },
    {
      "epoch": 1.61,
      "eval_loss": 2.4881107807159424,
      "eval_runtime": 222.4413,
      "eval_samples_per_second": 923.884,
      "eval_steps_per_second": 57.746,
      "step": 392000
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.416666666666667e-07,
      "loss": 2.5718,
      "step": 400000
    },
    {
      "epoch": 1.64,
      "eval_loss": 2.4960451126098633,
      "eval_runtime": 222.8741,
      "eval_samples_per_second": 922.09,
      "eval_steps_per_second": 57.633,
      "step": 400000
    },
    {
      "epoch": 1.67,
      "eval_loss": 2.490837574005127,
      "eval_runtime": 222.3679,
      "eval_samples_per_second": 924.189,
      "eval_steps_per_second": 57.765,
      "step": 408000
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.3893333333333335e-07,
      "loss": 2.5627,
      "step": 416000
    },
    {
      "epoch": 1.7,
      "eval_loss": 2.4970648288726807,
      "eval_runtime": 223.472,
      "eval_samples_per_second": 919.623,
      "eval_steps_per_second": 57.479,
      "step": 416000
    },
    {
      "epoch": 1.74,
      "eval_loss": 2.4916465282440186,
      "eval_runtime": 222.5109,
      "eval_samples_per_second": 923.595,
      "eval_steps_per_second": 57.728,
      "step": 424000
    },
    {
      "epoch": 1.77,
      "learning_rate": 3.3619999999999995e-07,
      "loss": 2.5641,
      "step": 432000
    },
    {
      "epoch": 1.77,
      "eval_loss": 2.4971389770507812,
      "eval_runtime": 222.1533,
      "eval_samples_per_second": 925.082,
      "eval_steps_per_second": 57.82,
      "step": 432000
    },
    {
      "epoch": 1.8,
      "eval_loss": 2.495426654815674,
      "eval_runtime": 223.2728,
      "eval_samples_per_second": 920.444,
      "eval_steps_per_second": 57.531,
      "step": 440000
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.3346666666666666e-07,
      "loss": 2.5633,
      "step": 448000
    },
    {
      "epoch": 1.84,
      "eval_loss": 2.485994815826416,
      "eval_runtime": 222.7264,
      "eval_samples_per_second": 922.702,
      "eval_steps_per_second": 57.672,
      "step": 448000
    },
    {
      "epoch": 1.87,
      "eval_loss": 2.4893651008605957,
      "eval_runtime": 223.4251,
      "eval_samples_per_second": 919.816,
      "eval_steps_per_second": 57.491,
      "step": 456000
    },
    {
      "epoch": 1.9,
      "learning_rate": 3.307333333333333e-07,
      "loss": 2.5676,
      "step": 464000
    },
    {
      "epoch": 1.9,
      "eval_loss": 2.489337205886841,
      "eval_runtime": 222.9423,
      "eval_samples_per_second": 921.808,
      "eval_steps_per_second": 57.616,
      "step": 464000
    },
    {
      "epoch": 1.93,
      "eval_loss": 2.4883553981781006,
      "eval_runtime": 223.2404,
      "eval_samples_per_second": 920.577,
      "eval_steps_per_second": 57.539,
      "step": 472000
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.28e-07,
      "loss": 2.5687,
      "step": 480000
    },
    {
      "epoch": 1.97,
      "eval_loss": 2.4921038150787354,
      "eval_runtime": 223.8809,
      "eval_samples_per_second": 917.943,
      "eval_steps_per_second": 57.374,
      "step": 480000
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.4873294830322266,
      "eval_runtime": 222.8771,
      "eval_samples_per_second": 922.078,
      "eval_steps_per_second": 57.633,
      "step": 488000
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.252666666666667e-07,
      "loss": 2.5633,
      "step": 496000
    },
    {
      "epoch": 2.03,
      "eval_loss": 2.4919497966766357,
      "eval_runtime": 222.6439,
      "eval_samples_per_second": 923.043,
      "eval_steps_per_second": 57.693,
      "step": 496000
    },
    {
      "epoch": 2.07,
      "eval_loss": 2.482137441635132,
      "eval_runtime": 222.747,
      "eval_samples_per_second": 922.616,
      "eval_steps_per_second": 57.666,
      "step": 504000
    },
    {
      "epoch": 2.1,
      "learning_rate": 3.2253333333333334e-07,
      "loss": 2.5547,
      "step": 512000
    },
    {
      "epoch": 2.1,
      "eval_loss": 2.490872621536255,
      "eval_runtime": 222.6765,
      "eval_samples_per_second": 922.908,
      "eval_steps_per_second": 57.685,
      "step": 512000
    },
    {
      "epoch": 2.13,
      "eval_loss": 2.4818356037139893,
      "eval_runtime": 223.7166,
      "eval_samples_per_second": 918.617,
      "eval_steps_per_second": 57.416,
      "step": 520000
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.198e-07,
      "loss": 2.5617,
      "step": 528000
    },
    {
      "epoch": 2.16,
      "eval_loss": 2.4854869842529297,
      "eval_runtime": 223.7715,
      "eval_samples_per_second": 918.392,
      "eval_steps_per_second": 57.402,
      "step": 528000
    },
    {
      "epoch": 2.2,
      "eval_loss": 2.48504638671875,
      "eval_runtime": 223.6654,
      "eval_samples_per_second": 918.828,
      "eval_steps_per_second": 57.43,
      "step": 536000
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.1706666666666665e-07,
      "loss": 2.5569,
      "step": 544000
    },
    {
      "epoch": 2.23,
      "eval_loss": 2.480282783508301,
      "eval_runtime": 222.7744,
      "eval_samples_per_second": 922.503,
      "eval_steps_per_second": 57.659,
      "step": 544000
    },
    {
      "epoch": 2.26,
      "eval_loss": 2.4775896072387695,
      "eval_runtime": 223.0018,
      "eval_samples_per_second": 921.562,
      "eval_steps_per_second": 57.6,
      "step": 552000
    },
    {
      "epoch": 2.29,
      "learning_rate": 3.1433333333333336e-07,
      "loss": 2.5535,
      "step": 560000
    },
    {
      "epoch": 2.29,
      "eval_loss": 2.4824471473693848,
      "eval_runtime": 223.1733,
      "eval_samples_per_second": 920.854,
      "eval_steps_per_second": 57.556,
      "step": 560000
    },
    {
      "epoch": 2.33,
      "eval_loss": 2.4821510314941406,
      "eval_runtime": 224.1586,
      "eval_samples_per_second": 916.806,
      "eval_steps_per_second": 57.303,
      "step": 568000
    },
    {
      "epoch": 2.36,
      "learning_rate": 3.116e-07,
      "loss": 2.5534,
      "step": 576000
    },
    {
      "epoch": 2.36,
      "eval_loss": 2.476337432861328,
      "eval_runtime": 223.4733,
      "eval_samples_per_second": 919.618,
      "eval_steps_per_second": 57.479,
      "step": 576000
    },
    {
      "epoch": 2.39,
      "eval_loss": 2.47969388961792,
      "eval_runtime": 224.2217,
      "eval_samples_per_second": 916.548,
      "eval_steps_per_second": 57.287,
      "step": 584000
    },
    {
      "epoch": 2.43,
      "learning_rate": 3.0886666666666667e-07,
      "loss": 2.5583,
      "step": 592000
    },
    {
      "epoch": 2.43,
      "eval_loss": 2.4872305393218994,
      "eval_runtime": 224.237,
      "eval_samples_per_second": 916.486,
      "eval_steps_per_second": 57.283,
      "step": 592000
    },
    {
      "epoch": 2.46,
      "eval_loss": 2.4812192916870117,
      "eval_runtime": 222.6272,
      "eval_samples_per_second": 923.113,
      "eval_steps_per_second": 57.697,
      "step": 600000
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.061333333333333e-07,
      "loss": 2.5545,
      "step": 608000
    },
    {
      "epoch": 2.49,
      "eval_loss": 2.474827527999878,
      "eval_runtime": 223.4042,
      "eval_samples_per_second": 919.902,
      "eval_steps_per_second": 57.497,
      "step": 608000
    },
    {
      "epoch": 2.52,
      "eval_loss": 2.4735865592956543,
      "eval_runtime": 224.1504,
      "eval_samples_per_second": 916.84,
      "eval_steps_per_second": 57.305,
      "step": 616000
    },
    {
      "epoch": 2.56,
      "learning_rate": 3.034e-07,
      "loss": 2.5561,
      "step": 624000
    },
    {
      "epoch": 2.56,
      "eval_loss": 2.4714128971099854,
      "eval_runtime": 223.2085,
      "eval_samples_per_second": 920.709,
      "eval_steps_per_second": 57.547,
      "step": 624000
    },
    {
      "epoch": 2.59,
      "eval_loss": 2.485759973526001,
      "eval_runtime": 222.8361,
      "eval_samples_per_second": 922.247,
      "eval_steps_per_second": 57.643,
      "step": 632000
    },
    {
      "epoch": 2.62,
      "learning_rate": 3.0066666666666663e-07,
      "loss": 2.5384,
      "step": 640000
    },
    {
      "epoch": 2.62,
      "eval_loss": 2.482938289642334,
      "eval_runtime": 223.4494,
      "eval_samples_per_second": 919.716,
      "eval_steps_per_second": 57.485,
      "step": 640000
    },
    {
      "epoch": 2.66,
      "eval_loss": 2.47662091255188,
      "eval_runtime": 222.9171,
      "eval_samples_per_second": 921.912,
      "eval_steps_per_second": 57.622,
      "step": 648000
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.9793333333333334e-07,
      "loss": 2.541,
      "step": 656000
    },
    {
      "epoch": 2.69,
      "eval_loss": 2.4835963249206543,
      "eval_runtime": 223.4062,
      "eval_samples_per_second": 919.894,
      "eval_steps_per_second": 57.496,
      "step": 656000
    },
    {
      "epoch": 2.72,
      "eval_loss": 2.465118408203125,
      "eval_runtime": 226.1239,
      "eval_samples_per_second": 908.838,
      "eval_steps_per_second": 56.805,
      "step": 664000
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.952e-07,
      "loss": 2.5439,
      "step": 672000
    },
    {
      "epoch": 2.75,
      "eval_loss": 2.4797005653381348,
      "eval_runtime": 224.1173,
      "eval_samples_per_second": 916.975,
      "eval_steps_per_second": 57.314,
      "step": 672000
    },
    {
      "epoch": 2.79,
      "eval_loss": 2.4702000617980957,
      "eval_runtime": 223.8532,
      "eval_samples_per_second": 918.057,
      "eval_steps_per_second": 57.381,
      "step": 680000
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.9246666666666665e-07,
      "loss": 2.5597,
      "step": 688000
    },
    {
      "epoch": 2.82,
      "eval_loss": 2.475144386291504,
      "eval_runtime": 223.4589,
      "eval_samples_per_second": 919.677,
      "eval_steps_per_second": 57.483,
      "step": 688000
    },
    {
      "epoch": 2.85,
      "eval_loss": 2.474367618560791,
      "eval_runtime": 222.9092,
      "eval_samples_per_second": 921.945,
      "eval_steps_per_second": 57.624,
      "step": 696000
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.897333333333333e-07,
      "loss": 2.5491,
      "step": 704000
    },
    {
      "epoch": 2.88,
      "eval_loss": 2.4756221771240234,
      "eval_runtime": 223.5443,
      "eval_samples_per_second": 919.325,
      "eval_steps_per_second": 57.461,
      "step": 704000
    },
    {
      "epoch": 2.92,
      "eval_loss": 2.4731247425079346,
      "eval_runtime": 223.5397,
      "eval_samples_per_second": 919.345,
      "eval_steps_per_second": 57.462,
      "step": 712000
    },
    {
      "epoch": 2.95,
      "learning_rate": 2.8699999999999996e-07,
      "loss": 2.5505,
      "step": 720000
    },
    {
      "epoch": 2.95,
      "eval_loss": 2.475615978240967,
      "eval_runtime": 223.941,
      "eval_samples_per_second": 917.697,
      "eval_steps_per_second": 57.359,
      "step": 720000
    },
    {
      "epoch": 2.98,
      "eval_loss": 2.4703986644744873,
      "eval_runtime": 224.1288,
      "eval_samples_per_second": 916.928,
      "eval_steps_per_second": 57.311,
      "step": 728000
    },
    {
      "epoch": 3.02,
      "learning_rate": 2.8426666666666667e-07,
      "loss": 2.5432,
      "step": 736000
    },
    {
      "epoch": 3.02,
      "eval_loss": 2.4762611389160156,
      "eval_runtime": 223.7009,
      "eval_samples_per_second": 918.682,
      "eval_steps_per_second": 57.42,
      "step": 736000
    },
    {
      "epoch": 3.05,
      "eval_loss": 2.4743261337280273,
      "eval_runtime": 224.407,
      "eval_samples_per_second": 915.791,
      "eval_steps_per_second": 57.24,
      "step": 744000
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.815333333333333e-07,
      "loss": 2.5485,
      "step": 752000
    },
    {
      "epoch": 3.08,
      "eval_loss": 2.4626660346984863,
      "eval_runtime": 224.0612,
      "eval_samples_per_second": 917.205,
      "eval_steps_per_second": 57.328,
      "step": 752000
    },
    {
      "epoch": 3.11,
      "eval_loss": 2.471444606781006,
      "eval_runtime": 223.4318,
      "eval_samples_per_second": 919.788,
      "eval_steps_per_second": 57.49,
      "step": 760000
    },
    {
      "epoch": 3.15,
      "learning_rate": 2.7880000000000003e-07,
      "loss": 2.5482,
      "step": 768000
    },
    {
      "epoch": 3.15,
      "eval_loss": 2.4684672355651855,
      "eval_runtime": 224.1026,
      "eval_samples_per_second": 917.035,
      "eval_steps_per_second": 57.317,
      "step": 768000
    },
    {
      "epoch": 3.18,
      "eval_loss": 2.4672694206237793,
      "eval_runtime": 224.9545,
      "eval_samples_per_second": 913.562,
      "eval_steps_per_second": 57.1,
      "step": 776000
    },
    {
      "epoch": 3.21,
      "learning_rate": 2.7606666666666664e-07,
      "loss": 2.5411,
      "step": 784000
    },
    {
      "epoch": 3.21,
      "eval_loss": 2.4726006984710693,
      "eval_runtime": 224.2901,
      "eval_samples_per_second": 916.269,
      "eval_steps_per_second": 57.27,
      "step": 784000
    },
    {
      "epoch": 3.25,
      "eval_loss": 2.476133108139038,
      "eval_runtime": 224.2831,
      "eval_samples_per_second": 916.297,
      "eval_steps_per_second": 57.271,
      "step": 792000
    },
    {
      "epoch": 3.28,
      "learning_rate": 2.733333333333333e-07,
      "loss": 2.5407,
      "step": 800000
    },
    {
      "epoch": 3.28,
      "eval_loss": 2.4611737728118896,
      "eval_runtime": 223.8119,
      "eval_samples_per_second": 918.226,
      "eval_steps_per_second": 57.392,
      "step": 800000
    },
    {
      "epoch": 3.31,
      "eval_loss": 2.4742894172668457,
      "eval_runtime": 224.8606,
      "eval_samples_per_second": 913.944,
      "eval_steps_per_second": 57.124,
      "step": 808000
    },
    {
      "epoch": 3.34,
      "learning_rate": 2.706e-07,
      "loss": 2.5307,
      "step": 816000
    },
    {
      "epoch": 3.34,
      "eval_loss": 2.469853401184082,
      "eval_runtime": 224.177,
      "eval_samples_per_second": 916.731,
      "eval_steps_per_second": 57.298,
      "step": 816000
    },
    {
      "epoch": 3.38,
      "eval_loss": 2.4721498489379883,
      "eval_runtime": 223.5611,
      "eval_samples_per_second": 919.256,
      "eval_steps_per_second": 57.456,
      "step": 824000
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.6786666666666666e-07,
      "loss": 2.5391,
      "step": 832000
    },
    {
      "epoch": 3.41,
      "eval_loss": 2.461381435394287,
      "eval_runtime": 224.3794,
      "eval_samples_per_second": 915.904,
      "eval_steps_per_second": 57.247,
      "step": 832000
    },
    {
      "epoch": 3.44,
      "eval_loss": 2.4641225337982178,
      "eval_runtime": 224.9887,
      "eval_samples_per_second": 913.423,
      "eval_steps_per_second": 57.092,
      "step": 840000
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.651333333333333e-07,
      "loss": 2.5378,
      "step": 848000
    },
    {
      "epoch": 3.47,
      "eval_loss": 2.4652435779571533,
      "eval_runtime": 225.299,
      "eval_samples_per_second": 912.166,
      "eval_steps_per_second": 57.013,
      "step": 848000
    },
    {
      "epoch": 3.51,
      "eval_loss": 2.4640512466430664,
      "eval_runtime": 224.7841,
      "eval_samples_per_second": 914.255,
      "eval_steps_per_second": 57.144,
      "step": 856000
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.624e-07,
      "loss": 2.5399,
      "step": 864000
    },
    {
      "epoch": 3.54,
      "eval_loss": 2.469067096710205,
      "eval_runtime": 225.0567,
      "eval_samples_per_second": 913.148,
      "eval_steps_per_second": 57.075,
      "step": 864000
    },
    {
      "epoch": 3.57,
      "eval_loss": 2.4611856937408447,
      "eval_runtime": 224.6227,
      "eval_samples_per_second": 914.912,
      "eval_steps_per_second": 57.185,
      "step": 872000
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.596666666666667e-07,
      "loss": 2.5412,
      "step": 880000
    },
    {
      "epoch": 3.61,
      "eval_loss": 2.469621419906616,
      "eval_runtime": 225.3239,
      "eval_samples_per_second": 912.065,
      "eval_steps_per_second": 57.007,
      "step": 880000
    },
    {
      "epoch": 3.64,
      "eval_loss": 2.4638073444366455,
      "eval_runtime": 224.7878,
      "eval_samples_per_second": 914.24,
      "eval_steps_per_second": 57.143,
      "step": 888000
    },
    {
      "epoch": 3.67,
      "learning_rate": 2.5693333333333333e-07,
      "loss": 2.5389,
      "step": 896000
    },
    {
      "epoch": 3.67,
      "eval_loss": 2.4658303260803223,
      "eval_runtime": 224.4987,
      "eval_samples_per_second": 915.417,
      "eval_steps_per_second": 57.216,
      "step": 896000
    },
    {
      "epoch": 3.7,
      "eval_loss": 2.4725189208984375,
      "eval_runtime": 226.711,
      "eval_samples_per_second": 906.484,
      "eval_steps_per_second": 56.658,
      "step": 904000
    },
    {
      "epoch": 3.74,
      "learning_rate": 2.542e-07,
      "loss": 2.5325,
      "step": 912000
    },
    {
      "epoch": 3.74,
      "eval_loss": 2.46415114402771,
      "eval_runtime": 225.3655,
      "eval_samples_per_second": 911.896,
      "eval_steps_per_second": 56.996,
      "step": 912000
    },
    {
      "epoch": 3.77,
      "eval_loss": 2.4599404335021973,
      "eval_runtime": 224.7087,
      "eval_samples_per_second": 914.562,
      "eval_steps_per_second": 57.163,
      "step": 920000
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.5146666666666664e-07,
      "loss": 2.5351,
      "step": 928000
    },
    {
      "epoch": 3.8,
      "eval_loss": 2.4616599082946777,
      "eval_runtime": 226.6966,
      "eval_samples_per_second": 906.542,
      "eval_steps_per_second": 56.662,
      "step": 928000
    },
    {
      "epoch": 3.84,
      "eval_loss": 2.464627265930176,
      "eval_runtime": 224.9933,
      "eval_samples_per_second": 913.405,
      "eval_steps_per_second": 57.091,
      "step": 936000
    },
    {
      "epoch": 3.87,
      "learning_rate": 2.4873333333333335e-07,
      "loss": 2.522,
      "step": 944000
    },
    {
      "epoch": 3.87,
      "eval_loss": 2.4665021896362305,
      "eval_runtime": 225.0978,
      "eval_samples_per_second": 912.981,
      "eval_steps_per_second": 57.064,
      "step": 944000
    },
    {
      "epoch": 3.9,
      "eval_loss": 2.4761972427368164,
      "eval_runtime": 224.2641,
      "eval_samples_per_second": 916.375,
      "eval_steps_per_second": 57.276,
      "step": 952000
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.46e-07,
      "loss": 2.5331,
      "step": 960000
    },
    {
      "epoch": 3.93,
      "eval_loss": 2.4668779373168945,
      "eval_runtime": 225.069,
      "eval_samples_per_second": 913.098,
      "eval_steps_per_second": 57.071,
      "step": 960000
    },
    {
      "epoch": 3.97,
      "eval_loss": 2.4549825191497803,
      "eval_runtime": 224.677,
      "eval_samples_per_second": 914.691,
      "eval_steps_per_second": 57.171,
      "step": 968000
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.4326666666666666e-07,
      "loss": 2.5276,
      "step": 976000
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.466226577758789,
      "eval_runtime": 224.6009,
      "eval_samples_per_second": 915.001,
      "eval_steps_per_second": 57.19,
      "step": 976000
    },
    {
      "epoch": 4.03,
      "eval_loss": 2.464536190032959,
      "eval_runtime": 224.76,
      "eval_samples_per_second": 914.353,
      "eval_steps_per_second": 57.15,
      "step": 984000
    },
    {
      "epoch": 4.06,
      "learning_rate": 2.405333333333333e-07,
      "loss": 2.5206,
      "step": 992000
    },
    {
      "epoch": 4.06,
      "eval_loss": 2.4586591720581055,
      "eval_runtime": 225.6548,
      "eval_samples_per_second": 910.727,
      "eval_steps_per_second": 56.923,
      "step": 992000
    },
    {
      "epoch": 4.1,
      "eval_loss": 2.47253680229187,
      "eval_runtime": 225.9081,
      "eval_samples_per_second": 909.706,
      "eval_steps_per_second": 56.859,
      "step": 1000000
    },
    {
      "epoch": 4.13,
      "learning_rate": 2.3779999999999997e-07,
      "loss": 2.5294,
      "step": 1008000
    },
    {
      "epoch": 4.13,
      "eval_loss": 2.458824634552002,
      "eval_runtime": 224.8489,
      "eval_samples_per_second": 913.991,
      "eval_steps_per_second": 57.127,
      "step": 1008000
    },
    {
      "epoch": 4.16,
      "eval_loss": 2.4591076374053955,
      "eval_runtime": 225.8271,
      "eval_samples_per_second": 910.032,
      "eval_steps_per_second": 56.88,
      "step": 1016000
    },
    {
      "epoch": 4.2,
      "learning_rate": 2.3506666666666668e-07,
      "loss": 2.5312,
      "step": 1024000
    },
    {
      "epoch": 4.2,
      "eval_loss": 2.4680891036987305,
      "eval_runtime": 226.319,
      "eval_samples_per_second": 908.055,
      "eval_steps_per_second": 56.756,
      "step": 1024000
    },
    {
      "epoch": 4.23,
      "eval_loss": 2.4624712467193604,
      "eval_runtime": 226.6472,
      "eval_samples_per_second": 906.74,
      "eval_steps_per_second": 56.674,
      "step": 1032000
    },
    {
      "epoch": 4.26,
      "learning_rate": 2.3233333333333334e-07,
      "loss": 2.525,
      "step": 1040000
    },
    {
      "epoch": 4.26,
      "eval_loss": 2.4659371376037598,
      "eval_runtime": 225.5287,
      "eval_samples_per_second": 911.237,
      "eval_steps_per_second": 56.955,
      "step": 1040000
    },
    {
      "epoch": 4.29,
      "eval_loss": 2.460909843444824,
      "eval_runtime": 225.5374,
      "eval_samples_per_second": 911.201,
      "eval_steps_per_second": 56.953,
      "step": 1048000
    },
    {
      "epoch": 4.33,
      "learning_rate": 2.2960000000000002e-07,
      "loss": 2.5318,
      "step": 1056000
    },
    {
      "epoch": 4.33,
      "eval_loss": 2.4571011066436768,
      "eval_runtime": 225.1138,
      "eval_samples_per_second": 912.916,
      "eval_steps_per_second": 57.06,
      "step": 1056000
    },
    {
      "epoch": 4.36,
      "eval_loss": 2.4581968784332275,
      "eval_runtime": 226.7154,
      "eval_samples_per_second": 906.467,
      "eval_steps_per_second": 56.657,
      "step": 1064000
    },
    {
      "epoch": 4.39,
      "learning_rate": 2.2686666666666667e-07,
      "loss": 2.5332,
      "step": 1072000
    },
    {
      "epoch": 4.39,
      "eval_loss": 2.456618547439575,
      "eval_runtime": 225.422,
      "eval_samples_per_second": 911.668,
      "eval_steps_per_second": 56.982,
      "step": 1072000
    },
    {
      "epoch": 4.43,
      "eval_loss": 2.4587738513946533,
      "eval_runtime": 226.533,
      "eval_samples_per_second": 907.197,
      "eval_steps_per_second": 56.703,
      "step": 1080000
    },
    {
      "epoch": 4.46,
      "learning_rate": 2.2413333333333333e-07,
      "loss": 2.5168,
      "step": 1088000
    },
    {
      "epoch": 4.46,
      "eval_loss": 2.4606146812438965,
      "eval_runtime": 226.9924,
      "eval_samples_per_second": 905.361,
      "eval_steps_per_second": 56.588,
      "step": 1088000
    },
    {
      "epoch": 4.49,
      "eval_loss": 2.4597506523132324,
      "eval_runtime": 228.0241,
      "eval_samples_per_second": 901.264,
      "eval_steps_per_second": 56.332,
      "step": 1096000
    },
    {
      "epoch": 4.52,
      "learning_rate": 2.214e-07,
      "loss": 2.5181,
      "step": 1104000
    },
    {
      "epoch": 4.52,
      "eval_loss": 2.454252004623413,
      "eval_runtime": 228.6537,
      "eval_samples_per_second": 898.783,
      "eval_steps_per_second": 56.177,
      "step": 1104000
    },
    {
      "epoch": 4.56,
      "eval_loss": 2.4619953632354736,
      "eval_runtime": 226.293,
      "eval_samples_per_second": 908.159,
      "eval_steps_per_second": 56.763,
      "step": 1112000
    },
    {
      "epoch": 4.59,
      "learning_rate": 2.1866666666666667e-07,
      "loss": 2.5246,
      "step": 1120000
    },
    {
      "epoch": 4.59,
      "eval_loss": 2.4638657569885254,
      "eval_runtime": 228.163,
      "eval_samples_per_second": 900.716,
      "eval_steps_per_second": 56.297,
      "step": 1120000
    },
    {
      "epoch": 4.62,
      "eval_loss": 2.4556171894073486,
      "eval_runtime": 228.3656,
      "eval_samples_per_second": 899.917,
      "eval_steps_per_second": 56.248,
      "step": 1128000
    },
    {
      "epoch": 4.65,
      "learning_rate": 2.1593333333333332e-07,
      "loss": 2.5318,
      "step": 1136000
    },
    {
      "epoch": 4.65,
      "eval_loss": 2.457075595855713,
      "eval_runtime": 227.349,
      "eval_samples_per_second": 903.941,
      "eval_steps_per_second": 56.499,
      "step": 1136000
    },
    {
      "epoch": 4.69,
      "eval_loss": 2.4636013507843018,
      "eval_runtime": 228.1517,
      "eval_samples_per_second": 900.76,
      "eval_steps_per_second": 56.3,
      "step": 1144000
    },
    {
      "epoch": 4.72,
      "learning_rate": 2.132e-07,
      "loss": 2.512,
      "step": 1152000
    },
    {
      "epoch": 4.72,
      "eval_loss": 2.4567556381225586,
      "eval_runtime": 228.1066,
      "eval_samples_per_second": 900.938,
      "eval_steps_per_second": 56.311,
      "step": 1152000
    },
    {
      "epoch": 4.75,
      "eval_loss": 2.4644010066986084,
      "eval_runtime": 228.916,
      "eval_samples_per_second": 897.753,
      "eval_steps_per_second": 56.112,
      "step": 1160000
    },
    {
      "epoch": 4.79,
      "learning_rate": 2.1046666666666666e-07,
      "loss": 2.5174,
      "step": 1168000
    },
    {
      "epoch": 4.79,
      "eval_loss": 2.4528720378875732,
      "eval_runtime": 228.2634,
      "eval_samples_per_second": 900.32,
      "eval_steps_per_second": 56.273,
      "step": 1168000
    },
    {
      "epoch": 4.82,
      "eval_loss": 2.4613921642303467,
      "eval_runtime": 228.3765,
      "eval_samples_per_second": 899.874,
      "eval_steps_per_second": 56.245,
      "step": 1176000
    },
    {
      "epoch": 4.85,
      "learning_rate": 2.0773333333333334e-07,
      "loss": 2.5196,
      "step": 1184000
    },
    {
      "epoch": 4.85,
      "eval_loss": 2.463758707046509,
      "eval_runtime": 227.0091,
      "eval_samples_per_second": 905.294,
      "eval_steps_per_second": 56.584,
      "step": 1184000
    },
    {
      "epoch": 4.88,
      "eval_loss": 2.453406572341919,
      "eval_runtime": 227.3538,
      "eval_samples_per_second": 903.921,
      "eval_steps_per_second": 56.498,
      "step": 1192000
    },
    {
      "epoch": 4.92,
      "learning_rate": 2.05e-07,
      "loss": 2.5248,
      "step": 1200000
    },
    {
      "epoch": 4.92,
      "eval_loss": 2.4553115367889404,
      "eval_runtime": 227.0142,
      "eval_samples_per_second": 905.274,
      "eval_steps_per_second": 56.582,
      "step": 1200000
    },
    {
      "epoch": 4.95,
      "eval_loss": 2.453683853149414,
      "eval_runtime": 226.8684,
      "eval_samples_per_second": 905.855,
      "eval_steps_per_second": 56.619,
      "step": 1208000
    },
    {
      "epoch": 4.98,
      "learning_rate": 2.0226666666666668e-07,
      "loss": 2.5201,
      "step": 1216000
    },
    {
      "epoch": 4.98,
      "eval_loss": 2.4578709602355957,
      "eval_runtime": 226.9695,
      "eval_samples_per_second": 905.452,
      "eval_steps_per_second": 56.594,
      "step": 1216000
    },
    {
      "epoch": 5.02,
      "eval_loss": 2.4524765014648438,
      "eval_runtime": 226.5657,
      "eval_samples_per_second": 907.066,
      "eval_steps_per_second": 56.694,
      "step": 1224000
    },
    {
      "epoch": 5.05,
      "learning_rate": 1.9953333333333333e-07,
      "loss": 2.5164,
      "step": 1232000
    },
    {
      "epoch": 5.05,
      "eval_loss": 2.4645235538482666,
      "eval_runtime": 227.3689,
      "eval_samples_per_second": 903.861,
      "eval_steps_per_second": 56.494,
      "step": 1232000
    },
    {
      "epoch": 5.08,
      "eval_loss": 2.447993040084839,
      "eval_runtime": 228.8072,
      "eval_samples_per_second": 898.18,
      "eval_steps_per_second": 56.139,
      "step": 1240000
    },
    {
      "epoch": 5.11,
      "learning_rate": 1.968e-07,
      "loss": 2.5186,
      "step": 1248000
    },
    {
      "epoch": 5.11,
      "eval_loss": 2.4605581760406494,
      "eval_runtime": 229.2086,
      "eval_samples_per_second": 896.607,
      "eval_steps_per_second": 56.041,
      "step": 1248000
    },
    {
      "epoch": 5.15,
      "eval_loss": 2.4623043537139893,
      "eval_runtime": 229.8264,
      "eval_samples_per_second": 894.197,
      "eval_steps_per_second": 55.89,
      "step": 1256000
    },
    {
      "epoch": 5.18,
      "learning_rate": 1.9406666666666667e-07,
      "loss": 2.5123,
      "step": 1264000
    },
    {
      "epoch": 5.18,
      "eval_loss": 2.456583261489868,
      "eval_runtime": 230.427,
      "eval_samples_per_second": 891.866,
      "eval_steps_per_second": 55.744,
      "step": 1264000
    },
    {
      "epoch": 5.21,
      "eval_loss": 2.464402437210083,
      "eval_runtime": 229.1803,
      "eval_samples_per_second": 896.717,
      "eval_steps_per_second": 56.048,
      "step": 1272000
    },
    {
      "epoch": 5.24,
      "learning_rate": 1.9133333333333333e-07,
      "loss": 2.5233,
      "step": 1280000
    },
    {
      "epoch": 5.24,
      "eval_loss": 2.457606792449951,
      "eval_runtime": 227.9824,
      "eval_samples_per_second": 901.429,
      "eval_steps_per_second": 56.342,
      "step": 1280000
    },
    {
      "epoch": 5.28,
      "eval_loss": 2.451943874359131,
      "eval_runtime": 229.374,
      "eval_samples_per_second": 895.96,
      "eval_steps_per_second": 56.0,
      "step": 1288000
    },
    {
      "epoch": 5.31,
      "learning_rate": 1.886e-07,
      "loss": 2.513,
      "step": 1296000
    },
    {
      "epoch": 5.31,
      "eval_loss": 2.456979513168335,
      "eval_runtime": 228.051,
      "eval_samples_per_second": 901.158,
      "eval_steps_per_second": 56.325,
      "step": 1296000
    },
    {
      "epoch": 5.34,
      "eval_loss": 2.462719202041626,
      "eval_runtime": 228.489,
      "eval_samples_per_second": 899.43,
      "eval_steps_per_second": 56.217,
      "step": 1304000
    },
    {
      "epoch": 5.38,
      "learning_rate": 1.8586666666666666e-07,
      "loss": 2.5226,
      "step": 1312000
    },
    {
      "epoch": 5.38,
      "eval_loss": 2.449977397918701,
      "eval_runtime": 227.9952,
      "eval_samples_per_second": 901.379,
      "eval_steps_per_second": 56.339,
      "step": 1312000
    },
    {
      "epoch": 5.41,
      "eval_loss": 2.4563188552856445,
      "eval_runtime": 227.2759,
      "eval_samples_per_second": 904.231,
      "eval_steps_per_second": 56.517,
      "step": 1320000
    },
    {
      "epoch": 5.44,
      "learning_rate": 1.8313333333333332e-07,
      "loss": 2.5222,
      "step": 1328000
    },
    {
      "epoch": 5.44,
      "eval_loss": 2.4521265029907227,
      "eval_runtime": 226.8418,
      "eval_samples_per_second": 905.962,
      "eval_steps_per_second": 56.625,
      "step": 1328000
    },
    {
      "epoch": 5.47,
      "eval_loss": 2.4591453075408936,
      "eval_runtime": 226.8374,
      "eval_samples_per_second": 905.98,
      "eval_steps_per_second": 56.626,
      "step": 1336000
    },
    {
      "epoch": 5.51,
      "learning_rate": 1.804e-07,
      "loss": 2.5191,
      "step": 1344000
    },
    {
      "epoch": 5.51,
      "eval_loss": 2.4508602619171143,
      "eval_runtime": 228.6931,
      "eval_samples_per_second": 898.628,
      "eval_steps_per_second": 56.167,
      "step": 1344000
    },
    {
      "epoch": 5.54,
      "eval_loss": 2.455850124359131,
      "eval_runtime": 228.2295,
      "eval_samples_per_second": 900.453,
      "eval_steps_per_second": 56.281,
      "step": 1352000
    },
    {
      "epoch": 5.57,
      "learning_rate": 1.7766666666666666e-07,
      "loss": 2.5243,
      "step": 1360000
    },
    {
      "epoch": 5.57,
      "eval_loss": 2.4501898288726807,
      "eval_runtime": 228.7596,
      "eval_samples_per_second": 898.367,
      "eval_steps_per_second": 56.151,
      "step": 1360000
    },
    {
      "epoch": 5.61,
      "eval_loss": 2.4514639377593994,
      "eval_runtime": 227.2719,
      "eval_samples_per_second": 904.247,
      "eval_steps_per_second": 56.518,
      "step": 1368000
    },
    {
      "epoch": 5.64,
      "learning_rate": 1.7493333333333334e-07,
      "loss": 2.5157,
      "step": 1376000
    },
    {
      "epoch": 5.64,
      "eval_loss": 2.4562854766845703,
      "eval_runtime": 227.9532,
      "eval_samples_per_second": 901.545,
      "eval_steps_per_second": 56.349,
      "step": 1376000
    },
    {
      "epoch": 5.67,
      "eval_loss": 2.452606678009033,
      "eval_runtime": 227.4532,
      "eval_samples_per_second": 903.527,
      "eval_steps_per_second": 56.473,
      "step": 1384000
    },
    {
      "epoch": 5.7,
      "learning_rate": 1.722e-07,
      "loss": 2.5162,
      "step": 1392000
    },
    {
      "epoch": 5.7,
      "eval_loss": 2.458620071411133,
      "eval_runtime": 228.2374,
      "eval_samples_per_second": 900.422,
      "eval_steps_per_second": 56.279,
      "step": 1392000
    },
    {
      "epoch": 5.74,
      "eval_loss": 2.458387613296509,
      "eval_runtime": 228.0105,
      "eval_samples_per_second": 901.318,
      "eval_steps_per_second": 56.335,
      "step": 1400000
    },
    {
      "epoch": 5.77,
      "learning_rate": 1.6946666666666668e-07,
      "loss": 2.5169,
      "step": 1408000
    },
    {
      "epoch": 5.77,
      "eval_loss": 2.454158067703247,
      "eval_runtime": 227.4312,
      "eval_samples_per_second": 903.614,
      "eval_steps_per_second": 56.479,
      "step": 1408000
    },
    {
      "epoch": 5.8,
      "eval_loss": 2.460242986679077,
      "eval_runtime": 228.5958,
      "eval_samples_per_second": 899.01,
      "eval_steps_per_second": 56.191,
      "step": 1416000
    },
    {
      "epoch": 5.84,
      "learning_rate": 1.6673333333333333e-07,
      "loss": 2.5127,
      "step": 1424000
    },
    {
      "epoch": 5.84,
      "eval_loss": 2.458707809448242,
      "eval_runtime": 228.0452,
      "eval_samples_per_second": 901.181,
      "eval_steps_per_second": 56.327,
      "step": 1424000
    },
    {
      "epoch": 5.87,
      "eval_loss": 2.452913284301758,
      "eval_runtime": 227.4908,
      "eval_samples_per_second": 903.377,
      "eval_steps_per_second": 56.464,
      "step": 1432000
    },
    {
      "epoch": 5.9,
      "learning_rate": 1.64e-07,
      "loss": 2.5144,
      "step": 1440000
    },
    {
      "epoch": 5.9,
      "eval_loss": 2.462021827697754,
      "eval_runtime": 229.4885,
      "eval_samples_per_second": 895.513,
      "eval_steps_per_second": 55.972,
      "step": 1440000
    },
    {
      "epoch": 5.93,
      "eval_loss": 2.450927972793579,
      "eval_runtime": 227.9748,
      "eval_samples_per_second": 901.459,
      "eval_steps_per_second": 56.344,
      "step": 1448000
    },
    {
      "epoch": 5.97,
      "learning_rate": 1.6126666666666667e-07,
      "loss": 2.5175,
      "step": 1456000
    },
    {
      "epoch": 5.97,
      "eval_loss": 2.4503204822540283,
      "eval_runtime": 227.5178,
      "eval_samples_per_second": 903.27,
      "eval_steps_per_second": 56.457,
      "step": 1456000
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.4545462131500244,
      "eval_runtime": 227.7963,
      "eval_samples_per_second": 902.165,
      "eval_steps_per_second": 56.388,
      "step": 1464000
    },
    {
      "epoch": 6.03,
      "learning_rate": 1.5853333333333332e-07,
      "loss": 2.5147,
      "step": 1472000
    },
    {
      "epoch": 6.03,
      "eval_loss": 2.4440090656280518,
      "eval_runtime": 227.8162,
      "eval_samples_per_second": 902.087,
      "eval_steps_per_second": 56.383,
      "step": 1472000
    },
    {
      "epoch": 6.06,
      "eval_loss": 2.457670211791992,
      "eval_runtime": 228.5245,
      "eval_samples_per_second": 899.291,
      "eval_steps_per_second": 56.208,
      "step": 1480000
    },
    {
      "epoch": 6.1,
      "learning_rate": 1.558e-07,
      "loss": 2.5128,
      "step": 1488000
    },
    {
      "epoch": 6.1,
      "eval_loss": 2.456602096557617,
      "eval_runtime": 230.1502,
      "eval_samples_per_second": 892.939,
      "eval_steps_per_second": 55.811,
      "step": 1488000
    },
    {
      "epoch": 6.13,
      "eval_loss": 2.449889659881592,
      "eval_runtime": 228.3041,
      "eval_samples_per_second": 900.159,
      "eval_steps_per_second": 56.263,
      "step": 1496000
    },
    {
      "epoch": 6.16,
      "learning_rate": 1.5306666666666666e-07,
      "loss": 2.5168,
      "step": 1504000
    },
    {
      "epoch": 6.16,
      "eval_loss": 2.4480044841766357,
      "eval_runtime": 228.2508,
      "eval_samples_per_second": 900.369,
      "eval_steps_per_second": 56.276,
      "step": 1504000
    },
    {
      "epoch": 6.2,
      "eval_loss": 2.4436299800872803,
      "eval_runtime": 229.3638,
      "eval_samples_per_second": 896.0,
      "eval_steps_per_second": 56.003,
      "step": 1512000
    },
    {
      "epoch": 6.23,
      "learning_rate": 1.5033333333333332e-07,
      "loss": 2.5225,
      "step": 1520000
    },
    {
      "epoch": 6.23,
      "eval_loss": 2.446739912033081,
      "eval_runtime": 228.4899,
      "eval_samples_per_second": 899.427,
      "eval_steps_per_second": 56.217,
      "step": 1520000
    },
    {
      "epoch": 6.26,
      "eval_loss": 2.4519920349121094,
      "eval_runtime": 228.2075,
      "eval_samples_per_second": 900.54,
      "eval_steps_per_second": 56.286,
      "step": 1528000
    },
    {
      "epoch": 6.29,
      "learning_rate": 1.476e-07,
      "loss": 2.5135,
      "step": 1536000
    },
    {
      "epoch": 6.29,
      "eval_loss": 2.4535210132598877,
      "eval_runtime": 228.7342,
      "eval_samples_per_second": 898.466,
      "eval_steps_per_second": 56.157,
      "step": 1536000
    },
    {
      "epoch": 6.33,
      "eval_loss": 2.4462831020355225,
      "eval_runtime": 229.9473,
      "eval_samples_per_second": 893.727,
      "eval_steps_per_second": 55.861,
      "step": 1544000
    },
    {
      "epoch": 6.36,
      "learning_rate": 1.4486666666666665e-07,
      "loss": 2.5161,
      "step": 1552000
    },
    {
      "epoch": 6.36,
      "eval_loss": 2.4556400775909424,
      "eval_runtime": 228.5872,
      "eval_samples_per_second": 899.044,
      "eval_steps_per_second": 56.193,
      "step": 1552000
    },
    {
      "epoch": 6.39,
      "eval_loss": 2.4604580402374268,
      "eval_runtime": 229.1233,
      "eval_samples_per_second": 896.941,
      "eval_steps_per_second": 56.062,
      "step": 1560000
    },
    {
      "epoch": 6.43,
      "learning_rate": 1.4213333333333334e-07,
      "loss": 2.5144,
      "step": 1568000
    },
    {
      "epoch": 6.43,
      "eval_loss": 2.4516451358795166,
      "eval_runtime": 229.9726,
      "eval_samples_per_second": 893.628,
      "eval_steps_per_second": 55.854,
      "step": 1568000
    },
    {
      "epoch": 6.46,
      "eval_loss": 2.4487648010253906,
      "eval_runtime": 229.4253,
      "eval_samples_per_second": 895.76,
      "eval_steps_per_second": 55.988,
      "step": 1576000
    },
    {
      "epoch": 6.49,
      "learning_rate": 1.3940000000000002e-07,
      "loss": 2.5209,
      "step": 1584000
    },
    {
      "epoch": 6.49,
      "eval_loss": 2.4525067806243896,
      "eval_runtime": 228.8527,
      "eval_samples_per_second": 898.001,
      "eval_steps_per_second": 56.128,
      "step": 1584000
    },
    {
      "epoch": 6.52,
      "eval_loss": 2.450185537338257,
      "eval_runtime": 230.8087,
      "eval_samples_per_second": 890.391,
      "eval_steps_per_second": 55.652,
      "step": 1592000
    },
    {
      "epoch": 6.56,
      "learning_rate": 1.3666666666666665e-07,
      "loss": 2.5102,
      "step": 1600000
    },
    {
      "epoch": 6.56,
      "eval_loss": 2.453780174255371,
      "eval_runtime": 229.4733,
      "eval_samples_per_second": 895.573,
      "eval_steps_per_second": 55.976,
      "step": 1600000
    },
    {
      "epoch": 6.59,
      "eval_loss": 2.4490787982940674,
      "eval_runtime": 229.059,
      "eval_samples_per_second": 897.192,
      "eval_steps_per_second": 56.077,
      "step": 1608000
    },
    {
      "epoch": 6.62,
      "learning_rate": 1.3393333333333333e-07,
      "loss": 2.5176,
      "step": 1616000
    },
    {
      "epoch": 6.62,
      "eval_loss": 2.452752113342285,
      "eval_runtime": 228.4962,
      "eval_samples_per_second": 899.402,
      "eval_steps_per_second": 56.215,
      "step": 1616000
    },
    {
      "epoch": 6.65,
      "eval_loss": 2.44599986076355,
      "eval_runtime": 228.5114,
      "eval_samples_per_second": 899.342,
      "eval_steps_per_second": 56.212,
      "step": 1624000
    },
    {
      "epoch": 6.69,
      "learning_rate": 1.312e-07,
      "loss": 2.5208,
      "step": 1632000
    },
    {
      "epoch": 6.69,
      "eval_loss": 2.4484992027282715,
      "eval_runtime": 230.1605,
      "eval_samples_per_second": 892.899,
      "eval_steps_per_second": 55.809,
      "step": 1632000
    },
    {
      "epoch": 6.72,
      "eval_loss": 2.451284646987915,
      "eval_runtime": 229.1401,
      "eval_samples_per_second": 896.875,
      "eval_steps_per_second": 56.057,
      "step": 1640000
    },
    {
      "epoch": 6.75,
      "learning_rate": 1.2846666666666667e-07,
      "loss": 2.5064,
      "step": 1648000
    },
    {
      "epoch": 6.75,
      "eval_loss": 2.451927900314331,
      "eval_runtime": 229.3071,
      "eval_samples_per_second": 896.222,
      "eval_steps_per_second": 56.017,
      "step": 1648000
    },
    {
      "epoch": 6.79,
      "eval_loss": 2.449305295944214,
      "eval_runtime": 231.2204,
      "eval_samples_per_second": 888.806,
      "eval_steps_per_second": 55.553,
      "step": 1656000
    },
    {
      "epoch": 6.82,
      "learning_rate": 1.2573333333333332e-07,
      "loss": 2.5111,
      "step": 1664000
    },
    {
      "epoch": 6.82,
      "eval_loss": 2.4505178928375244,
      "eval_runtime": 230.2462,
      "eval_samples_per_second": 892.566,
      "eval_steps_per_second": 55.788,
      "step": 1664000
    },
    {
      "epoch": 6.85,
      "eval_loss": 2.4501988887786865,
      "eval_runtime": 229.973,
      "eval_samples_per_second": 893.627,
      "eval_steps_per_second": 55.854,
      "step": 1672000
    },
    {
      "epoch": 6.88,
      "learning_rate": 1.23e-07,
      "loss": 2.5141,
      "step": 1680000
    },
    {
      "epoch": 6.88,
      "eval_loss": 2.4560253620147705,
      "eval_runtime": 229.6465,
      "eval_samples_per_second": 894.897,
      "eval_steps_per_second": 55.934,
      "step": 1680000
    },
    {
      "epoch": 6.92,
      "eval_loss": 2.4499940872192383,
      "eval_runtime": 229.0726,
      "eval_samples_per_second": 897.139,
      "eval_steps_per_second": 56.074,
      "step": 1688000
    },
    {
      "epoch": 6.95,
      "learning_rate": 1.2026666666666666e-07,
      "loss": 2.5089,
      "step": 1696000
    },
    {
      "epoch": 6.95,
      "eval_loss": 2.4512550830841064,
      "eval_runtime": 228.4897,
      "eval_samples_per_second": 899.428,
      "eval_steps_per_second": 56.217,
      "step": 1696000
    },
    {
      "epoch": 6.98,
      "eval_loss": 2.4418201446533203,
      "eval_runtime": 229.377,
      "eval_samples_per_second": 895.949,
      "eval_steps_per_second": 56.0,
      "step": 1704000
    },
    {
      "epoch": 7.02,
      "learning_rate": 1.1753333333333334e-07,
      "loss": 2.5174,
      "step": 1712000
    },
    {
      "epoch": 7.02,
      "eval_loss": 2.447690010070801,
      "eval_runtime": 231.0137,
      "eval_samples_per_second": 889.601,
      "eval_steps_per_second": 55.603,
      "step": 1712000
    },
    {
      "epoch": 7.05,
      "eval_loss": 2.450817584991455,
      "eval_runtime": 231.6212,
      "eval_samples_per_second": 887.268,
      "eval_steps_per_second": 55.457,
      "step": 1720000
    },
    {
      "epoch": 7.08,
      "learning_rate": 1.1480000000000001e-07,
      "loss": 2.5198,
      "step": 1728000
    },
    {
      "epoch": 7.08,
      "eval_loss": 2.448648691177368,
      "eval_runtime": 230.9308,
      "eval_samples_per_second": 889.92,
      "eval_steps_per_second": 55.623,
      "step": 1728000
    },
    {
      "epoch": 7.11,
      "eval_loss": 2.4577322006225586,
      "eval_runtime": 230.4865,
      "eval_samples_per_second": 891.636,
      "eval_steps_per_second": 55.73,
      "step": 1736000
    },
    {
      "epoch": 7.15,
      "learning_rate": 1.1206666666666666e-07,
      "loss": 2.4974,
      "step": 1744000
    },
    {
      "epoch": 7.15,
      "eval_loss": 2.4416255950927734,
      "eval_runtime": 229.8237,
      "eval_samples_per_second": 894.207,
      "eval_steps_per_second": 55.891,
      "step": 1744000
    },
    {
      "epoch": 7.18,
      "eval_loss": 2.4549336433410645,
      "eval_runtime": 229.7571,
      "eval_samples_per_second": 894.466,
      "eval_steps_per_second": 55.907,
      "step": 1752000
    },
    {
      "epoch": 7.21,
      "learning_rate": 1.0933333333333333e-07,
      "loss": 2.5016,
      "step": 1760000
    },
    {
      "epoch": 7.21,
      "eval_loss": 2.455679416656494,
      "eval_runtime": 230.1335,
      "eval_samples_per_second": 893.003,
      "eval_steps_per_second": 55.815,
      "step": 1760000
    },
    {
      "epoch": 7.24,
      "eval_loss": 2.4531571865081787,
      "eval_runtime": 231.3847,
      "eval_samples_per_second": 888.175,
      "eval_steps_per_second": 55.514,
      "step": 1768000
    },
    {
      "epoch": 7.28,
      "learning_rate": 1.066e-07,
      "loss": 2.5112,
      "step": 1776000
    },
    {
      "epoch": 7.28,
      "eval_loss": 2.445054531097412,
      "eval_runtime": 231.2999,
      "eval_samples_per_second": 888.5,
      "eval_steps_per_second": 55.534,
      "step": 1776000
    },
    {
      "epoch": 7.31,
      "eval_loss": 2.460723638534546,
      "eval_runtime": 230.196,
      "eval_samples_per_second": 892.761,
      "eval_steps_per_second": 55.8,
      "step": 1784000
    },
    {
      "epoch": 7.34,
      "learning_rate": 1.0386666666666667e-07,
      "loss": 2.5172,
      "step": 1792000
    },
    {
      "epoch": 7.34,
      "eval_loss": 2.4451537132263184,
      "eval_runtime": 231.1406,
      "eval_samples_per_second": 889.112,
      "eval_steps_per_second": 55.572,
      "step": 1792000
    },
    {
      "epoch": 7.38,
      "eval_loss": 2.4426777362823486,
      "eval_runtime": 230.7159,
      "eval_samples_per_second": 890.749,
      "eval_steps_per_second": 55.675,
      "step": 1800000
    },
    {
      "epoch": 7.41,
      "learning_rate": 1.0113333333333334e-07,
      "loss": 2.5089,
      "step": 1808000
    },
    {
      "epoch": 7.41,
      "eval_loss": 2.4511077404022217,
      "eval_runtime": 231.5975,
      "eval_samples_per_second": 887.359,
      "eval_steps_per_second": 55.463,
      "step": 1808000
    },
    {
      "epoch": 7.44,
      "eval_loss": 2.4440526962280273,
      "eval_runtime": 231.4447,
      "eval_samples_per_second": 887.944,
      "eval_steps_per_second": 55.499,
      "step": 1816000
    },
    {
      "epoch": 7.47,
      "learning_rate": 9.84e-08,
      "loss": 2.5136,
      "step": 1824000
    },
    {
      "epoch": 7.47,
      "eval_loss": 2.4492361545562744,
      "eval_runtime": 231.7181,
      "eval_samples_per_second": 886.896,
      "eval_steps_per_second": 55.434,
      "step": 1824000
    },
    {
      "epoch": 7.51,
      "eval_loss": 2.4523823261260986,
      "eval_runtime": 231.3659,
      "eval_samples_per_second": 888.247,
      "eval_steps_per_second": 55.518,
      "step": 1832000
    },
    {
      "epoch": 7.54,
      "learning_rate": 9.566666666666666e-08,
      "loss": 2.509,
      "step": 1840000
    },
    {
      "epoch": 7.54,
      "eval_loss": 2.451181411743164,
      "eval_runtime": 230.8127,
      "eval_samples_per_second": 890.376,
      "eval_steps_per_second": 55.651,
      "step": 1840000
    },
    {
      "epoch": 7.57,
      "eval_loss": 2.4528069496154785,
      "eval_runtime": 230.6096,
      "eval_samples_per_second": 891.16,
      "eval_steps_per_second": 55.7,
      "step": 1848000
    },
    {
      "epoch": 7.61,
      "learning_rate": 9.293333333333333e-08,
      "loss": 2.5157,
      "step": 1856000
    },
    {
      "epoch": 7.61,
      "eval_loss": 2.4439537525177,
      "eval_runtime": 233.382,
      "eval_samples_per_second": 880.573,
      "eval_steps_per_second": 55.039,
      "step": 1856000
    },
    {
      "epoch": 7.64,
      "eval_loss": 2.4401602745056152,
      "eval_runtime": 231.584,
      "eval_samples_per_second": 887.41,
      "eval_steps_per_second": 55.466,
      "step": 1864000
    },
    {
      "epoch": 7.67,
      "learning_rate": 9.02e-08,
      "loss": 2.5181,
      "step": 1872000
    },
    {
      "epoch": 7.67,
      "eval_loss": 2.4537830352783203,
      "eval_runtime": 230.4518,
      "eval_samples_per_second": 891.77,
      "eval_steps_per_second": 55.738,
      "step": 1872000
    },
    {
      "epoch": 7.7,
      "eval_loss": 2.4480724334716797,
      "eval_runtime": 229.9532,
      "eval_samples_per_second": 893.703,
      "eval_steps_per_second": 55.859,
      "step": 1880000
    },
    {
      "epoch": 7.74,
      "learning_rate": 8.746666666666667e-08,
      "loss": 2.5145,
      "step": 1888000
    },
    {
      "epoch": 7.74,
      "eval_loss": 2.4417428970336914,
      "eval_runtime": 231.464,
      "eval_samples_per_second": 887.87,
      "eval_steps_per_second": 55.495,
      "step": 1888000
    },
    {
      "epoch": 7.77,
      "eval_loss": 2.4512147903442383,
      "eval_runtime": 231.0711,
      "eval_samples_per_second": 889.38,
      "eval_steps_per_second": 55.589,
      "step": 1896000
    },
    {
      "epoch": 7.8,
      "learning_rate": 8.473333333333334e-08,
      "loss": 2.5013,
      "step": 1904000
    },
    {
      "epoch": 7.8,
      "eval_loss": 2.45603084564209,
      "eval_runtime": 231.877,
      "eval_samples_per_second": 886.289,
      "eval_steps_per_second": 55.396,
      "step": 1904000
    },
    {
      "epoch": 7.83,
      "eval_loss": 2.4508955478668213,
      "eval_runtime": 230.4147,
      "eval_samples_per_second": 891.913,
      "eval_steps_per_second": 55.747,
      "step": 1912000
    },
    {
      "epoch": 7.87,
      "learning_rate": 8.2e-08,
      "loss": 2.5064,
      "step": 1920000
    },
    {
      "epoch": 7.87,
      "eval_loss": 2.447256565093994,
      "eval_runtime": 231.4505,
      "eval_samples_per_second": 887.922,
      "eval_steps_per_second": 55.498,
      "step": 1920000
    },
    {
      "epoch": 7.9,
      "eval_loss": 2.457575559616089,
      "eval_runtime": 232.2387,
      "eval_samples_per_second": 884.908,
      "eval_steps_per_second": 55.309,
      "step": 1928000
    },
    {
      "epoch": 7.93,
      "learning_rate": 7.926666666666666e-08,
      "loss": 2.5068,
      "step": 1936000
    },
    {
      "epoch": 7.93,
      "eval_loss": 2.4460949897766113,
      "eval_runtime": 230.6448,
      "eval_samples_per_second": 891.024,
      "eval_steps_per_second": 55.692,
      "step": 1936000
    },
    {
      "epoch": 7.97,
      "eval_loss": 2.4451067447662354,
      "eval_runtime": 231.6713,
      "eval_samples_per_second": 887.076,
      "eval_steps_per_second": 55.445,
      "step": 1944000
    },
    {
      "epoch": 8.0,
      "learning_rate": 7.653333333333333e-08,
      "loss": 2.5152,
      "step": 1952000
    },
    {
      "epoch": 8.0,
      "eval_loss": 2.442117214202881,
      "eval_runtime": 231.4315,
      "eval_samples_per_second": 887.995,
      "eval_steps_per_second": 55.502,
      "step": 1952000
    },
    {
      "epoch": 8.03,
      "eval_loss": 2.4458179473876953,
      "eval_runtime": 230.6413,
      "eval_samples_per_second": 891.037,
      "eval_steps_per_second": 55.693,
      "step": 1960000
    },
    {
      "epoch": 8.06,
      "learning_rate": 7.38e-08,
      "loss": 2.5025,
      "step": 1968000
    },
    {
      "epoch": 8.06,
      "eval_loss": 2.4532368183135986,
      "eval_runtime": 230.9812,
      "eval_samples_per_second": 889.726,
      "eval_steps_per_second": 55.611,
      "step": 1968000
    },
    {
      "epoch": 8.1,
      "eval_loss": 2.4541139602661133,
      "eval_runtime": 231.1965,
      "eval_samples_per_second": 888.898,
      "eval_steps_per_second": 55.559,
      "step": 1976000
    },
    {
      "epoch": 8.13,
      "learning_rate": 7.106666666666667e-08,
      "loss": 2.5151,
      "step": 1984000
    },
    {
      "epoch": 8.13,
      "eval_loss": 2.4499058723449707,
      "eval_runtime": 231.2124,
      "eval_samples_per_second": 888.836,
      "eval_steps_per_second": 55.555,
      "step": 1984000
    },
    {
      "epoch": 8.16,
      "eval_loss": 2.4501264095306396,
      "eval_runtime": 231.2241,
      "eval_samples_per_second": 888.791,
      "eval_steps_per_second": 55.552,
      "step": 1992000
    },
    {
      "epoch": 8.2,
      "learning_rate": 6.833333333333332e-08,
      "loss": 2.5138,
      "step": 2000000
    },
    {
      "epoch": 8.2,
      "eval_loss": 2.444784641265869,
      "eval_runtime": 231.6831,
      "eval_samples_per_second": 887.031,
      "eval_steps_per_second": 55.442,
      "step": 2000000
    },
    {
      "epoch": 8.23,
      "eval_loss": 2.4562456607818604,
      "eval_runtime": 231.974,
      "eval_samples_per_second": 885.918,
      "eval_steps_per_second": 55.373,
      "step": 2008000
    },
    {
      "epoch": 8.26,
      "learning_rate": 6.56e-08,
      "loss": 2.5039,
      "step": 2016000
    },
    {
      "epoch": 8.26,
      "eval_loss": 2.4612646102905273,
      "eval_runtime": 234.4229,
      "eval_samples_per_second": 876.663,
      "eval_steps_per_second": 54.794,
      "step": 2016000
    },
    {
      "epoch": 8.29,
      "eval_loss": 2.4471163749694824,
      "eval_runtime": 233.3806,
      "eval_samples_per_second": 880.579,
      "eval_steps_per_second": 55.039,
      "step": 2024000
    },
    {
      "epoch": 8.33,
      "learning_rate": 6.286666666666666e-08,
      "loss": 2.5055,
      "step": 2032000
    },
    {
      "epoch": 8.33,
      "eval_loss": 2.445026159286499,
      "eval_runtime": 233.3418,
      "eval_samples_per_second": 880.725,
      "eval_steps_per_second": 55.048,
      "step": 2032000
    },
    {
      "epoch": 8.36,
      "eval_loss": 2.4492921829223633,
      "eval_runtime": 232.3875,
      "eval_samples_per_second": 884.342,
      "eval_steps_per_second": 55.274,
      "step": 2040000
    },
    {
      "epoch": 8.39,
      "learning_rate": 6.013333333333333e-08,
      "loss": 2.5085,
      "step": 2048000
    },
    {
      "epoch": 8.39,
      "eval_loss": 2.448164224624634,
      "eval_runtime": 233.4578,
      "eval_samples_per_second": 880.288,
      "eval_steps_per_second": 55.021,
      "step": 2048000
    },
    {
      "epoch": 8.42,
      "eval_loss": 2.4571895599365234,
      "eval_runtime": 235.4355,
      "eval_samples_per_second": 872.893,
      "eval_steps_per_second": 54.558,
      "step": 2056000
    },
    {
      "epoch": 8.46,
      "learning_rate": 5.7400000000000004e-08,
      "loss": 2.5114,
      "step": 2064000
    },
    {
      "epoch": 8.46,
      "eval_loss": 2.444307804107666,
      "eval_runtime": 234.4924,
      "eval_samples_per_second": 876.404,
      "eval_steps_per_second": 54.778,
      "step": 2064000
    },
    {
      "epoch": 8.49,
      "eval_loss": 2.445603132247925,
      "eval_runtime": 234.6223,
      "eval_samples_per_second": 875.919,
      "eval_steps_per_second": 54.748,
      "step": 2072000
    },
    {
      "epoch": 8.52,
      "learning_rate": 5.4666666666666666e-08,
      "loss": 2.5132,
      "step": 2080000
    },
    {
      "epoch": 8.52,
      "eval_loss": 2.4528441429138184,
      "eval_runtime": 234.3887,
      "eval_samples_per_second": 876.791,
      "eval_steps_per_second": 54.802,
      "step": 2080000
    },
    {
      "epoch": 8.56,
      "eval_loss": 2.449744939804077,
      "eval_runtime": 233.1003,
      "eval_samples_per_second": 881.638,
      "eval_steps_per_second": 55.105,
      "step": 2088000
    },
    {
      "epoch": 8.59,
      "learning_rate": 5.1933333333333335e-08,
      "loss": 2.5072,
      "step": 2096000
    },
    {
      "epoch": 8.59,
      "eval_loss": 2.4547877311706543,
      "eval_runtime": 232.2237,
      "eval_samples_per_second": 884.966,
      "eval_steps_per_second": 55.313,
      "step": 2096000
    },
    {
      "epoch": 8.62,
      "eval_loss": 2.4547617435455322,
      "eval_runtime": 232.0067,
      "eval_samples_per_second": 885.794,
      "eval_steps_per_second": 55.365,
      "step": 2104000
    },
    {
      "epoch": 8.65,
      "learning_rate": 4.92e-08,
      "loss": 2.504,
      "step": 2112000
    },
    {
      "epoch": 8.65,
      "eval_loss": 2.444261312484741,
      "eval_runtime": 232.4079,
      "eval_samples_per_second": 884.264,
      "eval_steps_per_second": 55.269,
      "step": 2112000
    },
    {
      "epoch": 8.69,
      "eval_loss": 2.445204734802246,
      "eval_runtime": 233.2645,
      "eval_samples_per_second": 881.017,
      "eval_steps_per_second": 55.066,
      "step": 2120000
    },
    {
      "epoch": 8.72,
      "learning_rate": 4.6466666666666666e-08,
      "loss": 2.5128,
      "step": 2128000
    },
    {
      "epoch": 8.72,
      "eval_loss": 2.4509565830230713,
      "eval_runtime": 233.1857,
      "eval_samples_per_second": 881.315,
      "eval_steps_per_second": 55.085,
      "step": 2128000
    },
    {
      "epoch": 8.75,
      "eval_loss": 2.447999954223633,
      "eval_runtime": 233.2452,
      "eval_samples_per_second": 881.09,
      "eval_steps_per_second": 55.071,
      "step": 2136000
    },
    {
      "epoch": 8.79,
      "learning_rate": 4.3733333333333335e-08,
      "loss": 2.5133,
      "step": 2144000
    },
    {
      "epoch": 8.79,
      "eval_loss": 2.4470479488372803,
      "eval_runtime": 234.9529,
      "eval_samples_per_second": 874.686,
      "eval_steps_per_second": 54.671,
      "step": 2144000
    },
    {
      "epoch": 8.82,
      "eval_loss": 2.4436631202697754,
      "eval_runtime": 234.9836,
      "eval_samples_per_second": 874.572,
      "eval_steps_per_second": 54.663,
      "step": 2152000
    },
    {
      "epoch": 8.85,
      "learning_rate": 4.1e-08,
      "loss": 2.5067,
      "step": 2160000
    },
    {
      "epoch": 8.85,
      "eval_loss": 2.444672107696533,
      "eval_runtime": 234.3233,
      "eval_samples_per_second": 877.036,
      "eval_steps_per_second": 54.817,
      "step": 2160000
    },
    {
      "epoch": 8.88,
      "eval_loss": 2.453118085861206,
      "eval_runtime": 233.5384,
      "eval_samples_per_second": 879.984,
      "eval_steps_per_second": 55.002,
      "step": 2168000
    },
    {
      "epoch": 8.92,
      "learning_rate": 3.8266666666666665e-08,
      "loss": 2.4996,
      "step": 2176000
    },
    {
      "epoch": 8.92,
      "eval_loss": 2.447479009628296,
      "eval_runtime": 235.7844,
      "eval_samples_per_second": 871.601,
      "eval_steps_per_second": 54.478,
      "step": 2176000
    },
    {
      "epoch": 8.95,
      "eval_loss": 2.4438347816467285,
      "eval_runtime": 233.6193,
      "eval_samples_per_second": 879.679,
      "eval_steps_per_second": 54.983,
      "step": 2184000
    },
    {
      "epoch": 8.98,
      "learning_rate": 3.5533333333333334e-08,
      "loss": 2.5123,
      "step": 2192000
    },
    {
      "epoch": 8.98,
      "eval_loss": 2.4552195072174072,
      "eval_runtime": 235.201,
      "eval_samples_per_second": 873.763,
      "eval_steps_per_second": 54.613,
      "step": 2192000
    },
    {
      "epoch": 9.01,
      "eval_loss": 2.4441311359405518,
      "eval_runtime": 234.6948,
      "eval_samples_per_second": 875.648,
      "eval_steps_per_second": 54.731,
      "step": 2200000
    },
    {
      "epoch": 9.05,
      "learning_rate": 3.28e-08,
      "loss": 2.5044,
      "step": 2208000
    },
    {
      "epoch": 9.05,
      "eval_loss": 2.4438366889953613,
      "eval_runtime": 233.1145,
      "eval_samples_per_second": 881.584,
      "eval_steps_per_second": 55.102,
      "step": 2208000
    },
    {
      "epoch": 9.08,
      "eval_loss": 2.453371286392212,
      "eval_runtime": 234.9783,
      "eval_samples_per_second": 874.592,
      "eval_steps_per_second": 54.665,
      "step": 2216000
    },
    {
      "epoch": 9.11,
      "learning_rate": 3.0066666666666665e-08,
      "loss": 2.5068,
      "step": 2224000
    },
    {
      "epoch": 9.11,
      "eval_loss": 2.449671745300293,
      "eval_runtime": 232.7881,
      "eval_samples_per_second": 882.82,
      "eval_steps_per_second": 55.179,
      "step": 2224000
    },
    {
      "epoch": 9.15,
      "eval_loss": 2.444044828414917,
      "eval_runtime": 233.2255,
      "eval_samples_per_second": 881.164,
      "eval_steps_per_second": 55.075,
      "step": 2232000
    },
    {
      "epoch": 9.18,
      "learning_rate": 2.7333333333333333e-08,
      "loss": 2.5165,
      "step": 2240000
    },
    {
      "epoch": 9.18,
      "eval_loss": 2.457695722579956,
      "eval_runtime": 234.2621,
      "eval_samples_per_second": 877.265,
      "eval_steps_per_second": 54.832,
      "step": 2240000
    },
    {
      "epoch": 9.21,
      "eval_loss": 2.4506990909576416,
      "eval_runtime": 232.4001,
      "eval_samples_per_second": 884.294,
      "eval_steps_per_second": 55.271,
      "step": 2248000
    },
    {
      "epoch": 9.24,
      "learning_rate": 2.46e-08,
      "loss": 2.5087,
      "step": 2256000
    },
    {
      "epoch": 9.24,
      "eval_loss": 2.4494166374206543,
      "eval_runtime": 233.0606,
      "eval_samples_per_second": 881.788,
      "eval_steps_per_second": 55.114,
      "step": 2256000
    },
    {
      "epoch": 9.28,
      "eval_loss": 2.4393150806427,
      "eval_runtime": 234.1923,
      "eval_samples_per_second": 877.527,
      "eval_steps_per_second": 54.848,
      "step": 2264000
    },
    {
      "epoch": 9.31,
      "learning_rate": 2.1866666666666667e-08,
      "loss": 2.5036,
      "step": 2272000
    },
    {
      "epoch": 9.31,
      "eval_loss": 2.4486756324768066,
      "eval_runtime": 233.5876,
      "eval_samples_per_second": 879.798,
      "eval_steps_per_second": 54.99,
      "step": 2272000
    },
    {
      "epoch": 9.34,
      "eval_loss": 2.442298173904419,
      "eval_runtime": 233.1053,
      "eval_samples_per_second": 881.619,
      "eval_steps_per_second": 55.104,
      "step": 2280000
    },
    {
      "epoch": 9.38,
      "learning_rate": 1.9133333333333333e-08,
      "loss": 2.5086,
      "step": 2288000
    },
    {
      "epoch": 9.38,
      "eval_loss": 2.4455623626708984,
      "eval_runtime": 232.7856,
      "eval_samples_per_second": 882.83,
      "eval_steps_per_second": 55.18,
      "step": 2288000
    },
    {
      "epoch": 9.41,
      "eval_loss": 2.449575185775757,
      "eval_runtime": 234.5471,
      "eval_samples_per_second": 876.199,
      "eval_steps_per_second": 54.765,
      "step": 2296000
    },
    {
      "epoch": 9.44,
      "learning_rate": 1.64e-08,
      "loss": 2.5034,
      "step": 2304000
    },
    {
      "epoch": 9.44,
      "eval_loss": 2.4498891830444336,
      "eval_runtime": 232.7935,
      "eval_samples_per_second": 882.8,
      "eval_steps_per_second": 55.178,
      "step": 2304000
    },
    {
      "epoch": 9.47,
      "eval_loss": 2.4432790279388428,
      "eval_runtime": 233.6332,
      "eval_samples_per_second": 879.627,
      "eval_steps_per_second": 54.979,
      "step": 2312000
    },
    {
      "epoch": 9.51,
      "learning_rate": 1.3666666666666667e-08,
      "loss": 2.5099,
      "step": 2320000
    },
    {
      "epoch": 9.51,
      "eval_loss": 2.4534084796905518,
      "eval_runtime": 233.6601,
      "eval_samples_per_second": 879.525,
      "eval_steps_per_second": 54.973,
      "step": 2320000
    },
    {
      "epoch": 9.54,
      "eval_loss": 2.4494857788085938,
      "eval_runtime": 233.4959,
      "eval_samples_per_second": 880.144,
      "eval_steps_per_second": 55.012,
      "step": 2328000
    },
    {
      "epoch": 9.57,
      "learning_rate": 1.0933333333333334e-08,
      "loss": 2.5065,
      "step": 2336000
    },
    {
      "epoch": 9.57,
      "eval_loss": 2.4510202407836914,
      "eval_runtime": 233.7041,
      "eval_samples_per_second": 879.36,
      "eval_steps_per_second": 54.963,
      "step": 2336000
    },
    {
      "epoch": 9.6,
      "eval_loss": 2.4512877464294434,
      "eval_runtime": 237.5888,
      "eval_samples_per_second": 864.982,
      "eval_steps_per_second": 54.064,
      "step": 2344000
    },
    {
      "epoch": 9.64,
      "learning_rate": 8.2e-09,
      "loss": 2.502,
      "step": 2352000
    },
    {
      "epoch": 9.64,
      "eval_loss": 2.451225996017456,
      "eval_runtime": 233.4207,
      "eval_samples_per_second": 880.427,
      "eval_steps_per_second": 55.029,
      "step": 2352000
    },
    {
      "epoch": 9.67,
      "eval_loss": 2.4469268321990967,
      "eval_runtime": 234.7938,
      "eval_samples_per_second": 875.279,
      "eval_steps_per_second": 54.708,
      "step": 2360000
    },
    {
      "epoch": 9.7,
      "learning_rate": 5.466666666666667e-09,
      "loss": 2.5043,
      "step": 2368000
    },
    {
      "epoch": 9.7,
      "eval_loss": 2.4544479846954346,
      "eval_runtime": 234.2869,
      "eval_samples_per_second": 877.173,
      "eval_steps_per_second": 54.826,
      "step": 2368000
    },
    {
      "epoch": 9.74,
      "eval_loss": 2.4492740631103516,
      "eval_runtime": 234.1805,
      "eval_samples_per_second": 877.571,
      "eval_steps_per_second": 54.851,
      "step": 2376000
    },
    {
      "epoch": 9.77,
      "learning_rate": 2.7333333333333334e-09,
      "loss": 2.5068,
      "step": 2384000
    },
    {
      "epoch": 9.77,
      "eval_loss": 2.453711748123169,
      "eval_runtime": 233.0608,
      "eval_samples_per_second": 881.787,
      "eval_steps_per_second": 55.114,
      "step": 2384000
    },
    {
      "epoch": 9.8,
      "eval_loss": 2.4386837482452393,
      "eval_runtime": 234.1662,
      "eval_samples_per_second": 877.625,
      "eval_steps_per_second": 54.854,
      "step": 2392000
    },
    {
      "epoch": 9.83,
      "learning_rate": 0.0,
      "loss": 2.5118,
      "step": 2400000
    },
    {
      "epoch": 9.83,
      "eval_loss": 2.4494030475616455,
      "eval_runtime": 234.0187,
      "eval_samples_per_second": 878.178,
      "eval_steps_per_second": 54.889,
      "step": 2400000
    },
    {
      "epoch": 9.83,
      "step": 2400000,
      "total_flos": 7.305293129309786e+17,
      "train_loss": 2.5438934391276042,
      "train_runtime": 220778.1092,
      "train_samples_per_second": 173.93,
      "train_steps_per_second": 10.871
    }
  ],
  "logging_steps": 16000,
  "max_steps": 2400000,
  "num_train_epochs": 10,
  "save_steps": 32000,
  "total_flos": 7.305293129309786e+17,
  "trial_name": null,
  "trial_params": null
}