We Are Going To Discuss About Pytorch RuntimeError: CUDA out of memory with a huge amount of free memory. So lets Start this Python Article.
Pytorch RuntimeError: CUDA out of memory with a huge amount of free memory
- How to solve Pytorch RuntimeError: CUDA out of memory with a huge amount of free memory
Your problem may be due to fragmentation of your GPU memory.You may want to empty your cached memory used by caching allocator.
import torch torch.cuda.empty_cache() - Pytorch RuntimeError: CUDA out of memory with a huge amount of free memory
Your problem may be due to fragmentation of your GPU memory.You may want to empty your cached memory used by caching allocator.
import torch torch.cuda.empty_cache()
Solution 1
Your problem may be due to fragmentation of your GPU memory.You may want to empty your cached memory used by caching allocator.
import torch
torch.cuda.empty_cache()
Original Author Erol Gelbul Of This Content
Solution 2
I was trying this command:
python3 val.py --weights ./weights/yolov5l-xs-1.pt --img 1996 --data ./data/VisDrone.yaml
and I have a 24G Titan video Card.
Then I reduced the image size and worked for me. to:
python3 val.py --weights ./weights/yolov5l-xs-1.pt --img 1280 --data ./data/VisDrone.yaml
Results:
Class Images Labels P R [email protected] [email protected]:.95: 100%|████████████████████████████████| 18/18 [00:50<00:00, 2.79s/it]
all 548 38759 0.653 0.537 0.584 0.375
pedestrian 548 8844 0.74 0.631 0.708 0.375
people 548 5125 0.677 0.506 0.574 0.258
bicycle 548 1287 0.541 0.377 0.41 0.213
car 548 14064 0.828 0.868 0.904 0.681
van 548 1975 0.636 0.566 0.601 0.453
truck 548 750 0.595 0.516 0.538 0.388
tricycle 548 1045 0.601 0.416 0.457 0.288
awning-tricycle 548 532 0.387 0.242 0.245 0.173
bus 548 251 0.782 0.653 0.725 0.565
motor 548 4886 0.744 0.598 0.674 0.355
Original Author Oscar Rangel Of This Content
Conclusion
So This is all About This Tutorial. Hope This Tutorial Helped You. Thank You.
