wwlCape / HAN

PyTorch code for our ECCV 2020 paper "Single Image Super-Resolution via a Holistic Attention Network"
Apache License 2.0
167 stars 32 forks source link

about self.gamma #12

Open fenghao195 opened 2 years ago

fenghao195 commented 2 years ago

Hi,I find that self.gamma is not trained,and requires_grad of tensor is True.model do not contain it