A PyTorch implementation of Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks
- Python 3 (tested with Python 3.7)
- PyTorch (tested with torch v1.3 and torchvision v0.4)
- Python packages as specified in requirements.txt
$ git clone https://github.com/RashedDoha/Unsupervised-Pixel-Level-Domain-Adaptation.git
$ cd Unsupervised-Pixel-Level-Domain-Adaptation/
$ sudo pip3 install -r requirements.txt
Training on CPU
python train.py
To train on a CUDA enabled GPU, use the --cuda flag: python train.py --cuda
To launch tensorboard make sure to set the --logdir
flag with the same directory as in the command line flag for train.py --tensorboard_logs
. (default: logs
)
To contribute to the project, please refer to the contribution guidelines