---
license: cc
dataset_info:
  features:
  - name: sketch
    dtype: image
  - name: image
    dtype: image
  - name: uri
    dtype: string
  splits:
  - name: train
    num_bytes: 85251273
    num_examples: 549
  download_size: 85034553
  dataset_size: 85251273
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
pretty_name: SketchFig
size_categories:
- n<1K
---

# Dataset Card for SketchFig
SketchFig is a dataset of scientific images with paired human-created
sketches collected on the [T<sub>E</sub>X Stack
Exchange](https://tex.stackexchange.com) platform. Visit the
[DeTi*k*Zify](https://github.com/potamides/DeTikZify) project for more
information.

## Usage
```python
from  datasets  import  load_dataset

ds = load_dataset("nllg/sketchfig", split="train")
for sample in ds:
    sample['image'].show()
    sample['sketch'].show()
```