{
  "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": 0.3873046040534973,
      "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": 608.0756406784058,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.40932849049568176,
      "kl": 0.0003732144832611084,
      "learning_rate": 2.5e-06,
      "loss": 0.0205,
      "reward": 0.591238864697516,
      "reward_std": 0.3592316168360412,
      "rewards/accuracy_reward": 0.591238864697516,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 611.1538192749024,
      "epoch": 0.17057569296375266,
      "grad_norm": 1.189740777015686,
      "kl": 0.01576576232910156,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0602,
      "reward": 0.6685268126428128,
      "reward_std": 0.3173162743449211,
      "rewards/accuracy_reward": 0.6685268126428128,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 604.9509216308594,
      "epoch": 0.255863539445629,
      "grad_norm": 0.22091668844223022,
      "kl": 0.006317520141601562,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0681,
      "reward": 0.7582589656114578,
      "reward_std": 0.2310322556644678,
      "rewards/accuracy_reward": 0.7582589656114578,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.3093933105469,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.23042617738246918,
      "kl": 0.003882598876953125,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0576,
      "reward": 0.7727678909897804,
      "reward_std": 0.19997056629508733,
      "rewards/accuracy_reward": 0.7727678909897804,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.6232421875,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.24932904541492462,
      "kl": 0.00549468994140625,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.037,
      "reward": 0.7665178909897804,
      "reward_std": 0.17505639586597682,
      "rewards/accuracy_reward": 0.7665178909897804,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.7830589294433,
      "epoch": 0.511727078891258,
      "grad_norm": 0.12253822386264801,
      "kl": 0.003799247741699219,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0402,
      "reward": 0.7544643223285675,
      "reward_std": 0.1821566427126527,
      "rewards/accuracy_reward": 0.7544643223285675,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 591.4761390686035,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.21455420553684235,
      "kl": 0.007158660888671875,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0349,
      "reward": 0.7428571835160256,
      "reward_std": 0.17305572014302015,
      "rewards/accuracy_reward": 0.7428571835160256,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.2694511413574,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.1763172149658203,
      "kl": 0.0064422607421875,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0448,
      "reward": 0.7700893267989158,
      "reward_std": 0.18249586317688227,
      "rewards/accuracy_reward": 0.7700893267989158,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.8049400329589,
      "epoch": 0.767590618336887,
      "grad_norm": 0.13369952142238617,
      "kl": 0.006565475463867187,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0355,
      "reward": 0.7477678924798965,
      "reward_std": 0.18774770591408013,
      "rewards/accuracy_reward": 0.7477678924798965,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.1364059448242,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.2092757225036621,
      "kl": 0.012014389038085938,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0392,
      "reward": 0.745312537252903,
      "reward_std": 0.18103930186480283,
      "rewards/accuracy_reward": 0.745312537252903,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.1323883056641,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.11179689317941666,
      "kl": 0.011909866333007812,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0323,
      "reward": 0.7830357506871224,
      "reward_std": 0.1971761178225279,
      "rewards/accuracy_reward": 0.7830357506871224,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.4685948689779,
      "epoch": 0.9893390191897654,
      "kl": 0.010700225830078125,
      "reward": 0.772693487505118,
      "reward_std": 0.18821328474829593,
      "rewards/accuracy_reward": 0.772693487505118,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.04097457230893959,
      "train_runtime": 12199.579,
      "train_samples_per_second": 0.615,
      "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
}