{
  "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": [
    {
      "completion_length": 608.3236770629883,
      "epoch": 0.017057569296375266,
      "grad_norm": 10.999775886535645,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0,
      "reward": 0.6250000298023224,
      "reward_std": 0.3478057999163866,
      "rewards/accuracy_reward": 0.6250000298023224,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "completion_length": 600.2413778305054,
      "epoch": 0.08528784648187633,
      "grad_norm": 5.341479778289795,
      "kl": 0.000471651554107666,
      "learning_rate": 2.5e-06,
      "loss": 0.0,
      "reward": 0.6138393152505159,
      "reward_std": 0.3516102097928524,
      "rewards/accuracy_reward": 0.6138393152505159,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "completion_length": 620.4908721923828,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.668684184551239,
      "kl": 0.008291149139404297,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0003,
      "reward": 0.6718750268220901,
      "reward_std": 0.31286490224301816,
      "rewards/accuracy_reward": 0.6718750268220901,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "completion_length": 628.9826179504395,
      "epoch": 0.255863539445629,
      "grad_norm": 0.13771334290504456,
      "kl": 0.005892181396484375,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0002,
      "reward": 0.7491071790456771,
      "reward_std": 0.23660710267722607,
      "rewards/accuracy_reward": 0.7491071790456771,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "completion_length": 618.5221313476562,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.15839262306690216,
      "kl": 0.0176788330078125,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0007,
      "reward": 0.7529018208384514,
      "reward_std": 0.22067742738872767,
      "rewards/accuracy_reward": 0.7529018208384514,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "completion_length": 614.5660995483398,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.11253199726343155,
      "kl": 0.007561492919921875,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0003,
      "reward": 0.7607143178582192,
      "reward_std": 0.1883007898926735,
      "rewards/accuracy_reward": 0.7607143178582192,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "completion_length": 613.0859634399415,
      "epoch": 0.511727078891258,
      "grad_norm": 0.8198254704475403,
      "kl": 0.005454635620117188,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0002,
      "reward": 0.7500000327825547,
      "reward_std": 0.19777565877884626,
      "rewards/accuracy_reward": 0.7500000327825547,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 604.8518127441406,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.24099703133106232,
      "kl": 0.005191421508789063,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0002,
      "reward": 0.7433036074042321,
      "reward_std": 0.18659685049206018,
      "rewards/accuracy_reward": 0.7433036074042321,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "completion_length": 591.693775177002,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.14078277349472046,
      "kl": 0.005743408203125,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0002,
      "reward": 0.7678571730852127,
      "reward_std": 0.1761675203219056,
      "rewards/accuracy_reward": 0.7678571730852127,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "completion_length": 607.4225692749023,
      "epoch": 0.767590618336887,
      "grad_norm": 0.26444485783576965,
      "kl": 0.0046844482421875,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0002,
      "reward": 0.7575893178582191,
      "reward_std": 0.1912793930619955,
      "rewards/accuracy_reward": 0.7575893178582191,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "completion_length": 610.6031555175781,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.2796413004398346,
      "kl": 0.004559326171875,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0002,
      "reward": 0.7430803939700127,
      "reward_std": 0.19141427483409643,
      "rewards/accuracy_reward": 0.7430803939700127,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "completion_length": 599.6714477539062,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.08973807841539383,
      "kl": 0.03614883422851563,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0014,
      "reward": 0.7792411088943482,
      "reward_std": 0.18675131760537625,
      "rewards/accuracy_reward": 0.7792411088943482,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "completion_length": 594.2611796061198,
      "epoch": 0.9893390191897654,
      "kl": 0.0211181640625,
      "reward": 0.7667411044239998,
      "reward_std": 0.16868622973561287,
      "rewards/accuracy_reward": 0.7667411044239998,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.0008640994389406744,
      "train_runtime": 7436.3725,
      "train_samples_per_second": 1.009,
      "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
}