Update Space

app.py

@@ -113,20 +113,8 @@ def respond(message, history, system_message, max_tokens, temperature, top_p):
     # 3) Generate the assistant response from your HF model.
     #    (This code streams token-by-token.)
     response = ""
-    for resp_chunk in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = resp_chunk.choices[0].delta.get("content", "")
-        response += token
-
-        # If you want to display the partial response with the EOU probability
-        # appended at the bottom, you can do so each step. For cleanliness,
-        # we'll do it in-line as a bracketed note at the end. 
-        yield response + f"\n\n[EOU Probability: {eou_prob:.4f}]"
+    
+    yield f"[EOU Probability: {eou_prob:.4f}]"
 
 # ------------------------------------------------
 # Gradio ChatInterface