Support imatrix-guided quantization for NPU CW

[rnwang04]

Description

1. Why the change?

work with https://github.com/intel-analytics/llm.cpp/pull/696

2. User API changes

• add a new parameter imatrix_file which default to None.
• Only when user pass imatrix_file & set IPEX_LLM_NPU_QUANTIZATION_OPT=1, this opt will be used.

3. Summary of the change

• Initial support of imatrix-guided quantizatio for NPU CW
• Will not expose this to users, just for internal test first

4. How to test?

• Unit test: Please manually trigger the PR Validation here by inputting the PR number (e.g., 1234). And paste your action link here once it has been successfully finished.

[jason-dai]

LGTM