{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9893390191897654,
  "eval_steps": 100,
  "global_step": 58,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 593.9107284545898,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.4269763231277466,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0167,
      "reward": 0.6294643059372902,
      "reward_std": 0.35917505994439125,
      "rewards/accuracy_reward": 0.6283482387661934,
      "rewards/format_reward": 0.0011160714784637094,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.0845737457275,
      "epoch": 0.08528784648187633,
      "grad_norm": 1.2334303855895996,
      "kl": 0.0003300309181213379,
      "learning_rate": 2.5e-06,
      "loss": 0.0188,
      "reward": 0.6121652042493224,
      "reward_std": 0.3490289170295,
      "rewards/accuracy_reward": 0.6116071688011289,
      "rewards/format_reward": 0.0005580357392318547,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 612.8330627441406,
      "epoch": 0.17057569296375266,
      "grad_norm": 4.521842956542969,
      "kl": 0.0027939796447753905,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0577,
      "reward": 0.6854911021888256,
      "reward_std": 0.2840955166146159,
      "rewards/accuracy_reward": 0.6854911021888256,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 604.9821670532226,
      "epoch": 0.255863539445629,
      "grad_norm": 1.5347368717193604,
      "kl": 0.015767669677734374,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.07,
      "reward": 0.7609375312924385,
      "reward_std": 0.21701927129179238,
      "rewards/accuracy_reward": 0.7609375312924385,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.6801628112793,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.6672903895378113,
      "kl": 0.005676651000976562,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0579,
      "reward": 0.7703125327825546,
      "reward_std": 0.20545081831514836,
      "rewards/accuracy_reward": 0.7703125327825546,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.0326194763184,
      "epoch": 0.42643923240938164,
      "grad_norm": 2323443.0,
      "kl": 85.62775764465331,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 3.4627,
      "reward": 0.761160746216774,
      "reward_std": 0.1914736282080412,
      "rewards/accuracy_reward": 0.761160746216774,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.8656478881836,
      "epoch": 0.511727078891258,
      "grad_norm": 31896284.0,
      "kl": 4184.925457000732,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 166.556,
      "reward": 0.7537946760654449,
      "reward_std": 0.1866186775267124,
      "rewards/accuracy_reward": 0.7537946760654449,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.7852882385254,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.18381251394748688,
      "kl": 175321907.9516266,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 6963359.2,
      "reward": 0.7421875357627868,
      "reward_std": 0.18768558986485004,
      "rewards/accuracy_reward": 0.7421875357627868,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.8495811462402,
      "epoch": 0.6823027718550106,
      "grad_norm": 187.1607208251953,
      "kl": 179490.89021034242,
      "learning_rate": 8.029152419343472e-07,
      "loss": 7215.3883,
      "reward": 0.7627232506871223,
      "reward_std": 0.18175358334556221,
      "rewards/accuracy_reward": 0.7627232506871223,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 598.5533767700196,
      "epoch": 0.767590618336887,
      "grad_norm": 170799.046875,
      "kl": 1.5941051483154296,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0995,
      "reward": 0.7484375357627868,
      "reward_std": 0.18508986476808786,
      "rewards/accuracy_reward": 0.7484375357627868,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 602.7493576049804,
      "epoch": 0.8528784648187633,
      "grad_norm": 23.201330184936523,
      "kl": 0.7593372344970704,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.071,
      "reward": 0.7455357477068901,
      "reward_std": 0.1892501600086689,
      "rewards/accuracy_reward": 0.7455357477068901,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 590.3078407287597,
      "epoch": 0.9381663113006397,
      "grad_norm": 1513.896484375,
      "kl": 0.4785778045654297,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.059,
      "reward": 0.7736607506871224,
      "reward_std": 0.185056865029037,
      "rewards/accuracy_reward": 0.7736607506871224,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.8786061604818,
      "epoch": 0.9893390191897654,
      "kl": 2.2987874348958335,
      "reward": 0.77194944024086,
      "reward_std": 0.1801375082383553,
      "rewards/accuracy_reward": 0.77194944024086,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 600926.2988391328,
      "train_runtime": 7660.7166,
      "train_samples_per_second": 0.979,
      "train_steps_per_second": 0.008
    }
  ],
  "logging_steps": 5,
  "max_steps": 58,
  "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": 16,
  "trial_name": null,
  "trial_params": null
}