Comment by searealist 2 days ago Just enable MTP on llama.cpp and you will get the same decode speeds. 0 comments searealist Reply No comments yet Contribute on Hacker News ↗
No comments yet
Contribute on Hacker News ↗