{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 40.0,
  "eval_steps": 500,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 2.5e-05,
      "loss": 1.1193,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.1183021068572998,
      "eval_runtime": 0.097,
      "eval_samples_per_second": 103.081,
      "eval_steps_per_second": 20.616,
      "step": 1
    },
    {
      "epoch": 2.0,
      "grad_norm": 19.4419002532959,
      "learning_rate": 5e-05,
      "loss": 1.1193,
      "step": 2
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.4163738191127777,
      "eval_runtime": 0.0967,
      "eval_samples_per_second": 103.452,
      "eval_steps_per_second": 20.69,
      "step": 2
    },
    {
      "epoch": 3.0,
      "grad_norm": 19.4419002532959,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.4152,
      "step": 3
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.4163738191127777,
      "eval_runtime": 0.0975,
      "eval_samples_per_second": 102.538,
      "eval_steps_per_second": 20.508,
      "step": 3
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.695544242858887,
      "learning_rate": 0.0001,
      "loss": 0.4152,
      "step": 4
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.8865498304367065,
      "eval_runtime": 0.0966,
      "eval_samples_per_second": 103.475,
      "eval_steps_per_second": 20.695,
      "step": 4
    },
    {
      "epoch": 5.0,
      "grad_norm": 12.695544242858887,
      "learning_rate": 0.000125,
      "loss": 0.8864,
      "step": 5
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.8865498304367065,
      "eval_runtime": 0.0977,
      "eval_samples_per_second": 102.389,
      "eval_steps_per_second": 20.478,
      "step": 5
    },
    {
      "epoch": 6.0,
      "grad_norm": 39.317420959472656,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.8864,
      "step": 6
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.6666438579559326,
      "eval_runtime": 0.0998,
      "eval_samples_per_second": 100.229,
      "eval_steps_per_second": 20.046,
      "step": 6
    },
    {
      "epoch": 7.0,
      "grad_norm": 39.317420959472656,
      "learning_rate": 0.000175,
      "loss": 1.6618,
      "step": 7
    },
    {
      "epoch": 7.0,
      "eval_loss": 1.6666438579559326,
      "eval_runtime": 0.0975,
      "eval_samples_per_second": 102.551,
      "eval_steps_per_second": 20.51,
      "step": 7
    },
    {
      "epoch": 8.0,
      "grad_norm": 44.225120544433594,
      "learning_rate": 0.0002,
      "loss": 1.6618,
      "step": 8
    },
    {
      "epoch": 8.0,
      "eval_loss": 3.048259973526001,
      "eval_runtime": 0.0983,
      "eval_samples_per_second": 101.777,
      "eval_steps_per_second": 20.355,
      "step": 8
    },
    {
      "epoch": 9.0,
      "grad_norm": 44.225120544433594,
      "learning_rate": 0.0001995184726672197,
      "loss": 3.0505,
      "step": 9
    },
    {
      "epoch": 9.0,
      "eval_loss": 3.048259973526001,
      "eval_runtime": 0.0993,
      "eval_samples_per_second": 100.727,
      "eval_steps_per_second": 20.145,
      "step": 9
    },
    {
      "epoch": 10.0,
      "grad_norm": 143.0501251220703,
      "learning_rate": 0.00019807852804032305,
      "loss": 3.0505,
      "step": 10
    },
    {
      "epoch": 10.0,
      "eval_loss": 2.2029616832733154,
      "eval_runtime": 0.097,
      "eval_samples_per_second": 103.098,
      "eval_steps_per_second": 20.62,
      "step": 10
    },
    {
      "epoch": 11.0,
      "grad_norm": 143.0501251220703,
      "learning_rate": 0.0001956940335732209,
      "loss": 2.2026,
      "step": 11
    },
    {
      "epoch": 11.0,
      "eval_loss": 2.2029616832733154,
      "eval_runtime": 0.0976,
      "eval_samples_per_second": 102.442,
      "eval_steps_per_second": 20.488,
      "step": 11
    },
    {
      "epoch": 12.0,
      "grad_norm": 27.46721076965332,
      "learning_rate": 0.0001923879532511287,
      "loss": 2.2026,
      "step": 12
    },
    {
      "epoch": 12.0,
      "eval_loss": 2.8971307277679443,
      "eval_runtime": 0.0976,
      "eval_samples_per_second": 102.486,
      "eval_steps_per_second": 20.497,
      "step": 12
    },
    {
      "epoch": 13.0,
      "grad_norm": 27.46721076965332,
      "learning_rate": 0.0001881921264348355,
      "loss": 2.9002,
      "step": 13
    },
    {
      "epoch": 13.0,
      "eval_loss": 2.8971307277679443,
      "eval_runtime": 0.0973,
      "eval_samples_per_second": 102.809,
      "eval_steps_per_second": 20.562,
      "step": 13
    },
    {
      "epoch": 14.0,
      "grad_norm": 49.42749786376953,
      "learning_rate": 0.00018314696123025454,
      "loss": 2.9002,
      "step": 14
    },
    {
      "epoch": 14.0,
      "eval_loss": 3.4301562309265137,
      "eval_runtime": 0.0984,
      "eval_samples_per_second": 101.612,
      "eval_steps_per_second": 20.322,
      "step": 14
    },
    {
      "epoch": 15.0,
      "grad_norm": 49.42749786376953,
      "learning_rate": 0.0001773010453362737,
      "loss": 3.4308,
      "step": 15
    },
    {
      "epoch": 15.0,
      "eval_loss": 3.4301562309265137,
      "eval_runtime": 0.0981,
      "eval_samples_per_second": 101.972,
      "eval_steps_per_second": 20.394,
      "step": 15
    },
    {
      "epoch": 16.0,
      "grad_norm": 46.989479064941406,
      "learning_rate": 0.00017071067811865476,
      "loss": 3.4308,
      "step": 16
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.4099053144454956,
      "eval_runtime": 0.0973,
      "eval_samples_per_second": 102.796,
      "eval_steps_per_second": 20.559,
      "step": 16
    },
    {
      "epoch": 17.0,
      "grad_norm": 46.989479064941406,
      "learning_rate": 0.00016343932841636456,
      "loss": 1.4133,
      "step": 17
    },
    {
      "epoch": 17.0,
      "eval_loss": 1.4099053144454956,
      "eval_runtime": 0.097,
      "eval_samples_per_second": 103.061,
      "eval_steps_per_second": 20.612,
      "step": 17
    },
    {
      "epoch": 18.0,
      "grad_norm": 19.974260330200195,
      "learning_rate": 0.00015555702330196023,
      "loss": 1.4133,
      "step": 18
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.7468957304954529,
      "eval_runtime": 0.0986,
      "eval_samples_per_second": 101.448,
      "eval_steps_per_second": 20.29,
      "step": 18
    },
    {
      "epoch": 19.0,
      "grad_norm": 19.974260330200195,
      "learning_rate": 0.0001471396736825998,
      "loss": 0.746,
      "step": 19
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.7468957304954529,
      "eval_runtime": 0.0972,
      "eval_samples_per_second": 102.877,
      "eval_steps_per_second": 20.575,
      "step": 19
    },
    {
      "epoch": 20.0,
      "grad_norm": 9.610711097717285,
      "learning_rate": 0.000138268343236509,
      "loss": 0.746,
      "step": 20
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.16501614451408386,
      "eval_runtime": 0.1002,
      "eval_samples_per_second": 99.82,
      "eval_steps_per_second": 19.964,
      "step": 20
    },
    {
      "epoch": 21.0,
      "grad_norm": 9.610711097717285,
      "learning_rate": 0.00012902846772544624,
      "loss": 0.1654,
      "step": 21
    },
    {
      "epoch": 21.0,
      "eval_loss": 0.16501614451408386,
      "eval_runtime": 0.0976,
      "eval_samples_per_second": 102.477,
      "eval_steps_per_second": 20.495,
      "step": 21
    },
    {
      "epoch": 22.0,
      "grad_norm": 3.4515697956085205,
      "learning_rate": 0.00011950903220161285,
      "loss": 0.1654,
      "step": 22
    },
    {
      "epoch": 22.0,
      "eval_loss": 0.14767812192440033,
      "eval_runtime": 0.097,
      "eval_samples_per_second": 103.11,
      "eval_steps_per_second": 20.622,
      "step": 22
    },
    {
      "epoch": 23.0,
      "grad_norm": 3.4515697956085205,
      "learning_rate": 0.0001098017140329561,
      "loss": 0.1474,
      "step": 23
    },
    {
      "epoch": 23.0,
      "eval_loss": 0.14767812192440033,
      "eval_runtime": 0.1041,
      "eval_samples_per_second": 96.05,
      "eval_steps_per_second": 19.21,
      "step": 23
    },
    {
      "epoch": 24.0,
      "grad_norm": 4.262996196746826,
      "learning_rate": 0.0001,
      "loss": 0.1474,
      "step": 24
    },
    {
      "epoch": 24.0,
      "eval_loss": 0.05527349188923836,
      "eval_runtime": 0.0983,
      "eval_samples_per_second": 101.742,
      "eval_steps_per_second": 20.348,
      "step": 24
    },
    {
      "epoch": 25.0,
      "grad_norm": 4.262996196746826,
      "learning_rate": 9.019828596704394e-05,
      "loss": 0.0569,
      "step": 25
    },
    {
      "epoch": 25.0,
      "eval_loss": 0.05527349188923836,
      "eval_runtime": 0.1002,
      "eval_samples_per_second": 99.795,
      "eval_steps_per_second": 19.959,
      "step": 25
    },
    {
      "epoch": 26.0,
      "grad_norm": 2.504621744155884,
      "learning_rate": 8.049096779838719e-05,
      "loss": 0.0569,
      "step": 26
    },
    {
      "epoch": 26.0,
      "eval_loss": 0.05637549236416817,
      "eval_runtime": 0.1041,
      "eval_samples_per_second": 96.026,
      "eval_steps_per_second": 19.205,
      "step": 26
    },
    {
      "epoch": 27.0,
      "grad_norm": 2.504621744155884,
      "learning_rate": 7.097153227455379e-05,
      "loss": 0.0575,
      "step": 27
    },
    {
      "epoch": 27.0,
      "eval_loss": 0.05637549236416817,
      "eval_runtime": 0.0973,
      "eval_samples_per_second": 102.725,
      "eval_steps_per_second": 20.545,
      "step": 27
    },
    {
      "epoch": 28.0,
      "grad_norm": 3.5743298530578613,
      "learning_rate": 6.173165676349103e-05,
      "loss": 0.0575,
      "step": 28
    },
    {
      "epoch": 28.0,
      "eval_loss": 0.003984498791396618,
      "eval_runtime": 0.0975,
      "eval_samples_per_second": 102.605,
      "eval_steps_per_second": 20.521,
      "step": 28
    },
    {
      "epoch": 29.0,
      "grad_norm": 3.5743298530578613,
      "learning_rate": 5.286032631740023e-05,
      "loss": 0.0041,
      "step": 29
    },
    {
      "epoch": 29.0,
      "eval_loss": 0.003984498791396618,
      "eval_runtime": 0.0979,
      "eval_samples_per_second": 102.094,
      "eval_steps_per_second": 20.419,
      "step": 29
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.1635422706604004,
      "learning_rate": 4.444297669803981e-05,
      "loss": 0.0041,
      "step": 30
    },
    {
      "epoch": 30.0,
      "eval_loss": 0.0021899878047406673,
      "eval_runtime": 0.098,
      "eval_samples_per_second": 102.061,
      "eval_steps_per_second": 20.412,
      "step": 30
    },
    {
      "epoch": 31.0,
      "grad_norm": 0.1635422706604004,
      "learning_rate": 3.6560671583635467e-05,
      "loss": 0.0022,
      "step": 31
    },
    {
      "epoch": 31.0,
      "eval_loss": 0.0021899878047406673,
      "eval_runtime": 0.0971,
      "eval_samples_per_second": 103.036,
      "eval_steps_per_second": 20.607,
      "step": 31
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.0847228392958641,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 0.0022,
      "step": 32
    },
    {
      "epoch": 32.0,
      "eval_loss": 0.0015291140880435705,
      "eval_runtime": 0.1006,
      "eval_samples_per_second": 99.357,
      "eval_steps_per_second": 19.871,
      "step": 32
    },
    {
      "epoch": 33.0,
      "grad_norm": 0.0847228392958641,
      "learning_rate": 2.26989546637263e-05,
      "loss": 0.0015,
      "step": 33
    },
    {
      "epoch": 33.0,
      "eval_loss": 0.0015291140880435705,
      "eval_runtime": 0.1027,
      "eval_samples_per_second": 97.387,
      "eval_steps_per_second": 19.477,
      "step": 33
    },
    {
      "epoch": 34.0,
      "grad_norm": 0.06535077840089798,
      "learning_rate": 1.6853038769745467e-05,
      "loss": 0.0015,
      "step": 34
    },
    {
      "epoch": 34.0,
      "eval_loss": 0.0011607444612309337,
      "eval_runtime": 0.0977,
      "eval_samples_per_second": 102.397,
      "eval_steps_per_second": 20.479,
      "step": 34
    },
    {
      "epoch": 35.0,
      "grad_norm": 0.06535077840089798,
      "learning_rate": 1.1807873565164506e-05,
      "loss": 0.0011,
      "step": 35
    },
    {
      "epoch": 35.0,
      "eval_loss": 0.0011607444612309337,
      "eval_runtime": 0.0971,
      "eval_samples_per_second": 102.985,
      "eval_steps_per_second": 20.597,
      "step": 35
    },
    {
      "epoch": 36.0,
      "grad_norm": 0.04596730321645737,
      "learning_rate": 7.612046748871327e-06,
      "loss": 0.0011,
      "step": 36
    },
    {
      "epoch": 36.0,
      "eval_loss": 0.0010076254839077592,
      "eval_runtime": 0.0972,
      "eval_samples_per_second": 102.913,
      "eval_steps_per_second": 20.583,
      "step": 36
    },
    {
      "epoch": 37.0,
      "grad_norm": 0.04596730321645737,
      "learning_rate": 4.305966426779118e-06,
      "loss": 0.001,
      "step": 37
    },
    {
      "epoch": 37.0,
      "eval_loss": 0.0010076254839077592,
      "eval_runtime": 0.097,
      "eval_samples_per_second": 103.097,
      "eval_steps_per_second": 20.619,
      "step": 37
    },
    {
      "epoch": 38.0,
      "grad_norm": 0.039849672466516495,
      "learning_rate": 1.921471959676957e-06,
      "loss": 0.001,
      "step": 38
    },
    {
      "epoch": 38.0,
      "eval_loss": 0.0009246615809388459,
      "eval_runtime": 0.0975,
      "eval_samples_per_second": 102.584,
      "eval_steps_per_second": 20.517,
      "step": 38
    },
    {
      "epoch": 39.0,
      "grad_norm": 0.039849672466516495,
      "learning_rate": 4.815273327803182e-07,
      "loss": 0.0009,
      "step": 39
    },
    {
      "epoch": 39.0,
      "eval_loss": 0.0009246615809388459,
      "eval_runtime": 0.099,
      "eval_samples_per_second": 101.052,
      "eval_steps_per_second": 20.21,
      "step": 39
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.037203624844551086,
      "learning_rate": 0.0,
      "loss": 0.0009,
      "step": 40
    },
    {
      "epoch": 40.0,
      "eval_loss": 0.0009156091255135834,
      "eval_runtime": 0.0977,
      "eval_samples_per_second": 102.33,
      "eval_steps_per_second": 20.466,
      "step": 40
    },
    {
      "epoch": 40.0,
      "step": 40,
      "total_flos": 3352730995261440.0,
      "train_loss": 0.9132046818675008,
      "train_runtime": 21.9272,
      "train_samples_per_second": 18.242,
      "train_steps_per_second": 1.824
    }
  ],
  "logging_steps": 1,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3352730995261440.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}