Landmarks Detection for Chest Image

Usage

1. training on single gpu:

 python train.py --config-file your_config_file

e.g. python train.py --config-file experiments/template/landmark_detection_template.yaml

2. training on multiple gpus (nodes):

python -m torch.distributed.launch \
   --nproc_per_node=4 \
   train_fp16.py --config-file your_config_file