{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.15458937198067632,
  "eval_steps": 2,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 653.5100730895996,
      "epoch": 0.015458937198067632,
      "grad_norm": 13904.753479263767,
      "kl": 9.926961135864257,
      "learning_rate": 2e-05,
      "loss": 0.3968,
      "reward": 0.6000000268220902,
      "reward_std": 0.33480377998203037,
      "rewards/accuracy_reward": 0.5995535984635353,
      "rewards/format_reward": 0.00044642859138548373,
      "step": 5
    },
    {
      "completion_length": 660.7125312805176,
      "epoch": 0.030917874396135265,
      "grad_norm": 197221.6943728757,
      "kl": 1280566.1007873535,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 51005.4313,
      "reward": 0.6585937816649675,
      "reward_std": 0.2747358553111553,
      "rewards/accuracy_reward": 0.6585937816649675,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "completion_length": 658.5587333679199,
      "epoch": 0.0463768115942029,
      "grad_norm": 5.735340194391574,
      "kl": 16.464306640625,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.6587,
      "reward": 0.6368303872644901,
      "reward_std": 0.2583037616685033,
      "rewards/accuracy_reward": 0.6368303872644901,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "completion_length": 665.5614128112793,
      "epoch": 0.06183574879227053,
      "grad_norm": 0.11683844185999785,
      "kl": 0.7914215087890625,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.0317,
      "reward": 0.6370535973459482,
      "reward_std": 0.25866390075534584,
      "rewards/accuracy_reward": 0.6370535973459482,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 638.1945602416993,
      "epoch": 0.07729468599033816,
      "grad_norm": 0.07460429310613971,
      "kl": 0.0471527099609375,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.0019,
      "reward": 0.6760044947266579,
      "reward_std": 0.2249570596963167,
      "rewards/accuracy_reward": 0.6760044947266579,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "completion_length": 623.0960060119629,
      "epoch": 0.0927536231884058,
      "grad_norm": 0.07393904563968409,
      "kl": 0.043280029296875,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.0017,
      "reward": 0.693080386891961,
      "reward_std": 0.219588840380311,
      "rewards/accuracy_reward": 0.693080386891961,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 607.3904300689697,
      "epoch": 0.10821256038647344,
      "grad_norm": 0.07179972285540803,
      "kl": 0.0819915771484375,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.0033,
      "reward": 0.7189732462167739,
      "reward_std": 0.2072824534960091,
      "rewards/accuracy_reward": 0.7189732462167739,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "completion_length": 601.369002532959,
      "epoch": 0.12367149758454106,
      "grad_norm": 0.0620520373411285,
      "kl": 0.03866424560546875,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.0015,
      "reward": 0.7085937805473804,
      "reward_std": 0.209117239061743,
      "rewards/accuracy_reward": 0.7085937805473804,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 607.5365215301514,
      "epoch": 0.1391304347826087,
      "grad_norm": 0.833460229793663,
      "kl": 0.03985748291015625,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.0016,
      "reward": 0.700892886519432,
      "reward_std": 0.2034242927096784,
      "rewards/accuracy_reward": 0.700892886519432,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "completion_length": 611.2590663909912,
      "epoch": 0.15458937198067632,
      "grad_norm": 0.06368324518689576,
      "kl": 0.03762359619140625,
      "learning_rate": 0.0,
      "loss": 0.0015,
      "reward": 0.7181919991970063,
      "reward_std": 0.21190554238855838,
      "rewards/accuracy_reward": 0.7181919991970063,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "epoch": 0.15458937198067632,
      "step": 50,
      "total_flos": 0.0,
      "train_loss": 5100.652998053767,
      "train_runtime": 11710.1509,
      "train_samples_per_second": 0.956,
      "train_steps_per_second": 0.004
    }
  ],
  "logging_steps": 5,
  "max_steps": 50,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 2,
  "trial_name": null,
  "trial_params": null
}