shengdewu/pwlq

Code for our paper at ECCV 2020: Post-Training Piecewise Linear Quantization for Deep Neural Networks

PWLQ

Updates

2020/07/16 - We are working on getting permission from our institution to release our source code. We will release it once we are granted approval.

2020/08/24 - Code released.

---

PyTorch Code for our paper at ECCV 2020 (oral presentation): Post-Training Piecewise Linear Quantization for Deep Neural Networks [Paper] [arXiv]

By Jun Fang, Ali Shafiee, Hamzah Abdel-Aziz, David Thorsley, Georgios Georgiadis, Joseph Hassoun

• Approach

• Performance

Requirements

The code was verified on Python-3.6+, PyTorch-1.2+.

Usage

Check PWLQ at quant/pwlq.py.

Run bash eval.sh to evaluate PWLQ on ImageNet. Results would be recorded at results/*.csv.

Results might be slightly different due to the randomness of calibration samples for activation ranges.

Citation

We appreciate it if you would please cite our paper:

@inproceedings{pwlq,
  title={Post-Training Piecewise Linear Quantization for Deep Neural Networks},
  author={Fang, Jun and Shafiee, Ali and Abdel-Aziz, Hamzah and Thorsley, David and Georgiadis, Georgios and Hassoun, Joseph},
  booktitle={ECCV},
  year={2020}
}