chunked/nltk_chunking/_generation_strategies/chunk_12.txt

When using pre-trained models for inference
within a [pipeline], the models call the PreTrainedModel.generate() method that applies a default generation
configuration under the hood.