OpenAI is too cheap to beat

1

Tesseract

> Does android even have native OCR? Tesseract? <a href="https://github.com/tesseract-ocr/tesseract">https://github.com/tesseract-ocr/tesseract</a>.

#OCR #Image Recognition #PDF Editor 72 social mentions

2

CommonCrawl

Common Crawl claims to have 82% of the tokens used to train GPT-3, and it's available to anyone. Add all the downloadable material at archive.org and you've got a formidable corpus. https://commoncrawl.org/.

#Search Engine #Web Scraping #Data Extraction 90 social mentions