Skip to content

Dataset · KeepaceAI

Chinese sports and health benchmark

KeepaceAI/MoveBench

Set of 809 questions to test model knowledge in fitness, sports, and health.

What's inside

Text benchmark with fewer than 1,000 records (809 questions specified in the description) in Chinese, created jointly with Capital University of Physical Education and Sports. Materials are designed to evaluate model capabilities in sports coaching and health recommendation scenarios. Only text is available.

What you can do with it

  • Evaluating LLM response quality in the role of a sports coach or health consultant
  • Checking domain knowledge of models in sports science in Chinese
Tasks
Question answering, Text generation
Languages
Chinese
Size
up to 1K rows
Downloads this month
831
Likes
18

How to load it

from datasets import load_dataset

ds = load_dataset("KeepaceAI/MoveBench")