Fixed: Training of AoANet Model

Wangt-CN / VC-R-CNN

[CVPR 2020] The official pytorch implementation of ``Visual Commonsense R-CNN''

MIT License

352 stars 61 forks source link

Fixed: Training of AoANet Model #15

Open alfred100p opened 3 years ago

alfred100p commented 3 years ago

There are two errors while training: wrong order of parameters caused the function call to not function properly in misc/utils.py.
dtype of a tensor was not casted to the expected type in models/Attmodel.py.