This repository provides a PyTorch implementation of SENet, a convolutional neural network architecture introduced in the paper Squeeze-and-Excitation Networks by Jie Hu et al., which won the ILSVRC 2017 classification competition. The implementation includes SE-ResNet variants (20, 56, 50, 101, 152) and SE-Inception-v3 with training scripts for CIFAR-10 and ImageNet datasets. Pretrained models are available via torch.hub.