diff options
| author | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2018-08-23 17:26:21 +0100 |
|---|---|---|
| committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:54:54 +0000 |
| commit | fbf3ecc1833b860019f965c88cda87ec9f44c3d5 (patch) | |
| tree | a839e4c12b27346731c1dfd9a58e6eec343e5569 /tests/benchmark/fixtures | |
| parent | 1fa17c2edc51e53c6bd388b332d825208f6562e8 (diff) | |
| download | ComputeLibrary-fbf3ecc1833b860019f965c88cda87ec9f44c3d5.tar.gz | |
COMPMID-1534 - Fix GEMM and Magnitude test for FP16
On GEMM we had accuracy issue
On Magnitude we have disabled the fp16 acceleration since we do not have feature parity with CL
and this function is not used for ML
Change-Id: Iaebe3bbbd2a9f45db0c714aa5ebaf48eb0b65741
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/145467
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Diffstat (limited to 'tests/benchmark/fixtures')
| -rw-r--r-- | tests/benchmark/fixtures/MagnitudeFixture.h | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/tests/benchmark/fixtures/MagnitudeFixture.h b/tests/benchmark/fixtures/MagnitudeFixture.h index f75540c66f..a1b8054529 100644 --- a/tests/benchmark/fixtures/MagnitudeFixture.h +++ b/tests/benchmark/fixtures/MagnitudeFixture.h @@ -41,7 +41,7 @@ class MagnitudeFixture : public framework::Fixture { public: template <typename...> - void setup(const TensorShape &shape, Format format, MagnitudeType magnitude_type, bool use_fp16) + void setup(const TensorShape &shape, Format format, MagnitudeType magnitude_type) { // Create tensors src1 = create_tensor<TensorType>(shape, format); @@ -49,7 +49,7 @@ public: dst = create_tensor<TensorType>(shape, format); // Create and configure function - magnitude_func.configure(&src1, &src2, &dst, magnitude_type, use_fp16); + magnitude_func.configure(&src1, &src2, &dst, magnitude_type); // Allocate tensors src1.allocator()->allocate(); |