Provide mirror for model files in research network like DFN X-WiN

[paulmenzel]

Download from the MPI for Molecular Genetics, the download rates are comparatively slow for storage.googleapis.com:

$ python scripts/download_model.py
Downloading stylegan2_lions_512_pytorch.pkl from https://storage.googleapis.com/self-distilled-stylegan/lions_512_pytorch.pkl
100%|█████████████████████████████| 364M/364M [00:12<00:00, 30.0MiB/s]
Downloading stylegan2_dogs_1024_pytorch.pkl from https://storage.googleapis.com/self-distilled-stylegan/dogs_1024_pytorch.pkl
100%|█████████████████████████████| 382M/382M [00:12<00:00, 30.1MiB/s]
Downloading stylegan2_horses_256_pytorch.pkl from https://storage.googleapis.com/self-distilled-stylegan/horses_256_pytorch.pkl
100%|█████████████████████████████| 357M/357M [00:11<00:00, 31.3MiB/s]
Downloading stylegan2_elephants_512_pytorch.pkl from https://storage.googleapis.com/self-distilled-stylegan/elephants_512_pytorch.pkl
100%|█████████████████████████████| 364M/364M [00:11<00:00, 32.1MiB/s]
Downloading stylegan2-ffhq-512x512.pkl from https://api.ngc.nvidia.com/v2/models/nvidia/research/stylegan2/versions/1/files/stylegan2-ffhq-512x512.pkl
100%|█████████████████████████████| 364M/364M [00:28<00:00, 12.7MiB/s]
Downloading stylegan2-afhqcat-512x512.pkl from https://api.ngc.nvidia.com/v2/models/nvidia/research/stylegan2/versions/1/files/stylegan2-afhqcat-512x512.pkl
100%|█████████████████████████████| 364M/364M [00:31<00:00, 11.7MiB/s]
Downloading stylegan2-car-config-f.pkl from http://d36zk2xti64re0.cloudfront.net/stylegan2/networks/stylegan2-car-config-f.pkl
100%|██████████████████████████████| 364M/364M [00:03<00:00, 117MiB/s]
Downloading stylegan2-cat-config-f.pkl from http://d36zk2xti64re0.cloudfront.net/stylegan2/networks/stylegan2-cat-config-f.pkl
100%|██████████████████████████████| 357M/357M [00:03<00:00, 115MiB/s]

It’d be great if you hosted the model files also on the systems of your institute.