alexwong edited a comment on issue #4497: [Relay] Add a PyTorch to Relay Parser URL: https://github.com/apache/incubator-tvm/pull/4497#issuecomment-585414848 > Do you think you can fix the padding issue? I haven't looked into what is going on, we may need to update cuda topi schedule. > > It is ok if you leave that to me, I can take a look in the follow up PR. I can try taking a look today but I don't have too much experience here and wouldn't want to block this going out if everything else seems okay. https://github.com/zxy844288792/tvm/blob/00fd0c45ac6a6c880c1f2c21eb79ca86b4778ba2/python/tvm/relay/op/nn/util.py#L43, not sure if obvious but the error is from converting 2d padding to 4d padding here (ie: just returning padding here fixes the cuda error). I spoke with @zxy844288792 and he agrees it is probably a schedule issue as #4787 PR changed the workloads and the tophub file (guessing where default tuned schedules are) are not updated yet (correct me if any of this seems off).
---------------------------------------------------------------- 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
