trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 48,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1e-07,
      "logits/chosen": -2.8401148319244385,
      "logits/rejected": -2.8160030841827393,
      "logps/chosen": -284.8484802246094,
      "logps/rejected": -263.6097106933594,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.83504027183137e-07,
      "logits/chosen": -2.793118476867676,
      "logits/rejected": -2.7711308002471924,
      "logps/chosen": -256.3919982910156,
      "logps/rejected": -223.95652770996094,
      "loss": 0.691,
      "rewards/accuracies": 0.5208333134651184,
      "rewards/chosen": 0.004140747711062431,
      "rewards/margins": 0.004918030463159084,
      "rewards/rejected": -0.0007772822282277048,
      "step": 10
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.643105808261596e-07,
      "logits/chosen": -2.8137385845184326,
      "logits/rejected": -2.8089663982391357,
      "logps/chosen": -276.8554382324219,
      "logps/rejected": -248.77206420898438,
      "loss": 0.674,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 0.026244645938277245,
      "rewards/margins": 0.03955433517694473,
      "rewards/rejected": -0.013309692963957787,
      "step": 20
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8676665440207977e-07,
      "logits/chosen": -2.7731940746307373,
      "logits/rejected": -2.769883155822754,
      "logps/chosen": -290.5604248046875,
      "logps/rejected": -260.43145751953125,
      "loss": 0.6543,
      "rewards/accuracies": 0.640625,
      "rewards/chosen": -0.0338541641831398,
      "rewards/margins": 0.07991626113653183,
      "rewards/rejected": -0.11377042531967163,
      "step": 30
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.1500545527530544e-08,
      "logits/chosen": -2.7984321117401123,
      "logits/rejected": -2.7847955226898193,
      "logps/chosen": -302.61767578125,
      "logps/rejected": -292.78704833984375,
      "loss": 0.6357,
      "rewards/accuracies": 0.684374988079071,
      "rewards/chosen": -0.08540390431880951,
      "rewards/margins": 0.14555008709430695,
      "rewards/rejected": -0.23095396161079407,
      "step": 40
    },
    {
      "epoch": 1.0,
      "step": 48,
      "total_flos": 0.0,
      "train_loss": 0.6586156586805979,
      "train_runtime": 1568.8191,
      "train_samples_per_second": 7.794,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 10,
  "max_steps": 48,
  "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
}