A simple pytorch implementation of UDA for image classification.
- Install
pip install -r requirements.txt
- Generate training data
python data/random_augmenter.py "sample_directory" "output_augment_directory" "number_of_copies_per_sample"
python data/random_augmenter.py data/images data/aug_images 100
- Training
Change to your model
Change config.py (attent at the paths in label_data, unlabel_data)
run train_uda.py