{
  "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": 0.39137744903564453,
      "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": 607.5834541320801,
      "epoch": 0.08528784648187633,
      "grad_norm": 1.0777415037155151,
      "kl": 0.00038692355155944824,
      "learning_rate": 2.5e-06,
      "loss": 0.0259,
      "reward": 0.5962611874565482,
      "reward_std": 0.36695866473019123,
      "rewards/accuracy_reward": 0.5962611874565482,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.7803886413574,
      "epoch": 0.17057569296375266,
      "grad_norm": 1.1746565103530884,
      "kl": 0.010793495178222656,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.048,
      "reward": 0.6758928880095482,
      "reward_std": 0.30583293717354537,
      "rewards/accuracy_reward": 0.6758928880095482,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 602.4096214294434,
      "epoch": 0.255863539445629,
      "grad_norm": 0.1635478436946869,
      "kl": 0.008692169189453125,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0619,
      "reward": 0.755357176065445,
      "reward_std": 0.22135366648435592,
      "rewards/accuracy_reward": 0.755357176065445,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.5116325378418,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.3694714605808258,
      "kl": 0.00950794219970703,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0543,
      "reward": 0.7654018238186836,
      "reward_std": 0.2084361480548978,
      "rewards/accuracy_reward": 0.7654018238186836,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.0078392028809,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.21624431014060974,
      "kl": 0.004351234436035157,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0513,
      "reward": 0.7598214641213417,
      "reward_std": 0.19899208266288043,
      "rewards/accuracy_reward": 0.7598214641213417,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.0727897644043,
      "epoch": 0.511727078891258,
      "grad_norm": 0.26695239543914795,
      "kl": 0.005370330810546875,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0455,
      "reward": 0.7566964611411094,
      "reward_std": 0.18958494029939174,
      "rewards/accuracy_reward": 0.7566964611411094,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.4431076049805,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.12381298094987869,
      "kl": 0.004366302490234375,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0353,
      "reward": 0.747098246216774,
      "reward_std": 0.18222492672502993,
      "rewards/accuracy_reward": 0.747098246216774,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.5830627441406,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.15292133390903473,
      "kl": 0.003928375244140625,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0394,
      "reward": 0.7667411059141159,
      "reward_std": 0.1841550744138658,
      "rewards/accuracy_reward": 0.7667411059141159,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.4895347595215,
      "epoch": 0.767590618336887,
      "grad_norm": 0.2122126966714859,
      "kl": 0.004204940795898437,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0305,
      "reward": 0.7553571775555611,
      "reward_std": 0.1827343128155917,
      "rewards/accuracy_reward": 0.7553571775555611,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.6477951049804,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.10254240781068802,
      "kl": 0.007993698120117188,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0327,
      "reward": 0.7506696820259094,
      "reward_std": 0.19283948028460146,
      "rewards/accuracy_reward": 0.7506696820259094,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 590.7453369140625,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.22994418442249298,
      "kl": 0.00672760009765625,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0323,
      "reward": 0.7854911103844643,
      "reward_std": 0.18603479415178298,
      "rewards/accuracy_reward": 0.7854911103844643,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.9811820983887,
      "epoch": 0.9893390191897654,
      "kl": 0.004169464111328125,
      "reward": 0.7682292014360428,
      "reward_std": 0.18959872238337994,
      "rewards/accuracy_reward": 0.7682292014360428,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.04025471127784715,
      "train_runtime": 12294.0923,
      "train_samples_per_second": 0.61,
      "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
}