{
  "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": 605.6372985839844,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.5198791027069092,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0297,
      "reward": 0.6272321715950966,
      "reward_std": 0.3432547841221094,
      "rewards/accuracy_reward": 0.6272321715950966,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.5561113357544,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.3226652443408966,
      "kl": 0.0007017850875854492,
      "learning_rate": 2.5e-06,
      "loss": 0.0077,
      "reward": 0.603794670663774,
      "reward_std": 0.35096015920862556,
      "rewards/accuracy_reward": 0.603794670663774,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 608.2205635070801,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.2746102213859558,
      "kl": 0.0061968803405761715,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0503,
      "reward": 0.6727678865194321,
      "reward_std": 0.3072200361639261,
      "rewards/accuracy_reward": 0.6727678865194321,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 600.5703376770019,
      "epoch": 0.255863539445629,
      "grad_norm": 0.15307798981666565,
      "kl": 0.005248260498046875,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0737,
      "reward": 0.758705398440361,
      "reward_std": 0.22059124447405337,
      "rewards/accuracy_reward": 0.758705398440361,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.4640838623047,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.13478825986385345,
      "kl": 0.004410362243652344,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0521,
      "reward": 0.7656250417232513,
      "reward_std": 0.19759314637631178,
      "rewards/accuracy_reward": 0.7656250417232513,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.369002532959,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.13584953546524048,
      "kl": 0.005138778686523437,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0484,
      "reward": 0.7676339626312256,
      "reward_std": 0.17082388242706656,
      "rewards/accuracy_reward": 0.7676339626312256,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 598.1558288574219,
      "epoch": 0.511727078891258,
      "grad_norm": 0.09639401733875275,
      "kl": 0.003318023681640625,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0395,
      "reward": 0.7537946760654449,
      "reward_std": 0.17525787679478527,
      "rewards/accuracy_reward": 0.7537946760654449,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.4710098266602,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.10184694081544876,
      "kl": 0.0056324005126953125,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0295,
      "reward": 0.7473214656114578,
      "reward_std": 0.1758826152421534,
      "rewards/accuracy_reward": 0.7473214656114578,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 579.4178863525391,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.0822824090719223,
      "kl": 0.005002212524414062,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0407,
      "reward": 0.7607143178582192,
      "reward_std": 0.17344756741076708,
      "rewards/accuracy_reward": 0.7607143178582192,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.2201156616211,
      "epoch": 0.767590618336887,
      "grad_norm": 0.08887068182229996,
      "kl": 0.004409027099609375,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0351,
      "reward": 0.7473214611411094,
      "reward_std": 0.18249810487031937,
      "rewards/accuracy_reward": 0.7473214611411094,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.4567222595215,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.09352638572454453,
      "kl": 0.004033279418945312,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0373,
      "reward": 0.747544676065445,
      "reward_std": 0.18660746775567533,
      "rewards/accuracy_reward": 0.747544676065445,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.4038238525391,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.512617290019989,
      "kl": 0.004796600341796875,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0337,
      "reward": 0.7707589581608772,
      "reward_std": 0.18231952451169492,
      "rewards/accuracy_reward": 0.7707589581608772,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 578.9475390116373,
      "epoch": 0.9893390191897654,
      "kl": 0.0038859049479166665,
      "reward": 0.7656250298023224,
      "reward_std": 0.1823850224415461,
      "rewards/accuracy_reward": 0.7656250298023224,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.04243174159963583,
      "train_runtime": 7171.9003,
      "train_samples_per_second": 1.046,
      "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
}