comaniac commented on issue #4995: [AutoTVM] Avoid using RPC for LocalRunner URL: https://github.com/apache/incubator-tvm/pull/4995#issuecomment-595998520 @merrymercy I tested the local runner on Nvidia T4 and observed the issue you mentioned, so I put the get device attributes logic to a separate process so that the master process will not access the GPU context at all. On the other hand, there's still an issue. The tuning process may still encounter initialization errors after a number of trials. I need to further investigate the root cause.
---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
