SLLM-multi-hop/AnimalQA
收藏Hugging Face2025-06-09 更新2025-11-01 收录
下载链接：
https://hf-mirror.com/datasets/SLLM-multi-hop/AnimalQA
下载链接
链接失效反馈官方服务：
资源简介：
SAKURA-AnimalQA数据集包含了Interspeech 2025论文中SAKURA基准的动物追踪部分的音频记录以及单跳和多跳问题和答案。数据集字段包括音频文件名、音频记录、音频属性标签（即发出声音的动物种类）、单跳问题指令、单跳问题答案、多跳问题指令和多跳问题答案。
The SAKURA-AnimalQA dataset contains the audio recordings and single/multi-hop questions/answers from the animal track of the SAKURA benchmark in the Interspeech 2025 paper. The dataset fields include the filename of the audio files, the audio recordings, the attribute labels (i.e., the types of animals making the sounds), single-hop question instructions, single-hop answers, multi-hop question instructions, and multi-hop answers.
提供机构：
SLLM-multi-hop
