Skip to content

Latest commit

 

History

History
53 lines (43 loc) · 1.21 KB

README.md

File metadata and controls

53 lines (43 loc) · 1.21 KB

Trip-ROMA: Self-Supervised Learning with Triplets and Random Mappings

This is a PyTorch implementation of the Trip-ROMA paper.

Dependencies

If you don't have python 3 environment:

conda create -n trip python=3.8
conda activate trip

Then install the required packages:

pip install -r requirements.txt

Train

  1. Config CONFIG, DATA_DIR and LOG_DIR in script/train.sh or script/train_ddp.sh. We use DDP for ImageNet and STL10 by default.
  2. Run 
    sh scripts/train.sh
    for single gpu training, or 
    sh scripts/train_ddp.sh
    for DDP training.

Test

  1. Config CONFIG, DATA_DIR and LOG_DIR in script/eval.sh.
  2. Run 
    sh scripts/eval.sh

Citation

@article{
li2023triproma,
title={Trip-{ROMA}: Self-Supervised Learning with Triplets and Random Mappings},
author={Wenbin Li and Xuesong Yang and Meihao Kong and Lei Wang and Jing Huo and Yang Gao and Jiebo Luo},
journal={Transactions on Machine Learning Research},
issn={2835-8856},
year={2023}
}

Acknowledgement

PatrickHua/SimSiam