trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9874476987447699,
  "eval_steps": 500,
  "global_step": 59,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 8.333333333333333e-08,
      "logits/chosen": -2.330733299255371,
      "logits/rejected": -2.316999912261963,
      "logps/chosen": -411.3883972167969,
      "logps/rejected": -413.6264343261719,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.930057285201027e-07,
      "logits/chosen": -2.288774251937866,
      "logits/rejected": -2.2626779079437256,
      "logps/chosen": -393.1912841796875,
      "logps/rejected": -471.7352294921875,
      "loss": 0.6677,
      "rewards/accuracies": 0.6180555820465088,
      "rewards/chosen": -0.07132196426391602,
      "rewards/margins": 0.05536883696913719,
      "rewards/rejected": -0.1266908049583435,
      "step": 10
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.187457503795526e-07,
      "logits/chosen": -1.9292032718658447,
      "logits/rejected": -1.8881103992462158,
      "logps/chosen": -504.60784912109375,
      "logps/rejected": -635.0429077148438,
      "loss": 0.5686,
      "rewards/accuracies": 0.753125011920929,
      "rewards/chosen": -1.2000805139541626,
      "rewards/margins": 0.6865085363388062,
      "rewards/rejected": -1.8865890502929688,
      "step": 20
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.8691164100062034e-07,
      "logits/chosen": -1.9375982284545898,
      "logits/rejected": -1.797930359840393,
      "logps/chosen": -505.9959411621094,
      "logps/rejected": -627.4420776367188,
      "loss": 0.5084,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.128210425376892,
      "rewards/margins": 0.88212651014328,
      "rewards/rejected": -2.0103371143341064,
      "step": 30
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.4248369943086995e-07,
      "logits/chosen": -1.9036223888397217,
      "logits/rejected": -1.7595268487930298,
      "logps/chosen": -421.82745361328125,
      "logps/rejected": -559.4238891601562,
      "loss": 0.5093,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.7564447522163391,
      "rewards/margins": 0.6973710060119629,
      "rewards/rejected": -1.4538156986236572,
      "step": 40
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.473909705816111e-08,
      "logits/chosen": -1.8024383783340454,
      "logits/rejected": -1.680045485496521,
      "logps/chosen": -437.82403564453125,
      "logps/rejected": -581.6019287109375,
      "loss": 0.4927,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.7874447107315063,
      "rewards/margins": 0.6873693466186523,
      "rewards/rejected": -1.4748141765594482,
      "step": 50
    },
    {
      "epoch": 0.99,
      "step": 59,
      "total_flos": 0.0,
      "train_loss": 0.5465719497809975,
      "train_runtime": 1938.7154,
      "train_samples_per_second": 7.883,
      "train_steps_per_second": 0.03
    }
  ],
  "logging_steps": 10,
  "max_steps": 59,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}