{
  "best_metric": 1.035351276397705,
  "best_model_checkpoint": "outputs/checkpoint-463",
  "epoch": 12.998650472334683,
  "eval_steps": 500,
  "global_step": 602,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4318488529014845,
      "grad_norm": 0.8678001165390015,
      "learning_rate": 1e-05,
      "loss": 2.3338,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.6083477139472961,
      "learning_rate": 2e-05,
      "loss": 2.2493,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 1.9349104166030884,
      "eval_runtime": 11.681,
      "eval_samples_per_second": 31.847,
      "eval_steps_per_second": 4.024,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.508277416229248,
      "learning_rate": 3e-05,
      "loss": 2.1059,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.5420729517936707,
      "learning_rate": 4e-05,
      "loss": 1.9279,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.6580256223678589,
      "eval_runtime": 11.7033,
      "eval_samples_per_second": 31.786,
      "eval_steps_per_second": 4.016,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 0.6726937294006348,
      "learning_rate": 5e-05,
      "loss": 1.8699,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 0.6740496158599854,
      "learning_rate": 4.985837000525343e-05,
      "loss": 1.7469,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 1.4485936164855957,
      "eval_runtime": 11.6969,
      "eval_samples_per_second": 31.803,
      "eval_steps_per_second": 4.018,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 1.0841119289398193,
      "learning_rate": 4.9435084745446666e-05,
      "loss": 1.7191,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 1.1675149202346802,
      "learning_rate": 4.873494021170953e-05,
      "loss": 1.5895,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 1.175584077835083,
      "learning_rate": 4.7765869321372836e-05,
      "loss": 1.5507,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 1.2826032638549805,
      "eval_runtime": 11.7024,
      "eval_samples_per_second": 31.788,
      "eval_steps_per_second": 4.016,
      "step": 185
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 1.4653639793395996,
      "learning_rate": 4.653885203484515e-05,
      "loss": 1.4345,
      "step": 200
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 1.7988545894622803,
      "learning_rate": 4.5067790948274094e-05,
      "loss": 1.4308,
      "step": 220
    },
    {
      "epoch": 4.987854251012146,
      "eval_loss": 1.1710984706878662,
      "eval_runtime": 11.7104,
      "eval_samples_per_second": 31.767,
      "eval_steps_per_second": 4.014,
      "step": 231
    },
    {
      "epoch": 5.182186234817814,
      "grad_norm": 1.7637828588485718,
      "learning_rate": 4.336935377157668e-05,
      "loss": 1.3828,
      "step": 240
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.7401658296585083,
      "learning_rate": 4.146278447662597e-05,
      "loss": 1.2948,
      "step": 260
    },
    {
      "epoch": 5.98110661268556,
      "eval_loss": 1.1053041219711304,
      "eval_runtime": 11.7036,
      "eval_samples_per_second": 31.785,
      "eval_steps_per_second": 4.016,
      "step": 277
    },
    {
      "epoch": 6.045883940620783,
      "grad_norm": 1.7593504190444946,
      "learning_rate": 3.9369685255360175e-05,
      "loss": 1.2478,
      "step": 280
    },
    {
      "epoch": 6.477732793522267,
      "grad_norm": 1.7287532091140747,
      "learning_rate": 3.711377175831626e-05,
      "loss": 1.1565,
      "step": 300
    },
    {
      "epoch": 6.909581646423752,
      "grad_norm": 1.9551547765731812,
      "learning_rate": 3.472060438683302e-05,
      "loss": 1.2179,
      "step": 320
    },
    {
      "epoch": 6.995951417004049,
      "eval_loss": 1.072460651397705,
      "eval_runtime": 11.7026,
      "eval_samples_per_second": 31.788,
      "eval_steps_per_second": 4.016,
      "step": 324
    },
    {
      "epoch": 7.341430499325236,
      "grad_norm": 2.535532236099243,
      "learning_rate": 3.2217298683490525e-05,
      "loss": 1.1197,
      "step": 340
    },
    {
      "epoch": 7.77327935222672,
      "grad_norm": 2.150266170501709,
      "learning_rate": 2.9632218102177862e-05,
      "loss": 1.1465,
      "step": 360
    },
    {
      "epoch": 7.989203778677463,
      "eval_loss": 1.0444090366363525,
      "eval_runtime": 11.6993,
      "eval_samples_per_second": 31.797,
      "eval_steps_per_second": 4.017,
      "step": 370
    },
    {
      "epoch": 8.205128205128204,
      "grad_norm": 2.408799409866333,
      "learning_rate": 2.6994652638827078e-05,
      "loss": 1.0879,
      "step": 380
    },
    {
      "epoch": 8.63697705802969,
      "grad_norm": 2.7303247451782227,
      "learning_rate": 2.433448696405563e-05,
      "loss": 1.0654,
      "step": 400
    },
    {
      "epoch": 8.982456140350877,
      "eval_loss": 1.0364046096801758,
      "eval_runtime": 11.6996,
      "eval_samples_per_second": 31.796,
      "eval_steps_per_second": 4.017,
      "step": 416
    },
    {
      "epoch": 9.068825910931174,
      "grad_norm": 2.2821261882781982,
      "learning_rate": 2.1681861817906954e-05,
      "loss": 1.0641,
      "step": 420
    },
    {
      "epoch": 9.50067476383266,
      "grad_norm": 2.2011351585388184,
      "learning_rate": 1.9066832503222128e-05,
      "loss": 1.0035,
      "step": 440
    },
    {
      "epoch": 9.932523616734143,
      "grad_norm": 2.181905508041382,
      "learning_rate": 1.651902834704924e-05,
      "loss": 1.036,
      "step": 460
    },
    {
      "epoch": 9.997300944669366,
      "eval_loss": 1.035351276397705,
      "eval_runtime": 11.6997,
      "eval_samples_per_second": 31.796,
      "eval_steps_per_second": 4.017,
      "step": 463
    },
    {
      "epoch": 10.364372469635628,
      "grad_norm": 2.7207131385803223,
      "learning_rate": 1.4067316988528617e-05,
      "loss": 0.9789,
      "step": 480
    },
    {
      "epoch": 10.796221322537113,
      "grad_norm": 2.6506447792053223,
      "learning_rate": 1.173947729700644e-05,
      "loss": 0.9563,
      "step": 500
    },
    {
      "epoch": 10.99055330634278,
      "eval_loss": 1.0434376001358032,
      "eval_runtime": 11.6958,
      "eval_samples_per_second": 31.806,
      "eval_steps_per_second": 4.019,
      "step": 509
    },
    {
      "epoch": 11.228070175438596,
      "grad_norm": 2.4060709476470947,
      "learning_rate": 9.561884626345205e-06,
      "loss": 1.0145,
      "step": 520
    },
    {
      "epoch": 11.65991902834008,
      "grad_norm": 2.9326512813568115,
      "learning_rate": 7.5592119716253855e-06,
      "loss": 0.9698,
      "step": 540
    },
    {
      "epoch": 11.983805668016194,
      "eval_loss": 1.0567598342895508,
      "eval_runtime": 11.7029,
      "eval_samples_per_second": 31.787,
      "eval_steps_per_second": 4.016,
      "step": 555
    },
    {
      "epoch": 12.091767881241566,
      "grad_norm": 2.6050772666931152,
      "learning_rate": 5.75415041425234e-06,
      "loss": 0.9381,
      "step": 560
    },
    {
      "epoch": 12.523616734143049,
      "grad_norm": 2.099339008331299,
      "learning_rate": 4.167152022937124e-06,
      "loss": 0.9417,
      "step": 580
    },
    {
      "epoch": 12.955465587044534,
      "grad_norm": 3.280305862426758,
      "learning_rate": 2.8161981235857143e-06,
      "loss": 0.9234,
      "step": 600
    },
    {
      "epoch": 12.998650472334683,
      "eval_loss": 1.0533791780471802,
      "eval_runtime": 11.6821,
      "eval_samples_per_second": 31.844,
      "eval_steps_per_second": 4.023,
      "step": 602
    }
  ],
  "logging_steps": 20,
  "max_steps": 690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 5.346992839930675e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}