Skip to content

Dataset · ecnu-icalk

Chinese multimodal math dataset

ecnu-icalk/cmm-math

A collection of school math problems with images in Chinese for training and evaluating multimodal models.

What's inside

The set includes 10 to 100 thousand examples (over 28 thousand problems) in Chinese, combining images and text. It contains problems of various types (tests, fill-in-the-blanks, detailed solutions) with step-by-step answers following the 12-grade school curriculum from elementary to high school.

What you can do with it

  • Evaluating multimodal model (LMM) capability to solve geometry and chart problems
  • SFT fine-tuning of vision-language models for educational tasks
Languages
Chinese
Size
10K–100K rows
Downloads this month
1.6K
Likes
37

How to load it

from datasets import load_dataset

ds = load_dataset("ecnu-icalk/cmm-math")