diff options
author | Mohammed Suhail Munshi <MohammedSuhail.Munshi@arm.com> | 2022-10-21 11:15:54 +0100 |
---|---|---|
committer | Suhail Munshi <MohammedSuhail.Munshi@arm.com> | 2022-11-15 11:29:21 +0000 |
commit | 97a609bb9b2ad22a2ebd54493cd1374f0992eb52 (patch) | |
tree | a04d04262f25b0f29965a2b2567f19e92f46f58a /tests/datasets/HOGMultiDetectionDataset.h | |
parent | b230a1f9bc53bf6b3bd74789e0085ff48336f348 (diff) | |
download | ComputeLibrary-97a609bb9b2ad22a2ebd54493cd1374f0992eb52.tar.gz |
Fix GemmLowp BatchMatMul Tests to use quantized Outputs
- Fix includes int8/uint8 quantized inputs
- Bias S32 value is limited to better allow detection of mismatches in gemmlowp kernel
Resolves: [COMPMID-5659]
Signed-off-by: Mohammed Suhail Munshi <MohammedSuhail.Munshi@arm.com>
Change-Id: Ie9cca430c6ab66253fe1d5252bd2c5396c7f38cf
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/8514
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'tests/datasets/HOGMultiDetectionDataset.h')
0 files changed, 0 insertions, 0 deletions