Backport of FQ+Mul transform to master (#2214)
authorTomasz Dołbniak <tomasz.dolbniak@intel.com>
Mon, 21 Sep 2020 10:21:27 +0000 (12:21 +0200)
committerGitHub <noreply@github.com>
Mon, 21 Sep 2020 10:21:27 +0000 (13:21 +0300)
commitdda6d9136bd2ce6b8f34ac048da6bbfc299371c3
tree2150f7e74b237511b1ae58d751fa2c9f08619d5f
parentc13ec24e1e27926eb0bc8e074666df74328f567c
Backport of FQ+Mul transform to master (#2214)

* Backport of FQ+Mul transform to master

* Accept any type of input to FQ in the transformation

* Test the fusion when all FQ inputs are non-const

* Fusion test when only one output limit is const

* Test passing the output of FQ to second input of Mul
inference-engine/src/transformations/include/transformations/common_optimizations/fq_mul_fusion.hpp [new file with mode: 0644]
inference-engine/src/transformations/src/transformations/common_optimizations/fq_mul_fusion.cpp [new file with mode: 0644]
inference-engine/src/transformations/src/transformations/convert_opset1_to_legacy/convert_opset1_to_legacy.cpp
inference-engine/tests/functional/inference_engine/transformations/fq_mul_fusion_test.cpp [new file with mode: 0644]