How to use multiple GPUs to train and test the network?

[lhc11171552]

I used two GPUs to train the network, and was an error： ValueError: Error initializing torch.distributed using env:// rendezvous: environment variable RANK expected, but not set

[michuanhaohao]

CUDA_VISIBLE_DEVICES=0,1,2,3 python -m torch.distributed.launch --nproc_per_node=4 --master_port 66666 train.py --config_file configs/VehicleID/vit_transreid_stride.yml MODEL.DIST_TRAIN True

Sometimes you may encounter some errors because of the cuda driver. Please run the shell again after killing CUDA pids (if necessary).

[michuanhaohao]

Two gpus:

CUDA_VISIBLE_DEVICES=0,1
--nproc_per_node=2
Batch size * 2

[lhc11171552]

Thanks,I will have a try