Closed whyhardt closed 3 months ago
Could not reproduce on single GPU (non-DDP training) Two approaches; Check each individually:
In my most recent trainings, I also didn't see this issue emerge. I will try to reproduce it explicitly with GPUs on Oscar and on Google Colab. If I can't reproduce it, I'll close the issue.
Seems fixed now (could not reproduce the issue).
Running gan training several times in a row seems to deplete some resources Exact steps were: