{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 64,
  "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": 281.7,
      "completions/max_terminated_length": 281.7,
      "completions/mean_length": 173.90625,
      "completions/mean_terminated_length": 173.90625,
      "completions/min_length": 60.1,
      "completions/min_terminated_length": 60.1,
      "entropy": 0.34981602281332014,
      "epoch": 0.3125,
      "frac_reward_zero_std": 0.2,
      "grad_norm": 3.5286595821380615,
      "learning_rate": 8.593749999999999e-07,
      "loss": -0.03454443216323853,
      "num_tokens": 363490.0,
      "reward": -1.9279687214642762,
      "reward_std": 6.467041900753975,
      "rewards/pulse_reward/mean": -1.9279687214642762,
      "rewards/pulse_reward/std": 6.467041908204555,
      "step": 10,
      "step_time": 35.20241980789724,
      "tools/call_frequency": 5.49375,
      "tools/failure_frequency": 0.019680690905079245
    },
    {
      "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": 294.3,
      "completions/max_terminated_length": 294.3,
      "completions/mean_length": 172.56875,
      "completions/mean_terminated_length": 172.56875,
      "completions/min_length": 50.3,
      "completions/min_terminated_length": 50.3,
      "entropy": 0.23517516404390335,
      "epoch": 0.625,
      "frac_reward_zero_std": 0.2625,
      "grad_norm": 3.8192903995513916,
      "learning_rate": 7.031249999999999e-07,
      "loss": 0.01356593519449234,
      "num_tokens": 726552.0,
      "reward": 0.09853125289082527,
      "reward_std": 0.583341383934021,
      "rewards/pulse_reward/mean": 0.09853125289082527,
      "rewards/pulse_reward/std": 0.583341383934021,
      "step": 20,
      "step_time": 39.57589099079487,
      "tools/call_frequency": 6.35,
      "tools/failure_frequency": 0.005996426660567522
    },
    {
      "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": 273.0,
      "completions/max_terminated_length": 273.0,
      "completions/mean_length": 149.115625,
      "completions/mean_terminated_length": 149.115625,
      "completions/min_length": 50.9,
      "completions/min_terminated_length": 50.9,
      "entropy": 0.1797773003578186,
      "epoch": 0.9375,
      "frac_reward_zero_std": 0.0875,
      "grad_norm": 5.692322254180908,
      "learning_rate": 5.46875e-07,
      "loss": 0.06958286762237549,
      "num_tokens": 1082109.0,
      "reward": 0.3879156395792961,
      "reward_std": 0.6661841243505477,
      "rewards/pulse_reward/mean": 0.3879156395792961,
      "rewards/pulse_reward/std": 0.6661841526627541,
      "step": 30,
      "step_time": 30.443527194001945,
      "tools/call_frequency": 5.58125,
      "tools/failure_frequency": 0.005842319130897522
    },
    {
      "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": 115.3,
      "completions/max_terminated_length": 115.3,
      "completions/mean_length": 70.528125,
      "completions/mean_terminated_length": 70.528125,
      "completions/min_length": 38.1,
      "completions/min_terminated_length": 38.1,
      "entropy": 0.1547975329682231,
      "epoch": 1.25,
      "frac_reward_zero_std": 0.1625,
      "grad_norm": 3.677471399307251,
      "learning_rate": 3.9062499999999997e-07,
      "loss": 0.02682075500488281,
      "num_tokens": 1412518.0,
      "reward": 1.0730250239372254,
      "reward_std": 0.7086882412433624,
      "rewards/pulse_reward/mean": 1.0730250239372254,
      "rewards/pulse_reward/std": 0.7086882352828979,
      "step": 40,
      "step_time": 11.361506971600466,
      "tools/call_frequency": 2.215625,
      "tools/failure_frequency": 0.0015384615398943424
    },
    {
      "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": 83.1,
      "completions/max_terminated_length": 83.1,
      "completions/mean_length": 65.421875,
      "completions/mean_terminated_length": 65.421875,
      "completions/min_length": 48.0,
      "completions/min_terminated_length": 48.0,
      "entropy": 0.12452723067253828,
      "epoch": 1.5625,
      "frac_reward_zero_std": 0.6625,
      "grad_norm": 3.609057664871216,
      "learning_rate": 2.3437499999999998e-07,
      "loss": -3.346521407365799e-05,
      "num_tokens": 1741293.0,
      "reward": 1.4852312803268433,
      "reward_std": 0.3715872406959534,
      "rewards/pulse_reward/mean": 1.4852312803268433,
      "rewards/pulse_reward/std": 0.3715872406959534,
      "step": 50,
      "step_time": 8.49522676919878,
      "tools/call_frequency": 2.0,
      "tools/failure_frequency": 0.0
    },
    {
      "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": 86.6,
      "completions/max_terminated_length": 86.6,
      "completions/mean_length": 65.471875,
      "completions/mean_terminated_length": 65.471875,
      "completions/min_length": 53.6,
      "completions/min_terminated_length": 53.6,
      "entropy": 0.11676975060254335,
      "epoch": 1.875,
      "frac_reward_zero_std": 0.8125,
      "grad_norm": 2.142204761505127,
      "learning_rate": 7.812499999999999e-08,
      "loss": -0.00026753861457109453,
      "num_tokens": 2070084.0,
      "reward": 1.538143789768219,
      "reward_std": 0.2553505107760429,
      "rewards/pulse_reward/mean": 1.538143789768219,
      "rewards/pulse_reward/std": 0.25535051375627515,
      "step": 60,
      "step_time": 8.590878555196104,
      "tools/call_frequency": 1.996875,
      "tools/failure_frequency": 0.0
    },
    {
      "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": 74.75,
      "completions/max_terminated_length": 74.75,
      "completions/mean_length": 65.3203125,
      "completions/mean_terminated_length": 65.3203125,
      "completions/min_length": 61.25,
      "completions/min_terminated_length": 61.25,
      "entropy": 0.11779927462339401,
      "epoch": 2.0,
      "frac_reward_zero_std": 0.9375,
      "num_tokens": 2201581.0,
      "reward": 1.57924222946167,
      "reward_std": 0.11742392182350159,
      "rewards/pulse_reward/mean": 1.57924222946167,
      "rewards/pulse_reward/std": 0.11742392182350159,
      "step": 64,
      "step_time": 7.905290340990177,
      "tools/call_frequency": 2.0078125,
      "tools/failure_frequency": 0.0,
      "total_flos": 0.0,
      "train_loss": 0.011853500996949151,
      "train_runtime": 1404.4837,
      "train_samples_per_second": 0.365,
      "train_steps_per_second": 0.046
    }
  ],
  "logging_steps": 10,
  "max_steps": 64,
  "num_input_tokens_seen": 2201581,
  "num_train_epochs": 2,
  "save_steps": 500,
  "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": 8,
  "trial_name": null,
  "trial_params": null
}