JittorCLS

A codebase for image classification based on Jittor

UPDATE

Add More models

Res2Net: A New Multi-scale Backbone Architecture [TPAMI'21]

Related links: [paper] [中译版全文] [github]

@article{gao2019res2net,
  title={Res2Net: A New Multi-scale Backbone Architecture},
  author={Gao, Shang-Hua and Cheng, Ming-Ming and Zhao, Kai and Zhang, Xin-Yu and Yang, Ming-Hsuan and Torr, Philip},
  journal={IEEE TPAMI},
  year={2021},
  doi={10.1109/TPAMI.2019.2938758}, 
}

Pyramid Pooling Transformer for Scene Understanding [TPAMI'22]

Related links: [paper] [中译版全文] [github]

@ARTICLE{wu2022p2t,
  author={Wu, Yu-Huan and Liu, Yun and Zhan, Xin and Cheng, Ming-Ming},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}, 
  title={{P2T}: Pyramid Pooling Transformer for Scene Understanding}, 
  year={2022},
  doi = {10.1109/tpami.2022.3202765},
}

Conv2Former: A Simple Transformer-Style ConvNet for Visual Recognition [TPAMI'24]

Related links: [paper] [中译版全文] [github]

@article{hou2024conv2former,
  title={Conv2Former: A Simple Transformer-Style ConvNet for Visual Recognition},
  author={Hou, Qibin and Lu, Cheng-Ze and Cheng, Ming-Ming and Feng, Jiashi},
  journal={IEEE TPAMI},
  year={2024},
  doi={10.1109/TPAMI.2024.3401450}, 
}

Getting Started

Install Jittor

sudo apt install python3.7-dev libomp-dev
python3.7 -m pip install jittor
python3.7 -m jittor.test.test_example
# If your computer contains an Nvidia graphics card, check the cudnn acceleration library
python3.7 -m jittor.test.test_cudnn_op

For more information on how to install jittor, you can check here.

Install OpenMPI

sudo apt install openmpi-bin openmpi-common libopenmpi-dev

To obtain more information about OpenMPI, you can check here.

Train

We provide scripts for single-machine single-gpu, single-machine multi-gpu training. Multi-gpu dependence can be referred to here

# Single GPU
bash train.sh

# Multiple GPUs
bash dist_train.sh

Test

# Single GPU
bash test.sh

# Multiple GPUs
bash dist_test.sh