Update download_model.py

download_model.py

@@ -29,19 +29,15 @@ for cfg in configs:
 if not model_path:
     raise RuntimeError("Could not find model.onnx in any repo type")
 
-# Save original first
 print("Saving original model...")
 shutil.copy(model_path, "./model/model_original.onnx")
 
-# Quantize to INT8
 print("Quantizing to INT8...")
 quantize_dynamic(
     model_input="./model/model_original.onnx",
     model_output="./model/model.onnx",
-    weight_type=QuantType.QUInt8,
-    optimize_model=True
+    weight_type=QuantType.QUInt8
 )
 
-# Remove original to save space
 os.remove("./model/model_original.onnx")
 print("✅ Quantized ONNX model saved to ./model/model.onnx")