Hiya,
Before Christmas I resolved to teach myself a bit more Python so I could start dabbling in AI/ML. To that end I decided to start a new version of the search engine for the EQ Archives.
Built with a Python backend and a heavily modified Elastic Search-UI front-end in React, the new engine has already indexed about a million documents and should be done a full index by the end of the month.
I have four RTX 3090s here generating embeddings for the first pass, and further LLM enrichment will take place in the second pass, though I've already done a few domains as a demo.
This engine allows you to search the EQ Archives for websites, newsgroup posts and Yahoo Groups mailing list posts. It links back to the source in the Wayback Machine and in my archive.
Feel free to try it out.
https://search-beta.eqarchives.org/