reason
The root cause is that the GPU configuration is not enough, resulting in an error
solution 1
Modify the yolov5 code, modify the file in yolov5\utils\datasets.py
Modify the parameter num_workers
to 0
But this will bring a problem, that is, the speed is much slower when training.
solution 2
- Ctrl+q search: "view advanced system settings"
- Click Advanced - Settings of Performance
- Click Advanced - Change
- Then click on the disk where the project is located - custom size (depending on how much free space you have on your disk, you can directly fill the remaining space. Because it only occupies virtual memory when running, it will not affect it actually, don't worry) - Settings (must be The clicked, otherwise it will not work!!) - OK
