Mousius commented on pull request #9129:
URL: https://github.com/apache/tvm/pull/9129#issuecomment-934398890


   > I guess I am open to trying this, but I feel a bit like we should 
publicize this in the forum in case anyone else is attached to the current 
setup. My example came from an internal OctoML ask of me. I think I feel like 
this because I'm not sure we have hard metrics to consult.
   
   I echo @jroesch's concerns here, unless we push the issue people will remain 
with what's comfortable and use the resources for other things. I'd be 
interested in trying it in the spirit of trying an improvement and see what 
issues arise? I actually have a GPU handy yet I haven't configured my local 
tests to use a GPU which meant I was hit by it after switching context in 
https://github.com/apache/tvm/pull/9190 via CI, as a guilty party I genuinely 
believe that being forced to properly set this up is the correct approach. 
   
   I don't mind raising a Discuss thread if it's necessary, but I'm concerned 
we're going to have to integrate everyone's personal development practices 
rather than optimising for the CI use case which the system is meant for.
   
   


-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


Reply via email to