I only queried it for a bit and it seems to do well compared to llama 3

→ View original post on X — @rasbt