Create app.py

app.py

@@ -0,0 +1,45 @@
+import gradio as gr
+import joblib
+import pandas as pd
+
+# Model ve scaler yükleniyor
+model = joblib.load("xgb_model.pkl")
+scaler = joblib.load("scaler.pkl")
+
+# Tahmin fonksiyonu
+def predict_quality(pressure, temp_x_pressure, fusion_metric):
+    input_df = pd.DataFrame([[pressure, temp_x_pressure, fusion_metric]],
+                            columns=["Pressure (kPa)", "Temperature x Pressure", "Material Fusion Metric"])
+    scaled_data = scaler.transform(input_df)
+    prediction = model.predict(scaled_data)
+    return f"🔍 Tahmin Edilen Kalite Skoru: {prediction[0]:.2f}"
+
+# Chatbot formatında cevap veren fonksiyon
+def respond(press, temp, fusion, chat_history):
+    result = predict_quality(press, temp, fusion)
+    chat_history.append({"role": "user", "content": "Özellikleri girdim."})
+    chat_history.append({"role": "assistant", "content": result})
+    return chat_history
+
+# Arayüz
+with gr.Blocks() as demo:
+    gr.Markdown("## 🛠️ Kalite Tahmini Chatbotu")
+
+    chatbot = gr.Chatbot(label="Kalite Tahmin Chatbotu", type="messages")
+    state = gr.State([])
+
+    with gr.Row():
+        pressure = gr.Number(label="Pressure (kPa)")
+        temp_x_pressure = gr.Number(label="Temperature x Pressure")
+        fusion_metric = gr.Number(label="Material Fusion Metric")
+
+    send_btn = gr.Button("Tahmin Et")
+
+    send_btn.click(
+        fn=respond,
+        inputs=[pressure, temp_x_pressure, fusion_metric, state],
+        outputs=[chatbot],
+        queue=False
+    ).then(lambda chat: chat, inputs=[chatbot], outputs=[state])
+
+demo.launch()