chunked/nltk_chunking/model_doc_clap/chunk_11.txt

Third, we perform comprehensive experiments to evaluate our model across three tasks: text-to-audio retrieval, zero-shot audio classification, and supervised audio classification.