This repo contains the reference source code in PyTorch of the FedoSSL algorithm. For more details please check our paper Towards Unbiased Training in Federated Open-world Semi-supervised Learning (ICML '23).
The code is built with the following libraries:
We use SimCLR for pretraining. The weights used in our paper can be downloaded in this link.
- To train on CIFAR-100, run
python fedossl_cifar.py --dataset cifar100 --labeled-num 50 --labeled-ratio 0.5
- To train on ImageNet-100, first use
gen_imagenet_list.py
to generate corresponding splitting lists, then run
python fedossl_imagenet.py --labeled-num 50 --labeled-ratio 0.5