metadata

language:
  - en
license: mit
size_categories:
  - 100K<n<1M
task_categories:
  - text-generation
pretty_name: SVGEN RJU
dataset_info:
  features:
    - name: description
      dtype: string
    - name: image
      struct:
        - name: bytes
          dtype: binary
        - name: path
          dtype: 'null'
    - name: svg
      dtype: string
    - name: uuid
      dtype: string
    - name: json
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 3151387335
      num_examples: 216275
  download_size: 1843399077
  dataset_size: 3151387335
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags:
  - art
  - code

SVGEN RJU - SVGEN 500k: Rasterized, JSONified, UUID'ed

I have selected every svg image from svgen that would rasterize under cairosvg, which is significantly less than a 1% failure rate. Under development.

Reasoning

This is the 1st of many SVG datasets I am collecting, extracting, and rasterizing in an attempt to produce a meaningfully helpful spatial reasoning and vertex manipulation model.

Usage

The rasterized images are in PNG format, as bytes. They may be extracted through the creation of an io.BytesIO instance on the bytes then running Image.open on the byte stream. They could also be written to disk first for convenience and storage.