Enhance llmc CI on GPU and XPU

[chensuyue]

Description

Enhance llmc CI on GPU and XPU

Type of Change

• Bug fix
• New feature
• Documentation update
• Performance improvement
• Code refactoring
• Other (please specify):

Related Issues

#1149

Fixes or relates to #

Checklist Before Submitting

• My code has been tested locally.
• Documentation has been updated as needed.
• New or updated tests are included where applicable.

[Copilot]

Pull request overview

Adds LLMC integration tests targeting Intel XPU execution and adjusts the CUDA test suite layout to support updated CI coverage.

Changes:

• Added new XPU integration tests for llmcompressor.oneshot with multiple AutoRoundModifier recipes.
• Added XPU multi-device coverage via device_ids.
• Removed the CUDA integration test stub that previously redirected to the CPU integration test.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 13 comments.

File	Description
test/test_xpu/test_llmc_integration.py	New XPU integration tests validating quantization config + targeted layer quantization after oneshot().
test/test_cuda/integrations/test_llmc_integration.py	Removes the CUDA integration test redirect to the CPU integration test module.

Comments suppressed due to low confidence (1)

test/test_cuda/integrations/test_llmc_integration.py:1

• Removing this CUDA integration test redirect likely stops CUDA CI from executing the LLMC integration coverage that previously ran via the CPU test module. If CUDA CI is still expected to validate LLMC integration, keep an equivalent CUDA entry point (e.g., a thin wrapper/import) or add a CUDA-specific test module so the integration tests remain discoverable under the CUDA test path.