[VitisAI] AMD NPU LLM Quantization - Add Windows + CUDA support for Quark Quantizer

[poganesh]

Describe your changes

This PR extends the Quark quantization support to enable Windows + CUDA workflows and improves quark quantization stability:
Key Changes:

• Added Windows + CUDA platform support for Quark quantization pass
• Added bfloat16 dtype support in quark quantization pass for improved AWQ quantization stability
• Tested on Windows 11 Pro with NVIDIA RTX 3090 + CUDA 13.0

Checklist before requesting a review

• Add unit tests for this change.
• Make sure all tests can pass.
• Update documents if necessary.
• Lint and apply fixes to your code by running lintrunner -a
• Is this a user-facing change? If yes, give a description of this change to be included in the release notes.

(Optional) Issue link