Code example for the ICLR 2018 oral paper I've taken the code commented it up a bit more, made it python 3 compatible, added some more options, and linked the options
- NVIDIA GPU + CUDA + CuDNN
- Tensorflow (GPU version)
- python 3
- tqdm
Download and generate CIFAR10 dataset:
cd dataSet/
python CIFAR10.py
Change your configurations in the file
gedit source/Option.py
Start training:
cd source/
python Top.py