Update app.py

app.py

@@ -1,55 +1,9 @@
 import streamlit as st
 from transformers import pipeline
 import torch
-import pandas as pd
+import time
 
-# ---- Page Configuration ----
-st.set_page_config(
-    page_title="Emotion Prediction App",
-    page_icon="🌟",
-    layout="centered",
-    initial_sidebar_state="expanded",
-)
-
-# ---- App Title ----
-st.title("🌟 Emotion Prediction App 🌈")
-st.subheader("Get emotional insights with tailored resources!")
-
-# ---- Function to Load Emotion Analysis Model ----
-@st.cache_resource
-def load_emotion_model():
-    try:
-        st.info("⏳ Loading the emotion analysis model, please wait...")
-        # Using a publicly available model for emotion analysis
-        emotion_analyzer = pipeline(
-            "text-classification",
-            model="bhadresh-savani/distilbert-base-uncased-emotion",  # A valid public model
-            device=0 if torch.cuda.is_available() else -1,  # Use GPU if available
-        )
-        st.success("✅ Model loaded successfully!")
-        return emotion_analyzer
-    except Exception as e:
-        st.error(f"⚠️ Error loading model: {e}")
-        return None
-
-# ---- Load the Model ----
-emotion_analyzer = load_emotion_model()
-
-# ---- Function for Predicting Emotion ----
-def predict_emotion(text):
-    if emotion_analyzer is None:
-        st.error("⚠️ Model not loaded. Please reload the app.")
-        return {"Error": "Emotion analyzer model not initialized. Please try again later."}
-    
-    try:
-        # Analyze emotions
-        result = emotion_analyzer([text])
-        return {res["label"]: round(res["score"], 4) for res in result}
-    except Exception as e:
-        st.error(f"⚠️ Prediction failed: {e}")
-        return {"Error": f"Prediction failed: {e}"}
-
-# ---- Enhanced Suggestion Database ----
+# Enhanced Suggestion Database (Now includes resources)
 suggestion_database = {
     "sadness": {
         "suggestions": ["Try a guided meditation", "Take a walk in nature", "Connect with a friend"],
@@ -77,11 +31,11 @@ suggestion_database = {
         "suggestions": ["Take a break", "Engage in a relaxing activity", "Spend time in nature"],
         "articles": [
             {"title": "Importance of Self-Care", "url": "https://example.com/selfcare1"},
-            {"title": "Stress Management Techniques", "url": "https://example.com/stress1"}
+            {"title": "Stress Management Techniques", "url": "https://example.com/stress1"},
         ],
         "videos": [
             {"title": "Relaxation Techniques", "url": "https://www.youtube.com/watch?v=relaxvideo1"},
-            {"title": "Mindfulness Exercises", "url": "https://www.youtube.com/watch?v=mindfulnessvideo1"}
+            {"title": "Mindfulness Exercises", "url": "https://www.youtube.com/watch?v=mindfulnessvideo1"},
         ]
     }
 }
@@ -91,85 +45,107 @@ def get_relevant_resources(emotion):
     resources = suggestion_database.get(emotion, {})
     return resources.get("suggestions", []), resources.get("articles", []), resources.get("videos", [])
 
-# Enhanced Suggestion Function
-def suggest_activity(emotion_analysis):
-    max_emotion = max(emotion_analysis, key=emotion_analysis.get) if emotion_analysis else "neutral"
-    suggestions, articles, videos = get_relevant_resources(max_emotion)
-    return {
-        "suggestions": suggestions,
-        "articles": articles,
-        "videos": videos,
-    }
+# Function to load the model with error handling and retries
+@st.cache_resource
+def load_model():
+    try:
+        st.write("Attempting to load the emotion analysis model...")
+        # Using a smaller model for quick load times
+        emotion_analyzer = pipeline("text-classification", model="j-hartmann/emotion-english-distilroberta", device=0 if torch.cuda.is_available() else -1)
+        st.write("Model loaded successfully!")
+        return emotion_analyzer
+    except Exception as e:
+        st.write(f"Error loading the model: {e}")
+        return None
+
+# Function to predict emotion for a single response
+def predict_emotion_single(response, emotion_analyzer):
+    if emotion_analyzer is None:
+        st.error("Model not loaded. Please try reloading the app.")
+        return {"Error": "Emotion analyzer model not initialized. Please check model loading."}
+
+    try:
+        result = emotion_analyzer([response])
+        return {res["label"]: round(res["score"], 4) for res in result}
+    except Exception as e:
+        st.error(f"Prediction failed: {e}")
+        return {"Error": f"Prediction failed: {e}"}
+
+# Streamlit App Layout
+st.title("Emotion Prediction App: Your Personal Wellness Assistant")
+
+st.write("**How it works:**")
+st.write("- Enter your thoughts or feelings.")
+st.write("- Our AI analyzes your text to predict your emotional state.")
+st.write("- Receive personalized suggestions to improve your well-being.")
 
-# ---- User Input Section ----
-st.write("### 🌺 Answer 3 Simple Questions!")
+# Define questions for the user
 questions = [
-    "How are you feeling right now?",
-    "What was the most significant emotion you experienced today?",
-    "What helps you feel better when you’re stressed?"
+    "How are you feeling today?",
+    "Describe your mood in a few words.",
+    "What was the most significant emotion you felt this week?"
 ]
 
+# Initialize a dictionary to store responses
 responses = {}
 
-# ---- Ask Questions and Analyze Responses ----
+# Initialize the emotion analysis model with retries
+emotion_analyzer = None
+max_retries = 3
+retry_delay = 5  # seconds
+
+# Try loading the model with retries
+for attempt in range(max_retries):
+    emotion_analyzer = load_model()
+    if emotion_analyzer:
+        break
+    if attempt < max_retries - 1:
+        st.warning(f"Retrying model load... Attempt {attempt + 2}/{max_retries}")
+        time.sleep(retry_delay)
+    else:
+        st.error("Model failed to load after multiple attempts. Please try again later.")
+
+# Function to handle responses and emotion analysis
 for i, question in enumerate(questions, start=1):
-    st.write(f"#### ❓ Question {i}: {question}")
-    user_response = st.text_input(f"Your answer to Q{i}:", key=f"q{i}")
-    
+    user_response = st.text_input(f"Question {i}: {question}")
     if user_response:
-        with st.spinner("Analyzing emotion... 🎭"):
-            analysis = predict_emotion(user_response)
-        responses[question] = {"Response": user_response, "Analysis": analysis}
-        
-        # Display Emotion Analysis
-        st.success(f"🎯 Emotion Analysis: {analysis}")
-        
-        # Display Activity Suggestion
-        if analysis:
-            max_emotion = max(analysis, key=analysis.get)
-            activity_suggestion = suggest_activity(analysis)
-            st.write(f"### 🧘 Suggested Activity: {activity_suggestion['suggestions'][0]}")
-            st.write(f"📖 Suggested Articles:")
-            for article in activity_suggestion['articles']:
-                st.markdown(f"- [{article['title']}]({article['url']})")
-            st.write(f"🎥 Suggested Videos:")
-            for video in activity_suggestion['videos']:
-                st.markdown(f"- [{video['title']}]({video['url']})")
-
-# ---- Display Summary Button ----
-if st.button("Get Summary"):
-    st.write("### 📊 Your Emotional Insights Summary")
-
-    # Prepare the summary data
-    summary_data = []
-    for i, (question, details) in enumerate(responses.items(), start=1):
-        max_emotion = max(details["Analysis"], key=details["Analysis"].get)
-        activity_suggestion = suggest_activity(details["Analysis"])
-        
-        summary_data.append({
-            "Question": question,
-            "Your Response": details["Response"],
-            "Emotion Analysis": max_emotion,
-            "Suggested Activity": activity_suggestion["suggestions"][0],
-            "Suggested Articles": ", ".join([article['title'] for article in activity_suggestion['articles']]),
-            "Suggested Videos": ", ".join([video['title'] for video in activity_suggestion['videos']]),
-        })
-
-    # Create a DataFrame from the summary data
-    df = pd.DataFrame(summary_data)
-
-    # Display the summary table
-    st.dataframe(df)  # or use st.table(df) for a static table
-
-    # Provide additional suggestions at the end of all questions
-    st.write("### 🌟 Final Suggestions")
-    st.write("Incorporating mindfulness, deep breathing, and spending time outdoors can help with emotional balance. Keep exploring activities that promote well-being! 🌱")
-
-# ---- Footer ----
-st.markdown(
-    """
-    ---
-    **Developed using 🤗 Transformers**  
-    Designed for a fun and intuitive experience! 🌟  
-    """
-)
+        analysis = predict_emotion_single(user_response, emotion_analyzer)
+        responses[question] = (user_response, analysis)
+        st.write(f"**Your Response**: {user_response}")
+        st.write(f"**Emotion Analysis**: {analysis}")
+
+        # Based on the emotion, suggest activities, articles, and videos
+        max_emotion = max(analysis, key=analysis.get) if analysis else "neutral"
+        suggestions, articles, videos = get_relevant_resources(max_emotion)
+
+        if suggestions:
+            st.write(f"### 🧘 Suggested Activity: {suggestions[0]}")
+        else:
+            st.write("### 🧘 No suggestions available at the moment.")
+
+        if articles:
+            st.write(f"### 📚 Suggested Articles:")
+            for article in articles:
+                st.write(f"[{article['title']}]({article['url']})")
+        else:
+            st.write("### 📚 No articles available at the moment.")
+
+        if videos:
+            st.write(f"### 🎥 Suggested Videos:")
+            for video in videos:
+                st.write(f"[{video['title']}]({video['url']})")
+        else:
+            st.write("### 🎥 No videos available at the moment.")
+
+# Provide button to clear input fields
+if st.button("Clear Responses"):
+    st.experimental_rerun()
+
+# Display results once all responses are filled
+if st.button("Submit Responses"):
+    if responses:
+        st.write("-- Emotion Analysis Results ---")
+        for i, (question, (response, analysis)) in enumerate(responses.items(), start=1):
+            st.write(f"**{question}**")
+            st.write(f"Response: {response}")
+            st.write(f"Emotion Analysis: {analysis}")