I wrote a post about the little toy embedding search I added to my website over the weekend. It’s pretty good (if you cherry-pick the good queries)!
https://sambreed.dev/writing/2023/lil-vector-search
This profile might be incomplete.
Open on mastodon.social "sam breed"Website:
Work:
Location:
Mill Valley, CA
Personal infoAbout:
partner at SHV, where I build prototypes and help new companies. mostly trail running; myriad typos, no complaints
Wall 2 posts
I wrote a post about the little toy embedding search I added to my website over the weekend. It’s pretty good (if you cherry-pick the good queries)! after kicking the tires on a few of the twitter archive tools, my favorite is by @darius: easy to customize, zero build step, good search. https://tinysubversions.com/twitter-archive/make-your-own/ I tweaked it some and deployed to a new domain, behold my bad tweets for time in memoriam: https://hellsite.sambreed.dev/?q=good%20dogs |
@sambreed Cool!
I'm trying to grok embeddings: you're basically creating the vector based on OpenAI's model right? is it feasible to do “locally” with your own model? Also, do you call OpenAI’s API to create the vector of the query?
(not opposed to calling the API, just curious how the vectors correlate)