Update app.py

app.py

@@ -5,31 +5,29 @@ client = InferenceClient("thviet79/model-QA-medical")
 
 def respond(
     message,
-    history: list[dict],  # Use list of dicts instead of tuples
+    history: list[tuple[str, str]],
     system_message,
     max_tokens,
     temperature,
     top_p,
 ):
-    # Prepare messages for the model
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
-        if val["role"] == "user":
-            messages.append({"role": "user", "content": val["content"]})
-        if val["role"] == "assistant":
-            messages.append({"role": "assistant", "content": val["content"]})
+        if val[0]:
+            messages.append({"role": "user", "content": val[0]})
+        if val[1]:
+            messages.append({"role": "assistant", "content": val[1]})
 
     messages.append({"role": "user", "content": message})
 
-    # Temporary placeholder response for testing
-    response = f"Test response for message: {message}"
-    
-    # Update history with assistant's response
-    history.append({"role": "user", "content": message})
-    history.append({"role": "assistant", "content": response})
-
-    return response, history
+    # Here, call the Hugging Face Inference API with the messages and other parameters
+    response = client.text_generation(
+        model="thviet79/model-QA-medical",
+        inputs=messages,
+        parameters={"max_tokens": max_tokens, "temperature": temperature, "top_p": top_p},
+    )
+    return response["generated_text"]  # Extract the model's response
 
 demo = gr.ChatInterface(
     respond,