Huanmin Dou

Hello , Is this package able to train a multi-label dataset with one-hot encoding?

2

It is a great package that improves my efficiency , When I test cifar10, for the one-hot label , I can use ```python label = torch.argmax(label,dim=1) ``` to transform one-hot...

Hello,I have some question about the dataset COCO

1

Thanks for sharing the code, but When I followed the instruction and got ready to test it on COCO I met some error ```python RuntimeError: stack expects each tensor to...

A Installation Problem

1

Hello, Thanks for sharing the code, I met some problems when I trying to install the LAVIS, Here are my step as follows ```shell conda create --name "LAVIS" python=3.12 pip...

thanks for sharing the code but I found that in the contrastive_memory_prism loss file it shows that ```python self.margin_window=self.margin_window[-1:] ``` before taking a sum operation, the self.margin_windows only contains one...

Update loss.py

4

I wonder if it is possible to convert the zero vectors in the matrix to one vectors? To avoid NaN loss, add the following two lines of code. I'm testing...