Already seeing RAG with Llama 3 using @ollama!
@ollama I’ve tested llama 3 8B q5 instruct with github.com/ollama-webui/o… And it’s really good