Update app.py

app.py

@@ -6,7 +6,6 @@ client = InferenceClient("Qwen/Qwen2.5-7B-Instruct")
 
 def respond(message, history):
 
-    # System message
     messages = [
         {
             "role": "system",
@@ -14,8 +13,11 @@ def respond(message, history):
         }
     ]
 
-    # Convert Gradio history format
-    for user_msg, bot_msg in history:
+    # history format for newer Gradio versions
+    for item in history:
+
+        user_msg = item["user"]
+        bot_msg = item["assistant"]
 
         messages.append({
             "role": "user",
@@ -27,16 +29,13 @@ def respond(message, history):
             "content": bot_msg
         })
 
-    # Add latest user message
     messages.append({
         "role": "user",
         "content": message
     })
 
-    # Store streamed response
     partial_response = ""
 
-    # Stream response
     stream = client.chat_completion(
         messages=messages,
         max_tokens=100,
@@ -45,12 +44,11 @@ def respond(message, history):
         stream=True
     )
 
-    # Receive tokens one by one
     for chunk in stream:
 
         token = chunk.choices[0].delta.content
 
-        if token is not None:
+        if token:
 
             partial_response += token