Adjust queue setup

app.py

@@ -378,4 +378,9 @@ with gr.Blocks(title="ChatGLM2-6B-int4", theme=gr.themes.Soft(text_size="sm")) a
 # demo.queue().launch(share=False, inbrowser=True)
 # demo.queue().launch(share=True, inbrowser=True, debug=True)
 
-demo.queue().launch(debug=True)
+# concurrency_count > 1 requires more memory, max_size: queue size
+# T4 medium: 30GB, model size: ~6G concurrency_count = 3
+# leave one for api access
+# reduce to 3 if OOM
+
+block.queue(concurrency_count=4, max_size=20).launch(debug=True)