{
  "best_global_step": 10,
  "best_metric": -0.00283643277361989,
  "best_model_checkpoint": null,
  "epoch": 2.96,
  "eval_steps": 5,
  "global_step": 18,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "grad_norm": 0.02929246425628662,
      "learning_rate": 7.777777777777778e-07,
      "loss": 0.0088,
      "step": 5
    },
    {
      "epoch": 0.8,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 246.5,
      "eval_kl": 3.5843253135681154e-05,
      "eval_loss": 0.025880733504891396,
      "eval_num_tokens": 30835.0,
      "eval_reward": 0.0375,
      "eval_reward_std": 0.075,
      "eval_rewards/equation_reward_func": 0.0125,
      "eval_rewards/format_reward_func": 0.025,
      "eval_runtime": 649.2069,
      "eval_samples_per_second": 0.031,
      "eval_steps_per_second": 0.008,
      "step": 5
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0,
      "step": 10
    },
    {
      "epoch": 1.6400000000000001,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 248.4,
      "eval_kl": 3.7136673927307126e-05,
      "eval_loss": -0.00283643277361989,
      "eval_num_tokens": 62155.0,
      "eval_reward": 0.025,
      "eval_reward_std": 0.05,
      "eval_rewards/equation_reward_func": 0.025,
      "eval_rewards/format_reward_func": 0.0,
      "eval_runtime": 648.6534,
      "eval_samples_per_second": 0.031,
      "eval_steps_per_second": 0.008,
      "step": 10
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.024362999945878983,
      "learning_rate": 2.222222222222222e-07,
      "loss": 0.005,
      "step": 15
    },
    {
      "epoch": 2.48,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 252.0,
      "eval_kl": 3.8695335388183596e-05,
      "eval_loss": 0.00723703857511282,
      "eval_num_tokens": 93368.0,
      "eval_reward": 0.0375,
      "eval_reward_std": 0.075,
      "eval_rewards/equation_reward_func": 0.0125,
      "eval_rewards/format_reward_func": 0.025,
      "eval_runtime": 648.8287,
      "eval_samples_per_second": 0.031,
      "eval_steps_per_second": 0.008,
      "step": 15
    }
  ],
  "logging_steps": 5,
  "max_steps": 18,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 15,
  "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": 4,
  "trial_name": null,
  "trial_params": null
}