{
  "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": 0.3905457556247711,
      "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": 600.2712335586548,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.7190936207771301,
      "kl": 0.00037989020347595215,
      "learning_rate": 2.5e-06,
      "loss": 0.0095,
      "reward": 0.5926339542493224,
      "reward_std": 0.3625427111983299,
      "rewards/accuracy_reward": 0.5926339542493224,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 609.6085060119628,
      "epoch": 0.17057569296375266,
      "grad_norm": 183480.84375,
      "kl": 5.480923271179199,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.3253,
      "reward": 0.6698661014437676,
      "reward_std": 0.3133059959858656,
      "rewards/accuracy_reward": 0.6698661014437676,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.4218978881836,
      "epoch": 0.255863539445629,
      "grad_norm": 0.16699491441249847,
      "kl": 0.009376716613769532,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0661,
      "reward": 0.7607143208384514,
      "reward_std": 0.21833304818719626,
      "rewards/accuracy_reward": 0.7607143208384514,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.3482398986816,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.2864713966846466,
      "kl": 0.007395362854003907,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0638,
      "reward": 0.7571428924798965,
      "reward_std": 0.21090923752635718,
      "rewards/accuracy_reward": 0.7571428924798965,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.333283996582,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.24418725073337555,
      "kl": 0.004486083984375,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0438,
      "reward": 0.7598214641213417,
      "reward_std": 0.18390203788876533,
      "rewards/accuracy_reward": 0.7598214641213417,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 600.6495803833008,
      "epoch": 0.511727078891258,
      "grad_norm": 0.14322561025619507,
      "kl": 0.005170440673828125,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0419,
      "reward": 0.7488839626312256,
      "reward_std": 0.20049505326896905,
      "rewards/accuracy_reward": 0.7488839626312256,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.189981842041,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.10154467821121216,
      "kl": 0.004907989501953125,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0362,
      "reward": 0.7575893223285675,
      "reward_std": 0.17536313086748123,
      "rewards/accuracy_reward": 0.7575893223285675,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.7109580993653,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.1248159408569336,
      "kl": 82.40533447265625,
      "learning_rate": 8.029152419343472e-07,
      "loss": 9.2884,
      "reward": 0.7660714700818062,
      "reward_std": 0.1841113742440939,
      "rewards/accuracy_reward": 0.7660714700818062,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 602.1176597595215,
      "epoch": 0.767590618336887,
      "grad_norm": 0.11193355917930603,
      "kl": 0.005156707763671875,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0337,
      "reward": 0.7497768238186836,
      "reward_std": 0.18463366031646727,
      "rewards/accuracy_reward": 0.7497768238186836,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.8102981567383,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.23557083308696747,
      "kl": 0.006516265869140625,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0351,
      "reward": 0.742857177555561,
      "reward_std": 0.18797835633158683,
      "rewards/accuracy_reward": 0.742857177555561,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.9042694091797,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.1702045351266861,
      "kl": 0.06582679748535156,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.04,
      "reward": 0.7823661074042321,
      "reward_std": 0.19186323396861554,
      "rewards/accuracy_reward": 0.7823661074042321,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 585.5672543843588,
      "epoch": 0.9893390191897654,
      "kl": 0.008692423502604166,
      "reward": 0.7600446765621504,
      "reward_std": 0.18551702642192444,
      "rewards/accuracy_reward": 0.7600446765621504,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.8614622223560281,
      "train_runtime": 12252.8253,
      "train_samples_per_second": 0.612,
      "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
}