chunked/nltk_chunking/_create_a_model/chunk_24.txt

Every model shares the base class [PreTrainedModel] and a few common methods like resizing input embeddings and pruning self-attention heads.