Add sentiment analysis Gradio app with Git LFS

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.keras filter=lfs diff=lfs merge=lfs -text

app.py

@@ -0,0 +1,182 @@
+import gradio as gr
+import tensorflow as tf
+from sentence_transformers import SentenceTransformer
+import numpy as np
+
+# Load models
+print("Loading models...")
+embedding_model = SentenceTransformer("alibayram/distilled-sentence-transformer-c400")
+sentiment_model = tf.keras.models.load_model('sentiment_v1.keras')
+print("Models loaded successfully!")
+
+def predict_sentiment(text):
+    """
+    Predict sentiment for the given text and return visual star representation
+    """
+    if not text.strip():
+        return "", "Lütfen bir metin girin."
+    
+    # Encode the text
+    encoded_text = embedding_model.encode([text])
+    
+    # Get prediction
+    prediction = sentiment_model.predict(encoded_text, verbose=0)
+    
+    # Get the predicted star rating (1-5)
+    predicted_class = np.argmax(prediction[0])
+    predicted_stars = predicted_class + 1  # Convert 0-4 to 1-5
+    
+    # Get confidence (probability of predicted class)
+    confidence = prediction[0][predicted_class] * 100
+    
+    # Create star visualization
+    stars_html = create_star_display(predicted_stars, prediction[0])
+    
+    # Create probability distribution display
+    prob_html = create_probability_display(prediction[0])
+    
+    return stars_html, prob_html
+
+def create_star_display(rating, probabilities):
+    """
+    Create HTML for star display with the predicted rating
+    """
+    confidence = probabilities[rating - 1] * 100
+    
+    stars_html = '<div style="text-align: center; padding: 20px;">'
+    stars_html += '<div style="font-size: 60px; margin-bottom: 10px;">'
+    
+    # Full stars
+    for i in range(rating):
+        stars_html += '⭐'
+    
+    # Empty stars
+    for i in range(5 - rating):
+        stars_html += '☆'
+    
+    stars_html += '</div>'
+    stars_html += f'<div style="font-size: 24px; font-weight: bold; color: #FF6B35; margin-top: 10px;">{rating} / 5 Yıldız</div>'
+    stars_html += f'<div style="font-size: 16px; color: #666; margin-top: 5px;">Güven: %{confidence:.1f}</div>'
+    stars_html += '</div>'
+    
+    return stars_html
+
+def create_probability_display(probabilities):
+    """
+    Create HTML for probability distribution display
+    """
+    html = '<div style="padding: 20px;">'
+    html += '<h3 style="text-align: center; color: #333; margin-bottom: 15px;">Olasılık Dağılımı</h3>'
+    
+    for i, prob in enumerate(probabilities):
+        stars = i + 1
+        percentage = prob * 100
+        
+        # Create a bar for each star rating
+        html += f'<div style="margin-bottom: 10px;">'
+        html += f'<div style="display: flex; align-items: center; margin-bottom: 5px;">'
+        html += f'<span style="min-width: 80px; font-weight: 500;">{stars} Yıldız:</span>'
+        html += f'<div style="flex: 1; background-color: #f0f0f0; border-radius: 10px; height: 25px; margin: 0 10px; overflow: hidden;">'
+        html += f'<div style="width: {percentage}%; background: linear-gradient(90deg, #FFD700, #FFA500); height: 100%; border-radius: 10px; transition: width 0.3s ease;"></div>'
+        html += '</div>'
+        html += f'<span style="min-width: 60px; text-align: right; font-weight: 500; color: #FF6B35;">%{percentage:.1f}</span>'
+        html += '</div>'
+        html += '</div>'
+    
+    html += '</div>'
+    return html
+
+# Custom CSS for better styling
+custom_css = """
+#title {
+    text-align: center;
+    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+    color: white;
+    padding: 30px;
+    border-radius: 15px;
+    margin-bottom: 20px;
+}
+
+#subtitle {
+    text-align: center;
+    color: #666;
+    margin-bottom: 30px;
+    font-size: 18px;
+}
+
+.input-box textarea {
+    border-radius: 10px !important;
+    border: 2px solid #e0e0e0 !important;
+    font-size: 16px !important;
+}
+
+.predict-button {
+    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
+    border: none !important;
+    border-radius: 10px !important;
+    padding: 12px 30px !important;
+    font-size: 18px !important;
+    font-weight: bold !important;
+    color: white !important;
+}
+
+#output-box {
+    border-radius: 15px;
+    border: 2px solid #e0e0e0;
+    background-color: #fafafa;
+}
+"""
+
+# Create Gradio interface
+with gr.Blocks(css=custom_css, theme=gr.themes.Soft()) as demo:
+    gr.HTML('<div id="title"><h1 style="margin: 0; font-size: 42px;">🌟 Duygu Analizi 🌟</h1></div>')
+    gr.HTML('<div id="subtitle">Cümlenizi yazın ve duygu analizini keşfedin!</div>')
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            input_text = gr.Textbox(
+                label="Metninizi Girin",
+                placeholder="Örnek: Çok beğendim, harika bir ürün!",
+                lines=5,
+                elem_classes="input-box"
+            )
+            predict_btn = gr.Button("🔮 Tahmin Et", elem_classes="predict-button", size="lg")
+            
+            # Example inputs
+            gr.Examples(
+                examples=[
+                    ["Çok iyi beğendim, harika bir ürün!"],
+                    ["Hiç güzel değil, kesinlikle tavsiye etmem"],
+                    ["Fena değil alınır ama emin de değilim"],
+                    ["Mükemmel! Beklentilerimin çok üzerinde"],
+                    ["Berbat bir deneyimdi, çok kötü"]
+                ],
+                inputs=input_text,
+                label="Örnek Cümleler"
+            )
+    
+    with gr.Row():
+        with gr.Column(scale=1):
+            star_output = gr.HTML(label="Tahmin Sonucu", elem_id="output-box")
+        
+        with gr.Column(scale=1):
+            prob_output = gr.HTML(label="Detaylı Analiz", elem_id="output-box")
+    
+    # Connect the button to the prediction function
+    predict_btn.click(
+        fn=predict_sentiment,
+        inputs=input_text,
+        outputs=[star_output, prob_output]
+    )
+    
+    # Also allow Enter key to trigger prediction
+    input_text.submit(
+        fn=predict_sentiment,
+        inputs=input_text,
+        outputs=[star_output, prob_output]
+    )
+    
+    gr.HTML('<div style="text-align: center; margin-top: 30px; color: #999; font-size: 14px;">Model: alibayram/distilled-sentence-transformer-c400</div>')
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,5 @@
+gradio
+tensorflow
+tf-keras
+sentence-transformers
+numpy

sentiment_v1.keras

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe6a93d8148ca6245014e44d5e0cbc6c586edb86e04d069deee787288978a391
+size 20505785