Hi Anirudh, Could you provide more exact data points regarding the ONNX usage and MXNet version? If no one is actively maintaining ONNX any more, I don't see a compelling reason for an engineer to spend quality time to fix an ONNX test in order for his/her PRs to move forward.
Lin On Mon, Oct 7, 2019 at 1:19 PM Skalicky, Sam <sska...@amazon.com.invalid> wrote: > Hi Chai, > > If there is no one maintaining MXNet-ONNX support (or no one currently > available to help debug issues), then we shouldn’t block forward progress > because of failing ONNX tests. > > It would be great if someone wanted to work with Chai to debug the failing > tests. But I do not see any forward plans/proposals to continue to develop > or even just maintain the current ONNX support. > > Anirudh, if you can point those who are willing to maintain the ONNX > support to the issue Chai mentioned that would be a good place to start. > But if not, we should help Chai continue the great work he’s doing by > disabling the failing tests (like we normally do for any failing/flaky > tests already) > > Sam > > > On Oct 7, 2019, at 12:45 PM, Anirudh Acharya <anirudhk...@gmail.com> > wrote: > > > > Hi Chaitanya, > > > > The last I checked( a couple of months back) there are a few > > customers/users of MXNet in Amazon who use ONNX in production. > > > > The last commit for ONNX module was on Aug 29th > > - b7cca015553d707cd1c4ce292826d7311309419c > > > > So IMO disabling any of the tests is not a good idea. > > > > > > Thanks > > Anirudh > > > > > > On Mon, Oct 7, 2019 at 12:27 PM Chaitanya Bapat <chai.ba...@gmail.com> > > wrote: > > > >> Hello MXNet community, > >> > >> I wanted to know if MXNet should continue support for ONNX. Is there > anyone > >> actively working on MXNet ONNX or maintaining it? > >> > >> If not, can we skip/disable the ONNX tests from the CI. > >> Reason - Whilst working on a Transpose operator PR [1], I encountered > >> failure for ONNX [2]. Given operator passes rest of the CI pipeline > tests. > >> I am able to reproduce the error. However, the root cause for ONNX model > >> failure couldn't be found. Moreover, there seems to be near zero > activity > >> as far as PR check-ins are concerned. > >> > >> How does ONNX fit in for MXNet going forward? > >> Thank you > >> Chai > >> > >> > >> [1] https://github.com/apache/incubator-mxnet/pull/16104 > >> [2] > >> > >> > http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/mxnet-validation%2Funix-cpu/detail/PR-16104/14/pipeline > >> > >> -- > >> *Chaitanya Prakash Bapat* > >> *+1 (973) 953-6299* > >> > >> [image: https://www.linkedin.com//in/chaibapat25] > >> <https://github.com/ChaiBapchya>[image: > https://www.facebook.com/chaibapat > >> ] > >> <https://www.facebook.com/chaibapchya>[image: > >> https://twitter.com/ChaiBapchya] <https://twitter.com/ChaiBapchya > >[image: > >> https://www.linkedin.com//in/chaibapat25] > >> <https://www.linkedin.com//in/chaibapchya/> > >> > >
