this is what opensource LLM inference
looks like in my head > pure chaos
> lousy & misplaced integrations > everything half-broken, somehow still runs we're so early, and there's a lot of work to do

→ View original post on X — @theahmadosman,