{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.0,
  "global_step": 14294,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.49,
      "learning_rate": 1.9347045380346067e-05,
      "loss": 0.1452,
      "step": 500
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.8694090760692135e-05,
      "loss": 0.1234,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9677488207817078,
      "eval_loss": 0.10579279065132141,
      "eval_runtime": 3.6577,
      "eval_samples_per_second": 2551.628,
      "eval_steps_per_second": 79.832,
      "step": 1021
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.80411361410382e-05,
      "loss": 0.0813,
      "step": 1500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.7388181521384265e-05,
      "loss": 0.0862,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9707489609718323,
      "eval_loss": 0.1058717668056488,
      "eval_runtime": 3.5844,
      "eval_samples_per_second": 2603.792,
      "eval_steps_per_second": 81.464,
      "step": 2042
    },
    {
      "epoch": 2.45,
      "learning_rate": 1.673522690173033e-05,
      "loss": 0.0514,
      "step": 2500
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.6082272282076395e-05,
      "loss": 0.0485,
      "step": 3000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9696775078773499,
      "eval_loss": 0.1178998202085495,
      "eval_runtime": 3.5784,
      "eval_samples_per_second": 2608.148,
      "eval_steps_per_second": 81.601,
      "step": 3063
    },
    {
      "epoch": 3.43,
      "learning_rate": 1.5429317662422463e-05,
      "loss": 0.0332,
      "step": 3500
    },
    {
      "epoch": 3.92,
      "learning_rate": 1.4776363042768528e-05,
      "loss": 0.0286,
      "step": 4000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9706417918205261,
      "eval_loss": 0.16240529716014862,
      "eval_runtime": 3.571,
      "eval_samples_per_second": 2613.581,
      "eval_steps_per_second": 81.771,
      "step": 4084
    },
    {
      "epoch": 4.41,
      "learning_rate": 1.4123408423114595e-05,
      "loss": 0.016,
      "step": 4500
    },
    {
      "epoch": 4.9,
      "learning_rate": 1.3470453803460662e-05,
      "loss": 0.0172,
      "step": 5000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.966570258140564,
      "eval_loss": 0.21753180027008057,
      "eval_runtime": 3.5709,
      "eval_samples_per_second": 2613.643,
      "eval_steps_per_second": 81.773,
      "step": 5105
    },
    {
      "epoch": 5.39,
      "learning_rate": 1.2817499183806728e-05,
      "loss": 0.0093,
      "step": 5500
    },
    {
      "epoch": 5.88,
      "learning_rate": 1.2164544564152792e-05,
      "loss": 0.0089,
      "step": 6000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9691417813301086,
      "eval_loss": 0.2306356430053711,
      "eval_runtime": 3.7481,
      "eval_samples_per_second": 2490.091,
      "eval_steps_per_second": 77.907,
      "step": 6126
    },
    {
      "epoch": 6.37,
      "learning_rate": 1.1511589944498858e-05,
      "loss": 0.0101,
      "step": 6500
    },
    {
      "epoch": 6.86,
      "learning_rate": 1.0858635324844923e-05,
      "loss": 0.006,
      "step": 7000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9708561301231384,
      "eval_loss": 0.23669685423374176,
      "eval_runtime": 3.5849,
      "eval_samples_per_second": 2603.453,
      "eval_steps_per_second": 81.454,
      "step": 7147
    },
    {
      "epoch": 7.35,
      "learning_rate": 1.020568070519099e-05,
      "loss": 0.0071,
      "step": 7500
    },
    {
      "epoch": 7.84,
      "learning_rate": 9.552726085537057e-06,
      "loss": 0.0059,
      "step": 8000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.968820333480835,
      "eval_loss": 0.2358114868402481,
      "eval_runtime": 3.5969,
      "eval_samples_per_second": 2594.746,
      "eval_steps_per_second": 81.181,
      "step": 8168
    },
    {
      "epoch": 8.33,
      "learning_rate": 8.899771465883122e-06,
      "loss": 0.0029,
      "step": 8500
    },
    {
      "epoch": 8.81,
      "learning_rate": 8.246816846229188e-06,
      "loss": 0.0018,
      "step": 9000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9684988856315613,
      "eval_loss": 0.2649618685245514,
      "eval_runtime": 3.5581,
      "eval_samples_per_second": 2622.995,
      "eval_steps_per_second": 82.065,
      "step": 9189
    },
    {
      "epoch": 9.3,
      "learning_rate": 7.593862226575253e-06,
      "loss": 0.0017,
      "step": 9500
    },
    {
      "epoch": 9.79,
      "learning_rate": 6.94090760692132e-06,
      "loss": 0.0044,
      "step": 10000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9711775183677673,
      "eval_loss": 0.24978068470954895,
      "eval_runtime": 3.5834,
      "eval_samples_per_second": 2604.532,
      "eval_steps_per_second": 81.488,
      "step": 10210
    },
    {
      "epoch": 10.28,
      "learning_rate": 6.287952987267385e-06,
      "loss": 0.0009,
      "step": 10500
    },
    {
      "epoch": 10.77,
      "learning_rate": 5.634998367613451e-06,
      "loss": 0.0035,
      "step": 11000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9695703387260437,
      "eval_loss": 0.27938899397850037,
      "eval_runtime": 3.5853,
      "eval_samples_per_second": 2603.116,
      "eval_steps_per_second": 81.443,
      "step": 11231
    },
    {
      "epoch": 11.26,
      "learning_rate": 4.9820437479595175e-06,
      "loss": 0.0005,
      "step": 11500
    },
    {
      "epoch": 11.75,
      "learning_rate": 4.329089128305583e-06,
      "loss": 0.0012,
      "step": 12000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.9726775884628296,
      "eval_loss": 0.2688906788825989,
      "eval_runtime": 3.5905,
      "eval_samples_per_second": 2599.329,
      "eval_steps_per_second": 81.325,
      "step": 12252
    },
    {
      "epoch": 12.24,
      "learning_rate": 3.676134508651649e-06,
      "loss": 0.0012,
      "step": 12500
    },
    {
      "epoch": 12.73,
      "learning_rate": 3.023179888997715e-06,
      "loss": 0.0006,
      "step": 13000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9711775183677673,
      "eval_loss": 0.28157007694244385,
      "eval_runtime": 3.5828,
      "eval_samples_per_second": 2604.928,
      "eval_steps_per_second": 81.5,
      "step": 13273
    },
    {
      "epoch": 13.22,
      "learning_rate": 2.370225269343781e-06,
      "loss": 0.0007,
      "step": 13500
    },
    {
      "epoch": 13.71,
      "learning_rate": 1.7172706496898467e-06,
      "loss": 0.0008,
      "step": 14000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.973320484161377,
      "eval_loss": 0.273930162191391,
      "eval_runtime": 3.5757,
      "eval_samples_per_second": 2610.101,
      "eval_steps_per_second": 81.662,
      "step": 14294
    }
  ],
  "max_steps": 15315,
  "num_train_epochs": 15,
  "total_flos": 3.0079908340301824e+16,
  "trial_name": null,
  "trial_params": null
}