{
  "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": 606.7053756713867,
      "epoch": 0.017057569296375266,
      "grad_norm": 0.37430307269096375,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0331,
      "reward": 0.623883955180645,
      "reward_std": 0.34485295601189137,
      "rewards/accuracy_reward": 0.623883955180645,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 616.130612373352,
      "epoch": 0.08528784648187633,
      "grad_norm": 1.1676669120788574,
      "kl": 0.000269472599029541,
      "learning_rate": 2.5e-06,
      "loss": 0.0261,
      "reward": 0.593191989697516,
      "reward_std": 0.36569571029394865,
      "rewards/accuracy_reward": 0.593191989697516,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.5042640686036,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.9632540941238403,
      "kl": 0.009089851379394531,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0512,
      "reward": 0.6821428880095481,
      "reward_std": 0.30126961767673494,
      "rewards/accuracy_reward": 0.6821428880095481,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.6660972595215,
      "epoch": 0.255863539445629,
      "grad_norm": 0.14580217003822327,
      "kl": 0.004628181457519531,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0639,
      "reward": 0.765178607404232,
      "reward_std": 0.2118765540421009,
      "rewards/accuracy_reward": 0.765178607404232,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.9198951721191,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.4212934970855713,
      "kl": 0.053232765197753905,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0517,
      "reward": 0.7651786103844642,
      "reward_std": 0.19744985159486533,
      "rewards/accuracy_reward": 0.7651786103844642,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.8506958007813,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.18875020742416382,
      "kl": 0.004035186767578125,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0467,
      "reward": 0.7680803954601287,
      "reward_std": 0.1883797325193882,
      "rewards/accuracy_reward": 0.7680803954601287,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.9201133728027,
      "epoch": 0.511727078891258,
      "grad_norm": 0.11470279842615128,
      "kl": 0.00387420654296875,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0429,
      "reward": 0.7524553909897804,
      "reward_std": 0.17808201136067509,
      "rewards/accuracy_reward": 0.7524553909897804,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 590.4598487854004,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.2764819860458374,
      "kl": 0.0034837722778320312,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0316,
      "reward": 0.7455357491970063,
      "reward_std": 0.1789832441136241,
      "rewards/accuracy_reward": 0.7455357491970063,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.9904289245605,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.10016006976366043,
      "kl": 0.004136276245117187,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0423,
      "reward": 0.7671875342726707,
      "reward_std": 0.18496839571744203,
      "rewards/accuracy_reward": 0.7671875342726707,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.3118553161621,
      "epoch": 0.767590618336887,
      "grad_norm": 0.9257429242134094,
      "kl": 0.00593109130859375,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0291,
      "reward": 0.7488839641213417,
      "reward_std": 0.19455855945125222,
      "rewards/accuracy_reward": 0.7488839641213417,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.4419929504395,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.24372941255569458,
      "kl": 0.004146194458007813,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0398,
      "reward": 0.7470982506871223,
      "reward_std": 0.17433594232425093,
      "rewards/accuracy_reward": 0.7470982506871223,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.7348480224609,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.08758081495761871,
      "kl": 0.007670402526855469,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.035,
      "reward": 0.783928607404232,
      "reward_std": 0.18272702796384693,
      "rewards/accuracy_reward": 0.783928607404232,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 580.6320686340332,
      "epoch": 0.9893390191897654,
      "kl": 0.0042463938395182295,
      "reward": 0.7734375397364298,
      "reward_std": 0.18152949400246143,
      "rewards/accuracy_reward": 0.7734375397364298,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.04230335579606993,
      "train_runtime": 7223.2167,
      "train_samples_per_second": 1.038,
      "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
}