Update app.py

app.py

@@ -5,9 +5,22 @@ import pandas as pd
 import numpy as np
 import plotly.express as px
 import time
-from streamlit_lottie import st_lottie
-import requests
-import json
+
+# Try to import streamlit_lottie, but provide fallback if it fails
+try:
+    from streamlit_lottie import st_lottie
+    import requests
+    def load_lottie_url(url: str):
+        try:
+            r = requests.get(url)
+            if r.status_code != 200:
+                return None
+            return r.json()
+        except:
+            return None
+    LOTTIE_AVAILABLE = True
+except ImportError:
+    LOTTIE_AVAILABLE = False
 
 # Page configuration
 st.set_page_config(
@@ -82,26 +95,28 @@ st.markdown("""
         text-align: center;
         color: #666;
     }
+    .brain-icon {
+        font-size: 5rem;
+        text-align: center;
+        margin-bottom: 1rem;
+        color: #5E35B1;
+    }
 </style>
 """, unsafe_allow_html=True)
 
-# Function to load Lottie animation
-def load_lottie_url(url: str):
-    r = requests.get(url)
-    if r.status_code != 200:
-        return None
-    return r.json()
-
-# Load animations
-brain_animation = load_lottie_url("https://assets9.lottiefiles.com/packages/lf20_twdne5i2.json")
-analyzing_animation = load_lottie_url("https://assets8.lottiefiles.com/private_files/lf30_p9aibxmu.json")
-
-# Define model and tokenizer paths from Hugging Face
-MODEL_PATH = "DrSyedFaizan/mindBERT"
-
 # Create sidebar
 with st.sidebar:
-    st_lottie(brain_animation, height=200, key="brain_animation")
+    # Use either Lottie or a simple icon
+    if LOTTIE_AVAILABLE:
+        # Fixed Lottie URLs that are reliable
+        brain_animation = load_lottie_url("https://lottie.host/2eb12c32-787a-46f7-ac20-34c166d1a285/UcEEbJlFVH.json")
+        if brain_animation:
+            st_lottie(brain_animation, height=200, key="brain_animation")
+        else:
+            st.markdown("<div class='brain-icon'>🧠</div>", unsafe_allow_html=True)
+    else:
+        st.markdown("<div class='brain-icon'>🧠</div>", unsafe_allow_html=True)
+    
     st.markdown("## About MindBERT")
     st.info(
         "MindBERT is a fine-tuned BERT model specifically designed to detect "
@@ -145,12 +160,19 @@ with tab1:
     # Model loading feedback
     @st.cache_resource
     def load_model():
-        tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
-        model = AutoModelForSequenceClassification.from_pretrained(MODEL_PATH)
-        return tokenizer, model
+        try:
+            tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
+            model = AutoModelForSequenceClassification.from_pretrained(MODEL_PATH)
+            return tokenizer, model, True
+        except Exception as e:
+            st.error(f"Error loading model: {str(e)}")
+            return None, None, False
+    
+    # Define model and tokenizer paths from Hugging Face
+    MODEL_PATH = "DrSyedFaizan/mindBERT"
     
     with st.spinner("Loading model..."):
-        tokenizer, model = load_model()
+        tokenizer, model, model_loaded = load_model()
     
     # Analysis button
     col1, col2, col3 = st.columns([1, 2, 1])
@@ -159,112 +181,124 @@ with tab1:
     
     # Prediction logic
     if analyze_button:
-        if user_input.strip():
-            # Show analyzing animation
+        if not model_loaded:
+            st.error("Model failed to load. Please try again later.")
+        elif not user_input.strip():
+            st.warning("Please enter some text for analysis.")
+        else:
+            # Show analyzing animation or spinner
             with st.spinner("Analyzing..."):
-                st_lottie(analyzing_animation, height=200, key="analyze_animation", speed=1.5)
-                
-                # Tokenize input
-                inputs = tokenizer(user_input, return_tensors="pt", truncation=True, padding=True)
-                
-                # Make prediction
-                with torch.no_grad():
-                    outputs = model(**inputs)
-                    logits = outputs.logits
-                    probabilities = torch.nn.functional.softmax(logits, dim=1)[0]
-                    predicted_class = torch.argmax(logits, dim=1).item()
-                
-                # Mapping predicted class to mental state with descriptions
-                label_map = {
-                    0: {"name": "Anxiety", "color": "#FFD54F", "description": "Characterized by excessive worry, fear, or nervousness."},
-                    1: {"name": "Bipolar", "color": "#FF7043", "description": "Featuring alternating periods of depression and mania or elevated mood."},
-                    2: {"name": "Depression", "color": "#4FC3F7", "description": "Persistent feelings of sadness, hopelessness, and loss of interest."},
-                    3: {"name": "Normal", "color": "#81C784", "description": "Balanced emotional state without significant mental health concerns."},
-                    4: {"name": "Personality Disorder", "color": "#9575CD", "description": "Persistent patterns of thinking and behavior that deviate from social norms."},
-                    5: {"name": "Stress", "color": "#FF8A65", "description": "Physical or emotional tension due to challenging circumstances."},
-                    6: {"name": "Suicidal", "color": "#F44336", "description": "Thoughts or intentions of self-harm or taking one's own life."}
-                }
-                
-                mental_state = label_map.get(predicted_class, {"name": "Unknown", "color": "#BDBDBD", "description": "Unable to classify the mental state."})
+                if LOTTIE_AVAILABLE:
+                    analyzing_animation = load_lottie_url("https://lottie.host/16c400ec-7d59-4c0c-a84b-56c9134cd673/20XZXacKUS.json")
+                    if analyzing_animation:
+                        st_lottie(analyzing_animation, height=200, key="analyze_animation", speed=1.5)
                 
-                # Create data for visualization
-                all_probs = {label_map[i]["name"]: prob.item() * 100 for i, prob in enumerate(probabilities)}
-                probs_df = pd.DataFrame(list(all_probs.items()), columns=["Mental State", "Confidence (%)"])
-                probs_df = probs_df.sort_values("Confidence (%)", ascending=False)
+                # Add a slight delay to show the animation
+                time.sleep(1)
                 
-                # Display results
-                st.markdown("<div class='result-box'>", unsafe_allow_html=True)
-                
-                # Primary result
-                col1, col2 = st.columns([1, 2])
-                with col1:
-                    st.markdown(f"<div class='metric-value' style='color:{mental_state['color']}'>{mental_state['name']}</div>", unsafe_allow_html=True)
-                    st.markdown("<div class='metric-label'>Primary Detection</div>", unsafe_allow_html=True)
-                
-                with col2:
-                    st.markdown(f"<div style='background-color:{mental_state['color']}20; padding:15px; border-radius:10px; border-left:5px solid {mental_state['color']}'>")
-                    st.markdown(f"<b>{mental_state['name']}</b>: {mental_state['description']}")
-                    st.markdown("</div>", unsafe_allow_html=True)
-                
-                # Confidence scores visualization
-                st.markdown("<h3 class='sub-header'>Confidence Analysis</h3>", unsafe_allow_html=True)
-                
-                # Create bar chart
-                fig = px.bar(
-                    probs_df, 
-                    x="Confidence (%)", 
-                    y="Mental State",
-                    orientation="h",
-                    color="Mental State",
-                    color_discrete_map={
-                        "Anxiety": "#FFD54F",
-                        "Bipolar": "#FF7043",
-                        "Depression": "#4FC3F7",
-                        "Normal": "#81C784",
-                        "Personality Disorder": "#9575CD",
-                        "Stress": "#FF8A65",
-                        "Suicidal": "#F44336",
-                        "Unknown": "#BDBDBD"
+                try:
+                    # Tokenize input
+                    inputs = tokenizer(user_input, return_tensors="pt", truncation=True, padding=True)
+                    
+                    # Make prediction
+                    with torch.no_grad():
+                        outputs = model(**inputs)
+                        logits = outputs.logits
+                        probabilities = torch.nn.functional.softmax(logits, dim=1)[0]
+                        predicted_class = torch.argmax(logits, dim=1).item()
+                    
+                    # Mapping predicted class to mental state with descriptions
+                    label_map = {
+                        0: {"name": "Anxiety", "color": "#FFD54F", "description": "Characterized by excessive worry, fear, or nervousness."},
+                        1: {"name": "Bipolar", "color": "#FF7043", "description": "Featuring alternating periods of depression and mania or elevated mood."},
+                        2: {"name": "Depression", "color": "#4FC3F7", "description": "Persistent feelings of sadness, hopelessness, and loss of interest."},
+                        3: {"name": "Normal", "color": "#81C784", "description": "Balanced emotional state without significant mental health concerns."},
+                        4: {"name": "Personality Disorder", "color": "#9575CD", "description": "Persistent patterns of thinking and behavior that deviate from social norms."},
+                        5: {"name": "Stress", "color": "#FF8A65", "description": "Physical or emotional tension due to challenging circumstances."},
+                        6: {"name": "Suicidal", "color": "#F44336", "description": "Thoughts or intentions of self-harm or taking one's own life."}
                     }
-                )
-                fig.update_layout(
-                    height=350,
-                    margin=dict(l=20, r=20, t=30, b=20),
-                    xaxis_title="Confidence (%)",
-                    yaxis_title="",
-                    yaxis=dict(autorange="reversed"),
-                    xaxis=dict(range=[0, 100])
-                )
-                st.plotly_chart(fig, use_container_width=True)
-                
-                # Warning for high-risk categories
-                if mental_state["name"] in ["Suicidal", "Depression"] and all_probs[mental_state["name"]] > 50:
-                    st.warning(
-                        "⚠️ **High-risk mental state detected.** If you or someone you know is experiencing "
-                        "suicidal thoughts, please seek immediate professional help or call the National "
-                        "Suicide Prevention Lifeline at 988 or 1-800-273-8255."
+                    
+                    mental_state = label_map.get(predicted_class, {"name": "Unknown", "color": "#BDBDBD", "description": "Unable to classify the mental state."})
+                    
+                    # Create data for visualization
+                    all_probs = {label_map[i]["name"]: prob.item() * 100 for i, prob in enumerate(probabilities)}
+                    probs_df = pd.DataFrame(list(all_probs.items()), columns=["Mental State", "Confidence (%)"])
+                    probs_df = probs_df.sort_values("Confidence (%)", ascending=False)
+                    
+                    # Display results
+                    st.markdown("<div class='result-box'>", unsafe_allow_html=True)
+                    
+                    # Primary result
+                    col1, col2 = st.columns([1, 2])
+                    with col1:
+                        st.markdown(f"<div class='metric-value' style='color:{mental_state['color']}'>{mental_state['name']}</div>", unsafe_allow_html=True)
+                        st.markdown("<div class='metric-label'>Primary Detection</div>", unsafe_allow_html=True)
+                    
+                    with col2:
+                        st.markdown(f"<div style='background-color:{mental_state['color']}20; padding:15px; border-radius:10px; border-left:5px solid {mental_state['color']}'>")
+                        st.markdown(f"<b>{mental_state['name']}</b>: {mental_state['description']}")
+                        st.markdown("</div>", unsafe_allow_html=True)
+                    
+                    # Confidence scores visualization
+                    st.markdown("<h3 class='sub-header'>Confidence Analysis</h3>", unsafe_allow_html=True)
+                    
+                    # Create bar chart
+                    fig = px.bar(
+                        probs_df, 
+                        x="Confidence (%)", 
+                        y="Mental State",
+                        orientation="h",
+                        color="Mental State",
+                        color_discrete_map={
+                            "Anxiety": "#FFD54F",
+                            "Bipolar": "#FF7043",
+                            "Depression": "#4FC3F7",
+                            "Normal": "#81C784",
+                            "Personality Disorder": "#9575CD",
+                            "Stress": "#FF8A65",
+                            "Suicidal": "#F44336",
+                            "Unknown": "#BDBDBD"
+                        }
                     )
-                
-                st.markdown("</div>", unsafe_allow_html=True)
-                
-                # Suggestion based on detected mental state
-                suggestion_map = {
-                    "Anxiety": "Consider breathing exercises, meditation, or consulting with a mental health professional about anxiety management techniques.",
-                    "Bipolar": "Regular sleep schedules and medication management with professional oversight can help stabilize mood swings.",
-                    "Depression": "Regular physical activity, social connection, and professional therapy can be beneficial for managing depression.",
-                    "Normal": "Continue maintaining a healthy lifestyle with regular exercise, good sleep habits, and social connections.",
-                    "Personality Disorder": "Long-term psychotherapy with a specialist in personality disorders is often recommended.",
-                    "Stress": "Stress reduction techniques such as mindfulness, time management, and setting boundaries can be helpful.",
-                    "Suicidal": "Please seek immediate professional help. Call the National Suicide Prevention Lifeline at 988 or 1-800-273-8255."
-                }
-                
-                st.markdown("<div class='result-box'>", unsafe_allow_html=True)
-                st.markdown("<h3 class='sub-header'>Suggestions</h3>", unsafe_allow_html=True)
-                st.info(suggestion_map.get(mental_state["name"], "Consider consulting with a mental health professional for personalized guidance."))
-                st.markdown("</div>", unsafe_allow_html=True)
-                
-        else:
-            st.warning("Please enter some text for analysis.")
+                    fig.update_layout(
+                        height=350,
+                        margin=dict(l=20, r=20, t=30, b=20),
+                        xaxis_title="Confidence (%)",
+                        yaxis_title="",
+                        yaxis=dict(autorange="reversed"),
+                        xaxis=dict(range=[0, 100])
+                    )
+                    st.plotly_chart(fig, use_container_width=True)
+                    
+                    # Warning for high-risk categories
+                    if mental_state["name"] in ["Suicidal", "Depression"] and all_probs[mental_state["name"]] > 50:
+                        st.warning(
+                            "⚠️ **High-risk mental state detected.** If you or someone you know is experiencing "
+                            "suicidal thoughts, please seek immediate professional help or call the National "
+                            "Suicide Prevention Lifeline at 988 or 1-800-273-8255."
+                        )
+                    
+                    st.markdown("</div>", unsafe_allow_html=True)
+                    
+                    # Suggestion based on detected mental state
+                    suggestion_map = {
+                        "Anxiety": "Consider breathing exercises, meditation, or consulting with a mental health professional about anxiety management techniques.",
+                        "Bipolar": "Regular sleep schedules and medication management with professional oversight can help stabilize mood swings.",
+                        "Depression": "Regular physical activity, social connection, and professional therapy can be beneficial for managing depression.",
+                        "Normal": "Continue maintaining a healthy lifestyle with regular exercise, good sleep habits, and social connections.",
+                        "Personality Disorder": "Long-term psychotherapy with a specialist in personality disorders is often recommended.",
+                        "Stress": "Stress reduction techniques such as mindfulness, time management, and setting boundaries can be helpful.",
+                        "Suicidal": "Please seek immediate professional help. Call the National Suicide Prevention Lifeline at 988 or 1-800-273-8255."
+                    }
+                    
+                    st.markdown("<div class='result-box'>", unsafe_allow_html=True)
+                    st.markdown("<h3 class='sub-header'>Suggestions</h3>", unsafe_allow_html=True)
+                    st.info(suggestion_map.get(mental_state["name"], "Consider consulting with a mental health professional for personalized guidance."))
+                    st.markdown("</div>", unsafe_allow_html=True)
+                    
+                except Exception as e:
+                    st.error(f"Error during analysis: {str(e)}")
+                    st.info("Please try again with different text or contact support if the issue persists.")
 
 with tab2:
     st.markdown("<h3 class='sub-header'>Mental Health Categories Explained</h3>", unsafe_allow_html=True)