chunked/nltk_chunking/_model_summary/chunk_33.txt

Encoder-decoder[[cv-encoder-decoder]]
Vision models commonly use an encoder (also known as a backbone) to extract important image features before passing them to a Transformer decoder.