{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2898550724637681,
  "eval_steps": 40,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.036231884057971016,
      "grad_norm": 61.66789827681496,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7321553230285645,
      "logits/rejected": -2.7100937366485596,
      "logps/chosen": -182.5845489501953,
      "logps/rejected": -189.55001831054688,
      "loss": 0.689,
      "rewards/accuracies": 0.3187499940395355,
      "rewards/chosen": -0.0021577859297394753,
      "rewards/margins": 0.005646524019539356,
      "rewards/rejected": -0.007804309483617544,
      "step": 5
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 44.96703657621111,
      "learning_rate": 1e-06,
      "logits/chosen": -2.753889799118042,
      "logits/rejected": -2.7519516944885254,
      "logps/chosen": -197.34320068359375,
      "logps/rejected": -184.00961303710938,
      "loss": 0.6275,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.030874451622366905,
      "rewards/margins": 0.18904080986976624,
      "rewards/rejected": -0.15816636383533478,
      "step": 10
    },
    {
      "epoch": 0.10869565217391304,
      "grad_norm": 51.54402636298773,
      "learning_rate": 9.996221126793764e-07,
      "logits/chosen": -2.694437265396118,
      "logits/rejected": -2.691904067993164,
      "logps/chosen": -203.14883422851562,
      "logps/rejected": -204.52386474609375,
      "loss": 0.5851,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.6205412745475769,
      "rewards/margins": 0.9350436925888062,
      "rewards/rejected": -0.314502477645874,
      "step": 15
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 35.11282014809116,
      "learning_rate": 9.984890219128145e-07,
      "logits/chosen": -2.609405517578125,
      "logits/rejected": -2.5795176029205322,
      "logps/chosen": -188.33395385742188,
      "logps/rejected": -192.52633666992188,
      "loss": 0.5144,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.8838651776313782,
      "rewards/margins": 1.275496244430542,
      "rewards/rejected": -0.39163118600845337,
      "step": 20
    },
    {
      "epoch": 0.18115942028985507,
      "grad_norm": 37.05591291167376,
      "learning_rate": 9.966024404228493e-07,
      "logits/chosen": -2.4429798126220703,
      "logits/rejected": -2.4225034713745117,
      "logps/chosen": -179.79977416992188,
      "logps/rejected": -179.25279235839844,
      "loss": 0.5036,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 0.40532931685447693,
      "rewards/margins": 0.8947975039482117,
      "rewards/rejected": -0.4894680976867676,
      "step": 25
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 33.27618771325657,
      "learning_rate": 9.939652198703783e-07,
      "logits/chosen": -2.319044589996338,
      "logits/rejected": -2.320253372192383,
      "logps/chosen": -188.62039184570312,
      "logps/rejected": -193.8806915283203,
      "loss": 0.4987,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.6481167078018188,
      "rewards/margins": 1.2183058261871338,
      "rewards/rejected": -0.5701891183853149,
      "step": 30
    },
    {
      "epoch": 0.2536231884057971,
      "grad_norm": 38.645073192304736,
      "learning_rate": 9.905813465442354e-07,
      "logits/chosen": -2.242053747177124,
      "logits/rejected": -2.2155580520629883,
      "logps/chosen": -203.94546508789062,
      "logps/rejected": -194.8705291748047,
      "loss": 0.5091,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.8871867060661316,
      "rewards/margins": 1.2737131118774414,
      "rewards/rejected": -0.3865264356136322,
      "step": 35
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 30.937642063399068,
      "learning_rate": 9.864559353357187e-07,
      "logits/chosen": -2.0998620986938477,
      "logits/rejected": -2.094836473464966,
      "logps/chosen": -182.6898956298828,
      "logps/rejected": -185.73983764648438,
      "loss": 0.4889,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.0410809516906738,
      "rewards/margins": 1.0222995281219482,
      "rewards/rejected": 0.018781563267111778,
      "step": 40
    },
    {
      "epoch": 0.2898550724637681,
      "eval_logits/chosen": -2.035081624984741,
      "eval_logits/rejected": -1.9875677824020386,
      "eval_logps/chosen": -188.77523803710938,
      "eval_logps/rejected": -200.09498596191406,
      "eval_loss": 0.4642000198364258,
      "eval_rewards/accuracies": 0.7943548560142517,
      "eval_rewards/chosen": 1.1544201374053955,
      "eval_rewards/margins": 1.3431099653244019,
      "eval_rewards/rejected": -0.18868987262248993,
      "eval_runtime": 248.6586,
      "eval_samples_per_second": 15.773,
      "eval_steps_per_second": 0.249,
      "step": 40
    }
  ],
  "logging_steps": 5,
  "max_steps": 414,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 40,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 471248375119872.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}