[Feature]Add support for models quantized with AutoRound (#17850)

Signed-off-by: wenhuach21 <wenhua.cheng@intel.com>

tests/quantization/test_auto_round.py

@@ -0,0 +1,30 @@
+# SPDX-License-Identifier: Apache-2.0
+"""Test model set-up and inference for quantized HF models supported
+ on the AutoRound.
+
+ Validating the configuration and printing results for manual checking.
+
+ Run `pytest tests/quantization/test_auto_round.py`.
+"""
+
+import pytest
+
+from vllm.platforms import current_platform
+
+MODELS = [
+    "OPEA/Qwen2.5-0.5B-Instruct-int4-sym-inc",  ##auto_round:auto_gptq
+    "Intel/Qwen2-0.5B-Instruct-int4-sym-AutoRound"  ##auto_round:auto_awq
+]
+
+
+@pytest.mark.skipif(not current_platform.is_cpu()
+                    and not current_platform.is_xpu()
+                    and not current_platform.is_cuda(),
+                    reason="only supports CPU/XPU/CUDA backend.")
+@pytest.mark.parametrize("model", MODELS)
+def test_auto_round(vllm_runner, model):
+    with vllm_runner(model) as llm:
+        output = llm.generate_greedy(["The capital of France is"],
+                                     max_tokens=8)
+    assert output
+    print(f"{output[0][1]}")