dan
banner
woolr.bsky.social
dan
@woolr.bsky.social
ML Scientist working on RAG, agents, embeddings, search and everything else in the modern AI sphere at Graft.

Los Angeles
Nice - yeah totally agree, the (sometimes miniscule) performance bump just isn't worth the increased latency irl
November 22, 2024 at 7:33 PM
What's the memory needed to finetune these kinda models?
November 22, 2024 at 4:26 PM