{
  "best_metric": 0.90346360206604,
  "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": 0.5469953417778015,
      "learning_rate": 2e-05,
      "loss": 1.9745,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.48892971873283386,
      "learning_rate": 4e-05,
      "loss": 1.8762,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 1.6047619581222534,
      "eval_runtime": 12.7198,
      "eval_samples_per_second": 29.246,
      "eval_steps_per_second": 3.695,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.5274947285652161,
      "learning_rate": 6e-05,
      "loss": 1.7913,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.6796801686286926,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.2794835567474365,
      "eval_runtime": 12.7229,
      "eval_samples_per_second": 29.239,
      "eval_steps_per_second": 3.694,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 0.9777641892433167,
      "learning_rate": 0.0001,
      "loss": 1.5142,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 1.0962361097335815,
      "learning_rate": 9.985329005918702e-05,
      "loss": 1.3395,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 1.0361310243606567,
      "eval_runtime": 12.6733,
      "eval_samples_per_second": 29.353,
      "eval_steps_per_second": 3.709,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 1.30574369430542,
      "learning_rate": 9.941402118901744e-05,
      "loss": 1.2941,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 1.4931496381759644,
      "learning_rate": 9.868477119388896e-05,
      "loss": 1.1092,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 1.4002817869186401,
      "learning_rate": 9.766981960274653e-05,
      "loss": 1.0864,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 0.932435154914856,
      "eval_runtime": 12.7232,
      "eval_samples_per_second": 29.238,
      "eval_steps_per_second": 3.694,
      "step": 185
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 1.869638442993164,
      "learning_rate": 9.637512255510475e-05,
      "loss": 0.9429,
      "step": 200
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 1.8662426471710205,
      "learning_rate": 9.480827784805278e-05,
      "loss": 0.9713,
      "step": 220
    },
    {
      "epoch": 4.987854251012146,
      "eval_loss": 0.90346360206604,
      "eval_runtime": 12.6762,
      "eval_samples_per_second": 29.346,
      "eval_steps_per_second": 3.708,
      "step": 231
    },
    {
      "epoch": 5.182186234817814,
      "grad_norm": 1.7677403688430786,
      "learning_rate": 9.297848034936006e-05,
      "loss": 0.9134,
      "step": 240
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.6769498586654663,
      "learning_rate": 9.089646803833589e-05,
      "loss": 0.8171,
      "step": 260
    },
    {
      "epoch": 5.98110661268556,
      "eval_loss": 0.9262576103210449,
      "eval_runtime": 12.7331,
      "eval_samples_per_second": 29.215,
      "eval_steps_per_second": 3.691,
      "step": 277
    },
    {
      "epoch": 6.045883940620783,
      "grad_norm": 1.4469077587127686,
      "learning_rate": 8.857445899109715e-05,
      "loss": 0.7771,
      "step": 280
    },
    {
      "epoch": 6.477732793522267,
      "grad_norm": 1.5637768507003784,
      "learning_rate": 8.602607968003935e-05,
      "loss": 0.6662,
      "step": 300
    },
    {
      "epoch": 6.909581646423752,
      "grad_norm": 1.5962625741958618,
      "learning_rate": 8.326628500827826e-05,
      "loss": 0.736,
      "step": 320
    },
    {
      "epoch": 6.995951417004049,
      "eval_loss": 1.029656171798706,
      "eval_runtime": 12.6598,
      "eval_samples_per_second": 29.384,
      "eval_steps_per_second": 3.713,
      "step": 324
    },
    {
      "epoch": 7.341430499325236,
      "grad_norm": 2.1342411041259766,
      "learning_rate": 8.03112705483319e-05,
      "loss": 0.6183,
      "step": 340
    },
    {
      "epoch": 7.77327935222672,
      "grad_norm": 1.800087809562683,
      "learning_rate": 7.717837750006106e-05,
      "loss": 0.6515,
      "step": 360
    },
    {
      "epoch": 7.989203778677463,
      "eval_loss": 1.0591753721237183,
      "eval_runtime": 12.6625,
      "eval_samples_per_second": 29.378,
      "eval_steps_per_second": 3.712,
      "step": 370
    },
    {
      "epoch": 8.205128205128204,
      "grad_norm": 1.779543399810791,
      "learning_rate": 7.388599092561315e-05,
      "loss": 0.5783,
      "step": 380
    },
    {
      "epoch": 8.63697705802969,
      "grad_norm": 2.137484550476074,
      "learning_rate": 7.045343185856701e-05,
      "loss": 0.5429,
      "step": 400
    },
    {
      "epoch": 8.982456140350877,
      "eval_loss": 1.1770082712173462,
      "eval_runtime": 12.6696,
      "eval_samples_per_second": 29.362,
      "eval_steps_per_second": 3.71,
      "step": 416
    },
    {
      "epoch": 9.068825910931174,
      "grad_norm": 1.5784409046173096,
      "learning_rate": 6.690084392042513e-05,
      "loss": 0.5434,
      "step": 420
    },
    {
      "epoch": 9.50067476383266,
      "grad_norm": 1.5823142528533936,
      "learning_rate": 6.32490751098331e-05,
      "loss": 0.4707,
      "step": 440
    },
    {
      "epoch": 9.932523616734143,
      "grad_norm": 1.8034058809280396,
      "learning_rate": 5.951955545823342e-05,
      "loss": 0.4917,
      "step": 460
    },
    {
      "epoch": 9.997300944669366,
      "eval_loss": 1.2135237455368042,
      "eval_runtime": 12.6734,
      "eval_samples_per_second": 29.353,
      "eval_steps_per_second": 3.709,
      "step": 463
    },
    {
      "epoch": 10.364372469635628,
      "grad_norm": 1.660635232925415,
      "learning_rate": 5.573417126992003e-05,
      "loss": 0.4104,
      "step": 480
    },
    {
      "epoch": 10.796221322537113,
      "grad_norm": 1.7648595571517944,
      "learning_rate": 5.191513668450178e-05,
      "loss": 0.4114,
      "step": 500
    },
    {
      "epoch": 10.99055330634278,
      "eval_loss": 1.3011534214019775,
      "eval_runtime": 12.6994,
      "eval_samples_per_second": 29.293,
      "eval_steps_per_second": 3.701,
      "step": 509
    },
    {
      "epoch": 11.228070175438596,
      "grad_norm": 1.8220391273498535,
      "learning_rate": 4.8084863315498234e-05,
      "loss": 0.4249,
      "step": 520
    },
    {
      "epoch": 11.65991902834008,
      "grad_norm": 1.9576756954193115,
      "learning_rate": 4.4265828730079987e-05,
      "loss": 0.386,
      "step": 540
    },
    {
      "epoch": 11.983805668016194,
      "eval_loss": 1.4017869234085083,
      "eval_runtime": 12.6911,
      "eval_samples_per_second": 29.312,
      "eval_steps_per_second": 3.703,
      "step": 555
    }
  ],
  "logging_steps": 20,
  "max_steps": 920,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 4.280269211892941e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}