You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I install the latest llama.cpp on macos and tried gemma-7b-it-GGUF-Q4, but the model still generate nonsense either in chat mode or completion mode. I have added the <bos> token to the prompt.