Skip to content

grascii/datasets

Repository files navigation

Gregg Shorthand Datasets

This repository contains the source images and build scripts for grascii's Gregg Shorthand datasets on HuggingFace.

Datasets

This dataset consists of images of shorthand forms from the 1916 Gregg Shorthand Dictionary1. Most of the images originated from the Gregg1916 dataset.

Contribution Notes

Images are taken from a scan of the dictionary at 175% zoom.

Contributing

Contributions are welcome!

Dataset Quality

If you notice an incorrect or low-quality element in the dataset, please open an issue to report the problem and/or open a pull request that resolves the issue. Problematic elements include but are not limited to:

  • misnamed images
    • a name does not match the word represented by an image
    • a name is misspelled
  • poorly cropped images
    • the shorthand form in the image is cut off
    • excessive white-space exists around the shorthand form
  • images with stray marks (such marks should be whited-out)
    • the image contains scan artifacts
    • the image contains parts of adjacent shorthand forms or text

If you notice an error in the Grascii form for an image, open an issue in the dictionaries repository instead.

Footnotes

  1. Gregg, John Robert. Gregg Shorthand Dictionary. Gregg Publishing Company, 1916.