Question about prompt (especially about reward model prompt)

llava-rlhf / LLaVA-RLHF

Aligning LMMs with Factually Augmented RLHF

https://llava-rlhf.github.io/

GNU General Public License v3.0

323 stars 25 forks source link

Question about prompt (especially about reward model prompt) #37

Open tunantu opened 4 weeks ago

tunantu commented 4 weeks ago

Hi authors, when I tested the reward model you given, I found the reward scores are always below 0. May I get your reward prompt when you trained this model and applied it?

Also, does the prompt in LLaVA you used is identical to the original LLaVA prompt?

Thanks.