Cadene/pretrained-models.pytorch
A collection of pretrained ConvNet architectures (ResNet, InceptionV4, NASNet, Xception, etc.) for PyTorch with a unified API.

Velocity · 7d
+2.7
★ / day
Trend
→steady
star history
This repository provides pretrained convolutional neural network models for PyTorch, including architectures like ResNeXt, ResNet, InceptionV4, InceptionResNetV2, Xception, DPN, NASNet, SENet, and PolyNet. It offers a unified API inspired by torchvision for easy model loading, feature extraction, and transfer learning. The goal is to help researchers reproduce results from papers and facilitate transfer learning setups.