Update Documentation to Make CUDA Call Explicit (#67973)
Summary:
I am clarifying in the docs to make the call to cudaStreamWaitEvent explicit.
Fixes https://github.com/pytorch/pytorch/issues/67866
Pull Request resolved: https://github.com/pytorch/pytorch/pull/67973
Reviewed By: mruberry
Differential Revision: D32620261
Pulled By: ngimel
fbshipit-source-id: 1fc8beb2062baaddb013ea4d7b10da2baa10f15e