Stefan Frisch

Had this issue as well. It can be ignored for now. The behavior might change later. OBJ_CLASS_IDS is intentionally different.

Hey @rsanandres-wandb thanks for the quick response. Our cluster is setup like this: - Frontend nodes: there are a couple of front-end nodes that are not used for compute-heavy workloads....

Well, this way works to manually create an artifact, but it is far from ideal. You can not use the filtering method on artifacts to filter for runs etc. Unless...

I can give a small update on this if you put `should_deduplicate=True` in your generation step, at least the trials won't be repeated.

You have to redo the normalization. The adversarial images will be between 0-1 (or your bounds). But your model expects something else. I had the same problem. This is easy...