Open LQF39466 opened 1 month ago
Hello. Thank you so much for reporting this issue and providing such a detailed description. We only run the code on single GPU per task. We did not try on multi-gpu situations. Thank you again, we will try to fix this bug.
I am able to run standard DiffPure on multiple GPUs and achieve satisfactory robust accuracy. However, when I try to execute the contrastive guided version with the same GPU configuration, I encounter an error where
self.optimizer
in thereset_drift_counter()
function receives an empty parameter list. Changing my GPU configuration to use onlycuda:0
resolved the issue, suggesting that the contrastive guided diffusion in this repository may only operate on a single GPU, specifically the first one in the system.Further investigation revealed that the
embed_net
in theContrastiveGuidedDiffusion
class seems to be fixed tocuda:0
, regardless of any modifications to the GPU configuration.I would like to understand whether this is a bug or an inherent limitation of the contrastive guidance implementation.