search

Chen-Rao / VD-Diff

Official implementation of Paper "Rethinking Video Deblurring with Wavelet-Aware Dynamic Transformer and Diffusion Model" (ECCV 2024)
MIT License
10 stars 0 forks source link

Single gpu #2

Open amthreeh opened 1 week ago

amthreeh commented 1 week ago

Why are you using only a single GPU? If you use DistributedDataParallel or DataParallel, does it slow down?

Chen-Rao commented 1 week ago

Due to limitations in laboratory computing resources at the time, I only used one GPU. It is feasible to use multiple GPUs during training.