Dataset · lmms-lab-encoder
GQA multimodal dataset for model evaluation
lmms-lab-encoder/GQA
Formatted version of the GQA visual-text dataset with over 10 million examples for the lmms-eval benchmark.
What's inside
The dataset contains image-text pairs aimed at testing visual reasoning. The volume ranges from 10 to 100 million records. Data is specifically prepared and formatted for one-click multimodal model evaluation via the lmms-eval tool.
What you can do with it
- Testing large multimodal models (LMMs) for understanding relationships in images
- Comparing models within the lmms-eval framework
- Size
- 10M–100M rows
- Downloads this month
- 38.3K
- Likes
- 39
How to load it
from datasets import load_dataset
ds = load_dataset("lmms-lab-encoder/GQA")