2 ms·
llama.cpp just got Falcon support (not yet merged), so you could run it on just RAM. Not too fast though.
by logicchains 3y ago
llama.cpp just got Falcon support (not yet merged), so you could run it on just RAM. Not too fast though.
- gardnr 3y agoIn case anyone wants to follow along: https://github.com/ggerganov/llama.cpp/issues/1602 https://github.com/ggerganov/llama.cpp/issues/1602