weiaicunzai/pytorch-cifar100
A PyTorch practice project implementing and training over 30 CNN architectures on the CIFAR-100 image classification dataset.

Velocity · 7d
+1.4
★ / day
Trend
→steady
star history
This repository provides implementations of classic convolutional neural network architectures including ResNet, VGG, DenseNet, GoogleNet, Inception variants, MobileNet, and others. Users can train these models on the CIFAR-100 dataset via command-line arguments, with optional tensorboard logging and warmup training strategies to prevent divergence during early training phases.