cdqncn

code on text-video qa

5

Dear authors, I was wondering if you could release the code on text-video qa(e.g., dataloader and how you process the videos). Thanks!

Hi, when I use gradcheckpoint for BertEncoder model, I met this issue. How do you solve it?

About the zeroshot video-text retrieval

2

Dear author, Thanks for your great work! I'd like to know something about the zeroshot video-text retrieval result, is it fine-tuning with coco on a 14M BLIP or a 129M...

如题

多张gpu没法加载

5

我在运行python web_demo.py --num_gpus 4的时候，执行到第49行代码的时候就报错cuda out of memory退出程序，执行不了后续代码。nvidia-smi监测显示只在第一张显卡上有效，其他显卡都没用上。 这种情况怎么处理呢