{
  "best_metric": 1.0736711025238037,
  "best_model_checkpoint": "outputs/checkpoint-741",
  "epoch": 16.993252361673413,
  "eval_steps": 500,
  "global_step": 787,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4318488529014845,
      "grad_norm": 1.7572729587554932,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.9109,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 1.7804350852966309,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.7779,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 2.4225194454193115,
      "eval_runtime": 12.3603,
      "eval_samples_per_second": 30.096,
      "eval_steps_per_second": 3.802,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 1.2646853923797607,
      "learning_rate": 1.2e-05,
      "loss": 2.4665,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.6016279458999634,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.0703,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.802353858947754,
      "eval_runtime": 12.3146,
      "eval_samples_per_second": 30.208,
      "eval_steps_per_second": 3.817,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 0.5685097575187683,
      "learning_rate": 2e-05,
      "loss": 2.0044,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 0.5021196603775024,
      "learning_rate": 1.9970658011837404e-05,
      "loss": 1.8962,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 1.635912537574768,
      "eval_runtime": 12.2905,
      "eval_samples_per_second": 30.267,
      "eval_steps_per_second": 3.824,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 0.7247435450553894,
      "learning_rate": 1.9882804237803487e-05,
      "loss": 1.8878,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 0.736103892326355,
      "learning_rate": 1.9736954238777793e-05,
      "loss": 1.7908,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 0.7256895899772644,
      "learning_rate": 1.9533963920549307e-05,
      "loss": 1.7571,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 1.50630784034729,
      "eval_runtime": 12.3056,
      "eval_samples_per_second": 30.23,
      "eval_steps_per_second": 3.819,
      "step": 185
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 0.9457058310508728,
      "learning_rate": 1.927502451102095e-05,
      "loss": 1.6802,
      "step": 200
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 1.0999541282653809,
      "learning_rate": 1.8961655569610557e-05,
      "loss": 1.6754,
      "step": 220
    },
    {
      "epoch": 4.987854251012146,
      "eval_loss": 1.4030449390411377,
      "eval_runtime": 12.3178,
      "eval_samples_per_second": 30.2,
      "eval_steps_per_second": 3.816,
      "step": 231
    },
    {
      "epoch": 5.182186234817814,
      "grad_norm": 0.947107195854187,
      "learning_rate": 1.8595696069872013e-05,
      "loss": 1.6317,
      "step": 240
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.141870141029358,
      "learning_rate": 1.8179293607667177e-05,
      "loss": 1.5657,
      "step": 260
    },
    {
      "epoch": 5.98110661268556,
      "eval_loss": 1.318428874015808,
      "eval_runtime": 12.2941,
      "eval_samples_per_second": 30.258,
      "eval_steps_per_second": 3.823,
      "step": 277
    },
    {
      "epoch": 6.045883940620783,
      "grad_norm": 1.174417495727539,
      "learning_rate": 1.7714891798219432e-05,
      "loss": 1.529,
      "step": 280
    },
    {
      "epoch": 6.477732793522267,
      "grad_norm": 1.2409827709197998,
      "learning_rate": 1.720521593600787e-05,
      "loss": 1.4535,
      "step": 300
    },
    {
      "epoch": 6.909581646423752,
      "grad_norm": 1.337241291999817,
      "learning_rate": 1.6653257001655652e-05,
      "loss": 1.4943,
      "step": 320
    },
    {
      "epoch": 6.995951417004049,
      "eval_loss": 1.251551866531372,
      "eval_runtime": 12.3013,
      "eval_samples_per_second": 30.241,
      "eval_steps_per_second": 3.821,
      "step": 324
    },
    {
      "epoch": 7.341430499325236,
      "grad_norm": 1.5739471912384033,
      "learning_rate": 1.6062254109666383e-05,
      "loss": 1.412,
      "step": 340
    },
    {
      "epoch": 7.77327935222672,
      "grad_norm": 1.3894844055175781,
      "learning_rate": 1.5435675500012212e-05,
      "loss": 1.4226,
      "step": 360
    },
    {
      "epoch": 7.989203778677463,
      "eval_loss": 1.2039802074432373,
      "eval_runtime": 12.2961,
      "eval_samples_per_second": 30.254,
      "eval_steps_per_second": 3.822,
      "step": 370
    },
    {
      "epoch": 8.205128205128204,
      "grad_norm": 1.5219025611877441,
      "learning_rate": 1.477719818512263e-05,
      "loss": 1.3758,
      "step": 380
    },
    {
      "epoch": 8.63697705802969,
      "grad_norm": 1.8435232639312744,
      "learning_rate": 1.4090686371713403e-05,
      "loss": 1.3446,
      "step": 400
    },
    {
      "epoch": 8.982456140350877,
      "eval_loss": 1.1595730781555176,
      "eval_runtime": 12.2686,
      "eval_samples_per_second": 30.321,
      "eval_steps_per_second": 3.831,
      "step": 416
    },
    {
      "epoch": 9.068825910931174,
      "grad_norm": 1.4322152137756348,
      "learning_rate": 1.3380168784085028e-05,
      "loss": 1.3416,
      "step": 420
    },
    {
      "epoch": 9.50067476383266,
      "grad_norm": 1.7712641954421997,
      "learning_rate": 1.264981502196662e-05,
      "loss": 1.2809,
      "step": 440
    },
    {
      "epoch": 9.932523616734143,
      "grad_norm": 1.577054500579834,
      "learning_rate": 1.1903911091646684e-05,
      "loss": 1.3033,
      "step": 460
    },
    {
      "epoch": 9.997300944669366,
      "eval_loss": 1.1354897022247314,
      "eval_runtime": 12.2627,
      "eval_samples_per_second": 30.336,
      "eval_steps_per_second": 3.833,
      "step": 463
    },
    {
      "epoch": 10.364372469635628,
      "grad_norm": 1.8544738292694092,
      "learning_rate": 1.1146834253984008e-05,
      "loss": 1.2552,
      "step": 480
    },
    {
      "epoch": 10.796221322537113,
      "grad_norm": 1.9725706577301025,
      "learning_rate": 1.0383027336900356e-05,
      "loss": 1.2186,
      "step": 500
    },
    {
      "epoch": 10.99055330634278,
      "eval_loss": 1.1109498739242554,
      "eval_runtime": 12.2456,
      "eval_samples_per_second": 30.378,
      "eval_steps_per_second": 3.838,
      "step": 509
    },
    {
      "epoch": 11.228070175438596,
      "grad_norm": 1.9104483127593994,
      "learning_rate": 9.616972663099648e-06,
      "loss": 1.2626,
      "step": 520
    },
    {
      "epoch": 11.65991902834008,
      "grad_norm": 2.256964921951294,
      "learning_rate": 8.853165746015997e-06,
      "loss": 1.224,
      "step": 540
    },
    {
      "epoch": 11.983805668016194,
      "eval_loss": 1.100693702697754,
      "eval_runtime": 12.1688,
      "eval_samples_per_second": 30.57,
      "eval_steps_per_second": 3.862,
      "step": 555
    },
    {
      "epoch": 12.091767881241566,
      "grad_norm": 2.1132662296295166,
      "learning_rate": 8.096088908353316e-06,
      "loss": 1.1934,
      "step": 560
    },
    {
      "epoch": 12.523616734143049,
      "grad_norm": 1.6598352193832397,
      "learning_rate": 7.350184978033386e-06,
      "loss": 1.175,
      "step": 580
    },
    {
      "epoch": 12.955465587044534,
      "grad_norm": 2.5442428588867188,
      "learning_rate": 6.619831215914974e-06,
      "loss": 1.1688,
      "step": 600
    },
    {
      "epoch": 12.998650472334683,
      "eval_loss": 1.08635675907135,
      "eval_runtime": 12.1828,
      "eval_samples_per_second": 30.535,
      "eval_steps_per_second": 3.858,
      "step": 602
    },
    {
      "epoch": 13.387314439946019,
      "grad_norm": 1.9692449569702148,
      "learning_rate": 5.9093136282866014e-06,
      "loss": 1.1366,
      "step": 620
    },
    {
      "epoch": 13.819163292847504,
      "grad_norm": 2.064366340637207,
      "learning_rate": 5.22280181487737e-06,
      "loss": 1.1971,
      "step": 640
    },
    {
      "epoch": 13.991902834008098,
      "eval_loss": 1.0831245183944702,
      "eval_runtime": 12.1763,
      "eval_samples_per_second": 30.551,
      "eval_steps_per_second": 3.86,
      "step": 648
    },
    {
      "epoch": 14.251012145748987,
      "grad_norm": 2.0344982147216797,
      "learning_rate": 4.56432449998779e-06,
      "loss": 1.162,
      "step": 660
    },
    {
      "epoch": 14.682860998650472,
      "grad_norm": 2.1550192832946777,
      "learning_rate": 3.937745890333623e-06,
      "loss": 1.1322,
      "step": 680
    },
    {
      "epoch": 14.98515519568151,
      "eval_loss": 1.0802310705184937,
      "eval_runtime": 12.1751,
      "eval_samples_per_second": 30.554,
      "eval_steps_per_second": 3.86,
      "step": 694
    },
    {
      "epoch": 15.114709851551957,
      "grad_norm": 2.1858975887298584,
      "learning_rate": 3.3467429983443477e-06,
      "loss": 1.1555,
      "step": 700
    },
    {
      "epoch": 15.54655870445344,
      "grad_norm": 2.1912343502044678,
      "learning_rate": 2.7947840639921308e-06,
      "loss": 1.1123,
      "step": 720
    },
    {
      "epoch": 15.978407557354926,
      "grad_norm": 2.083388328552246,
      "learning_rate": 2.2851082017805704e-06,
      "loss": 1.1694,
      "step": 740
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.0736711025238037,
      "eval_runtime": 12.1577,
      "eval_samples_per_second": 30.598,
      "eval_steps_per_second": 3.866,
      "step": 741
    },
    {
      "epoch": 16.41025641025641,
      "grad_norm": 2.217958927154541,
      "learning_rate": 1.820706392332824e-06,
      "loss": 1.0918,
      "step": 760
    },
    {
      "epoch": 16.842105263157894,
      "grad_norm": 2.019780397415161,
      "learning_rate": 1.4043039301279904e-06,
      "loss": 1.1445,
      "step": 780
    },
    {
      "epoch": 16.993252361673413,
      "eval_loss": 1.0765767097473145,
      "eval_runtime": 12.1611,
      "eval_samples_per_second": 30.589,
      "eval_steps_per_second": 3.865,
      "step": 787
    }
  ],
  "logging_steps": 20,
  "max_steps": 920,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 7.035222816423936e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}