Dataset · ckchaos
ChartDiff chart pair comparison benchmark
ckchaos/ChartDiff
Large-scale benchmark for comparative summarization of chart pairs by language models.
What's inside
The dataset contains 1K to 10K examples with images and text. It is designed to evaluate the ability of Vision-Language models to find differences and generate coherent comparative descriptions for chart pairs.
What you can do with it
- Evaluating and fine-tuning multimodal models on chart comparison tasks
- Developing automated visual data summarization systems
- Tasks
- Summarization, Image text → text, Image → text, Tabular → text
- Size
- 1K–10K rows
- Downloads this month
- 21.8K
- Likes
- 0
How to load it
from datasets import load_dataset
ds = load_dataset("ckchaos/ChartDiff")