{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.00582433796691776,
  "eval_steps": 500,
  "global_step": 300,
  "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
    },
    {
      "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": 0.00291216898345888,
      "frac_reward_zero_std": 0.66,
      "grad_norm": 0.02328324504196644,
      "kl": 0.2890375615656376,
      "learning_rate": 4.426283106939474e-06,
      "loss": 0.0003,
      "num_tokens": 666000.0,
      "reward": 0.425,
      "reward_std": 0.17928203105926513,
      "rewards/compute_reward/mean": 0.425,
      "rewards/compute_reward/std": 0.3971991038322449,
      "step": 150
    },
    {
      "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": 0.0038828919779451735,
      "frac_reward_zero_std": 0.62,
      "grad_norm": 5.6864848136901855,
      "kl": 0.496561706662178,
      "learning_rate": 3.7650844103029093e-06,
      "loss": 0.0005,
      "num_tokens": 888000.0,
      "reward": 0.475,
      "reward_std": 0.19928203105926515,
      "rewards/compute_reward/mean": 0.475,
      "rewards/compute_reward/std": 0.3720254868268967,
      "step": 200
    },
    {
      "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": 0.004853614972431467,
      "frac_reward_zero_std": 0.57,
      "grad_norm": 0.014993643388152122,
      "kl": 0.3951701292395592,
      "learning_rate": 2.9512978631264006e-06,
      "loss": 0.0004,
      "num_tokens": 1110000.0,
      "reward": 0.5025,
      "reward_std": 0.2273760414123535,
      "rewards/compute_reward/mean": 0.5025,
      "rewards/compute_reward/std": 0.4330417436361313,
      "step": 250
    },
    {
      "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": 0.00582433796691776,
      "frac_reward_zero_std": 0.67,
      "grad_norm": 7.204345703125,
      "kl": 0.5356883952021598,
      "learning_rate": 2.0830781332097446e-06,
      "loss": 0.0005,
      "num_tokens": 1332000.0,
      "reward": 0.3975,
      "reward_std": 0.17118802070617675,
      "rewards/compute_reward/mean": 0.3975,
      "rewards/compute_reward/std": 0.3806261843442917,
      "step": 300
    }
  ],
  "logging_steps": 50,
  "max_steps": 500,
  "num_input_tokens_seen": 1332000,
  "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
}