Chirag Pandya
Chirag Pandya
> @c-p-i-o do we have any actualy test for e2e usage in OSS? would be good to have some coverage We don't have tests for the analyzer portions in OSS....
> cc+ @shengfukevin @c-p-i-o Can you run: "@pytorchbot merge" This should get this PR into the main branch after CI passes.
Sorry for the delay. Are you able to add a test for this change?
Ignore the CI breakage for now. I'm trying to revive the CI for this repository.
> Hey @c-p-i-o can you please let me know what tests are failing? Also what kind of linter is used? Would just clang-format be enough to resolve lint errors? Sorry...
Grr. Some failures are internal to Meta when they try to build this change. Let me see if I can address these on the internal side.
Sorry about the delay here: Documentation for flight recorder is here: https://pytorch.org/tutorials/prototype/flight_recorder_tutorial.html
> I solved it by setting broadcast_buffers=False in DDP. But still wonder why this mattters. cc @wconstab, @kwen2501 - do you know off the top why this setting might matter?
Seems like there are some hard assumptions in gloo. Happy to take a PR if you have one to address this issue?
@venkatram-dev - do you want to resolve the conflicts and land this PR?