Exactly same as our study! Interesting!

[mu-cai]

Great work! It is pretty interesting that last week, we did an exactly same study,

https://github.com/WisconsinAIVision/ViP-LLaVA/blob/main/docs/study_llm_backbone.md

Our same empirical evaluation findings demonstrate the correctness of each other’s implementation. haha

Mu Cai

[Isaachhh]

We Bunny team also provide Bunny-Llama-3-8B-V and Bunny-Phi-3-V with much better performance~

https://github.com/BAAI-DCAI/Bunny

[mmaaz60]

Hi @mu-cai, @Isaachhh,

Its great to see such open-source efforts adapting recently released LLMs. Congratulations on your achievements!