Update functions.py

functions.py

@@ -81,7 +81,7 @@ def inference(link, upload, _asr_model):
 def sentiment_pipe(earnings_text):
     '''Determine the sentiment of the text'''
     
-    earnings_sentences = chunk_long_text(earnings_text,200,1)
+    earnings_sentences = chunk_long_text(earnings_text,150,1,1)
     earnings_sentiment = sent_pipe(earnings_sentences)
     
     return earnings_sentiment, earnings_sentences    
@@ -99,12 +99,11 @@ def clean_text(text):
     return text
        
 @st.experimental_memo(suppress_st_warning=True)
-def chunk_long_text(text,threshold,window_size=3):
+def chunk_long_text(text,threshold,window_size=3,stride=2):
     '''Preprocess text and chunk for semantic search and sentiment analysis'''
     
     #Convert cleaned text into sentences
     sentences = sent_tokenize(text)
-    
     out = []
 
     #Limit the length of each sentence to a threshold
@@ -121,12 +120,12 @@ def chunk_long_text(text,threshold,window_size=3):
     
     #Combine sentences into a window of size window_size
     for paragraph in [out]:
-        for start_idx in range(0, len(paragraph), window_size):
+        for start_idx in range(0, len(paragraph), stride):
             end_idx = min(start_idx+window_size, len(paragraph))
             passages.append(" ".join(paragraph[start_idx:end_idx]))
             
     return passages
-
+    
 @st.experimental_memo(suppress_st_warning=True)
 def chunk_and_preprocess_text(text,thresh=500):