RepVGG: Making VGG-style ConvNets Great Again

Code

• github.com/DingXiaoH/RepVGG
  Officialpytorch★ 3,466
• github.com/rwightman/pytorch-image-models
  pytorch★ 36,538
• github.com/xmu-xiaoma666/External-Attention-pytorch
  pytorch★ 12,169
• github.com/PaddlePaddle/PaddleClas
  paddle★ 5,788
• github.com/Deci-AI/super-gradients
  pytorch★ 5,017
• github.com/open-mmlab/mmclassification
  pytorch★ 3,839
• github.com/frgfm/Holocron
  pytorch★ 328
• github.com/ZJCV/ZCls
  pytorch★ 143
• github.com/upczww/TensorRT-RepVGG
  pytorch★ 81
• github.com/lmk123568/RepVGG_Tutorial
  none★ 49

Abstract

We present a simple but powerful architecture of convolutional neural network, which has a VGG-like inference-time body composed of nothing but a stack of 3x3 convolution and ReLU, while the training-time model has a multi-branch topology. Such decoupling of the training-time and inference-time architecture is realized by a structural re-parameterization technique so that the model is named RepVGG. On ImageNet, RepVGG reaches over 80% top-1 accuracy, which is the first time for a plain model, to the best of our knowledge. On NVIDIA 1080Ti GPU, RepVGG models run 83% faster than ResNet-50 or 101% faster than ResNet-101 with higher accuracy and show favorable accuracy-speed trade-off compared to the state-of-the-art models like EfficientNet and RegNet. The code and trained models are available at https://github.com/megvii-model/RepVGG.

Tasks

Benchmark Results

Reproductions