{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6153846153846154,
  "eval_steps": 500,
  "global_step": 48,
  "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": 0.0943359375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3681.1,
      "completions/mean_length": 1021.4728515625,
      "completions/mean_terminated_length": 701.1630737304688,
      "completions/min_length": 103.7,
      "completions/min_terminated_length": 103.7,
      "entropy": 0.17240249924361706,
      "epoch": 0.1282051282051282,
      "frac_reward_zero_std": 0.0515625,
      "grad_norm": 0.00012006821180025731,
      "learning_rate": 5.625e-06,
      "loss": -0.0001,
      "num_tokens": 10714829.0,
      "reward": -0.35699094235897066,
      "reward_std": 0.3353965669870377,
      "rewards/grpo_reward_function/mean": -0.35699094235897066,
      "rewards/grpo_reward_function/std": 0.7091993272304535,
      "sampling/importance_sampling_ratio/max": 0.1970108050853014,
      "sampling/importance_sampling_ratio/mean": 0.0012754969735397025,
      "sampling/importance_sampling_ratio/min": 0.0,
      "sampling/sampling_logp_difference/max": 1.831837797164917,
      "sampling/sampling_logp_difference/mean": 0.0368979848921299,
      "step": 10,
      "step_time": 179.07797421738505
    },
    {
      "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": 0.078125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3652.6,
      "completions/mean_length": 961.977734375,
      "completions/mean_terminated_length": 696.3083312988281,
      "completions/min_length": 86.3,
      "completions/min_terminated_length": 86.3,
      "entropy": 0.17260038182139398,
      "epoch": 0.2564102564102564,
      "frac_reward_zero_std": 0.04453125,
      "grad_norm": 0.00011115181306767518,
      "learning_rate": 9.942341621640558e-06,
      "loss": -0.0003,
      "num_tokens": 21190031.0,
      "reward": -0.3344299763441086,
      "reward_std": 0.32805776298046113,
      "rewards/grpo_reward_function/mean": -0.3344299763441086,
      "rewards/grpo_reward_function/std": 0.7096675753593444,
      "sampling/importance_sampling_ratio/max": 0.21866981824859977,
      "sampling/importance_sampling_ratio/mean": 0.0013687322207260877,
      "sampling/importance_sampling_ratio/min": 0.0,
      "sampling/sampling_logp_difference/max": 1.826985728740692,
      "sampling/sampling_logp_difference/mean": 0.03691292852163315,
      "step": 20,
      "step_time": 178.9407274030149
    },
    {
      "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": 0.0720703125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3748.0,
      "completions/mean_length": 916.4140625,
      "completions/mean_terminated_length": 670.6310546875,
      "completions/min_length": 91.7,
      "completions/min_terminated_length": 91.7,
      "entropy": 0.17158033028244973,
      "epoch": 0.38461538461538464,
      "frac_reward_zero_std": 0.046875,
      "grad_norm": 0.00029510410537975974,
      "learning_rate": 8.953878684688492e-06,
      "loss": -0.0,
      "num_tokens": 31354631.0,
      "reward": -0.3493777960538864,
      "reward_std": 0.3407163262367249,
      "rewards/grpo_reward_function/mean": -0.34937779903411864,
      "rewards/grpo_reward_function/std": 0.7120605766773224,
      "sampling/importance_sampling_ratio/max": 0.3080031927675009,
      "sampling/importance_sampling_ratio/mean": 0.002376599394483492,
      "sampling/importance_sampling_ratio/min": 0.0,
      "sampling/sampling_logp_difference/max": 2.0400707840919496,
      "sampling/sampling_logp_difference/mean": 0.03684176206588745,
      "step": 30,
      "step_time": 171.81590754203498
    },
    {
      "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": 0.0662109375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3509.1,
      "completions/mean_length": 912.234375,
      "completions/mean_terminated_length": 686.69970703125,
      "completions/min_length": 83.6,
      "completions/min_terminated_length": 83.6,
      "entropy": 0.16810744479298592,
      "epoch": 0.5128205128205128,
      "frac_reward_zero_std": 0.06328125,
      "grad_norm": 1.6541331856452166e-05,
      "learning_rate": 6.971779275566593e-06,
      "loss": -0.0001,
      "num_tokens": 41526511.0,
      "reward": -0.3015082895755768,
      "reward_std": 0.3079641282558441,
      "rewards/grpo_reward_function/mean": -0.301508292555809,
      "rewards/grpo_reward_function/std": 0.729301530122757,
      "sampling/importance_sampling_ratio/max": 0.3999401656910777,
      "sampling/importance_sampling_ratio/mean": 0.003650264849420637,
      "sampling/importance_sampling_ratio/min": 1.04037442665946e-40,
      "sampling/sampling_logp_difference/max": 1.7756866335868835,
      "sampling/sampling_logp_difference/mean": 0.0364519115537405,
      "step": 40,
      "step_time": 178.55749775655568
    }
  ],
  "logging_steps": 10,
  "max_steps": 78,
  "num_input_tokens_seen": 49637699,
  "num_train_epochs": 1,
  "save_steps": 16,
  "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": 16,
  "trial_name": null,
  "trial_params": null
}