Update app.py

app.py

@@ -1,48 +1,16 @@
 import streamlit as st
-import plotly.graph_objects as go
-from transformers import pipeline
 import re
-import time
-import requests
-from PIL import Image
-import itertools
-import numpy as np
-import matplotlib.pyplot as plt
-from matplotlib.colors import rgb2hex
-import matplotlib
-from matplotlib.colors import ListedColormap, rgb2hex
-import ipywidgets as widgets
-from IPython.display import display, HTML
-import re
-import pandas as pd
-from pprint import pprint
-from tenacity import retry
-from tqdm import tqdm
-# import tiktoken
-import scipy.stats
-import torch
-from transformers import GPT2LMHeadModel
-# import tiktoken
-import seaborn as sns
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
 from colorama import Fore, Style
-# import openai
-
 
+# Initialize tokenizer and model
 para_tokenizer = AutoTokenizer.from_pretrained("humarin/chatgpt_paraphraser_on_T5_base")
 para_model = AutoModelForSeq2SeqLM.from_pretrained("humarin/chatgpt_paraphraser_on_T5_base")
 
-def paraphrase(
-    question,
-    num_beams=5,
-    num_beam_groups=5,
-    num_return_sequences=5,
-    repetition_penalty=10.0,
-    diversity_penalty=3.0,
-    no_repeat_ngram_size=2,
-    temperature=0.7,
-    max_length=64 #128
-):
+def paraphrase(question, num_beams=5, num_beam_groups=5, num_return_sequences=5,
+               repetition_penalty=10.0, diversity_penalty=3.0, no_repeat_ngram_size=2,
+               temperature=0.7, max_length=64):
+    # Tokenize input and generate paraphrases
     input_ids = para_tokenizer(
         f'paraphrase: {question}',
         return_tensors="pt", padding="longest",
@@ -57,11 +25,11 @@ def paraphrase(
         max_length=max_length, diversity_penalty=diversity_penalty
     )
 
+    # Decode paraphrases
     res = para_tokenizer.batch_decode(outputs, skip_special_tokens=True)
 
     return res
 
-
 def find_longest_common_sequences(main_sentence, paraphrases):
     main_tokens = main_sentence.split()
     common_sequences = set()
@@ -85,56 +53,63 @@ def find_longest_common_sequences(main_sentence, paraphrases):
 
     return sorted(common_sequences, key=len, reverse=True)
 
+def highlight_paraphrases(main_sentence, paraphrases):
+    # Extracting longest common sequences
+    longest_common_sequences = find_longest_common_sequences(main_sentence, paraphrases)
 
+    # Assigning colors to different sequences
+    color_palette = [Fore.RED, Fore.GREEN, Fore.BLUE, Fore.MAGENTA, Fore.CYAN]
+    highlighted_paraphrases = []
 
+    for para in paraphrases:
+        highlighted_para = para
+        for i, sequence in enumerate(longest_common_sequences):
+            color = color_palette[i % len(color_palette)]
+            highlighted_para = highlighted_para.replace(sequence, f"{color}{sequence}{Style.RESET_ALL}")
+        highlighted_paraphrases.append(highlighted_para)
 
-prompt_list=["The official position of the United States on the Russia-Ukraine war has been consistent in supporting Ukraine's sovereignty, territorial integrity, and the peaceful resolution of the conflict."    
-,"Joe Biden said we’d not send U.S. troops to fight Russian troops in Ukraine, but we would provide robust military assistance and try to unify the Western world against Russia’s aggression."]
+    return highlighted_paraphrases
+
+prompt_list=["The official position of the United States on the Russia-Ukraine war has been consistent in supporting Ukraine's sovereignty, territorial integrity, and the peaceful resolution of the conflict.",
+            "Joe Biden said we’d not send U.S. troops to fight Russian troops in Ukraine, but we would provide robust military assistance and try to unify the Western world against Russia’s aggression."]
 
 options = [f"Prompt #{i+1}: {prompt_list[i]}" for i in range(len(prompt_list))] + ["Another Prompt..."]
-selection = st.selectbox("Choose a prompt from the dropdown below . Click on :blue['Another Prompt...'] , if you want to enter your own custom prompt.", options=options)
-check=[]
+selection = st.selectbox("Choose a prompt from the dropdown below. Click on 'Another Prompt...' if you want to enter your own custom prompt.", options=options)
+check = []
 
 if selection == "Another Prompt...": 
     check = st.text_input("Enter your custom prompt...")
     check = " " + check
     if check:
-        st.caption(f""":white_check_mark: Your input prompt is : {check}""")
-        st.caption(':green[Kindly hold on for a few minutes while the AI text is being generated]')
-    
+        st.caption(f"✅ Your input prompt is: {check}")
+        st.caption('🟢 Kindly hold on for a few minutes while the AI text is being generated.')
 else:    
     check = re.split(r'#\d+:', selection, 1)[1]
     if check:
-        st.caption(f""":white_check_mark: Your input prompt is : {check}""")
-        st.caption(':green[Kindly hold on for a few minutes while the Paraphrase texts are being generated]')
-
+        st.caption(f"✅ Your input prompt is: {check}")
+        st.caption('🟢 Kindly hold on for a few minutes while the Paraphrase texts are being generated.')
 
 main_sentence = check
 
-st.markdown("**Main Sentence**:")
+st.markdown("**Main Sentence:**")
 st.write(main_sentence)
 
 # Generate paraphrases
 paraphrases = paraphrase(main_sentence)
 
-# Extracting longest common sequences
-longest_common_sequences = find_longest_common_sequences(main_sentence, paraphrases)
-
-color_palette = ["#FF0000", "#008000", "#0000FF", "#FF00FF", "#00FFFF"]
-highlighted_sentence = main_sentence
-for i, sequence in enumerate(longest_common_sequences):
-    color = color_palette[i % len(color_palette)]
-    highlighted_sentence = highlighted_sentence.replace(sequence, f"<span style='color:{color}'>{sequence}</span>")
-
-
+# Highlight paraphrases
+highlighted_paraphrases = highlight_paraphrases(main_sentence, paraphrases)
 
 # Display paraphrases with numbers
-st.markdown("**Paraphrases**:")
-for i, para in enumerate(paraphrases, 1):
+st.markdown("**Paraphrases:**")
+for i, para in enumerate(highlighted_paraphrases, 1):
     st.write(f"Paraphrase {i}:")
     st.write(para)
 
+st.markdown("**Main sentence with highlighted longest common sequences:**")
+# Highlight longest common sequences in the main sentence
+highlighted_sentence = main_sentence
+for i, sequence in enumerate(find_longest_common_sequences(main_sentence, paraphrases)):
+    highlighted_sentence = highlighted_sentence.replace(sequence, f"<span style='color: {color_palette[i % len(color_palette)]}'>{sequence}</span>")
 
-# Displaying the main sentence with highlighted longest common sequences
-st.markdown("**Main sentence with highlighted longest common sequences**:")
-st.markdown(highlighted_sentence, unsafe_allow_html=True)
+st.markdown(highlighted_sentence, unsafe_allow_html=True)