{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.769230769230769,
  "eval_steps": 500,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 328.0,
      "learning_rate": 9.837012665064328e-06,
      "logits/chosen": -4.011571884155273,
      "logits/rejected": -3.9108498096466064,
      "logps/chosen": -6.5757951736450195,
      "logps/rejected": -6.684106349945068,
      "loss": 25.3313,
      "rewards/accuracies": 0.503125011920929,
      "rewards/chosen": -0.01468772441148758,
      "rewards/margins": -0.0018365973373875022,
      "rewards/rejected": -0.012851126492023468,
      "step": 5
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 336.0,
      "learning_rate": 2.2133278496394738e-05,
      "logits/chosen": -4.007956504821777,
      "logits/rejected": -3.892199993133545,
      "logps/chosen": -6.476508140563965,
      "logps/rejected": -6.814597129821777,
      "loss": 24.2428,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.011455019935965538,
      "rewards/margins": 0.009504524990916252,
      "rewards/rejected": -0.02095954492688179,
      "step": 10
    },
    {
      "epoch": 0.9230769230769231,
      "eval_logits/chosen": -3.8648571968078613,
      "eval_logits/rejected": -3.8175787925720215,
      "eval_logps/chosen": -6.708494663238525,
      "eval_logps/rejected": -6.904192924499512,
      "eval_loss": 23.919443130493164,
      "eval_rewards/accuracies": 0.6650000214576721,
      "eval_rewards/chosen": -0.021587923169136047,
      "eval_rewards/margins": 0.013099126517772675,
      "eval_rewards/rejected": -0.03468705341219902,
      "eval_runtime": 2.3045,
      "eval_samples_per_second": 86.785,
      "eval_steps_per_second": 86.785,
      "step": 12
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 316.0,
      "learning_rate": 3.442954432772515e-05,
      "logits/chosen": -3.931385040283203,
      "logits/rejected": -3.7742533683776855,
      "logps/chosen": -6.383437156677246,
      "logps/rejected": -7.210443019866943,
      "loss": 21.1059,
      "rewards/accuracies": 0.6767045259475708,
      "rewards/chosen": -0.013774263672530651,
      "rewards/margins": 0.052963126450777054,
      "rewards/rejected": -0.06673739105463028,
      "step": 15
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 294.0,
      "learning_rate": 4.6725810159055555e-05,
      "logits/chosen": -3.8227486610412598,
      "logits/rejected": -3.7290396690368652,
      "logps/chosen": -6.481485843658447,
      "logps/rejected": -7.9080963134765625,
      "loss": 18.1209,
      "rewards/accuracies": 0.815625011920929,
      "rewards/chosen": -0.00985046662390232,
      "rewards/margins": 0.11460874229669571,
      "rewards/rejected": -0.12445919215679169,
      "step": 20
    },
    {
      "epoch": 1.8461538461538463,
      "eval_logits/chosen": -3.344850778579712,
      "eval_logits/rejected": -3.3098304271698,
      "eval_logps/chosen": -6.919511318206787,
      "eval_logps/rejected": -7.514927864074707,
      "eval_loss": 22.132890701293945,
      "eval_rewards/accuracies": 0.6449999809265137,
      "eval_rewards/chosen": -0.04268946498632431,
      "eval_rewards/margins": 0.0530710332095623,
      "eval_rewards/rejected": -0.09576048702001572,
      "eval_runtime": 2.2925,
      "eval_samples_per_second": 87.242,
      "eval_steps_per_second": 87.242,
      "step": 24
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 360.0,
      "learning_rate": 5.902207599038597e-05,
      "logits/chosen": -3.528168201446533,
      "logits/rejected": -3.4186885356903076,
      "logps/chosen": -6.307492733001709,
      "logps/rejected": -8.121912956237793,
      "loss": 16.7445,
      "rewards/accuracies": 0.7718750238418579,
      "rewards/chosen": -0.0019027795642614365,
      "rewards/margins": 0.16127251088619232,
      "rewards/rejected": -0.1631752997636795,
      "step": 25
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 300.0,
      "learning_rate": 7.131834182171638e-05,
      "logits/chosen": -3.377274751663208,
      "logits/rejected": -3.336735963821411,
      "logps/chosen": -5.822231769561768,
      "logps/rejected": -8.159224510192871,
      "loss": 12.9372,
      "rewards/accuracies": 0.8737689852714539,
      "rewards/chosen": 0.05885304883122444,
      "rewards/margins": 0.20639033615589142,
      "rewards/rejected": -0.14753730595111847,
      "step": 30
    },
    {
      "epoch": 2.6923076923076925,
      "grad_norm": 380.0,
      "learning_rate": 8.361460765304679e-05,
      "logits/chosen": -3.175584077835083,
      "logits/rejected": -3.0830180644989014,
      "logps/chosen": -5.457265377044678,
      "logps/rejected": -8.095593452453613,
      "loss": 11.0462,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 0.08443113416433334,
      "rewards/margins": 0.236581951379776,
      "rewards/rejected": -0.15215083956718445,
      "step": 35
    },
    {
      "epoch": 2.769230769230769,
      "eval_logits/chosen": -2.8886406421661377,
      "eval_logits/rejected": -2.8580570220947266,
      "eval_logps/chosen": -7.031093120574951,
      "eval_logps/rejected": -7.9110846519470215,
      "eval_loss": 21.225357055664062,
      "eval_rewards/accuracies": 0.6800000071525574,
      "eval_rewards/chosen": -0.053847651928663254,
      "eval_rewards/margins": 0.08152848482131958,
      "eval_rewards/rejected": -0.13537615537643433,
      "eval_runtime": 2.2803,
      "eval_samples_per_second": 87.707,
      "eval_steps_per_second": 87.707,
      "step": 36
    }
  ],
  "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
}