ChineseNMT icon indicating copy to clipboard operation
ChineseNMT copied to clipboard

总是gpu错误

Open gamamasm opened this issue 2 years ago • 2 comments

CUDA error: invalid device ordinal CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect. For debugging consider passing CUDA_LAUNCH_BLOCKING=1.

gamamasm avatar Nov 08 '23 12:11 gamamasm

maybe you can turn to config.py, modify the device_id = [0, 1] according to your GPU situation, hope this can works.

bxy666666 avatar Nov 12 '23 05:11 bxy666666

我是在笔记本上运行。只有一张3060laptop。

我把os.environ['CUDA_VISIBLE_DEVICES'] = '2, 3' 改为了 os.environ['CUDA_VISIBLE_DEVICES'] = '0' 就可以运行了。 希望能够帮到你

WThirteen avatar Mar 28 '24 07:03 WThirteen