Add Additional Model Mappings for `AWQ` and `SmoothQuant`

[dsikka]

For AWQ, this requires updating this file: https://github.com/vllm-project/llm-compressor/blob/main/src/llmcompressor/modifiers/awq/mappings.py

For SmoothQuant, this requires updating : https://github.com/vllm-project/llm-compressor/blob/main/src/llmcompressor/modifiers/smoothquant/utils.py

[brian-dellabetta]

For an example PR adding Qwen and Phi mappings to AWQ, see #1440

[yugborana]

Hey. I am looking to work up on this issue.
Understanding the code base working...

[aladerran]

@yugborana Hi, any update?

I am happy to help or collaborate~

[brian-dellabetta]

Hi @yugborana and @aladerran , thanks for your interest in collaborating on this. An example PR adding mappings can be found here:

• AWQ Cohere, Mistral & Gemma mappings #1570

We just need to add them on a per-model basis. As an initial exercise, you could try adding any of the models in AutoAWQ not already included, perhaps the Qwen 2.5 VL or Deepseek v3 (which looks a little more complex)