[BUGFIX] GPTQ quantization compatibility for Qwen3 MOE models (AutoGPTQ and AutoRound-GPTQ) (#23994)

Signed-off-by: JartX <sagformas@epdcenter.es>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>

vllm/model_executor/layers/quantization/gptq_marlin.py

@@ -119,6 +119,9 @@ class GPTQMarlinConfig(QuantizationConfig):
 
         self.quant_type = self.TYPE_MAP[(weight_bits, is_sym)]
 
+        # used to identify GPTQ model quantized by autoround
+        self.autoround_version = full_config.get("autoround_version", "")
+
     def __repr__(self) -> str:
         return (f"GPTQMarlinConfig(quant_type={self.quant_type}, "
                 f"group_size={self.group_size}, "