search

OpenGVLab / Multi-Modality-Arena

Chatbot Arena meets multi-modality! Multi-Modality Arena allows you to benchmark vision-language models side-by-side while providing images as inputs. Supports MiniGPT-4, LLaMA-Adapter V2, LLaVA, BLIP-2, and many more!
463 stars 35 forks source link

mPLUG-Owl have been updated. #1

Closed MAGAer13 closed 1 year ago

MAGAer13 commented 1 year ago

Hi, just quick note that mPLUG-Owl has been updated with two different checkpoint (lora & ft). Would you mind replacing the online arena demo? Thanks!

MAGAer13 commented 1 year ago

Besides, I saw the leakage of "AI:" or "Assistant:" signal. Maybe try to use the following prompt and latest codebase to avoid it.

prompt = '''The following is a conversation between a curious human and AI assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.
Human: <image>
Human: {}
AI: '''.format(question)

Thanks again!

wqshao126 commented 1 year ago

Thanks for your suggestion. We have updated arena according to it.