token check

app.py

@@ -27,7 +27,7 @@ class StopOnTokens(StoppingCriteria):
         return False
 
 
-@spaces.GPU()
+@spaces.GPU(duration=120)
 def predict(history, prompt, max_length, top_p, temperature):
     stop = StopOnTokens()
     messages = []
@@ -79,6 +79,9 @@ with gr.Blocks() as demo:
             <a href="https://github.com/THUDM/LongWriter">🌐 Github</a> |
             <a href="https://arxiv.org/pdf/2408.07055">📜 arxiv </a>
         </div>
+        <div style="text-align: center; font-size: 15px; font-weight: bold; color: red; margin-bottom: 20px;">
+            ⚠️ This is just a basic demo. Due to the scheduling limitations of Zero GPU, the output length is restricted to under 4K. If you wish to experience the full capabilities of the model (output exceeding 10K), please deploy the model yourself. Thank you for your understanding.
+        </div>
         """
     )
     chatbot = gr.Chatbot()
@@ -94,7 +97,7 @@ with gr.Blocks() as demo:
             pBtn = gr.Button("Set Prompt")
         with gr.Column(scale=1):
             emptyBtn = gr.Button("Clear History")
-            max_length = gr.Slider(0, 10000000, value=10000, step=1.0, label="Maximum length", interactive=True)
+            max_length = gr.Slider(0, 128000, value=4096, step=1.0, label="Maximum length(Input + Output)", interactive=True)
             top_p = gr.Slider(0, 1, value=0.8, step=0.01, label="Top P", interactive=True)
             temperature = gr.Slider(0.01, 1, value=0.6, step=0.01, label="Temperature", interactive=True)