pytorch - 💡(How to fix) Fix DISABLED test_eager_equivalence_log1p_backend_inductor_default_cuda_float32 (__main__.TestOpInfoPropertiesCUDA) [1 comments, 2 participants]

TL;DR

The test failure on the main branch for the test_torchinductor_opinfo_properties.py test can likely be addressed by investigating and resolving the issues causing the failures in the recent examples provided.

Guidance

• Review the recent failure examples at https://torch-ci.com/failure?failureCaptures=%5B%22test%2Finductor%2Ftest_torchinductor_opinfo_properties.py%3A%3ATestOpInfoPropertiesCUDA%3A%3Atest_eager_equivalence_log1p_backend_inductor_default_cuda_float32%22%5D to understand the nature of the failures.
• Investigate potential issues with the test_torchinductor_opinfo_properties.py test, specifically the test_eager_equivalence_log1p_backend_inductor_default_cuda_float32 test case.
• Consider reaching out to the listed individuals (@jeffdaily, @sunway513, etc.) for additional guidance or support, as they may have insight into the test or the ROCm platform.

Notes

The provided information does not specify the exact cause of the test failure, so a thorough investigation is necessary to determine the root cause.

Recommendation

Apply workaround: Investigate and resolve the issues causing the test failures, as the provided information does not imply a straightforward fix or upgrade.