We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
e0780d0
Pretrained models for Stylized ImageNet (SIN), no pos-encoding, and different patch size experiments.