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

[GNA] Disable AddFakeQuantizeFusion (#5037)

Browse Source 
* issue 52034
   * as per-channel quantization not supported in FQ layer in GNAPlugin
This commit is contained in:
Krzysztof Bruniecki 2021-03-31 17:21:04 +02:00 committed by GitHub
parent 320009d3b2
commit 77d4fbc286
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
inference-engine/src/gna_plugin/gna_plugin.cpp
View File

@ -55,6 +55,7 @@
#include <transformations/common_optimizations/fq_reshape_fusion.hpp>
#include <transformations/common_optimizations/pull_transpose_through_fq.hpp>
#include <transformations/common_optimizations/relu_fake_quantize_fusion.hpp>
#include <transformations/common_optimizations/add_fake_quantize_fusion.hpp>
#if GNA_LIB_VER == 2
#include <gna2-model-api.h>
@ -635,6 +636,8 @@ void GNAPlugin::LoadNetwork(CNNNetwork & _network) {
pass_config->disable<ngraph::pass::FakeQuantizeReshapeFusion>();
pass_config->disable<ngraph::pass::PullTransposeThroughFQUp>();
pass_config->disable<ngraph::pass::ReluFakeQuantizeFusion>();
// Consider to enable after per-channel quantization on FakeQuantize layer is supported in GNAPlugin, see issue 52034
pass_config->disable<ngraph::pass::AddFakeQuantizeFusion>();
manager.run_passes(graph);
convertedNetwork = InferenceEngine::details::convertFunctionToICNNNetwork(graph, clonedNetwork);
}