Training source code of the RSNA Screening Mammography Breast Cancer Detection competition, 26th.
The original repo I've been used the whole time has some sensitive configs(like wandb/kaggle/neptune tokens), so I just mirrored some important files to this repo. Just to demonstrate how the work is done. It should change some codes to make the training work.
This is the first time I use the PyTorch-Tao in action, and I'm happy with most part of it, but some refactor still need to be done.