AWQ Modifier Support

[brian-dellabetta]

SUMMARY:
Addition of AWQModifier, based on AutoAWQ implementation.

Should be reviewed/merged in conjunction with neuralmagic/compressed-tensors#269

Replaces #181 and #824 (hence v3)

TEST PLAN:
Some unit tests included, but as this was mostly a port from AutoAWQ, we validated the code by ensuring we could reproduce the evaluation metrics in Table 4 of the paper. We achieve the following wikitext PPL scores:

Llama-2 7B Group 128:

1. Paper: 5.60
2. AutoAWQ: 5.615
3. This implementation: 5.612
4. we match what the paper reports for just RTN -- 5.73
5. We get reasonable results for channel-wise -- 6.788. AutoAWQ errors out for this (setting "q_group_size": -1 in the quant_config), and results not reported in paper.

Llama-2 13B Group 128:

1. We match the results of AutoAWQ and the results shown in the paper: 4.97
2. We match what the paper reports for just RTN -- 4.984

NOTE: We are excluding the clipping logic in this implementation, if we want to add it we should add it as another modifier, they are mutually exclusive and the data model for AWQ doesn't align well with clipping. That might be the reason for the slight deviation of results reported in the paper and in our implementation

[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.

[brian-dellabetta]

This is the one TODO. The logic for this in AutoAWQ is to only add this mapping if the shapes line up correctly (logic here). This is the case for the llama 2 models i've been testing on, but not all of the tiny llama models. Any suggestion on how best to handle for both cases?

[brian-dellabetta]

PPL is 5.607 for llama 2-7B with this included, 5.614 when it isn't.

[dsikka]

Should we add evals comparing to GPTQ?

[brian-dellabetta]

Using the latest commit at this time, I am getting the following results via lm-eval.

deepseek-ai/DeepSeek-R1-Distill-Llama-8B:
 dense:
   #gsm flexible-extract, strict-match
   gsm8k: .6619, .6490
   wititext ppl: 15.4498
 awq+quant sym:
   gsm8k: .6376, .6217
   wititext ppl: 18.8623
 quant sym:
   gsm8k: .6732, .6543
   wititext ppl: 16.7398
meta-llama/Llama-2-7b-hf:
 dense:
   gsm8k: .1342, .1342
   wititext ppl: 8.7587
 awq+quant sym:
   gsm8k: .1024, .1001
   wititext ppl: 9.194
 quant sym:
   gsm8k: .1183, .1152
   wititext ppl: 9.311