{
  "best_metric": 0.9229514598846436,
  "best_model_checkpoint": "outputs/checkpoint-231",
  "epoch": 11.983805668016194,
  "eval_steps": 500,
  "global_step": 555,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4318488529014845,
      "grad_norm": 1.8083888292312622,
      "learning_rate": 4e-05,
      "loss": 2.7223,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.509838342666626,
      "learning_rate": 8e-05,
      "loss": 2.063,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 1.6813331842422485,
      "eval_runtime": 12.0916,
      "eval_samples_per_second": 30.765,
      "eval_steps_per_second": 3.887,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.6181766390800476,
      "learning_rate": 0.00012,
      "loss": 1.8828,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.8454991579055786,
      "learning_rate": 0.00016,
      "loss": 1.6444,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.2503478527069092,
      "eval_runtime": 12.1403,
      "eval_samples_per_second": 30.642,
      "eval_steps_per_second": 3.871,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 1.1776914596557617,
      "learning_rate": 0.0002,
      "loss": 1.5242,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 1.091369390487671,
      "learning_rate": 0.00019943348002101371,
      "loss": 1.3296,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 1.028580904006958,
      "eval_runtime": 12.1164,
      "eval_samples_per_second": 30.702,
      "eval_steps_per_second": 3.879,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 1.0799137353897095,
      "learning_rate": 0.00019774033898178667,
      "loss": 1.2821,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 1.2310456037521362,
      "learning_rate": 0.00019493976084683813,
      "loss": 1.0677,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 1.1375850439071655,
      "learning_rate": 0.00019106347728549135,
      "loss": 1.0588,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 0.9346915483474731,
      "eval_runtime": 12.162,
      "eval_samples_per_second": 30.587,
      "eval_steps_per_second": 3.864,
      "step": 185
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 1.3716574907302856,
      "learning_rate": 0.0001861554081393806,
      "loss": 0.8876,
      "step": 200
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 1.3819169998168945,
      "learning_rate": 0.00018027116379309638,
      "loss": 0.9074,
      "step": 220
    },
    {
      "epoch": 4.987854251012146,
      "eval_loss": 0.9229514598846436,
      "eval_runtime": 12.1062,
      "eval_samples_per_second": 30.728,
      "eval_steps_per_second": 3.882,
      "step": 231
    },
    {
      "epoch": 5.182186234817814,
      "grad_norm": 1.327073097229004,
      "learning_rate": 0.00017347741508630672,
      "loss": 0.8266,
      "step": 240
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.140026569366455,
      "learning_rate": 0.00016585113790650388,
      "loss": 0.7142,
      "step": 260
    },
    {
      "epoch": 5.98110661268556,
      "eval_loss": 0.9703238606452942,
      "eval_runtime": 12.1046,
      "eval_samples_per_second": 30.732,
      "eval_steps_per_second": 3.883,
      "step": 277
    },
    {
      "epoch": 6.045883940620783,
      "grad_norm": 1.0288299322128296,
      "learning_rate": 0.0001574787410214407,
      "loss": 0.6626,
      "step": 280
    },
    {
      "epoch": 6.477732793522267,
      "grad_norm": 1.1120308637619019,
      "learning_rate": 0.00014845508703326504,
      "loss": 0.5305,
      "step": 300
    },
    {
      "epoch": 6.909581646423752,
      "grad_norm": 1.0534147024154663,
      "learning_rate": 0.00013888241754733208,
      "loss": 0.5869,
      "step": 320
    },
    {
      "epoch": 6.995951417004049,
      "eval_loss": 1.0680469274520874,
      "eval_runtime": 12.1954,
      "eval_samples_per_second": 30.503,
      "eval_steps_per_second": 3.854,
      "step": 324
    },
    {
      "epoch": 7.341430499325236,
      "grad_norm": 1.0005006790161133,
      "learning_rate": 0.0001288691947339621,
      "loss": 0.4619,
      "step": 340
    },
    {
      "epoch": 7.77327935222672,
      "grad_norm": 1.1516958475112915,
      "learning_rate": 0.00011852887240871145,
      "loss": 0.4722,
      "step": 360
    },
    {
      "epoch": 7.989203778677463,
      "eval_loss": 1.1098763942718506,
      "eval_runtime": 12.1616,
      "eval_samples_per_second": 30.588,
      "eval_steps_per_second": 3.865,
      "step": 370
    },
    {
      "epoch": 8.205128205128204,
      "grad_norm": 0.9738859534263611,
      "learning_rate": 0.00010797861055530831,
      "loss": 0.3932,
      "step": 380
    },
    {
      "epoch": 8.63697705802969,
      "grad_norm": 1.0977412462234497,
      "learning_rate": 9.733794785622253e-05,
      "loss": 0.3471,
      "step": 400
    },
    {
      "epoch": 8.982456140350877,
      "eval_loss": 1.191792607307434,
      "eval_runtime": 12.1511,
      "eval_samples_per_second": 30.614,
      "eval_steps_per_second": 3.868,
      "step": 416
    },
    {
      "epoch": 9.068825910931174,
      "grad_norm": 0.8212457895278931,
      "learning_rate": 8.672744727162781e-05,
      "loss": 0.3473,
      "step": 420
    },
    {
      "epoch": 9.50067476383266,
      "grad_norm": 1.007986068725586,
      "learning_rate": 7.626733001288851e-05,
      "loss": 0.2672,
      "step": 440
    },
    {
      "epoch": 9.932523616734143,
      "grad_norm": 1.0054905414581299,
      "learning_rate": 6.607611338819697e-05,
      "loss": 0.2752,
      "step": 460
    },
    {
      "epoch": 9.997300944669366,
      "eval_loss": 1.290297031402588,
      "eval_runtime": 12.1732,
      "eval_samples_per_second": 30.559,
      "eval_steps_per_second": 3.861,
      "step": 463
    },
    {
      "epoch": 10.364372469635628,
      "grad_norm": 0.9115879535675049,
      "learning_rate": 5.626926795411447e-05,
      "loss": 0.2023,
      "step": 480
    },
    {
      "epoch": 10.796221322537113,
      "grad_norm": 0.9712518453598022,
      "learning_rate": 4.695790918802576e-05,
      "loss": 0.2015,
      "step": 500
    },
    {
      "epoch": 10.99055330634278,
      "eval_loss": 1.3780165910720825,
      "eval_runtime": 12.1643,
      "eval_samples_per_second": 30.581,
      "eval_steps_per_second": 3.864,
      "step": 509
    },
    {
      "epoch": 11.228070175438596,
      "grad_norm": 0.9184095859527588,
      "learning_rate": 3.824753850538082e-05,
      "loss": 0.1934,
      "step": 520
    },
    {
      "epoch": 11.65991902834008,
      "grad_norm": 0.763605535030365,
      "learning_rate": 3.0236847886501542e-05,
      "loss": 0.1716,
      "step": 540
    },
    {
      "epoch": 11.983805668016194,
      "eval_loss": 1.4491263628005981,
      "eval_runtime": 12.1321,
      "eval_samples_per_second": 30.662,
      "eval_steps_per_second": 3.874,
      "step": 555
    }
  ],
  "logging_steps": 20,
  "max_steps": 690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 4.962667014222643e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}