TL;DR

The test test_scaled_matmul_cuda.py is failing on the main branch and needs to be investigated and fixed.

Guidance

• Review the recent examples of test failures on torch-ci.com to understand the error messages and patterns.
• Investigate the test_scaled_mm_deepseek_error_messages_bfloat16_lhs_block_128_rhs_block_1_M_256_N_256_K_256_cuda test case in test_scaled_matmul_cuda.py to identify the root cause of the failure.
• Check the test configuration and environment, especially the ROCm platform, to ensure it is set up correctly.
• Consider re-enabling the test and running it locally to reproduce and debug the issue.

Notes

The issue lacks detailed information about the error messages or the expected behavior of the test, making it difficult to provide a more specific fix or workaround.

Recommendation

Apply workaround: Re-enable the test and run it locally to reproduce and debug the issue, as this will provide more information about the root cause of the failure.