diff options
| author | Sang-Hoon Park <sang-hoon.park@arm.com> | 2020-03-13 14:56:05 +0000 |
|---|---|---|
| committer | Sang-Hoon Park <sang-hoon.park@arm.com> | 2020-04-07 09:00:09 +0000 |
| commit | 0d008f77b0085619c446d0ab5dc1228a80776706 (patch) | |
| tree | e1f6e91bf8da63e8ef98e11ab8eb6a6972a284f2 /arm_compute/core/NEON/NESymm.h | |
| parent | 4df2cf3177129d10500d30056bf8404418f703d6 (diff) | |
| download | ComputeLibrary-0d008f77b0085619c446d0ab5dc1228a80776706.tar.gz | |
COMPMID-3281: Implement QSYMM16 Layer Normalization for NEON QLSTM
- Reference kernel is modified to use the same algorithm as NEON kernel.
- NEON kernel is implemented.
- Tests for validation and run are added.
Change-Id: I3533bc2bd12c6e9cc75d837ecf193f74ceddf796
Signed-off-by: Sang-Hoon Park <sang-hoon.park@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/2948
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Diffstat (limited to 'arm_compute/core/NEON/NESymm.h')
| -rw-r--r-- | arm_compute/core/NEON/NESymm.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/arm_compute/core/NEON/NESymm.h b/arm_compute/core/NEON/NESymm.h index 0cc2a963cf..d6c5a7073a 100644 --- a/arm_compute/core/NEON/NESymm.h +++ b/arm_compute/core/NEON/NESymm.h @@ -239,7 +239,7 @@ inline qsymm16x8x2_t vquantize_qsymm16(const float32x4x4_t &qv, const UniformQua * * @return A neon vector holding the multiplied value */ -inline int32x4x2_t multiply_by_quantized_multipler_2row(int32x4x2_t input, int32_t qmul, int32_t shift) +inline int32x4x2_t multiply_by_quantized_multiplier_2row(int32x4x2_t input, int32_t qmul, int32_t shift) { const auto left_shift = shift > 0 ? shift : 0; const auto right_shift = shift > 0 ? 0 : -shift; |