Update src/transformers/pipelines/base.py

Co-authored-by: Pedro Cuenca <[email protected]>

src/transformers/pipelines/base.py

@@ -743,7 +743,7 @@ def build_pipeline_init_args(
     if has_tokenizer:
         docstring += r"""
         tokenizer ([`PreTrainedTokenizer`]):
-            The tokenizer that will be used by the pipeline to encode data for the model. This object inherits from
+            The tokenizer that will be used by the pipeline to encode or decode text data. This object inherits from
             [`PreTrainedTokenizer`]."""
     if has_feature_extractor:
         docstring += r"""