{
  "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": 1.1411,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.1278259754180908,
      "eval_runtime": 0.1561,
      "eval_samples_per_second": 64.049,
      "eval_steps_per_second": 12.81,
      "step": 1
    },
    {
      "epoch": 2.0,
      "grad_norm": 9.757247924804688,
      "learning_rate": 5e-05,
      "loss": 1.1201,
      "step": 2
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.7308214902877808,
      "eval_runtime": 0.157,
      "eval_samples_per_second": 63.676,
      "eval_steps_per_second": 12.735,
      "step": 2
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.757247924804688,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.7406,
      "step": 3
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7308214902877808,
      "eval_runtime": 0.1559,
      "eval_samples_per_second": 64.133,
      "eval_steps_per_second": 12.827,
      "step": 3
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.656477451324463,
      "learning_rate": 0.0001,
      "loss": 0.7483,
      "step": 4
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.1192941665649414,
      "eval_runtime": 0.1548,
      "eval_samples_per_second": 64.593,
      "eval_steps_per_second": 12.919,
      "step": 4
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.656477451324463,
      "learning_rate": 0.000125,
      "loss": 1.1127,
      "step": 5
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.1192941665649414,
      "eval_runtime": 0.1561,
      "eval_samples_per_second": 64.065,
      "eval_steps_per_second": 12.813,
      "step": 5
    },
    {
      "epoch": 6.0,
      "grad_norm": 18.55446434020996,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.0754,
      "step": 6
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.9689571261405945,
      "eval_runtime": 0.1555,
      "eval_samples_per_second": 64.312,
      "eval_steps_per_second": 12.862,
      "step": 6
    },
    {
      "epoch": 7.0,
      "grad_norm": 18.55446434020996,
      "learning_rate": 0.000175,
      "loss": 0.9708,
      "step": 7
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.9689571261405945,
      "eval_runtime": 0.1553,
      "eval_samples_per_second": 64.383,
      "eval_steps_per_second": 12.877,
      "step": 7
    },
    {
      "epoch": 8.0,
      "grad_norm": 13.187333106994629,
      "learning_rate": 0.0002,
      "loss": 1.0331,
      "step": 8
    },
    {
      "epoch": 8.0,
      "eval_loss": 5.132155418395996,
      "eval_runtime": 0.156,
      "eval_samples_per_second": 64.107,
      "eval_steps_per_second": 12.821,
      "step": 8
    },
    {
      "epoch": 9.0,
      "grad_norm": 13.187333106994629,
      "learning_rate": 0.0001995184726672197,
      "loss": 5.0832,
      "step": 9
    },
    {
      "epoch": 9.0,
      "eval_loss": 5.132155418395996,
      "eval_runtime": 0.1552,
      "eval_samples_per_second": 64.447,
      "eval_steps_per_second": 12.889,
      "step": 9
    },
    {
      "epoch": 10.0,
      "grad_norm": 140.08033752441406,
      "learning_rate": 0.00019807852804032305,
      "loss": 5.1665,
      "step": 10
    },
    {
      "epoch": 10.0,
      "eval_loss": 3.2012367248535156,
      "eval_runtime": 0.1572,
      "eval_samples_per_second": 63.616,
      "eval_steps_per_second": 12.723,
      "step": 10
    },
    {
      "epoch": 11.0,
      "grad_norm": 140.08033752441406,
      "learning_rate": 0.0001956940335732209,
      "loss": 3.2953,
      "step": 11
    },
    {
      "epoch": 11.0,
      "eval_loss": 3.2012367248535156,
      "eval_runtime": 0.1557,
      "eval_samples_per_second": 64.221,
      "eval_steps_per_second": 12.844,
      "step": 11
    },
    {
      "epoch": 12.0,
      "grad_norm": 122.30906677246094,
      "learning_rate": 0.0001923879532511287,
      "loss": 3.3383,
      "step": 12
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.2309796810150146,
      "eval_runtime": 0.1559,
      "eval_samples_per_second": 64.128,
      "eval_steps_per_second": 12.826,
      "step": 12
    },
    {
      "epoch": 13.0,
      "grad_norm": 122.30906677246094,
      "learning_rate": 0.0001881921264348355,
      "loss": 1.2275,
      "step": 13
    },
    {
      "epoch": 13.0,
      "eval_loss": 1.2309796810150146,
      "eval_runtime": 0.1566,
      "eval_samples_per_second": 63.87,
      "eval_steps_per_second": 12.774,
      "step": 13
    },
    {
      "epoch": 14.0,
      "grad_norm": 9.766098022460938,
      "learning_rate": 0.00018314696123025454,
      "loss": 1.2398,
      "step": 14
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.7279261946678162,
      "eval_runtime": 0.1553,
      "eval_samples_per_second": 64.403,
      "eval_steps_per_second": 12.881,
      "step": 14
    },
    {
      "epoch": 15.0,
      "grad_norm": 9.766098022460938,
      "learning_rate": 0.0001773010453362737,
      "loss": 0.7426,
      "step": 15
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.7279261946678162,
      "eval_runtime": 0.1568,
      "eval_samples_per_second": 63.763,
      "eval_steps_per_second": 12.753,
      "step": 15
    },
    {
      "epoch": 16.0,
      "grad_norm": 7.932348728179932,
      "learning_rate": 0.00017071067811865476,
      "loss": 0.7375,
      "step": 16
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.4009631276130676,
      "eval_runtime": 0.1564,
      "eval_samples_per_second": 63.956,
      "eval_steps_per_second": 12.791,
      "step": 16
    },
    {
      "epoch": 17.0,
      "grad_norm": 7.932348728179932,
      "learning_rate": 0.00016343932841636456,
      "loss": 0.4055,
      "step": 17
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.4009631276130676,
      "eval_runtime": 0.1554,
      "eval_samples_per_second": 64.342,
      "eval_steps_per_second": 12.868,
      "step": 17
    },
    {
      "epoch": 18.0,
      "grad_norm": 6.0723419189453125,
      "learning_rate": 0.00015555702330196023,
      "loss": 0.3971,
      "step": 18
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.15298452973365784,
      "eval_runtime": 0.1557,
      "eval_samples_per_second": 64.239,
      "eval_steps_per_second": 12.848,
      "step": 18
    },
    {
      "epoch": 19.0,
      "grad_norm": 6.0723419189453125,
      "learning_rate": 0.0001471396736825998,
      "loss": 0.1623,
      "step": 19
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.15298452973365784,
      "eval_runtime": 0.1553,
      "eval_samples_per_second": 64.4,
      "eval_steps_per_second": 12.88,
      "step": 19
    },
    {
      "epoch": 20.0,
      "grad_norm": 2.892711639404297,
      "learning_rate": 0.000138268343236509,
      "loss": 0.1554,
      "step": 20
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.09138597548007965,
      "eval_runtime": 0.1562,
      "eval_samples_per_second": 64.038,
      "eval_steps_per_second": 12.808,
      "step": 20
    },
    {
      "epoch": 21.0,
      "grad_norm": 2.892711639404297,
      "learning_rate": 0.00012902846772544624,
      "loss": 0.0876,
      "step": 21
    },
    {
      "epoch": 21.0,
      "eval_loss": 0.09138597548007965,
      "eval_runtime": 0.1555,
      "eval_samples_per_second": 64.296,
      "eval_steps_per_second": 12.859,
      "step": 21
    },
    {
      "epoch": 22.0,
      "grad_norm": 1.7072333097457886,
      "learning_rate": 0.00011950903220161285,
      "loss": 0.0933,
      "step": 22
    },
    {
      "epoch": 22.0,
      "eval_loss": 0.057169489562511444,
      "eval_runtime": 0.1554,
      "eval_samples_per_second": 64.345,
      "eval_steps_per_second": 12.869,
      "step": 22
    },
    {
      "epoch": 23.0,
      "grad_norm": 1.7072333097457886,
      "learning_rate": 0.0001098017140329561,
      "loss": 0.0621,
      "step": 23
    },
    {
      "epoch": 23.0,
      "eval_loss": 0.057169489562511444,
      "eval_runtime": 0.1566,
      "eval_samples_per_second": 63.85,
      "eval_steps_per_second": 12.77,
      "step": 23
    },
    {
      "epoch": 24.0,
      "grad_norm": 1.489992380142212,
      "learning_rate": 0.0001,
      "loss": 0.0561,
      "step": 24
    },
    {
      "epoch": 24.0,
      "eval_loss": 0.029277494177222252,
      "eval_runtime": 0.1557,
      "eval_samples_per_second": 64.208,
      "eval_steps_per_second": 12.842,
      "step": 24
    },
    {
      "epoch": 25.0,
      "grad_norm": 1.489992380142212,
      "learning_rate": 9.019828596704394e-05,
      "loss": 0.0273,
      "step": 25
    },
    {
      "epoch": 25.0,
      "eval_loss": 0.029277494177222252,
      "eval_runtime": 0.155,
      "eval_samples_per_second": 64.505,
      "eval_steps_per_second": 12.901,
      "step": 25
    },
    {
      "epoch": 26.0,
      "grad_norm": 0.7948157787322998,
      "learning_rate": 8.049096779838719e-05,
      "loss": 0.0276,
      "step": 26
    },
    {
      "epoch": 26.0,
      "eval_loss": 0.01596766524016857,
      "eval_runtime": 0.1574,
      "eval_samples_per_second": 63.526,
      "eval_steps_per_second": 12.705,
      "step": 26
    },
    {
      "epoch": 27.0,
      "grad_norm": 0.7948157787322998,
      "learning_rate": 7.097153227455379e-05,
      "loss": 0.0141,
      "step": 27
    },
    {
      "epoch": 27.0,
      "eval_loss": 0.01596766524016857,
      "eval_runtime": 0.1559,
      "eval_samples_per_second": 64.154,
      "eval_steps_per_second": 12.831,
      "step": 27
    },
    {
      "epoch": 28.0,
      "grad_norm": 0.47985532879829407,
      "learning_rate": 6.173165676349103e-05,
      "loss": 0.0167,
      "step": 28
    },
    {
      "epoch": 28.0,
      "eval_loss": 0.0055131325498223305,
      "eval_runtime": 0.1551,
      "eval_samples_per_second": 64.494,
      "eval_steps_per_second": 12.899,
      "step": 28
    },
    {
      "epoch": 29.0,
      "grad_norm": 0.47985532879829407,
      "learning_rate": 5.286032631740023e-05,
      "loss": 0.005,
      "step": 29
    },
    {
      "epoch": 29.0,
      "eval_loss": 0.0055131325498223305,
      "eval_runtime": 0.1555,
      "eval_samples_per_second": 64.319,
      "eval_steps_per_second": 12.864,
      "step": 29
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.2008497714996338,
      "learning_rate": 4.444297669803981e-05,
      "loss": 0.0056,
      "step": 30
    },
    {
      "epoch": 30.0,
      "eval_loss": 0.0027345926500856876,
      "eval_runtime": 0.1559,
      "eval_samples_per_second": 64.125,
      "eval_steps_per_second": 12.825,
      "step": 30
    },
    {
      "epoch": 31.0,
      "grad_norm": 0.2008497714996338,
      "learning_rate": 3.6560671583635467e-05,
      "loss": 0.0024,
      "step": 31
    },
    {
      "epoch": 31.0,
      "eval_loss": 0.0027345926500856876,
      "eval_runtime": 0.1559,
      "eval_samples_per_second": 64.15,
      "eval_steps_per_second": 12.83,
      "step": 31
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.06916889548301697,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 0.0026,
      "step": 32
    },
    {
      "epoch": 32.0,
      "eval_loss": 0.002255227416753769,
      "eval_runtime": 0.1552,
      "eval_samples_per_second": 64.439,
      "eval_steps_per_second": 12.888,
      "step": 32
    },
    {
      "epoch": 33.0,
      "grad_norm": 0.06916889548301697,
      "learning_rate": 2.26989546637263e-05,
      "loss": 0.0022,
      "step": 33
    },
    {
      "epoch": 33.0,
      "eval_loss": 0.002255227416753769,
      "eval_runtime": 0.1556,
      "eval_samples_per_second": 64.276,
      "eval_steps_per_second": 12.855,
      "step": 33
    },
    {
      "epoch": 34.0,
      "grad_norm": 0.0823979377746582,
      "learning_rate": 1.6853038769745467e-05,
      "loss": 0.0021,
      "step": 34
    },
    {
      "epoch": 34.0,
      "eval_loss": 0.0017385013634338975,
      "eval_runtime": 0.1552,
      "eval_samples_per_second": 64.422,
      "eval_steps_per_second": 12.884,
      "step": 34
    },
    {
      "epoch": 35.0,
      "grad_norm": 0.0823979377746582,
      "learning_rate": 1.1807873565164506e-05,
      "loss": 0.0015,
      "step": 35
    },
    {
      "epoch": 35.0,
      "eval_loss": 0.0017385013634338975,
      "eval_runtime": 0.1555,
      "eval_samples_per_second": 64.296,
      "eval_steps_per_second": 12.859,
      "step": 35
    },
    {
      "epoch": 36.0,
      "grad_norm": 0.05911393463611603,
      "learning_rate": 7.612046748871327e-06,
      "loss": 0.0016,
      "step": 36
    },
    {
      "epoch": 36.0,
      "eval_loss": 0.0014259777963161469,
      "eval_runtime": 0.1562,
      "eval_samples_per_second": 64.033,
      "eval_steps_per_second": 12.807,
      "step": 36
    },
    {
      "epoch": 37.0,
      "grad_norm": 0.05911393463611603,
      "learning_rate": 4.305966426779118e-06,
      "loss": 0.0013,
      "step": 37
    },
    {
      "epoch": 37.0,
      "eval_loss": 0.0014259777963161469,
      "eval_runtime": 0.1553,
      "eval_samples_per_second": 64.39,
      "eval_steps_per_second": 12.878,
      "step": 37
    },
    {
      "epoch": 38.0,
      "grad_norm": 0.043380431830883026,
      "learning_rate": 1.921471959676957e-06,
      "loss": 0.0012,
      "step": 38
    },
    {
      "epoch": 38.0,
      "eval_loss": 0.001306904829107225,
      "eval_runtime": 0.1552,
      "eval_samples_per_second": 64.441,
      "eval_steps_per_second": 12.888,
      "step": 38
    },
    {
      "epoch": 39.0,
      "grad_norm": 0.043380431830883026,
      "learning_rate": 4.815273327803182e-07,
      "loss": 0.0011,
      "step": 39
    },
    {
      "epoch": 39.0,
      "eval_loss": 0.001306904829107225,
      "eval_runtime": 0.155,
      "eval_samples_per_second": 64.517,
      "eval_steps_per_second": 12.903,
      "step": 39
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.03869909420609474,
      "learning_rate": 0.0,
      "loss": 0.0011,
      "step": 40
    },
    {
      "epoch": 40.0,
      "eval_loss": 0.0013022046769037843,
      "eval_runtime": 0.1569,
      "eval_samples_per_second": 63.717,
      "eval_steps_per_second": 12.743,
      "step": 40
    },
    {
      "epoch": 40.0,
      "step": 40,
      "total_flos": 4849620932886528.0,
      "train_loss": 0.7576449837215478,
      "train_runtime": 31.485,
      "train_samples_per_second": 12.704,
      "train_steps_per_second": 1.27
    }
  ],
  "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": 4849620932886528.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}