bookcorpus
Korpora
bookcorpus | Korpora | |
---|---|---|
3 | 1 | |
778 | 645 | |
- | 0.0% | |
3.1 | 0.0 | |
10 months ago | over 1 year ago | |
Python | Python | |
MIT License | Creative Commons Attribution 4.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
bookcorpus
- Show HN: New AI Dataset Based on LibGen and Sci-Hub
- Can chat GPT overtake Google if they play their cards right?
-
On the Danger of Stochastic Parrots [pdf]
The GPT-3 paper (section 2.2) mentions using two datasets referred to as "books1" and "books2", which are 12B and 55B byte pair encoded tokens each.
Project Gutenberg has 3B word tokens I believe, so it seems like it could be one of them, assuming the ratio of word tokens to byte-pair tokens is something like 3:12 to 3:55.
Another likely candidate alongside Gutenberg is libgen, apparently, and looks like there have been successful efforts to create a similar dataset called bookcorpus: https://github.com/soskek/bookcorpus/issues/27). The discussion on that github issue suggests bookcorpus is very similar to "books2", which would make gutenberg "books1"?
This might be why the paper is intentionally vague about the books used?
Korpora
-
Resources About Cross-Linguistic Relative Phoneme Frequency
- LDA (usually very expensive, but some options exist and in some cases you can google them to find them elsewhere for free): https://www.ldc.upenn.edu/ - Connecting with a university or looking at a linguistics lab's corpus holdings (some will host -- or freely acquired the corpus and therefore you can find it on the internet) - Some language-specific lists or collections: e.g. https://warwick.ac.uk/fac/soc/al/repository/staff/harrisontilly/corpora-for-workshop/, https://github.com/ko-nlp/Korpora , https://guides.uflib.ufl.edu/frenchlinguistics/corpora - Some larger overviews, which may contain links: e.g. https://www.clarin.eu/resource-families/corpora-academic-texts , https://libguides.reed.edu/linguistics/datasets-corpora - Some larger projects to create (often text-based) corpora for multiple languages (often for NLP): e.g. https://www.sketchengine.eu/documentation/tenten-corpora/
What are some alternatives?
awesome-chatgpt-prompts - This repo includes ChatGPT prompt curation to use ChatGPT better.
korean-word-ipa-dictionary - Dictionary of pairs of Korean word and IPA crawled from Wiktionary (Korean edition)
Replicate-Toronto-BookCorpus - This repository contains code to replicate the no-longer publicly available Toronto BookCorpus dataset
gum - Repository for the Georgetown University Multilayer Corpus (GUM)
instagram-scraper - scrapes medias, likes, followers, tags and all metadata. Inspired by instagram-php-scraper,bot
trafilatura - Python & command-line tool to gather text on the Web: web crawling/scraping, extraction of text, metadata, comments
japanese-words-to-vectors - Word2vec (word to vectors) approach for Japanese language using Gensim and Mecab.
open-discourse - Open Discourse is the first fully comprehensive corpus of the plenary proceedings of the federal German Parliament (Bundestag).
corpora - A collection of small corpuses of interesting data for the creation of bots and similar stuff.
stc - Distributed free search engine and AI tools that grant access to knowledge