Update app.py

app.py

@@ -1,115 +1,89 @@
 import streamlit as st
+from openai import OpenAI
+import os
 import requests
-import nltk
-from transformers import pipeline
-from rake_nltk import Rake
-from nltk.corpus import stopwords
-from fuzzywuzzy import fuzz
+from dotenv import load_dotenv
 
-st.title("Exploring Torch, Transformers, Rake, and Others analyzing Text")
+# Load environment variables
+load_dotenv()
 
-# Define the options for the dropdown menu, Selecting a remote txt file already created to analyze the text
-options = ['None','Apprecitation Letter', 'Regret Letter', 'Kindness Tale', 'Lost Melody Tale', 'Twitter Example 1', 'Twitter Example 2']
+# Initialize the client with HuggingFace
+client = OpenAI(
+    base_url="https://api-inference.huggingface.co/v1",
+    api_key=os.environ.get('HFSecret')  # Replace with your HuggingFace token
+)
 
-# Create a dropdown menu to select options
-selected_option = st.selectbox("Select a preset option", options)
-
-# Define URLs for different options
-url_option1 = "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Appreciation_Letter.txt"
-url_option2 = "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Regret_Letter.txt"
-url_option3 = "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Kindness_Tale.txt"
-url_option4 = "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Lost_Melody_Tale.txt"
-url_option5 = "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Twitter_Example_1.txt"
-url_option6 = "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Twitter_Example_2.txt"
+# Define the Llama 3 8B model
+repo_id = "meta-llama/Meta-Llama-3-8B-Instruct"
 
-# Function to fetch text content based on selected option
-def fetch_text_content(selected_option):
-    if selected_option == 'Apprecitation Letter':
-        return requests.get(url_option1).text
-    elif selected_option == 'Regret Letter':
-        return requests.get(url_option2).text
-    elif selected_option == 'Kindness Tale':
-        return requests.get(url_option3).text
-    elif selected_option == 'Lost Melody Tale':
-        return requests.get(url_option4).text
-    elif selected_option == 'Twitter Example 1':
-        return requests.get(url_option5).text
-    elif selected_option == 'Twitter Example 2':
-        return requests.get(url_option6).text
-    else:
-        return ""
-
-# Fetch text content based on selected option
-jd = fetch_text_content(selected_option)
+# Title of the App
+st.title("Text Analysis with Llama 3: Sentiment, Summarization, and Keyword Extraction")
 
-# Display text content in a text area
-#jd = st.text_area("Text File Content", text_content)
+# Dropdown options to choose a text file
+options = ['None', 'Appreciation Letter', 'Regret Letter', 'Kindness Tale', 'Lost Melody Tale', 'Twitter Example 1', 'Twitter Example 2']
 
+# Create a dropdown menu to select options
+selected_option = st.selectbox("Select a preset option", options)
 
-# Download NLTK resources
-nltk.download('punkt')
-nltk.download('stopwords')
+# Define URLs for different text options
+url_dict = {
+    'Appreciation Letter': "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Appreciation_Letter.txt",
+    'Regret Letter': "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Regret_Letter.txt",
+    'Kindness Tale': "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Kindness_Tale.txt",
+    'Lost Melody Tale': "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Lost_Melody_Tale.txt",
+    'Twitter Example 1': "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Twitter_Example_1.txt",
+    'Twitter Example 2': "https://raw.githubusercontent.com/peteciank/public_files/main/Transformers/Twitter_Example_2.txt"
+}
 
-# Initialize pipeline for sentiment analysis
-pipe_sent = pipeline('sentiment-analysis')
-# Initialize pipeline for summarization
-pipe_summ = pipeline("summarization", model="facebook/bart-large-cnn")
+# Function to fetch text content
+def fetch_text_content(option):
+    if option in url_dict:
+        response = requests.get(url_dict[option])
+        return response.text if response.status_code == 200 else "Error fetching the text"
+    return ""
 
-# Function to extract keywords and remove duplicates
-def extract_keywords(text):
-    r = Rake()
-    r.extract_keywords_from_text(text)
-    # Get all phrases scored
-    phrases_with_scores = r.get_ranked_phrases_with_scores()
-    # Filter out stopwords
-    stop_words = set(stopwords.words('english'))
-    keywords = []
-    for score, phrase in phrases_with_scores:
-        # Check if the phrase is not a stopword and add to the list
-        if phrase.lower() not in stop_words:
-            keywords.append((score, phrase))
-    # Sort keywords by score in descending order
-    keywords.sort(key=lambda x: x[0], reverse=True)
-    # Remove duplicates and merge similar keywords
-    unique_keywords = []
-    seen_phrases = set()
-    for score, phrase in keywords:
-        if phrase not in seen_phrases:
-            # Check if the phrase is similar to any of the seen phrases
-            similar_phrases = [seen_phrase for seen_phrase in seen_phrases if fuzz.ratio(phrase, seen_phrase) > 70]
-            if similar_phrases:
-                # If similar phrases are found, merge them into one phrase
-                merged_phrase = max([phrase] + similar_phrases, key=len)
-                unique_keywords.append((score, merged_phrase))
-            else:
-                unique_keywords.append((score, phrase))
-            seen_phrases.add(phrase)
-    return unique_keywords[:10]  # Return only the first 10 keywords
+# Fetch the selected text
+jd = fetch_text_content(selected_option)
 
+# Display fetched text
 text = st.text_area('Enter the text to analyze', jd)
 
+# Function to call Llama 3 for analysis
+def call_llama_analysis(task, text):
+    prompt = f"Perform {task} on the following text:\n\n{text}"
+    
+    # Call Llama 3 for the task
+    response = client.completions.create(
+        model=repo_id,
+        prompt=prompt,
+        max_tokens=3000,
+        temperature=0.5
+    )
+    
+    return response['choices'][0]['text']
+
+# Start analysis on button click
 if st.button("Start Analysis"):
-    with st.spinner("Analyzing Sentiment"):
-        with st.expander("Sentiment Analysis - ✅ Completed", expanded=False):
-            # Sentiment analysis
-            out_sentiment = pipe_sent(text)
-            # Display sentiment analysis result
-            sentiment_score = out_sentiment[0]['score']
-            sentiment_label = out_sentiment[0]['label']
-            sentiment_emoji = '😊' if sentiment_label == 'POSITIVE' else '😞'
-            sentiment_text = f"Sentiment Score: {sentiment_score}, Sentiment Label: {sentiment_label.capitalize()} {sentiment_emoji}"
-            st.write(sentiment_text)
+    with st.spinner("Analyzing Sentiment..."):
+        try:
+            sentiment_result = call_llama_analysis("sentiment analysis", text)
+            with st.expander("Sentiment Analysis - ✅ Completed", expanded=True):
+                st.write(sentiment_result)
+        except Exception as e:
+            st.error(f"Error in Sentiment Analysis: {str(e)}")
 
-    with st.spinner("Summarizing - This may take a while"):
-        with st.expander("Summarization - ✅ Completed", expanded=False):        
-            # Summarization
-            out_summ = pipe_summ(text)
-            summarized_text = out_summ[0]['summary_text']
-            st.write(summarized_text)
+    with st.spinner("Summarizing..."):
+        try:
+            summary_result = call_llama_analysis("summarization", text)
+            with st.expander("Summarization - ✅ Completed", expanded=True):
+                st.write(summary_result)
+        except Exception as e:
+            st.error(f"Error in Summarization: {str(e)}")
 
-    with st.spinner("Extracting Keywords"):
-        with st.expander("Keywords Extraction - ✅ Completed", expanded=False):
-            # Keyword extraction
-            keywords = extract_keywords(text)
-            keyword_list = [keyword[1] for keyword in keywords]
-            st.write(keyword_list)
+    with st.spinner("Extracting Keywords..."):
+        try:
+            keywords_result = call_llama_analysis("keyword extraction", text)
+            with st.expander("Keywords Extraction - ✅ Completed", expanded=True):
+                st.write(keywords_result)
+        except Exception as e:
+            st.error(f"Error in Keyword Extraction: {str(e)}")