Loading model with torch.hub.load()

WongKinYiu / yolov7

Implementation of paper - YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors

GNU General Public License v3.0

13.44k stars 4.23k forks source link

Loading model with torch.hub.load() #687

Open RK-BAKU opened 2 years ago

RK-BAKU commented 2 years ago

Loading model with torch.hub.load() always loads model as fp16 which is not supported on cuda thus results in slower inference. Is there any option in kwargs to disable half() during loading? Thanks in advance!

messier20 commented 2 years ago

@RK-BAKU how do you load the model with torch.hub.load()? For me it doesn't even load