diff options
| author | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2021-08-04 15:22:28 +0100 |
|---|---|---|
| committer | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2021-08-13 12:34:17 +0000 |
| commit | 37a4611dcbefc9ad19af803216d9e2cd5ec64697 (patch) | |
| tree | b479ebcd25e1326afa6e756c70ded4e87c0fa651 /src/runtime/CL/functions/CLFullyConnectedLayer.cpp | |
| parent | 7c2965760501ac96836b4f131d361266c99b91f2 (diff) | |
| download | ComputeLibrary-37a4611dcbefc9ad19af803216d9e2cd5ec64697.tar.gz | |
Add test for validating batched-GEMM on CPU
- Add new dataset for batched-GEMM
- Add test for running batched-GEMM without bias. Currently bias is not
supported in batched-GEMM
- Fix reference implementation to slide correctly the RHS tensor
Resolves COMPMID-4588
Change-Id: I20fcb5d9160f44292b7cc34570add911b1d732f6
Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/6040
Reviewed-by: SiCong Li <sicong.li@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/runtime/CL/functions/CLFullyConnectedLayer.cpp')
0 files changed, 0 insertions, 0 deletions