Wok

Have you tried setting `nproc_per_node` to `1`? ```bash python -m torch.distributed.run --nproc_per_node=1 train_retrieval.py \ --config ./configs/retrieval_coco.yaml \ --output_dir output/retrieval_coco ```

Sorry, I don't know the answers to these questions.

> The waifu2x neural network actually uses 2x2 pixel duplication for the input (nearest neighbor) Do you have a reference for this? If I quickly check the paper, I see...

I think donations have a special status, so that you can receive a small amount per year without much hassle. That is what I read about [`ArchiSteamFarm`](https://github.com/JustArchiNET/ArchiSteamFarm), whose author lives...

Relevant thread: - https://github.com/RamonUnch/AltSnap/issues/149

I don't think there is an opposition between attribute-specific channels and localized channels. Basically, if you have classifiers for the attribute of interest, then you can generate "a large number...

You don't need a classifier. Classifiers are used to "annotate a set of generated images", which allows to generate "a large number (1K) of positive examples". If you already have...

#1

#2 Also: 

You cropped the output log which detailed the error returned by NVCC. Other than that: - https://github.com/orpatashnik/StyleCLIP/issues/20#issuecomment-813103530 - https://github.com/NVlabs/stylegan2-ada/issues/91#issuecomment-842066449 - https://github.com/mikaelalafriz/lucid-sonic-dreams/issues/22#issuecomment-814968816