kfeng123 opened a new issue, #15365: URL: https://github.com/apache/tvm/issues/15365
Hello TVM community, In PR #15362, the following test case fails. PR #15362 improves a graph optimization pass, hence some compiled models may be optimized, and may use less runtime memory. It seems that the test case verify if the workspace_size of the compiled model is equal to a prespecified number. I am wonder how the specified number is obtained. I think this test case failure may be a result of graph optimization, not a real failure indeed. But I am not familiar with USMP. So I am not pretty sure. Can someone help verify if this test case is indeed correct or not? Thank you very much! ======================================= These tests were found to be flaky (intermittently failing on `main` or failed in a PR with unrelated changes). See [the docs](https://github.com/apache/tvm/blob/main/docs/contribute/ci.rst#handling-flaky-failures) for details. ### Tests(s) - `tests/python/relay/aot/test_crt_aot_usmp.py::test_tflite_model_u1_usecase[model_url1-greedy_by_conflicts-4845696-8468008]` ### Jenkins Links - https://ci.tlcpack.ai/job/tvm-cpu/job/PR-15362/1/display/redirect -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
