{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "grad_norm": 3.397108066667403,
      "learning_rate": 0.0005,
      "loss": 1.578,
      "step": 1
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.6268932148195803,
      "learning_rate": 0.0009330127018922195,
      "loss": 1.4849,
      "step": 5
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2847128566672604,
      "learning_rate": 0.0005868240888334653,
      "loss": 1.1777,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.4108675718307495,
      "eval_runtime": 111.5987,
      "eval_samples_per_second": 187.735,
      "eval_steps_per_second": 5.869,
      "step": 10
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.16816916470466492,
      "learning_rate": 0.0001786061951567303,
      "loss": 0.9659,
      "step": 15
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.13203755081634758,
      "learning_rate": 0.0,
      "loss": 0.9295,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.3907862901687622,
      "eval_runtime": 109.0816,
      "eval_samples_per_second": 192.067,
      "eval_steps_per_second": 6.005,
      "step": 20
    },
    {
      "epoch": 2.0,
      "step": 20,
      "total_flos": 1938397593600.0,
      "train_loss": 1.144122803211212,
      "train_runtime": 280.147,
      "train_samples_per_second": 9.131,
      "train_steps_per_second": 0.071
    }
  ],
  "logging_steps": 5,
  "max_steps": 20,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1938397593600.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}