{
  "best_metric": 1.2172602415084839,
  "best_model_checkpoint": "outputs/checkpoint-602",
  "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": 1.8696340322494507,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.0163,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.5766105651855469,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.9853,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 1.8159902095794678,
      "eval_runtime": 12.9003,
      "eval_samples_per_second": 28.837,
      "eval_steps_per_second": 3.643,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.4374605715274811,
      "learning_rate": 6e-06,
      "loss": 1.9778,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.46712979674339294,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.8764,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.6934317350387573,
      "eval_runtime": 12.8941,
      "eval_samples_per_second": 28.85,
      "eval_steps_per_second": 3.645,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 0.49692320823669434,
      "learning_rate": 1e-05,
      "loss": 1.86,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 0.46815183758735657,
      "learning_rate": 9.978795707151492e-06,
      "loss": 1.7959,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 1.5927985906600952,
      "eval_runtime": 12.8879,
      "eval_samples_per_second": 28.864,
      "eval_steps_per_second": 3.647,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 0.5940990447998047,
      "learning_rate": 9.915362677420045e-06,
      "loss": 1.8186,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 0.5652334690093994,
      "learning_rate": 9.810238931821139e-06,
      "loss": 1.7388,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 0.5389769077301025,
      "learning_rate": 9.664316100229578e-06,
      "loss": 1.7176,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 1.5169272422790527,
      "eval_runtime": 12.8974,
      "eval_samples_per_second": 28.843,
      "eval_steps_per_second": 3.644,
      "step": 185
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 0.6361340880393982,
      "learning_rate": 9.478831858827105e-06,
      "loss": 1.6774,
      "step": 200
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 0.766372561454773,
      "learning_rate": 9.255359432483106e-06,
      "loss": 1.6827,
      "step": 220
    },
    {
      "epoch": 4.987854251012146,
      "eval_loss": 1.4580748081207275,
      "eval_runtime": 12.9055,
      "eval_samples_per_second": 28.825,
      "eval_steps_per_second": 3.642,
      "step": 231
    },
    {
      "epoch": 5.182186234817814,
      "grad_norm": 0.6636700630187988,
      "learning_rate": 8.995794251106295e-06,
      "loss": 1.6551,
      "step": 240
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 0.7402545809745789,
      "learning_rate": 8.702337873144343e-06,
      "loss": 1.6079,
      "step": 260
    },
    {
      "epoch": 5.98110661268556,
      "eval_loss": 1.4053566455841064,
      "eval_runtime": 12.8973,
      "eval_samples_per_second": 28.843,
      "eval_steps_per_second": 3.644,
      "step": 277
    },
    {
      "epoch": 6.045883940620783,
      "grad_norm": 0.7764894366264343,
      "learning_rate": 8.37747931258788e-06,
      "loss": 1.6001,
      "step": 280
    },
    {
      "epoch": 6.477732793522267,
      "grad_norm": 0.8082602024078369,
      "learning_rate": 8.023973927857857e-06,
      "loss": 1.5414,
      "step": 300
    },
    {
      "epoch": 6.909581646423752,
      "grad_norm": 0.8577279448509216,
      "learning_rate": 7.644820051634813e-06,
      "loss": 1.5751,
      "step": 320
    },
    {
      "epoch": 6.995951417004049,
      "eval_loss": 1.3543951511383057,
      "eval_runtime": 12.9006,
      "eval_samples_per_second": 28.836,
      "eval_steps_per_second": 3.643,
      "step": 324
    },
    {
      "epoch": 7.341430499325236,
      "grad_norm": 1.0512497425079346,
      "learning_rate": 7.243233559849179e-06,
      "loss": 1.5131,
      "step": 340
    },
    {
      "epoch": 7.77327935222672,
      "grad_norm": 0.8688454627990723,
      "learning_rate": 6.822620595531286e-06,
      "loss": 1.5286,
      "step": 360
    },
    {
      "epoch": 7.989203778677463,
      "eval_loss": 1.3146252632141113,
      "eval_runtime": 12.9044,
      "eval_samples_per_second": 28.827,
      "eval_steps_per_second": 3.642,
      "step": 370
    },
    {
      "epoch": 8.205128205128204,
      "grad_norm": 0.9062187671661377,
      "learning_rate": 6.386548678869644e-06,
      "loss": 1.4984,
      "step": 380
    },
    {
      "epoch": 8.63697705802969,
      "grad_norm": 1.1049658060073853,
      "learning_rate": 5.938716448513819e-06,
      "loss": 1.4733,
      "step": 400
    },
    {
      "epoch": 8.982456140350877,
      "eval_loss": 1.2819702625274658,
      "eval_runtime": 12.9073,
      "eval_samples_per_second": 28.821,
      "eval_steps_per_second": 3.641,
      "step": 416
    },
    {
      "epoch": 9.068825910931174,
      "grad_norm": 0.8910373449325562,
      "learning_rate": 5.4829222907675895e-06,
      "loss": 1.4715,
      "step": 420
    },
    {
      "epoch": 9.50067476383266,
      "grad_norm": 1.2289608716964722,
      "learning_rate": 5.0230321227507595e-06,
      "loss": 1.4326,
      "step": 440
    },
    {
      "epoch": 9.932523616734143,
      "grad_norm": 1.0093016624450684,
      "learning_rate": 4.562946602783637e-06,
      "loss": 1.4576,
      "step": 460
    },
    {
      "epoch": 9.997300944669366,
      "eval_loss": 1.2573164701461792,
      "eval_runtime": 12.8922,
      "eval_samples_per_second": 28.855,
      "eval_steps_per_second": 3.646,
      "step": 463
    },
    {
      "epoch": 10.364372469635628,
      "grad_norm": 1.175689697265625,
      "learning_rate": 4.10656804610652e-06,
      "loss": 1.4222,
      "step": 480
    },
    {
      "epoch": 10.796221322537113,
      "grad_norm": 1.1676870584487915,
      "learning_rate": 3.6577673265456296e-06,
      "loss": 1.3863,
      "step": 500
    },
    {
      "epoch": 10.99055330634278,
      "eval_loss": 1.2392951250076294,
      "eval_runtime": 12.9835,
      "eval_samples_per_second": 28.652,
      "eval_steps_per_second": 3.62,
      "step": 509
    },
    {
      "epoch": 11.228070175438596,
      "grad_norm": 1.2346034049987793,
      "learning_rate": 3.220351044856247e-06,
      "loss": 1.4357,
      "step": 520
    },
    {
      "epoch": 11.65991902834008,
      "grad_norm": 1.4561930894851685,
      "learning_rate": 2.7980292422118282e-06,
      "loss": 1.414,
      "step": 540
    },
    {
      "epoch": 11.983805668016194,
      "eval_loss": 1.226759672164917,
      "eval_runtime": 12.9423,
      "eval_samples_per_second": 28.743,
      "eval_steps_per_second": 3.631,
      "step": 555
    },
    {
      "epoch": 12.091767881241566,
      "grad_norm": 1.2641685009002686,
      "learning_rate": 2.3943839326842096e-06,
      "loss": 1.3907,
      "step": 560
    },
    {
      "epoch": 12.523616734143049,
      "grad_norm": 0.9569095373153687,
      "learning_rate": 2.012838721613447e-06,
      "loss": 1.3752,
      "step": 580
    },
    {
      "epoch": 12.955465587044534,
      "grad_norm": 1.5854405164718628,
      "learning_rate": 1.6566297675557392e-06,
      "loss": 1.3707,
      "step": 600
    },
    {
      "epoch": 12.998650472334683,
      "eval_loss": 1.2172602415084839,
      "eval_runtime": 12.9465,
      "eval_samples_per_second": 28.734,
      "eval_steps_per_second": 3.63,
      "step": 602
    }
  ],
  "logging_steps": 20,
  "max_steps": 782,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 17,
  "save_steps": 500,
  "total_flos": 4.634294855388365e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}