{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.6,
  "eval_steps": 5,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "grad_norm": 0.7334153552614141,
      "learning_rate": 0.0001,
      "loss": 1.115,
      "step": 1
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.5975179564042915,
      "learning_rate": 0.00018660254037844388,
      "loss": 0.9142,
      "step": 5
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.722756564617157,
      "eval_runtime": 3.0808,
      "eval_samples_per_second": 6.492,
      "eval_steps_per_second": 1.623,
      "step": 5
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.4159151923493577,
      "learning_rate": 0.00011736481776669306,
      "loss": 0.6834,
      "step": 10
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.6278184056282043,
      "eval_runtime": 1.8619,
      "eval_samples_per_second": 10.742,
      "eval_steps_per_second": 2.685,
      "step": 10
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.40842948448715766,
      "learning_rate": 3.5721239031346066e-05,
      "loss": 0.5454,
      "step": 15
    },
    {
      "epoch": 1.2,
      "eval_loss": 0.6100303530693054,
      "eval_runtime": 1.8606,
      "eval_samples_per_second": 10.749,
      "eval_steps_per_second": 2.687,
      "step": 15
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.3737806540766406,
      "learning_rate": 0.0,
      "loss": 0.5243,
      "step": 20
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.6069896221160889,
      "eval_runtime": 1.8906,
      "eval_samples_per_second": 10.579,
      "eval_steps_per_second": 2.645,
      "step": 20
    },
    {
      "epoch": 1.6,
      "step": 20,
      "total_flos": 1591906369536.0,
      "train_loss": 0.6768446683883667,
      "train_runtime": 82.5503,
      "train_samples_per_second": 1.938,
      "train_steps_per_second": 0.242
    }
  ],
  "logging_steps": 5,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1591906369536.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}