K-H-Ismail/ConvNeXt-dcls

A ConvNet for the 2020s

Dilated Convolution with Learnable Spacings

This is an implementation of the experiments using ConvNeXt-dcls models described in [arXiv].

To install DCLS please follow the link: DCLS.

Please check INSTALL.md for installation instructions.

Changed files

• main.py : minor
• engine.py : add clamping, rep loss, and scaling the lr for positions
• optim_factory.py : add scaling for the lr of positions
• utils.py : define rep loss.
• run_with_submitit.py : adapt to cluster specificities

Added files

• hostlists.py
• configs for ConvNeXt-dcls backbones in semantic segmentation
• configs for ConvNeXt-dcls backbones in object detection
• models/convnext_dcls.py : the new model !

Added Backbones

• ConvNeXt-T-dcls
• ConvNeXt-S-dcls
• ConvNeXt-B-dcls
• Fine tuned models for object detection and semantic segmentation using the three backbones above.

A ConvNet for the 2020s

Official PyTorch implementation of ConvNeXt, from the following paper:

A ConvNet for the 2020s. CVPR 2022.
Zhuang Liu, Hanzi Mao, Chao-Yuan Wu, Christoph Feichtenhofer, Trevor Darrell and Saining Xie
Facebook AI Research, UC Berkeley
[arXiv][video]

---

We propose ConvNeXt, a pure ConvNet model constructed entirely from standard ConvNet modules. ConvNeXt is accurate, efficient, scalable and very simple in design.

Catalog

• ImageNet-1K Training Code
• ImageNet-22K Pre-training Code
• ImageNet-1K Fine-tuning Code
• Downstream Transfer (Detection, Segmentation) Code
• Image Classification [Colab] and Web Demo
• Fine-tune on CIFAR with Weights & Biases logging [Colab]

Results and Pre-trained Models

ImageNet-1K trained models

name	resolution	acc@1	#params	FLOPs	model
ConvNeXt-T	224x224	82.1	28M	4.5G	model
ConvNeXt-T-dcls	224x224	82.5	28M	5.0G	model
ConvNeXt-S	224x224	83.1	50M	8.7G	model
ConvNeXt-S-dcls	224x224	83.7	50M	9.5G	model
ConvNeXt-B	224x224	83.8	89M	15.4G	model
ConvNeXt-B-dcls	224x224	84.1	89M	16.5G	model
ConvNeXt-B	384x384	85.1	89M	45.0G	model
ConvNeXt-L	224x224	84.3	198M	34.4G	model
ConvNeXt-L	384x384	85.5	198M	101.0G	model

ImageNet-22K trained models

name	resolution	acc@1	#params	FLOPs	22k model	1k model
ConvNeXt-T	224x224	82.9	29M	4.5G	model	model
ConvNeXt-T	384x384	84.1	29M	13.1G	-	model
ConvNeXt-S	224x224	84.6	50M	8.7G	model	model
ConvNeXt-S	384x384	85.8	50M	25.5G	-	model
ConvNeXt-B	224x224	85.8	89M	15.4G	model	model
ConvNeXt-B	384x384	86.8	89M	47.0G	-	model
ConvNeXt-L	224x224	86.6	198M	34.4G	model	model
ConvNeXt-L	384x384	87.5	198M	101.0G	-	model
ConvNeXt-XL	224x224	87.0	350M	60.9G	model	model
ConvNeXt-XL	384x384	87.8	350M	179.0G	-	model

ImageNet-1K trained models (isotropic)

name	resolution	acc@1	#params	FLOPs	model
ConvNeXt-S	224x224	78.7	22M	4.3G	model
ConvNeXt-B	224x224	82.0	87M	16.9G	model
ConvNeXt-L	224x224	82.6	306M	59.7G	model

Installation

Please check INSTALL.md for installation instructions.

Evaluation

We give an example evaluation command for a ImageNet-22K pre-trained, then ImageNet-1K fine-tuned ConvNeXt-B:

Single-GPU

python main.py --model convnext_dcls_base --eval true \
--resume https://zenodo.org/record/7112021/files/convnext_dcls_base_1k_224_ema.pth \
--input_size 224 --drop_path 0.2 \
--data_path /path/to/imagenet-1k

Multi-GPU

python -m torch.distributed.launch --nproc_per_node=8 main.py \
--model convnext_dcls_base --eval true \
--resume https://zenodo.org/record/7112021/files/convnext_dcls_base_1k_224_ema.pth \
--input_size 224 --drop_path 0.2 \
--data_path /path/to/imagenet-1k

This should give for the convNeXt-B-dcls model

* Acc@1 84.094 Acc@5 96.958 loss 0.653

• For evaluating other model variants, change --model, --resume, --input_size accordingly. You can get the url to pre-trained models from the tables above.
• Setting model-specific --drop_path is not strictly required in evaluation, as the DropPath module in timm behaves the same during evaluation; but it is required in training. See TRAINING.md or our paper for the values used for different models.

Training

See TRAINING.md for training and fine-tuning instructions.

Acknowledgement

This repository is built using the timm library, DeiT and BEiT repositories.

License

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

Citation

If you find this repository helpful, please consider citing:

@Article{liu2022convnet,
  author  = {Zhuang Liu and Hanzi Mao and Chao-Yuan Wu and Christoph Feichtenhofer and Trevor Darrell and Saining Xie},
  title   = {A ConvNet for the 2020s},
  journal = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year    = {2022},
}