search

ifzhang / ByteTrack

[ECCV 2022] ByteTrack: Multi-Object Tracking by Associating Every Detection Box
MIT License
4.8k stars 906 forks source link

Is there a way to run demo_track.py with multiple GPU? #224

Open ashjostan opened 2 years ago

ashjostan commented 2 years ago

Currently the args for device are 'gpu' and 'cpu'. Is there a way to run the script with multiple gpu? I have tried the following and they do not work-

Option1: Inputting GPU ranks directly: --device '0,1'

Option2: Including the following code snippet in the script before import torch- os.environ["CUDA_DEVICE_ORDER"]="PCI_BUS_ID" os.environ["CUDA_VISIBLE_DEVICES"]="0,1"

Thanks for the repo and would appreciate any help!