{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9898666666666667,
  "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": 582.459849357605,
      "epoch": 0.017066666666666667,
      "grad_norm": 0.31589704751968384,
      "learning_rate": 0.0,
      "loss": -0.1186,
      "num_tokens": 653232.0,
      "reward": 0.3515625159488991,
      "reward_std": 0.3448667535558343,
      "rewards/accuracy_reward": 0.35156250034924597,
      "sce_advantage": 29.60546875,
      "sce_advantage_std": 1.6962890625,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 609.3683314323425,
      "epoch": 0.08533333333333333,
      "grad_norm": 0.3136809468269348,
      "learning_rate": 2e-06,
      "loss": -0.1096,
      "num_tokens": 3367024.0,
      "reward": 0.345982157450635,
      "reward_std": 0.342531890142709,
      "rewards/accuracy_reward": 0.34598214336438105,
      "sce_advantage": 29.724609375,
      "sce_advantage_std": 1.6939697265625,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 647.798466873169,
      "epoch": 0.17066666666666666,
      "grad_norm": 0.31411269307136536,
      "learning_rate": 2.9754298604207156e-06,
      "loss": -0.0326,
      "num_tokens": 6928708.0,
      "reward": 0.5000000235857442,
      "reward_std": 0.31452907202765346,
      "rewards/accuracy_reward": 0.500744047272019,
      "sce_advantage": 32.82578125,
      "sce_advantage_std": 1.568798828125,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 762.5181144714355,
      "epoch": 0.256,
      "grad_norm": 0.42305925488471985,
      "learning_rate": 2.8281840384798147e-06,
      "loss": 0.0293,
      "num_tokens": 11004595.0,
      "reward": 0.6053571700351312,
      "reward_std": 0.21672922340221704,
      "rewards/accuracy_reward": 0.605357142095454,
      "sce_advantage": 38.546875,
      "sce_advantage_std": 1.0988525390625,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 825.0319566726685,
      "epoch": 0.3413333333333333,
      "grad_norm": 0.12070361524820328,
      "learning_rate": 2.5606601717798212e-06,
      "loss": 0.0629,
      "num_tokens": 15349561.0,
      "reward": 0.6325893149245531,
      "reward_std": 0.21432666098698974,
      "rewards/accuracy_reward": 0.6325892861466855,
      "sce_advantage": 41.025,
      "sce_advantage_std": 1.011767578125,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 832.1634309768676,
      "epoch": 0.4266666666666667,
      "grad_norm": 0.11953950673341751,
      "learning_rate": 2.1970847580656528e-06,
      "loss": 0.0529,
      "num_tokens": 19747280.0,
      "reward": 0.6272321719443426,
      "reward_std": 0.20688999248668552,
      "rewards/accuracy_reward": 0.6272321433527395,
      "sce_advantage": 42.6671875,
      "sce_advantage_std": 0.94627685546875,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 824.6692317962646,
      "epoch": 0.512,
      "grad_norm": 0.12637147307395935,
      "learning_rate": 1.7703825567208588e-06,
      "loss": 0.0526,
      "num_tokens": 24100554.0,
      "reward": 0.6368303868919611,
      "reward_std": 0.19028520658612252,
      "rewards/accuracy_reward": 0.6368303560186177,
      "sce_advantage": 43.525,
      "sce_advantage_std": 0.9552490234375,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 848.5489213943481,
      "epoch": 0.5973333333333334,
      "grad_norm": 0.12432567030191422,
      "learning_rate": 1.3191949796170155e-06,
      "loss": 0.0577,
      "num_tokens": 28565891.0,
      "reward": 0.5912946681957691,
      "reward_std": 0.19803369138389826,
      "rewards/accuracy_reward": 0.591294644260779,
      "sce_advantage": 43.93125,
      "sce_advantage_std": 0.91083984375,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 837.3781648635864,
      "epoch": 0.6826666666666666,
      "grad_norm": 0.5495607256889343,
      "learning_rate": 8.843807918208651e-07,
      "loss": 0.0474,
      "num_tokens": 32990269.0,
      "reward": 0.6285714600235224,
      "reward_std": 0.1882201772648841,
      "rewards/accuracy_reward": 0.6285714281257242,
      "sce_advantage": 44.7953125,
      "sce_advantage_std": 0.8768310546875,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 813.5683387756347,
      "epoch": 0.768,
      "grad_norm": 0.13694342970848083,
      "learning_rate": 5.053160126388076e-07,
      "loss": 0.0488,
      "num_tokens": 37331289.0,
      "reward": 0.6265625279862433,
      "reward_std": 0.18500199196860195,
      "rewards/accuracy_reward": 0.6265624999534338,
      "sce_advantage": 45.015625,
      "sce_advantage_std": 0.9173583984375,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 829.90628490448,
      "epoch": 0.8533333333333334,
      "grad_norm": 0.6106674671173096,
      "learning_rate": 2.163280915478289e-07,
      "loss": 0.0456,
      "num_tokens": 41709642.0,
      "reward": 0.6316964592551813,
      "reward_std": 0.21558557036332787,
      "rewards/accuracy_reward": 0.6316964287543669,
      "sce_advantage": 45.315625,
      "sce_advantage_std": 0.8802734375,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 836.8087438583374,
      "epoch": 0.9386666666666666,
      "grad_norm": 0.17513859272003174,
      "learning_rate": 4.358727386092198e-08,
      "loss": 0.0335,
      "num_tokens": 46128522.0,
      "reward": 0.5850446717813611,
      "reward_std": 0.20032540340907873,
      "rewards/accuracy_reward": 0.5865327379666269,
      "sce_advantage": 45.3296875,
      "sce_advantage_std": 0.87066650390625,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 833.650455156962,
      "epoch": 0.9898666666666667,
      "num_tokens": 48759009.0,
      "reward": 0.5859375276292363,
      "reward_std": 0.2100959747719268,
      "rewards/accuracy_reward": 0.5859375,
      "sce_advantage": 45.328125,
      "sce_advantage_std": 0.9546305338541666,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.024251502135704303,
      "train_runtime": 40823.4006,
      "train_samples_per_second": 0.184,
      "train_steps_per_second": 0.001
    }
  ],
  "logging_steps": 5,
  "max_steps": 58,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}