diff options
| author | Jonathan Deakin <jonathan.deakin@arm.com> | 2024-01-24 09:15:38 +0000 |
|---|---|---|
| committer | Radu Salavat <radu.salavat@arm.com> | 2024-04-15 13:52:31 +0000 |
| commit | a668f9f8a4eab405df0fe8dd58e7d9425bcf9640 (patch) | |
| tree | db16e6af9289897557a58755b88d2c337dcb8650 /docs | |
| parent | 34bdffb288d6367cb6dca652ebed60c450854039 (diff) | |
| download | ComputeLibrary-a668f9f8a4eab405df0fe8dd58e7d9425bcf9640.tar.gz | |
Add s8f32 kernels and dynamic QuantizationInfo
- Add support for QASYMM_SIGNED*QASYMM8_SIGNED->F32 in
CpuGemmLowpMatrixMultiplyCore
- Add s8f32 kernel using existing s8->s32 kernels with a new
DequantizeFloat OutputStage, the structure is similar to Requantize32
but the opposite way around.
- Add SME s8f32 kernels with integrated support for DequantizeFloat.
- Add scale to CpuGemmLowpOffsetContributionKernel.
- Add virtual dequantize scale to gemm_common, only implemented for
gemm_interleaved.
- Update year to 2024 in generate_build_files.
- Add dynamic flag to QuantizationInfo which signals to operators that
it can change after configuration
- Add support for dynamic quantization in NEGEMMLowpMatrixMultiplyCore
- Add dynamic quantization fixture by extending
GEMMLowpGenericMatrixMultiplyCoreValidationFixture
- Add GEMMLowpDequantizedMatrixMultiplyValidationFixture
- Store k (number of cols of A) rather than k_offset in the offset
contribution kernels so that we can recompute it when the other
offsets change
relates to: ONCPUML-1444 MLINFSW-439
Co-authored-by: Milos Puzovic <Milos.Puzovic@arm.com>
Co-authored-by: David Mansell <David.Mansell@arm.com>
Change-Id: I58a3acf2c09289a303e52eea6b336a696a5bc8da
Signed-off-by: Jonathan Deakin <jonathan.deakin@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/11022
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'docs')
| -rw-r--r-- | docs/user_guide/operator_list.dox | 3 | ||||
| -rw-r--r-- | docs/user_guide/release_version_and_change_log.dox | 3 |
2 files changed, 5 insertions, 1 deletions
diff --git a/docs/user_guide/operator_list.dox b/docs/user_guide/operator_list.dox index 36275e68bf..e7f1823f8b 100644 --- a/docs/user_guide/operator_list.dox +++ b/docs/user_guide/operator_list.dox @@ -1,5 +1,5 @@ /// -/// Copyright (c) 2021-2023,2024 Arm Limited. +/// Copyright (c) 2021-2024 Arm Limited. /// /// SPDX-License-Identifier: MIT /// @@ -1773,6 +1773,7 @@ where N = batches, C = channels, H = height, W = width, D = depth <tr><td>QASYMM8_SIGNED<td>QASYMM8_SIGNED<td>S32<td>S32 <tr><td>QASYMM8_SIGNED<td>QSYMM8_PER_CHANNEL<td>S32<td>S32 <tr><td>QASYMM8_SIGNED<td>QSYMM8<td>S32<td>S32 + <tr><td>QASYMM8_SIGNED<td>QASYMM8_SIGNED<td>F32<td>F32 </table> <tr> <td>CLGEMMLowpMatrixMultiplyCore diff --git a/docs/user_guide/release_version_and_change_log.dox b/docs/user_guide/release_version_and_change_log.dox index 9da4956c43..3737dbfc33 100644 --- a/docs/user_guide/release_version_and_change_log.dox +++ b/docs/user_guide/release_version_and_change_log.dox @@ -49,6 +49,9 @@ v24.04 Public major release - Performance optimizations: - Optimize @ref NESoftmaxLayer for axis != 0 by natively supporting higher axes up to axis 3. - Add support for in place accumulation to CPU GEMM kernels. + - Add low-precision Int8 * Int8 -> FP32 CPU GEMM which dequantizes after multiplication + - Add is_dynamic flag to QuantizationInfo to signal to operators that it may change after configuration + v24.02.1 Public patch release - Fix performance regression in fixed-format kernels |