---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 42914147
    num_examples: 80
  download_size: 42914147
  dataset_size: 42914147
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: odc-by
---


This dataset consists of 80 extra difficult webpages from Github Pages, which challenges SoTA multimodal LLMs on converting visual designs into code implementations.

Each example is a pair of source HTML and screenshot ({id}.html and {id}.png).
See the "easy" version of the Design2Code testset [here](https://huggingface.co/datasets/SALT-NLP/Design2Code-hf)

Note that all images in these webpages are replaced by a placeholder image (rick.jpg)

Please refer to their [project page](https://salt-nlp.github.io/Design2Code/) and their [paper](arxiv.org/abs/2403.03163) for more information.