Skip to content
95.16% on CIFAR10 with PyTorch
Branch: master
Clone or download
Pull request Compare This branch is even with kuangliu:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

Train CIFAR10 with PyTorch

I'm playing with PyTorch on the CIFAR10 dataset.

Pros & cons


  • Built-in data loading and augmentation, very nice!
  • Training is fast, maybe even a little bit faster.
  • Very memory efficient!


  • No progress bar, sad :(
  • No built-in log.


Model Acc.
VGG16 92.64%
ResNet18 93.02%
ResNet50 93.62%
ResNet101 93.75%
MobileNetV2 94.43%
ResNeXt29(32x4d) 94.73%
ResNeXt29(2x64d) 94.82%
DenseNet121 95.04%
PreActResNet18 95.11%
DPN92 95.16%

Learning rate adjustment

I manually change the lr during training:

  • 0.1 for epoch [0,150)
  • 0.01 for epoch [150,250)
  • 0.001 for epoch [250,350)

Resume the training with python --resume --lr=0.01

You can’t perform that action at this time.