PyTorch pruning example for ResNet. ResNet18 pre-trained on CIFAR-10 dataset maintains the same prediction accuracy with 50x compression after pruning.
$ docker build -f docker/pytorch.Dockerfile --no-cache --tag=pytorch:1.13.0 .
$ docker run -it --rm --gpus device=0 -v $(pwd):/mnt pytorch:1.13.0
$ python pretrain.py
$ python prune.py