{
  "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": 498.510066986084,
      "epoch": 0.017057569296375266,
      "grad_norm": 1.116628856964355,
      "kl": 0.0,
      "learning_rate": 5e-07,
      "loss": 0.0115,
      "reward": 0.6171875260770321,
      "reward_std": 0.437016986310482,
      "rewards/accuracy_reward": 0.17633929289877415,
      "rewards/format_reward": 0.4408482313156128,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 487.5962829589844,
      "epoch": 0.08528784648187633,
      "grad_norm": 329.581274808389,
      "kl": 1.5817211270332336,
      "learning_rate": 2.5e-06,
      "loss": 0.0633,
      "reward": 0.7832031613215804,
      "reward_std": 0.3882951531559229,
      "rewards/accuracy_reward": 0.17801340064033866,
      "rewards/format_reward": 0.6051897583529353,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 471.0462287902832,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.22688607924492205,
      "kl": 0.0134246826171875,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0205,
      "reward": 1.2479911297559738,
      "reward_std": 0.33266205713152885,
      "rewards/accuracy_reward": 0.3131696585565805,
      "rewards/format_reward": 0.9348214715719223,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 441.45515899658204,
      "epoch": 0.255863539445629,
      "grad_norm": 0.3322143554658319,
      "kl": 0.01697235107421875,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.012,
      "reward": 1.4209821969270706,
      "reward_std": 0.33965970352292063,
      "rewards/accuracy_reward": 0.45714287683367727,
      "rewards/format_reward": 0.9638393223285675,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 421.6131874084473,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.20995700940835846,
      "kl": 0.037835693359375,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0117,
      "reward": 1.646428656578064,
      "reward_std": 0.2756752146407962,
      "rewards/accuracy_reward": 0.6725446745753288,
      "rewards/format_reward": 0.973883967101574,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 423.54466247558594,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.2012153869368259,
      "kl": 0.037371826171875,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0144,
      "reward": 1.6178572177886963,
      "reward_std": 0.27076737955212593,
      "rewards/accuracy_reward": 0.645758955180645,
      "rewards/format_reward": 0.9720982536673546,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 439.773233795166,
      "epoch": 0.511727078891258,
      "grad_norm": 1.0763397890575324,
      "kl": 0.0529998779296875,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0176,
      "reward": 1.668526867032051,
      "reward_std": 0.21837877184152604,
      "rewards/accuracy_reward": 0.6995536059141159,
      "rewards/format_reward": 0.9689732551574707,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 446.2640869140625,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.15301980687370056,
      "kl": 8.364321899414062,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.8027,
      "reward": 1.6857143670320511,
      "reward_std": 0.2085555238649249,
      "rewards/accuracy_reward": 0.7165178850293159,
      "rewards/format_reward": 0.9691964700818062,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 428.8069389343262,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.16199556630311315,
      "kl": 0.0596527099609375,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0204,
      "reward": 1.6937500774860381,
      "reward_std": 0.1925298016052693,
      "rewards/accuracy_reward": 0.7194196820259094,
      "rewards/format_reward": 0.9743303924798965,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 430.70425872802736,
      "epoch": 0.767590618336887,
      "grad_norm": 1.6843827012960852,
      "kl": 0.0576385498046875,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0227,
      "reward": 1.6881697207689286,
      "reward_std": 0.2063945535570383,
      "rewards/accuracy_reward": 0.7098214611411094,
      "rewards/format_reward": 0.9783482521772384,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 423.26809921264646,
      "epoch": 0.8528784648187633,
      "grad_norm": 5.03664692812433,
      "kl": 2.7302520751953123,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.1931,
      "reward": 1.6832590103149414,
      "reward_std": 0.20804516496136785,
      "rewards/accuracy_reward": 0.7024553880095482,
      "rewards/format_reward": 0.9808036014437675,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 423.15091094970705,
      "epoch": 0.9381663113006397,
      "grad_norm": 2.620417003259849,
      "kl": 0.3049896240234375,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0318,
      "reward": 1.7154018610715867,
      "reward_std": 0.2086696395650506,
      "rewards/accuracy_reward": 0.7316964611411094,
      "rewards/format_reward": 0.9837053924798965,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 413.19120534261066,
      "epoch": 0.9893390191897654,
      "kl": 0.060206095377604164,
      "reward": 1.6845238904158275,
      "reward_std": 0.19806722179055214,
      "rewards/accuracy_reward": 0.6994047885139784,
      "rewards/format_reward": 0.985119087000688,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.10446281573767292,
      "train_runtime": 6557.3192,
      "train_samples_per_second": 1.144,
      "train_steps_per_second": 0.009
    }
  ],
  "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
}