（fast-reid）多GPU训练出现RuntimeError: Address already in use解决

https://github.com/JDAI-CV/fast-reid

0.环境 

ubuntu16.04
cuda9.0
python3.6
torch==1.1.0
torchvision==0.3.0
Cython
yacs
tensorboard
future
termcolor
sklearn
tqdm
opencv-python==4.1.0.25
matplotlib
scikit-image
faiss-gpu==1.6.3
tabulate
gdown

1.多次使用多GPU出现错误

RuntimeError: Address already in use

2.解决

usage: train_net.py [-h] [--config-file FILE] [--resume] [--eval-only]
                    [--num-gpus NUM_GPUS] [--num-machines NUM_MACHINES]
                    [--machine-rank MACHINE_RANK] [--dist-url DIST_URL]

 

使用--dist-url参数指定路径，每次使用的都不一样就可以了。

CUDA_VISIBLE_DEVICES='0,1' python ./tools/train_net.py \
 --config-file ./projects/DistillReID/configs-bagtricks-ibn-market1501/bagtricks_R50-ibn.yml \
 --num-gpus 2 --dist-url tcp://127.0.0.1:50001

参考

1.AdelaiDet/issues/149