Fixed issue with Transformers

Author: AlexKoff88

optimum/intel/openvino/configuration.py

@@ -84,7 +84,7 @@ def __init__(
         compression: Union[List[Dict], Dict, None] = None,
         input_info: Optional[List] = None,
         save_onnx_model: bool = False,
-        quantization_config: Optional[QuantizationConfigMixin] = None,
+        weight_quantization_config: Optional[QuantizationConfigMixin] = None,
         **kwargs,
     ):
         super().__init__()
@@ -93,7 +93,7 @@ def __init__(
         self.save_onnx_model = save_onnx_model
         self._enable_standard_onnx_export_option()
         self.optimum_version = kwargs.pop("optimum_version", None)
-        self.quantization_config = quantization_config
+        self.weight_quantization_config = weight_quantization_config
 
     def add_input_info(self, model_inputs: Dict, force_batch_one: bool = False):
         self.input_info = [

optimum/intel/openvino/quantization.py

@@ -318,7 +318,7 @@ def _quantize_ovcausallm(
         save_directory.mkdir(parents=True, exist_ok=True)
 
         if weights_only:
-            quantization_config = None if ov_config is None else ov_config.quantization_config
+            quantization_config = None if ov_config is None else ov_config.weight_quantization_config
             if quantization_config is None:
                 # Use default 8-bit compression
                 self.model.model = nncf.compress_weights(self.model.model)