{
  "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": 0.39027485251426697,
      "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": 605.248908996582,
      "epoch": 0.08528784648187633,
      "grad_norm": 0.3570092022418976,
      "kl": 0.000306546688079834,
      "learning_rate": 2.5e-06,
      "loss": 0.0194,
      "reward": 0.6010044934228063,
      "reward_std": 0.3602098235860467,
      "rewards/accuracy_reward": 0.6010044934228063,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 615.1498031616211,
      "epoch": 0.17057569296375266,
      "grad_norm": 179.0993194580078,
      "kl": 0.05141096115112305,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.054,
      "reward": 0.6808036074042321,
      "reward_std": 0.29721736945211885,
      "rewards/accuracy_reward": 0.6808036074042321,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.5736854553222,
      "epoch": 0.255863539445629,
      "grad_norm": 0.4736057221889496,
      "kl": 0.9787862777709961,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.1261,
      "reward": 0.7497768238186836,
      "reward_std": 0.22287233918905258,
      "rewards/accuracy_reward": 0.7497768238186836,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.9386444091797,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.17763952910900116,
      "kl": 0.003890800476074219,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0682,
      "reward": 0.7674107491970062,
      "reward_std": 0.20648193154484035,
      "rewards/accuracy_reward": 0.7674107491970062,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.0500282287597,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.10052420943975449,
      "kl": 0.004927825927734375,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0488,
      "reward": 0.7645089656114579,
      "reward_std": 0.18550358079373835,
      "rewards/accuracy_reward": 0.7645089656114579,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.4877532958984,
      "epoch": 0.511727078891258,
      "grad_norm": 0.11159181594848633,
      "kl": 0.0230743408203125,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0407,
      "reward": 0.7553571775555611,
      "reward_std": 0.18435099869966506,
      "rewards/accuracy_reward": 0.7553571775555611,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 584.1875205993653,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.541372537612915,
      "kl": 0.003916549682617188,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0298,
      "reward": 0.7553571790456772,
      "reward_std": 0.17309768963605165,
      "rewards/accuracy_reward": 0.7553571790456772,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 577.2609634399414,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.1626732498407364,
      "kl": 0.018689346313476563,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0389,
      "reward": 0.7714286103844643,
      "reward_std": 0.18289219457656145,
      "rewards/accuracy_reward": 0.7714286103844643,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.4105178833008,
      "epoch": 0.767590618336887,
      "grad_norm": 0.3321472704410553,
      "kl": 0.004584503173828125,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.038,
      "reward": 0.743750037252903,
      "reward_std": 0.18981666448526086,
      "rewards/accuracy_reward": 0.743750037252903,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.6306045532226,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.27629345655441284,
      "kl": 0.025023269653320312,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.041,
      "reward": 0.7352678954601288,
      "reward_std": 0.19160795491188765,
      "rewards/accuracy_reward": 0.7352678954601288,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.5085067749023,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.22553016245365143,
      "kl": 0.023009109497070312,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0393,
      "reward": 0.7868303924798965,
      "reward_std": 0.1940290305763483,
      "rewards/accuracy_reward": 0.7868303924798965,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.0536079406738,
      "epoch": 0.9893390191897654,
      "kl": 0.0041840871175130205,
      "reward": 0.7626488506793976,
      "reward_std": 0.19432423853625855,
      "rewards/accuracy_reward": 0.7626488506793976,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.048550591330397234,
      "train_runtime": 12478.5367,
      "train_samples_per_second": 0.601,
      "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
}