diff options
| author | Omar Al Khatib <omar.alkhatib@arm.com> | 2024-04-23 16:26:56 +0100 |
|---|---|---|
| committer | Omar Al Khatib <omar.alkhatib@arm.com> | 2024-05-08 12:16:10 +0000 |
| commit | c1575b2c12a4cee3a60c711fe6521025a814b159 (patch) | |
| tree | dd88487711f631a83989dec5c7e058d29cc076ec /src/cpu/kernels/softmax/list.h | |
| parent | 2fea13593a4753316ae488edf489cb4b00150153 (diff) | |
| download | ComputeLibrary-c1575b2c12a4cee3a60c711fe6521025a814b159.tar.gz | |
Add SME2 implementation of Softmax for QASYMM8 and QASYMM8_SIGNED.
Resolves: [COMPMID-6917]
Change-Id: Id8b96efd29f6c61dd43a371341c6e1fe087953e9
Signed-off-by: Omar Al Khatib <omar.alkhatib@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/11509
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/cpu/kernels/softmax/list.h')
| -rw-r--r-- | src/cpu/kernels/softmax/list.h | 41 |
1 files changed, 34 insertions, 7 deletions
diff --git a/src/cpu/kernels/softmax/list.h b/src/cpu/kernels/softmax/list.h index 1bb8ed50f0..7bbb265022 100644 --- a/src/cpu/kernels/softmax/list.h +++ b/src/cpu/kernels/softmax/list.h @@ -28,9 +28,10 @@ namespace arm_compute { namespace cpu { -#define DECLARE_SOFTMAX_KERNEL(func_name) \ - template <bool IS_LOG> \ - void func_name(const ITensor *in, void *const tmp, ITensor *out, const float beta, int axis, const Window &window) +#define DECLARE_SOFTMAX_KERNEL(func_name) \ + template <bool IS_LOG> \ + void func_name(const ITensor *in, void *const tmp, ITensor *out, const float beta, int axis, const Window &window, \ + const float *lut_ptr) DECLARE_SOFTMAX_KERNEL(neon_fp32_softmax); DECLARE_SOFTMAX_KERNEL(neon_fp16_softmax); @@ -39,11 +40,37 @@ DECLARE_SOFTMAX_KERNEL(neon_qasymm8_signed_softmax); #ifdef ARM_COMPUTE_ENABLE_SME2 -void sme2_fp32_softmax( - const ITensor *in, void *const tmp, ITensor *out, const float beta, int axis, const Window &window); +void sme2_fp32_softmax(const ITensor *in, + void *const tmp, + ITensor *out, + const float beta, + int axis, + const Window &window, + const float *lut_ptr); -void sme2_fp16_softmax( - const ITensor *in, void *const tmp, ITensor *out, const float beta, int axis, const Window &window); +void sme2_fp16_softmax(const ITensor *in, + void *const tmp, + ITensor *out, + const float beta, + int axis, + const Window &window, + const float *lut_ptr); + +void sme2_qasymm8_softmax_lut_512VL(const ITensor *in, + void *const tmp, + ITensor *out, + const float beta, + int axis, + const Window &window, + const float *lut_ptr); + +void sme2_qasymm8_signed_softmax_lut_512VL(const ITensor *in, + void *const tmp, + ITensor *out, + const float beta, + int axis, + const Window &window, + const float *lut_ptr); #endif // ARM_COMPUTE_ENABLE_SME2 |