Model card references bfloat16, but the example is using float16/32, so which one is it?
· Sign up or log in to comment
