{
  "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": 594.6719055175781,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.3901163935661316,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": -0.0015,
      "reward": 0.5937500223517418,
      "reward_std": 0.36258383840322495,
      "rewards/accuracy_reward": 0.5937500223517418,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 609.8289909362793,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.5891452431678772,
      "kl": 0.0003406107425689697,
      "learning_rate": 2.5e-06,
      "loss": 0.028,
      "reward": 0.5943080652505159,
      "reward_std": 0.35135408770293,
      "rewards/accuracy_reward": 0.5943080652505159,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 622.6977935791016,
      "epoch": 0.17057569296375266,
      "grad_norm": 4.026884078979492,
      "kl": 0.009809780120849609,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0682,
      "reward": 0.6770089581608772,
      "reward_std": 0.3084997434169054,
      "rewards/accuracy_reward": 0.6770089581608772,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 604.310961151123,
      "epoch": 0.255863539445629,
      "grad_norm": 0.44674399495124817,
      "kl": 30.434170150756835,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 1.5901,
      "reward": 0.763392886519432,
      "reward_std": 0.21235014032572508,
      "rewards/accuracy_reward": 0.763392886519432,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.2562767028809,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.13691851496696472,
      "kl": 0.004651832580566406,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0554,
      "reward": 0.7613839641213417,
      "reward_std": 0.21757735572755338,
      "rewards/accuracy_reward": 0.7613839641213417,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.6230163574219,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.20785608887672424,
      "kl": 0.02872943878173828,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0575,
      "reward": 0.7502232536673545,
      "reward_std": 0.2002022891305387,
      "rewards/accuracy_reward": 0.7502232536673545,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.0156471252442,
      "epoch": 0.511727078891258,
      "grad_norm": 0.1560502052307129,
      "kl": 0.00396270751953125,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0352,
      "reward": 0.7459821790456772,
      "reward_std": 0.18388189654797316,
      "rewards/accuracy_reward": 0.7459821790456772,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.2908729553222,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.11733976751565933,
      "kl": 0.04864578247070313,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0487,
      "reward": 0.7290178924798966,
      "reward_std": 0.19413371523842216,
      "rewards/accuracy_reward": 0.7290178924798966,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.8761444091797,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.21926052868366241,
      "kl": 0.005985260009765625,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.049,
      "reward": 0.7500000327825547,
      "reward_std": 0.19939510114490985,
      "rewards/accuracy_reward": 0.7500000327825547,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.2828376770019,
      "epoch": 0.767590618336887,
      "grad_norm": 0.22137849032878876,
      "kl": 0.004445266723632812,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.043,
      "reward": 0.7392857477068902,
      "reward_std": 0.21266973707824946,
      "rewards/accuracy_reward": 0.7392857477068902,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 607.684846496582,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.12936115264892578,
      "kl": 0.010895919799804688,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0298,
      "reward": 0.7388393193483352,
      "reward_std": 0.19434305084869266,
      "rewards/accuracy_reward": 0.7388393193483352,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.9553848266602,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.17488008737564087,
      "kl": 0.0073909759521484375,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0423,
      "reward": 0.7638393193483353,
      "reward_std": 0.2086773581802845,
      "rewards/accuracy_reward": 0.7638393193483353,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.5217819213867,
      "epoch": 0.9893390191897654,
      "kl": 0.0069802602132161455,
      "reward": 0.7615327710906664,
      "reward_std": 0.1866085703174273,
      "rewards/accuracy_reward": 0.7615327710906664,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.1775998589502455,
      "train_runtime": 12344.1744,
      "train_samples_per_second": 0.608,
      "train_steps_per_second": 0.005
    }
  ],
  "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
}