---

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.

Introduction

Environment setup

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

Experiments

Toy

Go to folder toy_experiments, and follow the instructions therein.

multiMNIST image classification

Go to folder multiMNIST, and follow the instructions therein.

ASR

Go to folder ASR, and follow the instructions therein.

License

MIT license

Citation

@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}
}

Ackowledgement

• 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.