> Local inference for chats sucks.
/r/SillyTavernAI would disagree with you.
Many people who use ST have a "serious" nvidia card.
We are talking about NPUs here.
Many people who use ST have a "serious" nvidia card.
We are talking about NPUs here.