Continual Hyperparameter Selection Framework. Compares 11 state-of-the-art Lifelong Learning methods and 4 baselines. Official Codebase of "A continual learning survey: Defying forgetting in classification tasks." in IEEE TPAMI.
This is the original source code for the Continual Learning survey paper "A continual learning survey: Defying forgetting in classification tasks" published at TPAMI [TPAMI paper] [Open-Access paper].
This work allows comparing the state-of-the-art in a fair fashion using the Continual Hyperparameter Framework, which sets the hyperparameters dynamically based on the stability-plasticity dilemma. This addresses the longstanding problem in literature to set hyperparameters for different methods in a fair fashion, using ONLY the current task data (hence without using iid validation data, which is not available in continual learning).
The code contains a generalizing framework for 11 SOTA methods and 4 baselines in Pytorch. Implemented task-incremental methods are
SI | EWC | MAS | mean/mode-IMM | LWF | EBLL | PackNet | HAT | GEM | iCaRL
These are compared with 4 baselines:
Joint | Finetuning | Finetuning-FM | Finetuning-PM
This source code is released under a Attribution-NonCommercial 4.0 International license, find out more about it in the LICENSE file.
Reproducibility: Results from the paper can be obtained from src/main_'dataset'.sh. Full pipeline example in src/main_tinyimagenet.sh .
Pipeline: Constructing a custom pipeline typically requires the following steps.
conda create --name <ENV-NAME> python=3.7
conda activate <ENV-NAME>
# Main packages
conda install -c conda-forge matplotlib tqdm
conda install pytorch torchvision cudatoolkit=10.2 -c pytorch
# For GEM QP
conda install -c omnia quadprog
# For PackNet: torchnet
pip install git+https://github.com/pytorch/tnt.git@master
@ARTICLE{delange2021clsurvey,
author={M. {Delange} and R. {Aljundi} and M. {Masana} and S. {Parisot} and X. {Jia} and A. {Leonardis} and G. {Slabaugh} and T. {Tuytelaars}},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
title={A continual learning survey: Defying forgetting in classification tasks},
year={2021},volume={},number={},pages={1-1},
doi={10.1109/TPAMI.2021.3057446}}