trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0019414459889725867,
  "eval_steps": 500,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 512.0,
      "completions/clipped_ratio": 1.0,
      "completions/max_length": 512.0,
      "completions/max_terminated_length": 0.0,
      "completions/mean_length": 512.0,
      "completions/mean_terminated_length": 0.0,
      "completions/min_length": 512.0,
      "completions/min_terminated_length": 0.0,
      "epoch": 0.0009707229944862934,
      "frac_reward_zero_std": 0.61,
      "grad_norm": 0.02464698627591133,
      "kl": 0.03792417001328431,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0,
      "num_tokens": 222000.0,
      "reward": 0.43,
      "reward_std": 0.20350852847099304,
      "rewards/compute_reward/mean": 0.43,
      "rewards/compute_reward/std": 0.3326409870386124,
      "step": 50
    },
    {
      "completion_length": 512.0,
      "completions/clipped_ratio": 1.0,
      "completions/max_length": 512.0,
      "completions/max_terminated_length": 0.0,
      "completions/mean_length": 512.0,
      "completions/mean_terminated_length": 0.0,
      "completions/min_length": 512.0,
      "completions/min_terminated_length": 0.0,
      "epoch": 0.0019414459889725867,
      "frac_reward_zero_std": 0.67,
      "grad_norm": 0.020398980006575584,
      "kl": 0.27587527528405187,
      "learning_rate": 4.855143631968242e-06,
      "loss": 0.0003,
      "num_tokens": 444000.0,
      "reward": 0.4275,
      "reward_std": 0.17582903623580934,
      "rewards/compute_reward/mean": 0.4275,
      "rewards/compute_reward/std": 0.40157180368900297,
      "step": 100
    }
  ],
  "logging_steps": 50,
  "max_steps": 500,
  "num_input_tokens_seen": 444000,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}