Dataset · lhpku20010120
Multimodal School Education Instructions
lhpku20010120/Omni-Edu
Nearly 70,000 English and Chinese examples for training educational LLMs and VLMs.
What's inside
Includes 69,999 training examples for SFT: 60,951 specialized educational samples and 9,048 general-purpose ones. The data includes texts and images for text generation, question answering, and image-text analysis tasks. Materials cover the K-12 school curriculum, student knowledge diagnosis, and methodical support in English and Chinese.
What you can do with it
- Train or fine-tune educational LLMs and VLM tutors
- Create systems for diagnosing knowledge gaps in students
- Test models on school subjects and visual materials
- Tasks
- Text generation, Question answering, Image text → text
- Languages
- English, Chinese
- Size
- 10K–100K rows
- Downloads this month
- 52.6K
- Likes
- 6
Models trained on it
- prithivMLmods/Omni-Edu-4B-GGUFImage text → text 4.5K
- lhpku20010120/Omni-Edu-4BImage text → text 1.1K
- lhpku20010120/Omni-Edu-9BImage text → text 480
- lhpku20010120/Omni-Edu-27BImage text → text 431
- lhpku20010120/Omni-Edu-9B-FP8Image text → text 360
- lhpku20010120/Omni-Edu-4B-FP8Image text → text 339
- lhpku20010120/Omni-Edu-27B-FP8Image text → text 316
How to load it
from datasets import load_dataset
ds = load_dataset("lhpku20010120/Omni-Edu")