This is the official implementation of ICLR paper Designing BERT for Convolutional Networks: Sparse and Hierarchical Masked Modeling, which can pretrain any CNN (e.g., ResNet) in a BERT-style self-supervised manner. We've tried our best to make the codebase clean, short, easy to read, state-of-the-art, and only rely on minimal dependencies.
SparK_demo_22s_4k_wo_bages.1.mp4
- A brief introduction (in English) is available on our ICLR poster page! [
📹Recorded Video, Poster, and Slides]. - On May. 11th another livestream on OpenMMLab & ReadPaper (bilibili)! [
📹Recorded Video] - On Apr. 27th (UTC+8 8pm) another livestream would be held at OpenMMLab (bilibili)!
- On Mar. 22nd (UTC+8 8pm) another livestream would be held at 极市平台 (bilibili)! [
📹Recorded Video] - The share on TechBeat (将门创投) is scheduled on Mar. 16th (UTC+8 8pm) too! [
📹Recorded Video] - We are honored to be invited by Synced ("机器之心机动组 视频号" on WeChat) to give a talk about SparK on Feb. 27th (UTC+0 11am, UTC+8 7pm), welcome! [
📹Recorded Video] - This work got accepted to ICLR 2023 as a Spotlight (notable-top-25%).
- Other articles: [
Synced] [DeepAI] [TheGradient] [Bytedance] [CVers[QbitAI(量子位)] [BAAI(智源)] [机器之心机动组] [极市平台] [ReadPaper笔记]
Check pretrain/viz_reconstruction.ipynb for visualizing the reconstruction of SparK pretrained models, like:
We also provide pretrain/viz_spconv.ipynb that shows the "mask pattern vanishing" issue of dense conv layers.
See our paper for more analysis, discussions, and evaluations.
catalog
- Pretraining code
- Pretraining toturial for customized CNN model (Tutorial for pretraining your own CNN model)
- Pretraining toturial for customized dataset (Tutorial for pretraining your own dataset)
- Pretraining Colab visualization playground (reconstruction, sparse conv)
- Finetuning code
- Weights & visualization playground in
huggingface - Weights in
timm
Note: for network definitions, we directly use timm.models.ResNet and official ConvNeXt.
reso.: the image resolution; acc@1: ImageNet-1K finetuned acc (top-1)
| arch. | reso. | acc@1 | #params | flops | weights (self-supervised, without SparK's decoder) |
|---|---|---|---|---|---|
| ResNet50 | 224 | 80.6 | 26M | 4.1G | resnet50_1kpretrained_timm_style.pth |
| ResNet101 | 224 | 82.2 | 45M | 7.9G | resnet101_1kpretrained_timm_style.pth |
| ResNet152 | 224 | 82.7 | 60M | 11.6G | resnet152_1kpretrained_timm_style.pth |
| ResNet200 | 224 | 83.1 | 65M | 15.1G | resnet200_1kpretrained_timm_style.pth |
| ConvNeXt-S | 224 | 84.1 | 50M | 8.7G | convnextS_1kpretrained_official_style.pth |
| ConvNeXt-B | 224 | 84.8 | 89M | 15.4G | convnextB_1kpretrained_official_style.pth |
| ConvNeXt-L | 224 | 85.4 | 198M | 34.4G | convnextL_1kpretrained_official_style.pth |
| ConvNeXt-L | 384 | 86.0 | 198M | 101.0G | convnextL_384_1kpretrained_official_style.pth |
Pretrained weights (with SparK's UNet-style decoder; can be used to reconstruct images)
| arch. | reso. | acc@1 | #params | flops | weights (self-supervised, with SparK's decoder) |
|---|---|---|---|---|---|
| ResNet50 | 224 | 80.6 | 26M | 4.1G | res50_withdecoder_1kpretrained_spark_style.pth |
| ResNet101 | 224 | 82.2 | 45M | 7.9G | res101_withdecoder_1kpretrained_spark_style.pth |
| ResNet152 | 224 | 82.7 | 60M | 11.6G | res152_withdecoder_1kpretrained_spark_style.pth |
| ResNet200 | 224 | 83.1 | 65M | 15.1G | res200_withdecoder_1kpretrained_spark_style.pth |
| ConvNeXt-S | 224 | 84.1 | 50M | 8.7G | cnxS224_withdecoder_1kpretrained_spark_style.pth |
| ConvNeXt-L | 384 | 86.0 | 198M | 101.0G | cnxL384_withdecoder_1kpretrained_spark_style.pth |
We highly recommended you to use torch==1.10.0, torchvision==0.11.1, and timm==0.5.4 for reproduction.
Check INSTALL.md to install all pip dependencies.
- Loading pretrained model weights in 3 lines
# download our weights `resnet50_1kpretrained_timm_style.pth` first
import torch, timm
res50, state = timm.create_model('resnet50'), torch.load('resnet50_1kpretrained_timm_style.pth', 'cpu')
res50.load_state_dict(state.get('module', state), strict=False) # just in case the model weights are actually saved in state['module']-
Pretraining
- any ResNet or ConvNeXt on ImageNet-1k: see pretrain/
- your own CNN model: see pretrain/, especially pretrain/models/custom.py
-
Finetuning
- any ResNet or ConvNeXt on ImageNet-1k: check downstream_imagenet/ for subsequent instructions.
- ResNets on COCO: see downstream_d2/
- ConvNeXts on COCO: see downstream_mmdet/
We referred to these useful codebases:
This project is under the MIT license. See LICENSE for more details.
If you found this project useful, you can kindly give us a star ⭐, or cite us in your work 📖:
@Article{tian2023designing,
author = {Keyu Tian and Yi Jiang and Qishuai Diao and Chen Lin and Liwei Wang and Zehuan Yuan},
title = {Designing BERT for Convolutional Networks: Sparse and Hierarchical Masked Modeling},
journal = {arXiv:2301.03580},
year = {2023},
}