boheumd / MA-LMM

(2024CVPR) MA-LMM: Memory-Augmented Large Multimodal Model for Long-Term Video Understanding
https://boheumd.github.io/MA-LMM/
MIT License
221 stars 26 forks source link

zero shot demo #6

Closed xjtulien closed 5 months ago

xjtulien commented 5 months ago

I want to simply run the project demo, just like the jupyter example of blip in the lavis library. How can I do it?

mvsoom commented 5 months ago

Me too :)

boheumd commented 5 months ago

Hi, we are working on this and plan to release it soon.

zdaiot commented 5 months ago

@boheumd hello,any progress, please?

boheumd commented 5 months ago

@xjtulien @mvsoom @zdaiot Hello, the demo is now released in the latest version. Thank you for your patience.

mvsoom commented 5 months ago

Hi @boheumd, thank you so much for the quick release of the demo. It works perfectly on a A10.