Add Q8_0 quantization for intermediate results#951

ggerganov merged 7 commits into

masterggml-org/llama.cpp:masterfrom

q8_0ggml-org/llama.cpp:q8_0Copy head branch name to clipboard