{
  "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": 0.39008715748786926,
      "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": 605.4260902404785,
      "epoch": 0.08528784648187633,
      "grad_norm": 1.0314300060272217,
      "kl": 0.00019499659538269043,
      "learning_rate": 2.5e-06,
      "loss": 0.0241,
      "reward": 0.6032366305589676,
      "reward_std": 0.3533420516178012,
      "rewards/accuracy_reward": 0.6032366305589676,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.3067253112793,
      "epoch": 0.17057569296375266,
      "grad_norm": 0.17284299433231354,
      "kl": 0.005330562591552734,
      "learning_rate": 2.956412726139078e-06,
      "loss": 0.069,
      "reward": 0.6810268200933933,
      "reward_std": 0.3049736712127924,
      "rewards/accuracy_reward": 0.6810268200933933,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 606.5038230895996,
      "epoch": 0.255863539445629,
      "grad_norm": 0.15556961297988892,
      "kl": 0.05679264068603516,
      "learning_rate": 2.7836719084521715e-06,
      "loss": 0.0754,
      "reward": 0.7587053939700127,
      "reward_std": 0.22350936699658633,
      "rewards/accuracy_reward": 0.7587053939700127,
      "rewards/format_reward": 0.0,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 584.5225677490234,
      "epoch": 0.3411513859275053,
      "grad_norm": 0.2594759166240692,
      "kl": 0.005410385131835937,
      "learning_rate": 2.4946839873611927e-06,
      "loss": 0.0497,
      "reward": 0.7783482521772385,
      "reward_std": 0.19184589060023427,
      "rewards/accuracy_reward": 0.7783482521772385,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.5861892700195,
      "epoch": 0.42643923240938164,
      "grad_norm": 0.1489226222038269,
      "kl": 0.007462882995605468,
      "learning_rate": 2.1156192081791355e-06,
      "loss": 0.0395,
      "reward": 0.7607143223285675,
      "reward_std": 0.18709109332412482,
      "rewards/accuracy_reward": 0.7607143223285675,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 602.0279273986816,
      "epoch": 0.511727078891258,
      "grad_norm": 0.17840267717838287,
      "kl": 0.008930206298828125,
      "learning_rate": 1.6808050203829845e-06,
      "loss": 0.0342,
      "reward": 0.7479911029338837,
      "reward_std": 0.1881474507972598,
      "rewards/accuracy_reward": 0.7479911029338837,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 595.2819473266602,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.19424788653850555,
      "kl": 0.007519149780273437,
      "learning_rate": 1.2296174432791415e-06,
      "loss": 0.0401,
      "reward": 0.7430803917348385,
      "reward_std": 0.19397471882402897,
      "rewards/accuracy_reward": 0.7430803917348385,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 580.3154273986817,
      "epoch": 0.6823027718550106,
      "grad_norm": 0.1329621970653534,
      "kl": 0.009706497192382812,
      "learning_rate": 8.029152419343472e-07,
      "loss": 0.0334,
      "reward": 0.7716518208384514,
      "reward_std": 0.18079692414030432,
      "rewards/accuracy_reward": 0.7716518208384514,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.7227966308594,
      "epoch": 0.767590618336887,
      "grad_norm": 0.16628068685531616,
      "kl": 0.005841827392578125,
      "learning_rate": 4.3933982822017883e-07,
      "loss": 0.0443,
      "reward": 0.7444196790456772,
      "reward_std": 0.209815969876945,
      "rewards/accuracy_reward": 0.7444196790456772,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 600.9355201721191,
      "epoch": 0.8528784648187633,
      "grad_norm": 0.13309380412101746,
      "kl": 0.006610107421875,
      "learning_rate": 1.718159615201853e-07,
      "loss": 0.0371,
      "reward": 0.7435268238186836,
      "reward_std": 0.1858388701453805,
      "rewards/accuracy_reward": 0.7435268238186836,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.021898651123,
      "epoch": 0.9381663113006397,
      "grad_norm": 0.74787837266922,
      "kl": 0.007943344116210938,
      "learning_rate": 2.4570139579284723e-08,
      "loss": 0.0361,
      "reward": 0.7767857477068901,
      "reward_std": 0.19855766519904136,
      "rewards/accuracy_reward": 0.7767857477068901,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.1831906636556,
      "epoch": 0.9893390191897654,
      "kl": 0.006526947021484375,
      "reward": 0.7529762213428816,
      "reward_std": 0.1964037980263432,
      "rewards/accuracy_reward": 0.7529762213428816,
      "rewards/format_reward": 0.0,
      "step": 58,
      "total_flos": 0.0,
      "train_loss": 0.04403019331170824,
      "train_runtime": 12949.6635,
      "train_samples_per_second": 0.579,
      "train_steps_per_second": 0.004
    }
  ],
  "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
}