pytorch-parallel
pytorch-parallel copied to clipboard
How to debug CUDA codes?
I use the python script to call CUDA codes. Could you provide any suggestions for debugging CUDA codes?
It annoys me a lot, too. I will push some cases about debugging, maybe days later. As for other suggestions, if u can do a hard job, unit tests help a lot.
Thanks for your reply. I also have noticed that you delivered the message in your zhihu article (https://zhuanlan.zhihu.com/p/48463543) you will provide ideas on how to debug cuda codes. I am looking forward to your cases deeply.