soskek/bookcorpus

Crawl BookCorpus

47
/ 100
Emerging

This project helps natural language processing (NLP) researchers and practitioners create their own large-scale text dataset from books. It takes URLs from Smashwords.com as input and outputs a collection of text files, each containing a book's content. The output can then be further processed into a single file with one sentence per line, or even tokenized for use in training language models or sentence encoders.

853 stars. No commits in the last 6 months.

Use this if you need to build a substantial collection of free book texts for unsupervised learning in natural language processing and prefer to control the data acquisition process yourself.

Not ideal if you need an exact, bit-for-bit replica of the original BookCorpus dataset, as this tool creates a similar but potentially larger and slightly different collection.

natural-language-processing text-dataset-creation language-model-training corpus-linguistics unsupervised-learning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 21 / 25

How are scores calculated?

Stars

853

Forks

112

Language

Python

License

MIT

Last pushed

Jul 14, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/nlp/soskek/bookcorpus"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.