{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.6923076923076925,
  "eval_steps": 500,
  "global_step": 48,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 6.9375,
      "learning_rate": 4.708290280058699e-07,
      "logits/chosen": -4.012143611907959,
      "logits/rejected": -3.9117088317871094,
      "logps/chosen": -27.54085350036621,
      "logps/rejected": -28.105239868164062,
      "loss": 0.6996,
      "rewards/accuracies": 0.49687498807907104,
      "rewards/chosen": -0.060307908803224564,
      "rewards/margins": -0.006133540067821741,
      "rewards/rejected": -0.05417437106370926,
      "step": 5
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 7.375,
      "learning_rate": 1.0593653130132072e-06,
      "logits/chosen": -4.017362117767334,
      "logits/rejected": -3.906172513961792,
      "logps/chosen": -27.55423355102539,
      "logps/rejected": -28.075714111328125,
      "loss": 0.7001,
      "rewards/accuracies": 0.4593749940395355,
      "rewards/chosen": -0.05914067476987839,
      "rewards/margins": -0.008280659094452858,
      "rewards/rejected": -0.05086001753807068,
      "step": 10
    },
    {
      "epoch": 0.9230769230769231,
      "eval_logits/chosen": -3.926405906677246,
      "eval_logits/rejected": -3.88484263420105,
      "eval_logps/chosen": -28.034067153930664,
      "eval_logps/rejected": -30.699705123901367,
      "eval_loss": 0.6927853226661682,
      "eval_rewards/accuracies": 0.5350000262260437,
      "eval_rewards/chosen": 0.0003482414176687598,
      "eval_rewards/margins": 0.0007382288458757102,
      "eval_rewards/rejected": -0.00038998748641461134,
      "eval_runtime": 2.3408,
      "eval_samples_per_second": 85.439,
      "eval_steps_per_second": 85.439,
      "step": 12
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 7.34375,
      "learning_rate": 1.6479015980205448e-06,
      "logits/chosen": -3.9834656715393066,
      "logits/rejected": -3.840578556060791,
      "logps/chosen": -27.485614776611328,
      "logps/rejected": -28.10161781311035,
      "loss": 0.6972,
      "rewards/accuracies": 0.49621209502220154,
      "rewards/chosen": -0.06247218698263168,
      "rewards/margins": -0.0015615280717611313,
      "rewards/rejected": -0.0609106607735157,
      "step": 15
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 7.75,
      "learning_rate": 2.236437883027882e-06,
      "logits/chosen": -4.032500743865967,
      "logits/rejected": -3.943988800048828,
      "logps/chosen": -27.166378021240234,
      "logps/rejected": -27.592472076416016,
      "loss": 0.6971,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.056408680975437164,
      "rewards/margins": -0.001882375800050795,
      "rewards/rejected": -0.05452630668878555,
      "step": 20
    },
    {
      "epoch": 1.8461538461538463,
      "eval_logits/chosen": -3.9240834712982178,
      "eval_logits/rejected": -3.8822319507598877,
      "eval_logps/chosen": -28.04541778564453,
      "eval_logps/rejected": -30.720582962036133,
      "eval_loss": 0.6926065683364868,
      "eval_rewards/accuracies": 0.5550000071525574,
      "eval_rewards/chosen": -0.00017397981719113886,
      "eval_rewards/margins": 0.0010972179006785154,
      "eval_rewards/rejected": -0.0012711978051811457,
      "eval_runtime": 2.306,
      "eval_samples_per_second": 86.731,
      "eval_steps_per_second": 86.731,
      "step": 24
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 8.6875,
      "learning_rate": 2.8249741680352196e-06,
      "logits/chosen": -3.9794921875,
      "logits/rejected": -3.852832078933716,
      "logps/chosen": -27.62384033203125,
      "logps/rejected": -28.47451400756836,
      "loss": 0.6936,
      "rewards/accuracies": 0.5062500238418579,
      "rewards/chosen": -0.0576423816382885,
      "rewards/margins": 0.006730812601745129,
      "rewards/rejected": -0.0643731951713562,
      "step": 25
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 8.3125,
      "learning_rate": 3.4135104530425568e-06,
      "logits/chosen": -4.015521049499512,
      "logits/rejected": -3.958658218383789,
      "logps/chosen": -27.597198486328125,
      "logps/rejected": -28.10772132873535,
      "loss": 0.7022,
      "rewards/accuracies": 0.4940341114997864,
      "rewards/chosen": -0.05957310274243355,
      "rewards/margins": -0.011498005129396915,
      "rewards/rejected": -0.04807509481906891,
      "step": 30
    },
    {
      "epoch": 2.6923076923076925,
      "grad_norm": 8.0,
      "learning_rate": 4.002046738049894e-06,
      "logits/chosen": -3.974912643432617,
      "logits/rejected": -3.8793323040008545,
      "logps/chosen": -27.144947052001953,
      "logps/rejected": -27.934412002563477,
      "loss": 0.6963,
      "rewards/accuracies": 0.515625,
      "rewards/chosen": -0.06426174938678741,
      "rewards/margins": 0.0010266598546877503,
      "rewards/rejected": -0.06528842449188232,
      "step": 35
    },
    {
      "epoch": 2.769230769230769,
      "eval_logits/chosen": -3.918003797531128,
      "eval_logits/rejected": -3.875520944595337,
      "eval_logps/chosen": -28.07503890991211,
      "eval_logps/rejected": -30.783376693725586,
      "eval_loss": 0.6909805536270142,
      "eval_rewards/accuracies": 0.574999988079071,
      "eval_rewards/chosen": -0.0031362976878881454,
      "eval_rewards/margins": 0.004414236638695002,
      "eval_rewards/rejected": -0.007550534792244434,
      "eval_runtime": 2.3146,
      "eval_samples_per_second": 86.408,
      "eval_steps_per_second": 86.408,
      "step": 36
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 7.8125,
      "learning_rate": 3.9861897027011495e-06,
      "logits/chosen": -4.010741710662842,
      "logits/rejected": -3.8427538871765137,
      "logps/chosen": -27.8968448638916,
      "logps/rejected": -28.60370445251465,
      "loss": 0.6927,
      "rewards/accuracies": 0.5026515126228333,
      "rewards/chosen": -0.07123393565416336,
      "rewards/margins": 0.00865729060024023,
      "rewards/rejected": -0.07989122718572617,
      "step": 40
    },
    {
      "epoch": 3.4615384615384617,
      "grad_norm": 7.53125,
      "learning_rate": 3.4829202374186716e-06,
      "logits/chosen": -4.0228962898254395,
      "logits/rejected": -3.9077537059783936,
      "logps/chosen": -27.418827056884766,
      "logps/rejected": -28.112506866455078,
      "loss": 0.6957,
      "rewards/accuracies": 0.5062500238418579,
      "rewards/chosen": -0.06953997910022736,
      "rewards/margins": 0.000948831788264215,
      "rewards/rejected": -0.0704888105392456,
      "step": 45
    },
    {
      "epoch": 3.6923076923076925,
      "eval_logits/chosen": -3.9120235443115234,
      "eval_logits/rejected": -3.869976758956909,
      "eval_logps/chosen": -28.11831283569336,
      "eval_logps/rejected": -30.877779006958008,
      "eval_loss": 0.6903982758522034,
      "eval_rewards/accuracies": 0.6000000238418579,
      "eval_rewards/chosen": -0.005011914297938347,
      "eval_rewards/margins": 0.005597901530563831,
      "eval_rewards/rejected": -0.010609814897179604,
      "eval_runtime": 2.311,
      "eval_samples_per_second": 86.541,
      "eval_steps_per_second": 86.541,
      "step": 48
    }
  ],
  "logging_steps": 5,
  "max_steps": 65,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}