chunked/nltk_chunking/_task_summary/chunk_87.txt

For a task like image captioning, the model learns relationships between image embeddings and text embeddings.