IntenseWebs/openvino
3
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix performance on resnet50 quantized models (#7670)

Browse Source 
* Fix performance on resnet50 quantized models

LP transformations won't work on the model unless the last 4 inputs to FakeQuantize
are constants. In order to meet that requirement, we need to perform constant folding
for those inputs in QuantizeLinear ONNX operator.

Ticket: 65375

* fix "Cannot find blob with name: y" exception during onnx_model_quant_conv_linear

* remove linking with onnx_ngraph_frontend

* fix exclude path
This commit is contained in:
Mateusz Tabaka
2021-10-13 09:18:37 +02:00
committed by GitHub
parent 0d020974f9
commit db527fff41
14 changed files with 401 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
cmake/test_model_zoo.cmake
View File

@@ -66,6 +66,11 @@ ov_model_convert("${CMAKE_CURRENT_SOURCE_DIR}/ngraph/test"
"${CMAKE_RUNTIME_OUTPUT_DIRECTORY}/test_model_zoo/ngraph"
onnx_out_files)
set(rel_path "inference-engine/tests/functional/plugin/shared/models")
ov_model_convert("${OpenVINO_SOURCE_DIR}/${rel_path}"
"${CMAKE_RUNTIME_OUTPUT_DIRECTORY}/test_model_zoo/func_tests/models"
ft_out_files)
set(rel_path "inference-engine/tests/functional/inference_engine/onnx_reader")
ov_model_convert("${OpenVINO_SOURCE_DIR}/${rel_path}"
"${CMAKE_RUNTIME_OUTPUT_DIRECTORY}/test_model_zoo/onnx_reader"
@@ -116,6 +121,7 @@ if(ENABLE_TESTS)
endif()
add_custom_target(test_model_zoo DEPENDS ${onnx_out_files}
${ft_out_files}
${ie_onnx_out_files}
${ie_serialize_out_files}
${ie_onnx_import_out_files})