{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.171730940548895,
  "global_step": 64000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.9977639200449365e-05,
      "loss": 3.6579,
      "step": 500
    },
    {
      "epoch": 0.0,
      "learning_rate": 4.995527840089873e-05,
      "loss": 3.4615,
      "step": 1000
    },
    {
      "epoch": 0.0,
      "learning_rate": 4.993291760134809e-05,
      "loss": 3.3921,
      "step": 1500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.991055680179745e-05,
      "loss": 3.3552,
      "step": 2000
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.9888196002246816e-05,
      "loss": 3.3133,
      "step": 2500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.986583520269618e-05,
      "loss": 3.2847,
      "step": 3000
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.984347440314554e-05,
      "loss": 3.261,
      "step": 3500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.9821113603594904e-05,
      "loss": 3.2291,
      "step": 4000
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.979875280404427e-05,
      "loss": 3.2023,
      "step": 4500
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.977639200449363e-05,
      "loss": 3.1846,
      "step": 5000
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.975403120494299e-05,
      "loss": 3.1743,
      "step": 5500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9731670405392355e-05,
      "loss": 3.1475,
      "step": 6000
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.970930960584171e-05,
      "loss": 3.1315,
      "step": 6500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.968694880629108e-05,
      "loss": 3.1389,
      "step": 7000
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9664588006740444e-05,
      "loss": 3.0986,
      "step": 7500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.9642227207189806e-05,
      "loss": 3.0938,
      "step": 8000
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.961986640763916e-05,
      "loss": 3.0844,
      "step": 8500
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.959750560808853e-05,
      "loss": 3.0931,
      "step": 9000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9575144808537895e-05,
      "loss": 3.0665,
      "step": 9500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.955278400898726e-05,
      "loss": 3.0465,
      "step": 10000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.953042320943661e-05,
      "loss": 3.0343,
      "step": 10500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.950806240988598e-05,
      "loss": 3.036,
      "step": 11000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9485701610335346e-05,
      "loss": 3.0273,
      "step": 11500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.94633408107847e-05,
      "loss": 3.0231,
      "step": 12000
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9440980011234064e-05,
      "loss": 3.0177,
      "step": 12500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9418619211683434e-05,
      "loss": 2.9969,
      "step": 13000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9396258412132796e-05,
      "loss": 3.0019,
      "step": 13500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.937389761258215e-05,
      "loss": 2.9867,
      "step": 14000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9351536813031515e-05,
      "loss": 2.9825,
      "step": 14500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9329176013480885e-05,
      "loss": 2.9799,
      "step": 15000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.930681521393025e-05,
      "loss": 2.9578,
      "step": 15500
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9284454414379603e-05,
      "loss": 2.949,
      "step": 16000
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.9262093614828966e-05,
      "loss": 2.9598,
      "step": 16500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9239732815278336e-05,
      "loss": 2.9568,
      "step": 17000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.921737201572769e-05,
      "loss": 2.9395,
      "step": 17500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9195011216177054e-05,
      "loss": 2.9499,
      "step": 18000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.917265041662642e-05,
      "loss": 2.9316,
      "step": 18500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.915028961707579e-05,
      "loss": 2.9356,
      "step": 19000
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.912792881752514e-05,
      "loss": 2.9105,
      "step": 19500
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9105568017974505e-05,
      "loss": 2.9277,
      "step": 20000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.908320721842387e-05,
      "loss": 2.9224,
      "step": 20500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.906084641887324e-05,
      "loss": 2.9136,
      "step": 21000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.9038485619322594e-05,
      "loss": 2.9109,
      "step": 21500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.9016124819771956e-05,
      "loss": 2.896,
      "step": 22000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.899376402022132e-05,
      "loss": 2.8996,
      "step": 22500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.897140322067068e-05,
      "loss": 2.9059,
      "step": 23000
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.8949042421120045e-05,
      "loss": 2.8849,
      "step": 23500
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.892668162156941e-05,
      "loss": 2.8876,
      "step": 24000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.890432082201877e-05,
      "loss": 2.887,
      "step": 24500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.888196002246813e-05,
      "loss": 2.8687,
      "step": 25000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.8859599222917496e-05,
      "loss": 2.8734,
      "step": 25500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.883723842336686e-05,
      "loss": 2.8745,
      "step": 26000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.881487762381622e-05,
      "loss": 2.8904,
      "step": 26500
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.8792516824265584e-05,
      "loss": 2.8652,
      "step": 27000
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.8770156024714947e-05,
      "loss": 2.8715,
      "step": 27500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.874779522516431e-05,
      "loss": 2.8643,
      "step": 28000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.872543442561367e-05,
      "loss": 2.8483,
      "step": 28500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.8703073626063035e-05,
      "loss": 2.8537,
      "step": 29000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.86807128265124e-05,
      "loss": 2.8456,
      "step": 29500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.865835202696176e-05,
      "loss": 2.8494,
      "step": 30000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.863599122741112e-05,
      "loss": 2.8352,
      "step": 30500
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.8613630427860486e-05,
      "loss": 2.8388,
      "step": 31000
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.859126962830985e-05,
      "loss": 2.8381,
      "step": 31500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.856890882875921e-05,
      "loss": 2.82,
      "step": 32000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.8546548029208574e-05,
      "loss": 2.8428,
      "step": 32500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.852418722965794e-05,
      "loss": 2.8436,
      "step": 33000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.85018264301073e-05,
      "loss": 2.826,
      "step": 33500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.8479465630556655e-05,
      "loss": 2.8274,
      "step": 34000
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.8457104831006025e-05,
      "loss": 2.8338,
      "step": 34500
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.843474403145539e-05,
      "loss": 2.8228,
      "step": 35000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.841238323190475e-05,
      "loss": 2.8078,
      "step": 35500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.8390022432354106e-05,
      "loss": 2.8198,
      "step": 36000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.8367661632803476e-05,
      "loss": 2.8088,
      "step": 36500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.834530083325284e-05,
      "loss": 2.8287,
      "step": 37000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.83229400337022e-05,
      "loss": 2.7993,
      "step": 37500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.830057923415156e-05,
      "loss": 2.8223,
      "step": 38000
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.827821843460093e-05,
      "loss": 2.8084,
      "step": 38500
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.825585763505029e-05,
      "loss": 2.8119,
      "step": 39000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.8233496835499646e-05,
      "loss": 2.8131,
      "step": 39500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.821113603594901e-05,
      "loss": 2.7907,
      "step": 40000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.818877523639838e-05,
      "loss": 2.8071,
      "step": 40500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.816641443684774e-05,
      "loss": 2.7969,
      "step": 41000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.81440536372971e-05,
      "loss": 2.8045,
      "step": 41500
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.812169283774646e-05,
      "loss": 2.7991,
      "step": 42000
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.809933203819583e-05,
      "loss": 2.7845,
      "step": 42500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.807697123864519e-05,
      "loss": 2.8144,
      "step": 43000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.805461043909455e-05,
      "loss": 2.764,
      "step": 43500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.803224963954391e-05,
      "loss": 2.7744,
      "step": 44000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.800988883999328e-05,
      "loss": 2.7857,
      "step": 44500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.7987528040442636e-05,
      "loss": 2.8017,
      "step": 45000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.7965167240892e-05,
      "loss": 2.7958,
      "step": 45500
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.794280644134136e-05,
      "loss": 2.7766,
      "step": 46000
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.792044564179073e-05,
      "loss": 2.7614,
      "step": 46500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.789808484224009e-05,
      "loss": 2.7739,
      "step": 47000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.787572404268945e-05,
      "loss": 2.7736,
      "step": 47500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.785336324313882e-05,
      "loss": 2.7683,
      "step": 48000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.783100244358818e-05,
      "loss": 2.7765,
      "step": 48500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.780864164403754e-05,
      "loss": 2.752,
      "step": 49000
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.77862808444869e-05,
      "loss": 2.7666,
      "step": 49500
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.776392004493627e-05,
      "loss": 2.7481,
      "step": 50000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.7741559245385626e-05,
      "loss": 2.7724,
      "step": 50500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.771919844583499e-05,
      "loss": 2.7773,
      "step": 51000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.769683764628435e-05,
      "loss": 2.7659,
      "step": 51500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.767447684673372e-05,
      "loss": 2.7473,
      "step": 52000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.765211604718308e-05,
      "loss": 2.761,
      "step": 52500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.762975524763244e-05,
      "loss": 2.7289,
      "step": 53000
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.76073944480818e-05,
      "loss": 2.7657,
      "step": 53500
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.758503364853117e-05,
      "loss": 2.7719,
      "step": 54000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.756267284898053e-05,
      "loss": 2.7581,
      "step": 54500
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.754031204942989e-05,
      "loss": 2.7548,
      "step": 55000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7517951249879253e-05,
      "loss": 2.7509,
      "step": 55500
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7495590450328616e-05,
      "loss": 2.738,
      "step": 56000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.747322965077798e-05,
      "loss": 2.752,
      "step": 56500
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.745086885122734e-05,
      "loss": 2.7373,
      "step": 57000
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7428508051676704e-05,
      "loss": 2.7541,
      "step": 57500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.740614725212607e-05,
      "loss": 2.7332,
      "step": 58000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.738378645257543e-05,
      "loss": 2.713,
      "step": 58500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.736142565302479e-05,
      "loss": 2.7534,
      "step": 59000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.7339064853474155e-05,
      "loss": 2.753,
      "step": 59500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.731670405392352e-05,
      "loss": 2.7357,
      "step": 60000
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.729434325437288e-05,
      "loss": 2.7286,
      "step": 60500
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.7271982454822244e-05,
      "loss": 2.7137,
      "step": 61000
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.7249621655271606e-05,
      "loss": 2.73,
      "step": 61500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.722726085572097e-05,
      "loss": 2.7375,
      "step": 62000
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.720490005617033e-05,
      "loss": 2.7257,
      "step": 62500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.7182539256619695e-05,
      "loss": 2.7162,
      "step": 63000
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.716017845706906e-05,
      "loss": 2.7421,
      "step": 63500
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.713781765751842e-05,
      "loss": 2.7255,
      "step": 64000
    }
  ],
  "max_steps": 1118028,
  "num_train_epochs": 3,
  "total_flos": 1.4482346926915584e+17,
  "trial_name": null,
  "trial_params": null
}