{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.8,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 794.7712326049805,
      "epoch": 0.2,
      "grad_norm": 0.1648624688386917,
      "kl": 0.0,
      "learning_rate": 1e-06,
      "loss": 0.0129,
      "reward": 0.1897321529686451,
      "reward_std": 0.19452152773737907,
      "rewards/accuracy_reward": 0.16183036379516125,
      "rewards/format_reward": 0.027901787078008056,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 788.7571487426758,
      "epoch": 1.2,
      "grad_norm": 0.2475416511297226,
      "kl": 0.0007533133029937744,
      "learning_rate": 2.9392394604217463e-06,
      "loss": 0.024,
      "reward": 0.16434152494184673,
      "reward_std": 0.19889171607792377,
      "rewards/accuracy_reward": 0.13504464935977012,
      "rewards/format_reward": 0.029296876542503014,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 780.9437911987304,
      "epoch": 2.4,
      "grad_norm": 0.29467591643333435,
      "kl": 0.030749130249023437,
      "learning_rate": 2.3109612261833968e-06,
      "loss": 0.0394,
      "reward": 0.23705358430743217,
      "reward_std": 0.29603362139314415,
      "rewards/accuracy_reward": 0.13928571976721288,
      "rewards/format_reward": 0.09776786174625159,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 736.5603805541992,
      "epoch": 3.6,
      "grad_norm": 0.3751678168773651,
      "kl": 0.06479034423828126,
      "learning_rate": 1.2865277425900725e-06,
      "loss": 0.0791,
      "reward": 0.4093750208616257,
      "reward_std": 0.43266602158546447,
      "rewards/accuracy_reward": 0.12991072016302496,
      "rewards/format_reward": 0.2794642996042967,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 690.4044540405273,
      "epoch": 4.8,
      "grad_norm": 0.3264864385128021,
      "kl": 0.0665924072265625,
      "learning_rate": 3.6637563846861275e-07,
      "loss": 0.0782,
      "reward": 0.5441964529454708,
      "reward_std": 0.4915049530565739,
      "rewards/accuracy_reward": 0.118750005424954,
      "rewards/format_reward": 0.4254464492201805,
      "step": 20
    },
    {
      "epoch": 4.8,
      "step": 20,
      "total_flos": 0.0,
      "train_loss": 0.054626915324479344,
      "train_runtime": 6323.8598,
      "train_samples_per_second": 0.499,
      "train_steps_per_second": 0.004
    }
  ],
  "logging_steps": 5,
  "max_steps": 25,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}