AWQ Qwen and Phi mappings

[brian-dellabetta]

SUMMARY:
I wanted to create a PR showing users how they can add more mappings to AWQ to account for more models. Turns out qwen has the exact same as Llama, so I added one for Phi as well.

TEST PLAN:
examples/awq/llama_example.py works on this branch for

MODEL_ID = "microsoft/Phi-4-mini-reasoning"

and

recipe = [
    AWQModifier(
        ignore=["lm_head"],
        scheme="W4A16_ASYM",
        targets=["Linear"],
        mappings=AWQ_MAPPING_REGISTRY["Phi"],
    ),
]

when ".v_proj" in layer_name and ".o_proj" in balance_name checks are in skipping logic

TODOs:

• rebase off branch that fixes this issue

[github-actions]

👋 Hi! Thank you for contributing to llm-compressor. Please add the ready label when the PR is ready for review.

Note: This is required to complete the testing suite, please only add the label once the PR is code complete and local testing has been performed.

[kylesayrs]

neat