You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2021/09/02 18:30:28 UTC

[GitHub] [tvm] tqchen commented on pull request #8492: [AutoScheduler] Use PopenPool instead of multiprocessing.pool

tqchen commented on pull request #8492:
URL: https://github.com/apache/tvm/pull/8492#issuecomment-911949552


   we can serialize callable functions(through cloud pickle), but the requirement is that the callable functions can only call functions in the tvm namespace and should be available in https://github.com/apache/tvm/blob/main/python/tvm/exec/popen_worker.py
   
   It does mean that if we call functions that are not by default imported, we might need to do
   
   ```c++
   def custom_func():
       # import here so things are available in the callee side
       import dep1
       dep1.xyz
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org