Running with llama.cpp on Macbook M2 Pro

#1
by huksley - opened

Hi, I am trying to run poro-34b.Q4_K_M.gguf and it gives gibberish and an error "ggml_metal_graph_compute: command buffer 5 failed with status 5"

Which quantization works on MacBook M2 Pro 32Gb?

Sign up or log in to comment