{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9893390191897654,
  "eval_steps": 500,
  "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": 1023.0747833251953,
      "epoch": 0.017057569296375266,
      "grad_norm": 199.001220703125,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": -0.0012,
      "reward": 1.0267857536673546,
      "reward_std": 0.5617785826325417,
      "rewards/accuracy_reward": 0.6428571715950966,
      "rewards/format_reward": 0.3839285895228386,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.8353843688965,
      "epoch": 0.08528784648187633,
      "grad_norm": 2.830706834793091,
      "kl": 0.36637283861637115,
      "learning_rate": 2.5e-06,
      "loss": 0.0146,
      "reward": 0.9720982648432255,
      "reward_std": 0.5626517869532108,
      "rewards/accuracy_reward": 0.6007254756987095,
      "rewards/format_reward": 0.37137278355658054,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.2163040161133,
      "epoch": 0.17057569296375266,
      "grad_norm": 6271158.0,
      "kl": 191.4503131866455,
      "learning_rate": 2.956412726139078e-06,
      "loss": 7.6701,
      "reward": 1.2633929237723351,
      "reward_std": 0.5011446818709373,
      "rewards/accuracy_reward": 0.6506696730852127,
      "rewards/format_reward": 0.6127232387661934,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.2537994384766,
      "epoch": 0.255863539445629,
      "grad_norm": 6.252068996429443,
      "kl": 0.01535625457763672,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0006,
      "reward": 1.6102679371833801,
      "reward_std": 0.3114002957940102,
      "rewards/accuracy_reward": 0.7345982462167739,
      "rewards/format_reward": 0.8756696864962578,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.5511184692383,
      "epoch": 0.3411513859275053,
      "grad_norm": 36.48051452636719,
      "kl": 402.4727096557617,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 16.1525,
      "reward": 1.686160784959793,
      "reward_std": 0.2836891669780016,
      "rewards/accuracy_reward": 0.7502232521772385,
      "rewards/format_reward": 0.9359375432133674,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.8459838867187,
      "epoch": 0.42643923240938164,
      "grad_norm": 2.089581251144409,
      "kl": 44.92520294189453,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 1.7938,
      "reward": 1.5060268551111222,
      "reward_std": 0.3427631925791502,
      "rewards/accuracy_reward": 0.7473214641213417,
      "rewards/format_reward": 0.7587053954601288,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.6879486083984,
      "epoch": 0.511727078891258,
      "grad_norm": 7.333752155303955,
      "kl": 0.26961822509765626,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0106,
      "reward": 1.4774554312229156,
      "reward_std": 0.3182775381952524,
      "rewards/accuracy_reward": 0.732589314877987,
      "rewards/format_reward": 0.744866107404232,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.8511169433593,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.42886048555374146,
      "kl": 0.18978424072265626,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0073,
      "reward": 1.5037947058677674,
      "reward_std": 0.30398814380168915,
      "rewards/accuracy_reward": 0.7343750342726707,
      "rewards/format_reward": 0.7694196790456772,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.497770690918,
      "epoch": 0.6823027718550106,
      "grad_norm": 1.5065158605575562,
      "kl": 0.02209930419921875,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0001,
      "reward": 1.5272321969270706,
      "reward_std": 0.29541820362210275,
      "rewards/accuracy_reward": 0.7482143193483353,
      "rewards/format_reward": 0.7790178909897805,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.5520126342774,
      "epoch": 0.767590618336887,
      "grad_norm": 0.33261430263519287,
      "kl": 0.47607269287109377,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0184,
      "reward": 1.5303572088479995,
      "reward_std": 0.2931534621864557,
      "rewards/accuracy_reward": 0.7526786044239998,
      "rewards/format_reward": 0.777678607404232,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.6607192993164,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.32912492752075195,
      "kl": 0.01863250732421875,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0002,
      "reward": 1.5263393640518188,
      "reward_std": 0.313052724301815,
      "rewards/accuracy_reward": 0.7390625327825546,
      "rewards/format_reward": 0.7872768223285675,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1024.0,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.36546963453292847,
      "kl": 0.01671295166015625,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0007,
      "reward": 1.5573661357164383,
      "reward_std": 0.29526747241616247,
      "rewards/accuracy_reward": 0.7718750312924385,
      "rewards/format_reward": 0.7854911088943481,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1023.4776814778646,
      "epoch": 0.9893390191897654,
      "kl": 0.011998494466145834,
      "reward": 1.574776868025462,
      "reward_std": 0.2935656998306513,
      "rewards/accuracy_reward": 0.7533482437332472,
      "rewards/format_reward": 0.8214286093910536,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 2.2127183382268916,
      "train_runtime": 20476.5318,
      "train_samples_per_second": 0.366,
      "train_steps_per_second": 0.003
    }
  ],
  "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
}