numBery commited on
Commit
f195951
1 Parent(s): 5d54a10

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +17 -1
app.py CHANGED
@@ -164,7 +164,23 @@ def run_short_extraction(article, number_of_paraphrases):
164
  st.error('Error running Extraction Pipeline. Please contact admin')
165
  logger.error(traceback.format_exc())
166
 
167
-
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
168
 
169
  kw_extractor, t5_model, t5_tokenizer = load_base_model()
170
 
 
164
  st.error('Error running Extraction Pipeline. Please contact admin')
165
  logger.error(traceback.format_exc())
166
 
167
+
168
+
169
+
170
+ def check_document_length(article, number_of_paraphrases):
171
+ total_tokens = len(t5_tokenizer.encode(article))
172
+ st.info(f'Token Counts for Encoded Document: {total_tokens}')
173
+
174
+ if total_tokens >= 512:
175
+ st.info('Running Extraction for Long Document')
176
+ t5_keywords_df, original_keywords_df, unique_keywords_df, total_end = run_long_extraction(article, number_of_paraphrases)
177
+ else:
178
+ st.info('Running Extraction for Short Document')
179
+ t5_keywords_df, original_keywords_df, unique_keywords_df, total_end = run_short_extraction(article, number_of_paraphrases)
180
+
181
+ return t5_keywords_df, original_keywords_df, unique_keywords_df, total_end
182
+
183
+
184
 
185
  kw_extractor, t5_model, t5_tokenizer = load_base_model()
186