fix: propagate GLWE config on hybrid MLP unit test

[alinealfa]

Notes

Fix #927 - Address test_hybrid_converter.py failure on MacOS

This fix propagates the GLWE configuration in the hybrid MLP unit test.
Review needed to confirm if this aligns with the test's original intended behavior.

Testing

Executed the make tasks as outlined in contributing.md. For pytest, used pytest_no_flaky due to a local unrelated test_brevitas_qat.py test failure, which is known to be flaky.

[cla-bot]

Thank you for your pull request. We require contributors to sign our Contributor License Agreement / Terms and Conditions, and we don't seem to have the users @alinealfa on file. In order for us to review and merge your code, please sign:

• For individual contribution: our CLA
• for Bounty submission, if you are an individual: our T&C
• for Bounty submission, if you are a company: our T&C
  to get yourself added.

If you already signed one of this document, just wait to be added to the bot config.

[aquint-zama]

@cla-bot check

[cla-bot]

The cla-bot has been summoned, and re-checked this pull request!

[andrei-stoian-zama]

Thanks for the refactoring here!

[alinealfa]

Thanks @andrei-stoian-zama. I updated the commit as the CI was enforcing signature.

[andrei-stoian-zama]

@alinealfa , as we require signed commits only from approved Zama employees, I needed to re-base your PR . It's here: #935 and I'll close this one.

[alinealfa]

@alinealfa , as we require signed commits only from approved Zama employees, I needed to re-base your PR . It's here: #935 and I'll close this one.

Got it. So should external contributors not sign the commit? Actually the Contributing doc does not indicate need of signature. I ended up adding so we could get the CI/tests unblocked.