{
  "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": 1.1215301752090454,
      "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": 487.4813299179077,
      "epoch": 0.08528784648187633,
      "grad_norm": 476.3103332519531,
      "kl": 1.7723130583763123,
      "learning_rate": 2.5e-06,
      "loss": 0.0822,
      "reward": 0.7912946743890643,
      "reward_std": 0.3657265743240714,
      "rewards/accuracy_reward": 0.1872209922876209,
      "rewards/format_reward": 0.6040736874565482,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 466.69912643432616,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.41791579127311707,
      "kl": 0.012205886840820312,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.0203,
      "reward": 1.2589286297559739,
      "reward_std": 0.32850122936069964,
      "rewards/accuracy_reward": 0.3185267999768257,
      "rewards/format_reward": 0.9404018238186836,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 444.567431640625,
      "epoch": 0.255863539445629,
      "grad_norm": 0.6771596670150757,
      "kl": 0.02176055908203125,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0059,
      "reward": 1.4082589864730835,
      "reward_std": 0.3335796441882849,
      "rewards/accuracy_reward": 0.441294664144516,
      "rewards/format_reward": 0.9669643238186836,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 422.7493499755859,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.23769104480743408,
      "kl": 0.026959228515625,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.01,
      "reward": 1.4937500715255738,
      "reward_std": 0.3376178216189146,
      "rewards/accuracy_reward": 0.5156250238418579,
      "rewards/format_reward": 0.9781250357627869,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 433.67725372314453,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.1763259768486023,
      "kl": 0.0300628662109375,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0157,
      "reward": 1.5979911386966705,
      "reward_std": 0.2915887963026762,
      "rewards/accuracy_reward": 0.6209821671247482,
      "rewards/format_reward": 0.9770089671015739,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 461.1984573364258,
      "epoch": 0.511727078891258,
      "grad_norm": 0.1527547538280487,
      "kl": 0.0370269775390625,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0155,
      "reward": 1.6714286535978318,
      "reward_std": 0.2001216158270836,
      "rewards/accuracy_reward": 0.6986607477068901,
      "rewards/format_reward": 0.9727678924798966,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 458.88328018188474,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.13492096960544586,
      "kl": 0.03984375,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0205,
      "reward": 1.6863840162754058,
      "reward_std": 0.19958442291244866,
      "rewards/accuracy_reward": 0.7189732484519482,
      "rewards/format_reward": 0.9674107521772385,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 435.2419822692871,
      "epoch": 0.6823027718550106,
      "grad_norm": 1.2669559717178345,
      "kl": 0.0412078857421875,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0147,
      "reward": 1.7008929342031478,
      "reward_std": 0.18920395569875836,
      "rewards/accuracy_reward": 0.7238839611411094,
      "rewards/format_reward": 0.9770089611411095,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 436.82925872802736,
      "epoch": 0.767590618336887,
      "grad_norm": 0.38188719749450684,
      "kl": 0.047900390625,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0174,
      "reward": 1.6915179312229156,
      "reward_std": 0.19770997650921346,
      "rewards/accuracy_reward": 0.712276816368103,
      "rewards/format_reward": 0.9792410984635354,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 426.6221176147461,
      "epoch": 0.8528784648187633,
      "grad_norm": 1.376158356666565,
      "kl": 0.202972412109375,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0264,
      "reward": 1.6868304401636123,
      "reward_std": 0.19814990404993296,
      "rewards/accuracy_reward": 0.7000000312924385,
      "rewards/format_reward": 0.986830385029316,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 429.37256622314453,
      "epoch": 0.9381663113006397,
      "grad_norm": 2.8599460124969482,
      "kl": 0.0670989990234375,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0186,
      "reward": 1.722991144657135,
      "reward_std": 0.19592140736058355,
      "rewards/accuracy_reward": 0.7395089641213417,
      "rewards/format_reward": 0.9834821745753288,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 415.95802815755206,
      "epoch": 0.9893390191897654,
      "kl": 0.0960235595703125,
      "reward": 1.7020090073347092,
      "reward_std": 0.19463430003573498,
      "rewards/accuracy_reward": 0.7127976529300213,
      "rewards/format_reward": 0.9892113382617632,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.020674003962555837,
      "train_runtime": 6417.2756,
      "train_samples_per_second": 1.169,
      "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
}