AWQ QuantizationMixin + SequentialPipeline (#1426)

SUMMARY:
- Add QuantizationMixin to AWQModifier so we don't have redundant inputs
(num_bits, symmetric, group_size)
- Move AWQModifier to sequential pipelines, to avoid huge memory
requirements of caching all activations at once.

Regression test results are acceptable, results are all roughly the
same, and within stderr, see test plan below.

Resolves #1409 
Resolves #1369 
Related to #1383
Related to #1406 
Related to #1368 
Related to #1410 

More improvements split into #1435

TEST PLAN:
- [x] Rerun tests to validate
No regression in tests, comparing against those reported in [original
AWQ
PR](#1177 (comment)).
All gsm8k results are within stderr:

| Type            | gsm8k       | wikitext
| ------          | ------      | ----- 
| Old AWQ+QuantModifier Sym          | .1054, .1069     | 9.1931 
| New AWQ+QuantMixin Sym        | .1077, .1084 | 9.1841
| Old AWQ+QuantModifier Asym             | .1274, .1281 | 9.0281
| New AWQ+QuantMixin Asym        | .1312, .1350 | 9.0288

---------

Signed-off-by: Brian Dellabetta <[email protected]>
Co-authored-by: Kyle Sayers <[email protected]>

Author: brian-dellabetta

examples/awq/llama_example.py

@@ -1,17 +1,8 @@
-import lm_eval
-from compressed_tensors.quantization import (
-    QuantizationArgs,
-    QuantizationScheme,
-    QuantizationStrategy,
-    QuantizationType,
-)
 from datasets import load_dataset
-from lm_eval.utils import make_table
 from transformers import AutoModelForCausalLM, AutoTokenizer
 
 from llmcompressor import oneshot
 from llmcompressor.modifiers.awq import AWQModifier
-from llmcompressor.modifiers.quantization import QuantizationModifier
 
 # Select model and load it.
 MODEL_ID = "meta-llama/Meta-Llama-3-8B-Instruct"
@@ -61,23 +52,7 @@ def tokenize(sample):
 
 # Configure the quantization algorithm to run.
 recipe = [
-    AWQModifier(bits=4, symmetric=False),
-    QuantizationModifier(
-        ignore=["lm_head"],
-        config_groups={
-            "group_0": QuantizationScheme(
-                targets=["Linear"],
-                weights=QuantizationArgs(
-                    num_bits=4,
-                    type=QuantizationType.INT,
-                    dynamic=False,
-                    symmetric=False,
-                    strategy=QuantizationStrategy.GROUP,
-                    group_size=128,
-                ),
-            )
-        },
-    ),
+    AWQModifier(ignore=["lm_head"], scheme="W4A16_ASYM", targets=["Linear"]),
 ]
 
 # Apply algorithms.
@@ -101,21 +76,3 @@ def tokenize(sample):
 SAVE_DIR = MODEL_ID.split("/")[-1] + "-awq-asym"
 model.save_pretrained(SAVE_DIR, save_compressed=True)
 tokenizer.save_pretrained(SAVE_DIR)
-
-#
-# 2) Evaluate model on wikitext perplexity
-#
-
-results = lm_eval.simple_evaluate(
-    model="hf",
-    model_args={
-        "pretrained": SAVE_DIR,
-        "add_bos_token": True,
-        "dtype": "bfloat16",
-        "gpu_memory_utilization": 0.5,
-    },
-    tasks=["wikitext"],
-    num_fewshot=5,
-    batch_size="auto",
-)
-print(make_table(results))