Update extract_abs.py

extract_abs.py

@@ -141,7 +141,8 @@ def parse_info(sentences:List[str], model_outputs:List[List[Union[Dict[str,str],
     for output in model_outputs:
         #This abstracts the labels so that models with different types and numbers of labels can be used.
         for label in labels:
-            output_dict[label]+=[entity_dict['word'] for entity_dict in output if entity_dict['entity_group'] ==label]
+                                #This sub removes the ## which denotes that the token is not a the beginning of a word
+            output_dict[label]+=[re.sub('##','',entity_dict['word']) for entity_dict in output if entity_dict['entity_group'] ==label]
                 
     if 'DIS' not in output_dict.keys() and extract_diseases:
         output_dict['DIS'] = []