Rename app_2.py to app_3.py

app_2.py → app_3.py

@@ -131,38 +131,10 @@ def extract_arguments(text, tokenizer, model, beam_search=True):
     
     tokens = tokenizer.convert_ids_to_tokens(inputs["input_ids"][0])
     token_ids = inputs["input_ids"][0]
-    offset_mapping = inputs["offset_mapping"][0].tolist()
-    
-    for i, (token, word_id) in enumerate(zip(tokens, word_ids)):
-        st.write(f"Token {i}: {token}, Word ID: {word_id}")
-
-    st.write("Token & offset:")
-    for i, (token, offset) in enumerate(zip(tokens, offset_mapping)):
-        st.write(f"Token {i}: {token}, Offset: {offset}")
+    #offset_mapping = inputs["offset_mapping"][0].tolist()
     
 
-    st.write("Token Positions, IDs, and Corresponding Tokens:")
-    for position, (token_id, token) in enumerate(zip(token_ids, tokens)):
-        st.write(f"Position: {position}, ID: {token_id}, Token: {token}")
 
-    st.write(f"Start Cause 1: {start_cause1}, End Cause: {end_cause1}")
-    st.write(f"Start Effect 1: {start_effect1}, End Cause: {end_effect1}")
-    st.write(f"Start Signal: {start_signal}, End Signal: {end_signal}")
-
-    def extract_span(start, end):
-        return tokenizer.convert_tokens_to_string(tokens[start:end+1]) if start is not None and end is not None else ""
-
-    cause1 = extract_span(start_cause1, end_cause1)
-    cause2 = extract_span(start_cause2, end_cause2)
-    effect1 = extract_span(start_effect1, end_effect1)
-    effect2 = extract_span(start_effect2, end_effect2)
-    if has_signal:
-        signal = extract_span(start_signal, end_signal)
-    if not has_signal:
-        signal = 'NA'
-    list1 = [start_cause1, end_cause1, start_effect1, end_effect1, start_signal, end_signal]
-    list2 = [start_cause2, end_cause2, start_effect2, end_effect2, start_signal, end_signal]
-    #return cause1, cause2, effect1, effect2, signal, list1, list2
     return start_cause1, end_cause1, start_cause2, end_cause2, start_effect1, end_effect1, start_effect2, end_effect2, start_signal, end_signal
 
 
@@ -170,7 +142,9 @@ def extract_arguments(text, tokenizer, model, beam_search=True):
 def mark_text_by_position(original_text, start_token, end_token, color):
     """Marks text in the original string based on character positions."""
     # Inserts tags into the original text based on token offsets.
-
+    
+    offset_mapping = inputs["offset_mapping"][0].tolist()
+    
     start_idx, end_idx = offset_mapping[start_token][0], offset_mapping[end_token][1]
      
     if start_idx is not None and end_idx is not None and start_idx <= end_idx:
@@ -212,55 +186,5 @@ if st.button("Extract"):
             st.markdown(f"**Effect:** {effect_text2}", unsafe_allow_html=True)
             st.markdown(f"**Signal:** {signal_text}", unsafe_allow_html=True)
 
-    else:
-        st.warning("Please enter some text before extracting.")
-
-
-        
-
-    
-if st.button("Extract1"):
-    if input_text:
-        start_cause_id, end_cause_id, start_effect_id, end_effect_id, start_signal_id, end_signal_id = extract_arguments(input_text, tokenizer, model, beam_search=beam_search)
-
-        cause_text = mark_text_by_word_ids(input_text, inputs["input_ids"][0], start_cause_id, end_cause_id, "#FFD700")  # Gold for cause
-        effect_text = mark_text_by_word_ids(input_text, inputs["input_ids"][0], start_effect_id, end_effect_id, "#90EE90")  # Light green for effect
-        signal_text = mark_text_by_word_ids(input_text, inputs["input_ids"][0], start_signal_id, end_signal_id, "#FF6347")  # Tomato red for signal
-
-        st.markdown(f"**Cause:**<br>{cause_text}", unsafe_allow_html=True)
-        st.markdown(f"**Effect:**<br>{effect_text}", unsafe_allow_html=True)
-        st.markdown(f"**Signal:**<br>{signal_text}", unsafe_allow_html=True)
-    else:
-        st.warning("Please enter some text before extracting.")
-
-
-
-
-if st.button("Extract1"):
-    if input_text:
-        start_cause1, end_cause1, start_cause2, end_cause2, start_effect1, end_effect1, start_effect2, end_effect2, start_signal, end_signal = extract_arguments(input_text, tokenizer, model, beam_search=beam_search)
-
-        # Convert text to tokenized format
-        tokenized_input = tokenizer.tokenize(input_text)
-        
-        cause_text1 = mark_text_by_tokens(tokenizer, tokenized_input, start_cause1, end_cause1, "#FFD700")  # Gold for cause
-        effect_text1 = mark_text_by_tokens(tokenizer, tokenized_input, start_effect1, end_effect1, "#90EE90")  # Light green for effect
-        signal_text = mark_text_by_tokens(tokenizer, tokenized_input, start_signal, end_signal, "#FF6347")  # Tomato red for signal
-
-        # Display first relation
-        st.markdown(f"<strong>Relation 1:</strong>", unsafe_allow_html=True)
-        st.markdown(f"**Cause:** {cause_text1}", unsafe_allow_html=True)
-        st.markdown(f"**Effect:** {effect_text1}", unsafe_allow_html=True)
-        st.markdown(f"**Signal:** {signal_text}", unsafe_allow_html=True)
-
-        # Display second relation if beam search is enabled
-        if beam_search:
-            cause_text2 = mark_text_by_tokens(tokenizer, tokenized_input, start_cause2, end_cause2, "#FFD700")
-            effect_text2 = mark_text_by_tokens(tokenizer, tokenized_input, start_effect2, end_effect2, "#90EE90")
-
-            st.markdown(f"<strong>Relation 2:</strong>", unsafe_allow_html=True)
-            st.markdown(f"**Cause:** {cause_text2}", unsafe_allow_html=True)
-            st.markdown(f"**Effect:** {effect_text2}", unsafe_allow_html=True)
-            st.markdown(f"**Signal:** {signal_text}", unsafe_allow_html=True)
     else:
         st.warning("Please enter some text before extracting.")