xla: relax error message assertion in dot_general test (allow backend…

[Rahulchaube1]

This PR relaxes the error message assertion in the test_raises_error_on_non_xla_tensor test by changing from an exact string match (assertEqual) to a substring match (assertIn). This allows for backend-specific context to be included in error messages while still validating the core error content.This PR relaxes the error message assertion in the test_raises_error_on_non_xla_tensor test by changing from an exact string match (assertEqual) to a substring match (assertIn). This allows for backend-specific context to be included in error messages while still validating the core error content.… context)xla: relax error message assertion in dot_general test (allow backend context)Update test_dot_general.py

[Rahulchaube1]

Thanks for updating the error message assertion in test_raises_error_on_non_xla_tensor. Switching from an exact match to a substring check makes the test more robust for backend-specific messages while still ensuring the intended error is caught. I’ve reviewed the code and it looks good to me. Just one linter check is failing—could you take a look? Otherwise, I support merging this once the checks pass!

[qihqi]

Please run yapf -i -r *.py test/ scripts/ torch_xla/ benchmarks/ torchax/ to make the linter pass. I enabled auto-merge it should merge automatically once CI passes. Thanks. @Rahulchaube1