Better to mention the prompt used in the original demo in the `readme.md`.
#4
by
liyucheng
- opened
I tried the model with LlamaForCausalLM. It doesn't act like an assistant. Should use a leading prompt in real applicaiton.
The prompt in the original demo is:
"A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions."
We should mention this in the model card, or it's hard for users to realize the issues is in prompting.
lmzheng
changed discussion status to
closed