Update app.py

app.py

@@ -283,7 +283,7 @@ def init_agent():
     )
     sampling_params = SamplingParams(
         temperature=0.2,
-        max_tokens=256,  # Reduced for faster streaming
+        max_tokens=256,
         stop=["</s>", "[INST]"],
     )
     log_system_usage("After Load")
@@ -375,7 +375,7 @@ Patient Record Excerpt (Chunk {0} of {1}):
                                             history[-1] = {"role": "assistant", "content": chunk_response}
                                             threading.Thread(target=log_response_partial, args=(chunk_response,)).start()
                                             yield history, None, ""
-                                            await asyncio.sleep(0.01)  # Prevent UI blocking
+                                            await asyncio.sleep(0.01)
                     
                             if chunk_response:
                                 combined_response += f"--- Analysis for Chunk {batch_idx + 1} ---\n{chunk_response}\n"
@@ -398,8 +398,8 @@ Patient Record Excerpt (Chunk {0} of {1}):
                 threading.Thread(target=log_response_partial, args=(f"Error: {str(e)}",)).start()
                 yield history, None, f"Error occurred during analysis: {str(e)}"
 
-        send_btn.click(analyze, inputs=[msg_input, gr.State([]), file_upload], outputs=[chatbot, download_output, final_summary], _js="() => {return {streaming: true}}")
-        msg_input.submit(analyze, inputs=[msg_input, gr.State([]), file_upload], outputs=[chatbot, download_output, final_summary], _js="() => {return {streaming: true}}")
+        send_btn.click(analyze, inputs=[msg_input, gr.State([]), file_upload], outputs=[chatbot, download_output, final_summary])
+        msg_input.submit(analyze, inputs=[msg_input, gr.State([]), file_upload], outputs=[chatbot, download_output, final_summary])
     return demo
 
 if __name__ == "__main__":