Skip to content

This is the formal code implementation of the CVPR 2024 paper 'Traceable Federated Continual Learning'.

License

Notifications You must be signed in to change notification settings

P0werWeirdo/TagFCL

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Official Pytorch Implementation for TagFed

This is the implementation code of the CVPR 2024 paper "Traceable Federated Continual Learning"

Framework:

overview

Prerequisites:

* python == 3.6
* torch == 1.2.0
* numpy
* PIL
* torchvision == 0.4.0
* cv2
* scipy == 1.5.2
* sklearn == 0.24.1

Datasets:

  • CIFAR100: You don't need to do anything before running the experiments on CIFAR100 dataset.

  • Imagenet-Subset (Mini-Imagenet): Please manually download the on Imagenet-Subset (Mini-Imagenet) dataset from the official websites, and place it in './train'.

  • Tiny-Imagenet: Please manually download the on Tiny-Imagenet dataset from the official websites, and place it in './tiny-imagenet-200'.

Training:

  • Please check the detailed arguments in './src/option.py'.
python main.py

Performance:

performance

Acknowledge:

We would like to thank the following related open source works:

  1. [CVPR-2022] Federated Class-Incremental Learning [Code]
  2. [NeurIPS-2019] Compacting, Picking and Growing for Unforgetting Continual Learning Code

About

This is the formal code implementation of the CVPR 2024 paper 'Traceable Federated Continual Learning'.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages