diff options
| author | Francesco Petrogalli <francesco.petrogalli@arm.com> | 2022-06-30 10:22:01 +0000 |
|---|---|---|
| committer | Francesco Petrogalli <francesco.petrogalli@arm.com> | 2022-07-19 09:26:27 +0000 |
| commit | 553f6953fe3bdfad53c11c25f305a16d79d83b24 (patch) | |
| tree | 73642b948b79662096f593458c6138d2f7f48ec6 /src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp | |
| parent | 99c46475daf277aa53e6747f9e41209f418fed33 (diff) | |
| download | ComputeLibrary-553f6953fe3bdfad53c11c25f305a16d79d83b24.tar.gz | |
[ONCPUML-951] Variable weight support for Convolution.
API changes for NEGEMMConvolutionLayer and CpuGemmConv2d
Built with:
scons neon=1 opencl=0 os=linux arch=armv8.2-a multi_isa=1 \
build=native -j32 Werror=false validation_tests=1 build_dir=opt \
standalone=1 asserts=1 experimental_fixed_format_kernels=1 .
Tested with:
./build/opt/tests/arm_compute_validation
Hardware where the test executable was run:
Neoverse N1
Test coverage:
* NEGEMMConvolutionLayer, CpuGemmConv2d
* NHWC (the only one supported by the fixed-format kernels)
* F16, F32
* Shapes: RunSmall
Change-Id: I4fd3e495a7cbf61210ea02d37440ba9652934e99
Signed-off-by: Francesco Petrogalli <francesco.petrogalli@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/7632
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp')
| -rw-r--r-- | src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp b/src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp index 95139c2bf6..be7a4ee570 100644 --- a/src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp +++ b/src/core/NEON/kernels/arm_gemm/gemm_quint8.cpp @@ -197,7 +197,7 @@ const GemmImplementation<uint8_t, uint8_t, Requantize32> *gemm_implementation_li } template UniqueGemmCommon<uint8_t, uint8_t> gemm<uint8_t, uint8_t, Requantize32>(const GemmArgs &args, const Requantize32 &os); -template bool has_opt_gemm<uint8_t, uint8_t, Requantize32>(const GemmArgs &args, const Requantize32 &os); +template bool has_opt_gemm<uint8_t, uint8_t, Requantize32>(WeightFormat &weight_format, const GemmArgs &args, const Requantize32 &os); template std::vector<KernelDescription> get_compatible_kernels<uint8_t, uint8_t, Requantize32>(const GemmArgs &args, const Requantize32 &os); } // namespace arm_gemm |