diff --git a/trl/trainer/dpo_trainer.py b/trl/trainer/dpo_trainer.py index f7f211a414..95ee673982 100644 --- a/trl/trainer/dpo_trainer.py +++ b/trl/trainer/dpo_trainer.py @@ -146,7 +146,7 @@ def __init__( args: TrainingArguments = None, data_collator: Optional[DataCollator] = None, label_pad_token_id: int = -100, - padding_value: int = 0, + padding_value: Optional[int] = None, truncation_mode: str = "keep_end", train_dataset: Optional[Dataset] = None, eval_dataset: Optional[Union[Dataset, Dict[str, Dataset]]] = None,