TabLib

Text embeddings are a critical piece of many pipelines, from search, to RAG, to vector databases and more. Most embedding models are BERT/Transformer-based and typically have short context lengths (e.g., 512). That’s only about two pages of text, but documents can be very long – books, legal cases, TV screenplays, code repositories, etc can be tens... See more

Tabby

Opensource, self-hosted AI coding assistant

Meta AI released LLaMA ... and they included a paper which described exactly what it was trained on. It was 5TB of data.

2/3 of it was from Common Crawl. It had content from GitHub, Wikipedia, ArXiv, StackExchange and something called “Books”.

What’s Books? 4.5% of the training data was books. Part of this was Project Gutenberg, which is public dom

... See more