{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9893390191897654,
  "eval_steps": 500,
  "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": 1024.0,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.09009335935115814,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0,
      "reward": 0.2321428693830967,
      "reward_std": 0.3139922395348549,
      "rewards/accuracy_reward": 0.2321428693830967,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.07332542538642883,
      "kl": 9.924173355102539e-05,
      "learning_rate": 2.5e-06,
      "loss": 0.0,
      "reward": 0.22935269074514508,
      "reward_std": 0.30691308877430856,
      "rewards/accuracy_reward": 0.22935269074514508,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.0727960467338562,
      "kl": 0.0027547836303710937,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0001,
      "reward": 0.4187500193715096,
      "reward_std": 0.33326763547956945,
      "rewards/accuracy_reward": 0.4187500193715096,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.255863539445629,
      "grad_norm": 0.06532283872365952,
      "kl": 0.17901992797851562,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0072,
      "reward": 0.7390625298023223,
      "reward_std": 0.17226867079734803,
      "rewards/accuracy_reward": 0.7390625298023223,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.08123523741960526,
      "kl": 0.020281982421875,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0008,
      "reward": 0.7357143208384513,
      "reward_std": 0.16036716918461025,
      "rewards/accuracy_reward": 0.7357143208384513,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.42643923240938164,
      "grad_norm": 98653831168.0,
      "kl": 209715200.01299286,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 8388608.0,
      "reward": 0.731919676065445,
      "reward_std": 0.15736169517040252,
      "rewards/accuracy_reward": 0.731919676065445,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.511727078891258,
      "grad_norm": 13029.787109375,
      "kl": 67.636962890625,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 2.7023,
      "reward": 0.7205357491970062,
      "reward_std": 0.1578733256086707,
      "rewards/accuracy_reward": 0.7205357491970062,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.39150699973106384,
      "kl": 2.8228424072265623,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.1129,
      "reward": 0.7111607402563095,
      "reward_std": 0.16545333191752434,
      "rewards/accuracy_reward": 0.7111607402563095,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.6823027718550106,
      "grad_norm": 91.16580963134766,
      "kl": 28502426.75253143,
      "learning_rate": 8.029152419343472e-07,
      "loss": 1146119.4,
      "reward": 0.7203125312924386,
      "reward_std": 0.16900291871279477,
      "rewards/accuracy_reward": 0.7203125312924386,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.767590618336887,
      "grad_norm": 552.0439453125,
      "kl": 116332.38398895264,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 4652.1254,
      "reward": 0.7176339611411094,
      "reward_std": 0.1728838672861457,
      "rewards/accuracy_reward": 0.7176339611411094,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.14540241658687592,
      "kl": 801.6095245361328,
      "learning_rate": 1.718159615201853e-07,
      "loss": 32.0004,
      "reward": 0.7058035969734192,
      "reward_std": 0.1750278390944004,
      "rewards/accuracy_reward": 0.7058035969734192,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.9381663113006397,
      "grad_norm": 4.781561851501465,
      "kl": 22835.21565475464,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 914.4983,
      "reward": 0.7439732491970062,
      "reward_std": 0.16849460527300836,
      "rewards/accuracy_reward": 0.7439732491970062,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.9893390191897654,
      "kl": 1372.5967407226562,
      "reward": 0.7388393208384514,
      "reward_std": 0.16757233006258807,
      "rewards/accuracy_reward": 0.7388393208384514,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 822442.3794969221,
      "train_runtime": 20401.2733,
      "train_samples_per_second": 0.368,
      "train_steps_per_second": 0.003
    }
  ],
  "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
}