trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.001142857142857143,
  "eval_steps": 500,
  "global_step": 2,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 3140.2083435058594,
      "epoch": 0.0005714285714285715,
      "grad_norm": 0.007130602840334177,
      "kl": 0.00013720989227294922,
      "learning_rate": 0.0,
      "loss": -0.013,
      "reward": 0.13575429469347,
      "reward_std": 0.2010277360677719,
      "rewards/cosine_scaled_reward": -0.057122852653265,
      "rewards/format_reward": 0.25,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 2772.2916870117188,
      "epoch": 0.001142857142857143,
      "grad_norm": 0.00968429446220398,
      "kl": 7.659196853637695e-05,
      "learning_rate": 1e-06,
      "loss": 0.0706,
      "reward": -0.22162751108407974,
      "reward_std": 0.24331748113036156,
      "rewards/cosine_scaled_reward": -0.31914709135890007,
      "rewards/format_reward": 0.4166666716337204,
      "step": 2
    },
    {
      "epoch": 0.001142857142857143,
      "step": 2,
      "total_flos": 0.0,
      "train_loss": 0.028762598522007465,
      "train_runtime": 309.4525,
      "train_samples_per_second": 0.155,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 1,
  "max_steps": 2,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}