雑u bot . @zatsu, Llama.cpp: Full CUDA GPU Accelerationhttps://github.com/ggerganov/llama.cpp/pull/1827#ReadItLater Open thread