This is a markdown file test.
- Python (3.7.10)
- Pytorch (1.7.1)
- torchvision (0.8.2)
- CUDA
- Numpy
datasets
├──cifar100
PADE-master
├── README.md
├── ...
python train_cifar.py --lr 0.1 --epochs 200 --lam 0.5 --t 1. --pa True --resume None
python test_all_cifar.py --resume log/cifar100_resnet32_3_0.01_4/checkpoint.best.pth.tar --num_experts 3 --num_classes 100 --pa True