Fix an option name in QDQQuantizer (#187)

Author: stevenlix

Diff for: quantization/nlp/bert/trt/e2e_tensorrt_bert_example.py

@@ -274,7 +274,7 @@ def get_op_nodes_not_followed_by_specific_op(model, op1, op2):
         [], #nodes_to_quantize
         nodes_to_exclude,
         op_types_to_quantize,
-        {'ActivationSymmetric' : True, 'AddQDQPairToWeight' : True, 'OpTypesToExcludeOutputQuantizatioin': op_types_to_quantize, 'DedicatedQDQPair': True, 'QDQOpTypePerChannelSupportToAxis': {'MatMul': 1} }) #extra_options
+        {'ActivationSymmetric' : True, 'AddQDQPairToWeight' : True, 'OpTypesToExcludeOutputQuantization': op_types_to_quantize, 'DedicatedQDQPair': True, 'QDQOpTypePerChannelSupportToAxis': {'MatMul': 1} }) #extra_options
     quantizer.quantize_model()
     quantizer.model.save_model_to_file(qdq_model_path, False)
     print("QDQ model is saved to ", qdq_model_path)