This repository contains the code for the paper's experiments: "FERERO: A Flexible Framework for Preference-Guided Multi-Objective Learning".

We continuously update this repo.

1. Use the following command to install the dependencies

conda create -n ferero python=3.8
conda activate ferero
conda install pytorch torchvision==0.9.0 torchaudio pytorch-cuda=11.7 -c pytorch -c nvidia
conda install numpy scipy seaborn tqdm autograd==1.3
conda install -c conda-forge cvxpy cvxopt matplotlib-label-lines

Go to folder toy_experiments, and follow the instructions therein.

Go to folder multiMNIST, and follow the instructions therein.

Go to folder ASR, and follow the instructions therein.

MIT license

@inproceedings{chen2024FERERO,
  title={FERERO: A Flexible Framework for Preference-Guided Multi-Objective Learning},
  author={Chen, Lisha and Saif, AFM and Shen, Yanning and Chen, Tianyi},
  booktitle={Advances in Neural Information Processing Systems},
  year={2024}
}

• The toy example, multi-MNIST classification, and emotion recognition experiments use the code from PMTL and EPO.
• The multi-lingual ASR experiment uses the code from M2ASR as a baseline.

We thank the authors for providing the code and data. Please cite their works and ours if you use the code or data.