Setting use_onnx_custom_quantizer_ops to true in tutorials pytorch ON…

…NX export (#866) Co-authored-by: Ofir Gordon <[email protected]>
sony · Nov 22, 2023 · f926b03 · f926b03
1 parent e921060
commit f926b03
Showing 1 changed file with 2 additions and 1 deletion.
diff --git a/tutorials/quick_start/pytorch_fw/quant.py b/tutorials/quick_start/pytorch_fw/quant.py
@@ -139,7 +139,8 @@ def quantize(model: nn.Module,
         _, onnx_file_path = tempfile.mkstemp('.onnx') # Path of exported model
         mct.exporter.pytorch_export_model(model=quantized_model, save_model_path=onnx_file_path,
                                           repr_dataset=representative_data_gen, target_platform_capabilities=tpc,
-                                          serialization_format=mct.exporter.PytorchExportSerializationFormat.ONNX)
+                                          serialization_format=mct.exporter.PytorchExportSerializationFormat.ONNX,
+                                          use_onnx_custom_quantizer_ops=True)
 
 
     return quantized_model, QuantInfo(user_info=quantization_info, tpc_info=tpc.get_info(), quantization_workflow=workflow, mp_weights_compression=mp_wcr)