{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 53,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "debug/losses": 0.342240571975708,
      "debug/policy_weights": 0.4937487840652466,
      "debug/raw_losses": 0.6931471824645996,
      "epoch": 0.02,
      "learning_rate": 8.333333333333333e-08,
      "logits/chosen": -2.8462421894073486,
      "logits/rejected": -2.8283610343933105,
      "logps/chosen": -274.7393798828125,
      "logps/rejected": -204.42575073242188,
      "loss": 0.3624,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "debug/losses": 0.3810771703720093,
      "debug/policy_weights": 0.5504893660545349,
      "debug/raw_losses": 0.6920116543769836,
      "epoch": 0.19,
      "learning_rate": 4.911172937635942e-07,
      "logits/chosen": -2.852349281311035,
      "logits/rejected": -2.83735990524292,
      "logps/chosen": -306.01458740234375,
      "logps/rejected": -295.93804931640625,
      "loss": 0.3763,
      "rewards/accuracies": 0.4861111044883728,
      "rewards/chosen": 0.0012468346394598484,
      "rewards/margins": 0.0023373025469481945,
      "rewards/rejected": -0.0010904677910730243,
      "step": 10
    },
    {
      "debug/losses": 0.3667130768299103,
      "debug/policy_weights": 0.5492504835128784,
      "debug/raw_losses": 0.6667538285255432,
      "epoch": 0.38,
      "learning_rate": 3.982949361823388e-07,
      "logits/chosen": -2.8518126010894775,
      "logits/rejected": -2.872077226638794,
      "logps/chosen": -296.2869567871094,
      "logps/rejected": -332.9769592285156,
      "loss": 0.374,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.017367612570524216,
      "rewards/margins": 0.05825704336166382,
      "rewards/rejected": -0.04088941961526871,
      "step": 20
    },
    {
      "debug/losses": 0.35409680008888245,
      "debug/policy_weights": 0.5632873177528381,
      "debug/raw_losses": 0.6232098340988159,
      "epoch": 0.57,
      "learning_rate": 2.416462557480814e-07,
      "logits/chosen": -2.824850559234619,
      "logits/rejected": -2.8103976249694824,
      "logps/chosen": -312.3518981933594,
      "logps/rejected": -323.0265197753906,
      "loss": 0.3551,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.014037144370377064,
      "rewards/margins": 0.1827748715877533,
      "rewards/rejected": -0.1687377393245697,
      "step": 30
    },
    {
      "debug/losses": 0.34752941131591797,
      "debug/policy_weights": 0.550245463848114,
      "debug/raw_losses": 0.6313939690589905,
      "epoch": 0.75,
      "learning_rate": 8.859303711029939e-08,
      "logits/chosen": -2.79345965385437,
      "logits/rejected": -2.797208547592163,
      "logps/chosen": -275.5638122558594,
      "logps/rejected": -348.8089294433594,
      "loss": 0.3454,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.0044286223128438,
      "rewards/margins": 0.20287349820137024,
      "rewards/rejected": -0.20730212330818176,
      "step": 40
    },
    {
      "debug/losses": 0.3295789361000061,
      "debug/policy_weights": 0.5445331335067749,
      "debug/raw_losses": 0.5927887558937073,
      "epoch": 0.94,
      "learning_rate": 5.009573740853313e-09,
      "logits/chosen": -2.829876184463501,
      "logits/rejected": -2.8419814109802246,
      "logps/chosen": -307.4209899902344,
      "logps/rejected": -348.12298583984375,
      "loss": 0.3378,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0008583018789067864,
      "rewards/margins": 0.31627127528190613,
      "rewards/rejected": -0.3154129683971405,
      "step": 50
    },
    {
      "epoch": 1.0,
      "step": 53,
      "total_flos": 0.0,
      "train_loss": 0.3564033879424041,
      "train_runtime": 425.1442,
      "train_samples_per_second": 15.877,
      "train_steps_per_second": 0.125
    }
  ],
  "logging_steps": 10,
  "max_steps": 53,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}