James Futhey @jamesfuthey.com · Apr 22 1 / 2
Ayyyyy... quantized embeddings are getting computed for every post in real time on cpu time (🤩) and I am still able to run clustering on my M1 Max 🙌 LLM-assisted labelling is doing it's thing after rewriting the prompt for 30 minutes. Hoping to see clusters stabilize with <1 week of posts.
8 likes 2 replies
?
Replies
I have 2 clustering pipelines now, one is behavioral, this one is semantic. Additionally I have this keyword expansion method in prod to try to filter and boost posts based on topic. keyword method is brutish. behavioral pipeline tends to typecast creators. semantic should be gold standard
There is also an interesting idea for image blobs -> github.com/darwinium-co...