CL_Visualization
DEV NoTEs
Typical command
python main.py --scenario_name SpuriousFeatures --num_tasks 2 --name_algo ib_irm --dataset CIFAR10 --fast --lr 0.1 --seed 1664 --OutLayer Linear
This repository is a research repository there is no warranty on the current state of the code.
Nevertheless, this repository contains code to reproduce some paper experiments. In the "Reproducibility" folder you can find information to reproduce results for the papers:
- "Regularization Shortcomings for Continual Learning" - Timothée Lesort, Andrei Stoian, David Filliat
- "Continual Learning in Deep Networks: an Analysis of the Last Layer" - Timothée Lesort, Thomas George, Irina Rish