zhouyizhuang-megvii / RepLKNet

RepLKNet (CVPR 2022)

This is the official MegEngine implementation of RepLKNet, from the following CVPR-2022 paper:

Scaling Up Your Kernels to 31x31: Revisiting Large Kernel Design in CNNs.

The paper is released on arXiv: https://arxiv.org/abs/2203.06717.

Other implementations

framework	link
PyTorch (official)	https://github.com/DingXiaoH/RepLKNet-pytorch
Tensorflow	re-implementations are welcomed
PaddlePaddle	re-implementations are welcomed
...

Catalog

• Model code
• MegEngine pretrained models
• MegEngine training code
• MegEngine downstream models
• MegEngine downstream code

Results and Pre-trained Models

ImageNet-1K Models

name	resolution	acc	#params	FLOPs	download

ImageNet-22K Models

name	resolution	acc	#params	FLOPs	22K model	1K model

MegData-73M Models

name	resolution	acc@1	#params	FLOPs	MegData-73M model	1K model

Installation of MegEngine

Please check the HomePage.

Evaluation

Training

License

This project is released under the MIT license. Please see the LICENSE file for more information.