About the lora config alpha and rank

[lucasjinreal]

Hello, in llava original paper, the lora model uses alpha=256 and rank=128, why did u using aplha=4? Have u tried using original config?

[gaowei724]

@ermu2001 Hi,thx for your great work. I noticed that the lora config in config_pllava_nframe.py is {alpha=32, rank=128}

but in eval.sh the lora config is {alpha=4, rank=128}, why the config of the inference is different from the trainning config.

[lucasjinreal]

I remmber somewhere used 4.. maybe I miss look