Hacker News new | ask | show | jobs
by gehsty 206 days ago
Only in the size of model it can run, not speed of token generation.