trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4,
  "eval_steps": 500,
  "global_step": 3,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 605.5,
      "completions/max_terminated_length": 605.5,
      "completions/mean_length": 214.5625,
      "completions/mean_terminated_length": 214.5625,
      "completions/min_length": 9.0,
      "completions/min_terminated_length": 9.0,
      "epoch": 0.26666666666666666,
      "grad_norm": 14.364421844482422,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": -0.0457,
      "num_tokens": 7529.0,
      "reward": 0.27101069688796997,
      "reward_std": 0.3605286180973053,
      "rewards/concensus_correctness_reward_func/mean": 0.0,
      "rewards/concensus_correctness_reward_func/std": 0.0,
      "rewards/consensus_reward_func/mean": 0.125,
      "rewards/consensus_reward_func/std": 0.3535533845424652,
      "rewards/cumulative_reward_2/mean": 0.0,
      "rewards/cumulative_reward_2/std": 0.0,
      "rewards/final_correctness_reward_func/mean": 0.125,
      "rewards/final_correctness_reward_func/std": 0.3535533845424652,
      "rewards/question_recreation_reward_func/mean": 0.02001071721315384,
      "rewards/question_recreation_reward_func/std": 0.009133741725236177,
      "rewards/soft_format_reward_func/mean": 0.0,
      "rewards/soft_format_reward_func/std": 0.0,
      "rewards/strict_format_reward_func/mean": 0.0,
      "rewards/strict_format_reward_func/std": 0.0,
      "rewards/xmlcount_reward_func/mean": 0.0010000000474974513,
      "rewards/xmlcount_reward_func/std": 0.002828427357599139,
      "step": 2
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 321.0,
      "completions/max_terminated_length": 321.0,
      "completions/mean_length": 167.625,
      "completions/mean_terminated_length": 167.625,
      "completions/min_length": 19.0,
      "completions/min_terminated_length": 19.0,
      "epoch": 0.4,
      "kl": 0.0005115475723869167,
      "num_tokens": 10918.0,
      "reward": 0.16452272236347198,
      "reward_std": 0.054086603224277496,
      "rewards/concensus_correctness_reward_func/mean": 0.0,
      "rewards/concensus_correctness_reward_func/std": 0.0,
      "rewards/consensus_reward_func/mean": 0.0,
      "rewards/consensus_reward_func/std": 0.0,
      "rewards/cumulative_reward_2/mean": 0.0,
      "rewards/cumulative_reward_2/std": 0.0,
      "rewards/final_correctness_reward_func/mean": 0.0,
      "rewards/final_correctness_reward_func/std": 0.0,
      "rewards/question_recreation_reward_func/mean": 0.16452273726463318,
      "rewards/question_recreation_reward_func/std": 0.05140342935919762,
      "rewards/soft_format_reward_func/mean": 0.0,
      "rewards/soft_format_reward_func/std": 0.0,
      "rewards/strict_format_reward_func/mean": 0.0,
      "rewards/strict_format_reward_func/std": 0.0,
      "rewards/xmlcount_reward_func/mean": 0.0,
      "rewards/xmlcount_reward_func/std": 0.0,
      "step": 3,
      "total_flos": 0.0,
      "train_loss": 0.06536803146203359,
      "train_runtime": 3637.9329,
      "train_samples_per_second": 0.007,
      "train_steps_per_second": 0.001
    }
  ],
  "logging_steps": 2,
  "max_steps": 3,
  "num_input_tokens_seen": 10918,
  "num_train_epochs": 1,
  "save_steps": 25,
  "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": 2,
  "trial_name": null,
  "trial_params": null
}