{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 3816,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9968553459119497,
      "grad_norm": 2.179837942123413,
      "learning_rate": 0.000818351406877476,
      "loss": 0.2751,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6880645161290323,
      "eval_loss": 0.1752808839082718,
      "eval_runtime": 2.6716,
      "eval_samples_per_second": 1160.358,
      "eval_steps_per_second": 24.33,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 1.8881250619888306,
      "learning_rate": 0.0009058169107598648,
      "loss": 0.2375,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5232258064516129,
      "eval_loss": 0.24496081471443176,
      "eval_runtime": 2.6251,
      "eval_samples_per_second": 1180.886,
      "eval_steps_per_second": 24.761,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 1.2267760038375854,
      "learning_rate": 0.0008155768225414873,
      "loss": 0.2098,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6074193548387097,
      "eval_loss": 0.1944843977689743,
      "eval_runtime": 2.6042,
      "eval_samples_per_second": 1190.407,
      "eval_steps_per_second": 24.96,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 1.5767661333084106,
      "learning_rate": 0.0007253367343231098,
      "loss": 0.1752,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6370967741935484,
      "eval_loss": 0.1834734082221985,
      "eval_runtime": 2.6137,
      "eval_samples_per_second": 1186.055,
      "eval_steps_per_second": 24.869,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.8071340918540955,
      "learning_rate": 0.0006350966461047324,
      "loss": 0.139,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6964516129032258,
      "eval_loss": 0.13570688664913177,
      "eval_runtime": 2.6164,
      "eval_samples_per_second": 1184.832,
      "eval_steps_per_second": 24.843,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 1.5833699703216553,
      "learning_rate": 0.0005448565578863549,
      "loss": 0.1062,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7351612903225806,
      "eval_loss": 0.11580219119787216,
      "eval_runtime": 2.6685,
      "eval_samples_per_second": 1161.719,
      "eval_steps_per_second": 24.359,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.5633471012115479,
      "learning_rate": 0.0004546164696679774,
      "loss": 0.0789,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7838709677419354,
      "eval_loss": 0.09835980832576752,
      "eval_runtime": 2.7074,
      "eval_samples_per_second": 1145.011,
      "eval_steps_per_second": 24.008,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.38575279712677,
      "learning_rate": 0.0003643763814495999,
      "loss": 0.0603,
      "step": 2536
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8106451612903226,
      "eval_loss": 0.08627542108297348,
      "eval_runtime": 2.6659,
      "eval_samples_per_second": 1162.852,
      "eval_steps_per_second": 24.382,
      "step": 2544
    },
    {
      "epoch": 8.971698113207546,
      "grad_norm": 0.26294469833374023,
      "learning_rate": 0.0002741362932312224,
      "loss": 0.044,
      "step": 2853
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.837741935483871,
      "eval_loss": 0.07514873147010803,
      "eval_runtime": 2.6544,
      "eval_samples_per_second": 1167.86,
      "eval_steps_per_second": 24.487,
      "step": 2862
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.11992543935775757,
      "learning_rate": 0.00018389620501284493,
      "loss": 0.0351,
      "step": 3170
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8509677419354839,
      "eval_loss": 0.06701228022575378,
      "eval_runtime": 2.6604,
      "eval_samples_per_second": 1165.252,
      "eval_steps_per_second": 24.433,
      "step": 3180
    },
    {
      "epoch": 10.965408805031446,
      "grad_norm": 0.1402372270822525,
      "learning_rate": 9.365611679446747e-05,
      "loss": 0.0285,
      "step": 3487
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.8722580645161291,
      "eval_loss": 0.06256449222564697,
      "eval_runtime": 2.6512,
      "eval_samples_per_second": 1169.302,
      "eval_steps_per_second": 24.518,
      "step": 3498
    },
    {
      "epoch": 11.962264150943396,
      "grad_norm": 0.114226795732975,
      "learning_rate": 3.4160285760899993e-06,
      "loss": 0.0247,
      "step": 3804
    }
  ],
  "logging_steps": 317,
  "max_steps": 3816,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 1000000000.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 989687612074644.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.4357555942520456,
    "fp16": false,
    "learning_rate": 0.0009784075180017773,
    "lr_scheduler": "cosine_with_restarts",
    "num_train_epochs": 12,
    "temperature": 5,
    "warmup_steps": 379,
    "weight_decay": 0.29576423114734285
  }
}