fix code snippet

README.md

@@ -45,8 +45,8 @@ The calibration dataloader is the train dataloader. The default calibration samp
 ### Load with optimum:
 
 ```python
-from optimum.intel.neural_compressor.quantization import IncQuantizedModelForTokenClassification
-int8_model = IncQuantizedModelForTokenClassification.from_pretrained(
-    'Intel/distilbert-base-uncased-finetuned-conll03-english-int8-static',
-)
+from optimum.intel import INCModelForTokenClassification
+
+model_id = "Intel/distilbert-base-uncased-finetuned-conll03-english-int8-static"
+int8_model = INCModelForTokenClassification.from_pretrained(model_id)
 ```