[TEST][FLAKY] frontend/pytorch/test_forward.py::test_forward_nms

[tqchen]

https://ci.tlcpack.ai/job/tvm/job/main/19/execution/node/380/log/

Could due to too close ties, perhaps we should construct nms test-cases differently that have no clear ties

[tqchen]

cc @masahi @Laurawly

[masahi]

cc @yongwww

[yongwww]

@masahi @tqchen I failed to reproduce it on my environment. The flaky issue was introduced due to the accumulation of tiny difference of computations, especially for a large number of input boxes. I will send a pr to loosen the tolerance.

[masahi]

@yongwww maybe setting rng manual seed via torch.manual_seed(...) help.

[masahi]

Still has this problem, @yongwww please take a look

https://ci.tlcpack.ai/blue/organizations/jenkins/tvm/detail/PR-7824/1/pipeline

[denise-k]

Hi folks, is this still active?

[masahi]

I was able to reproduce the error after 30 runs, it is fixed in #9140