chunked/nltk_chunking/model_doc_openai-gpt/chunk_3.txt

Although large unlabeled text corpora are abundant,
labeled data for learning these specific tasks is scarce, making it challenging for discriminatively trained models to
perform adequately.