BlueDragon

This is a bug you may encounter if you want to compile with CPU only flag on. The backward function for argmax layer was not implemented for CPU users, this...

@yuyay Your solution works just fine. However, I still hope that more efficient ways can be found, probably making the DataLoader be aware of the memory allocated by pynvvl in...

Right now I find it hard to manage the memory when using pynvvl in PyTorch dataloader. The problem is when you dlpack a cupy array and use from_dlpack to convert...