CUDA kernel errors

[Fujiaoji]

I meet the issue with "CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect" when I run run_celeba64.py with eh data_resize_celeba.py for auto encoder when running in my own dataset. Do you have any idea? At first I think it is because of cuda or the driver, but I reinstall the driver and it occurs again and destroy my coda again. Thanks