Update app.py

app.py

@@ -121,9 +121,11 @@ def generate_response_stream(prompt, history, max_length, temperature, top_p, to
 
         params = og.GeneratorParams(model)
         params.set_search_options(**search_options)
-        params.input_ids = input_tokens
+        # FIX: Use the set_input_ids method instead of direct attribute assignment
+        params.set_input_ids(input_tokens)
 
         start_time = time.time()
+        # Create generator AFTER setting parameters including input_ids
         generator = og.Generator(model, params)
         model_status = "Generating..." # Update status indicator
         logging.info("Streaming response...")
@@ -157,6 +159,11 @@ def generate_response_stream(prompt, history, max_length, temperature, top_p, to
         logging.info(f"Generation complete. Tokens: {token_count}, Total Time: {total_time:.2f}s, TTFT: {ttft:.2f}ms, TPS: {tps:.2f}")
         model_status = f"Model Ready ({EXECUTION_PROVIDER.upper()} / {model_variant_name})" # Reset status
 
+    except AttributeError as ae:
+         # Catch potential future API changes or issues during generation setup
+         logging.error(f"AttributeError during generation setup: {ae}", exc_info=True)
+         model_status = f"Generation Setup Error: {ae}"
+         yield f"\n\nSorry, an error occurred setting up generation: {ae}"
     except Exception as e:
         logging.error(f"Error during generation: {e}", exc_info=True)
         model_status = f"Error during generation: {e}"