From f92ec6e31b279e88addbd0a112e247c797435ce9 Mon Sep 17 00:00:00 2001 From: nickfraser Date: Tue, 16 Jul 2024 18:17:32 +0100 Subject: [PATCH] Fix (export/qonnx): Fixed symbolic kwargs order. (#988) --- src/brevitas/export/onnx/qonnx/handler.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/brevitas/export/onnx/qonnx/handler.py b/src/brevitas/export/onnx/qonnx/handler.py index effe9a9a2..c42714346 100644 --- a/src/brevitas/export/onnx/qonnx/handler.py +++ b/src/brevitas/export/onnx/qonnx/handler.py @@ -72,9 +72,9 @@ def prepare_for_export(self, module: WeightQuantProxyFromInjector): 'scale': first_qweight.scale, 'zero_point': first_qweight.zero_point, 'bit_width': first_qweight.bit_width, - 'signed': first_qweight.signed, # narrow_range is not a property of the QuantTensor, take it from the proxy instead 'narrow_range': module.is_narrow_range, + 'signed': first_qweight.signed, # override rounding mode since quantization has been pre-applied 'rounding_mode': 'ROUND'}