Update app.py

app.py

@@ -4,22 +4,29 @@ from huggingface_hub import InferenceClient
 client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
 
 def respond(message, history):
-    
+    response = ""
+
     messages = [{"role": "system", "content": "You are a friendly chatbot."}]
-    
+
     if history:
         messages.extend(history)
-        
+
     messages.append({"role": "user", "content": message})
-    
-    response = client.chat_completion(
+
+    stream = client.chat_completion(
         messages,
-        max_tokens=100
+        max_tokens=100,
+        temperature=1.2,
+        stream=True
     )
-    
-    return response['choices'][0]['message']['content'].strip()
+
+    for message in stream:
+        token = message.choices[0].delta.content
+
+        if token is not None:
+            response += token
+            yield response
 
 chatbot = gr.ChatInterface(respond, type="messages")
 
 chatbot.launch()
-