diff options
author | Ryan OShea <ryan.oshea3@arm.com> | 2022-11-07 16:20:48 +0000 |
---|---|---|
committer | ryan.oshea3 <ryan.oshea3@arm.com> | 2022-11-16 15:22:50 +0000 |
commit | 31441595009182c985dacbedc70c41ee6664d070 (patch) | |
tree | 248a85295aeff4022c9b395fc97748b0a0aa6b35 /src/backends/neon/NeonWorkloadFactory.cpp | |
parent | bd18eab07a8f30492de1e462b1815189014cb8d5 (diff) | |
download | armnn-31441595009182c985dacbedc70c41ee6664d070.tar.gz |
IVGCVSW-7214 Disable BF16-Turbo-Mode and remove conversion layers
- Remove Bf16ToFp32 Conversion Layer
- Remove Fp32ToBf16 Conversion Layer
- Remove B16 Conversion tests
* Throw exception if m_ReduceFp32ToBf16 optimzer option is set to true
* Provide comments to enable fast math in order to use bf16
* Update docs to inform users to enable fast math for bf16
Execute Network Changes
* Require bf16_turbo_mode to also have fast_math_enabled set to true
- Remove setting m_ReduceFp32ToBf16 optimizer option
Signed-off-by: Ryan OShea <ryan.oshea3@arm.com>
Change-Id: Ibaa6da9d29c96a1ce32ff5196b0847fde9f04a1c
Diffstat (limited to 'src/backends/neon/NeonWorkloadFactory.cpp')
-rw-r--r-- | src/backends/neon/NeonWorkloadFactory.cpp | 26 |
1 files changed, 0 insertions, 26 deletions
diff --git a/src/backends/neon/NeonWorkloadFactory.cpp b/src/backends/neon/NeonWorkloadFactory.cpp index d5a7c684d3..dccd4a3a36 100644 --- a/src/backends/neon/NeonWorkloadFactory.cpp +++ b/src/backends/neon/NeonWorkloadFactory.cpp @@ -194,24 +194,12 @@ std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateWorkload(LayerType type, auto constantQueueDescriptor = PolymorphicDowncast<const ConstantQueueDescriptor*>(&descriptor); return std::make_unique<NeonConstantWorkload>(*constantQueueDescriptor, info); } - case LayerType::ConvertBf16ToFp32 : - { - auto convertBf16ToFp32QueueDescriptor - = PolymorphicDowncast<const ConvertBf16ToFp32QueueDescriptor*>(&descriptor); - return std::make_unique<NeonConvertBf16ToFp32Workload>(*convertBf16ToFp32QueueDescriptor, info); - } case LayerType::ConvertFp16ToFp32 : { auto convertFp16ToFp32QueueDescriptor = PolymorphicDowncast<const ConvertFp16ToFp32QueueDescriptor*>(&descriptor); return std::make_unique<NeonConvertFp16ToFp32Workload>(*convertFp16ToFp32QueueDescriptor, info); } - case LayerType::ConvertFp32ToBf16 : - { - auto convertFp32ToBf16QueueDescriptor - = PolymorphicDowncast<const ConvertFp32ToBf16QueueDescriptor*>(&descriptor); - return std::make_unique<NeonConvertFp32ToBf16Workload>(*convertFp32ToBf16QueueDescriptor, info); - } case LayerType::ConvertFp32ToFp16 : { auto convertFp32ToFp16QueueDescriptor @@ -655,13 +643,6 @@ std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateConstant(const ConstantQue return std::make_unique<NeonConstantWorkload>(descriptor, info); } -std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateConvertBf16ToFp32( - const ConvertBf16ToFp32QueueDescriptor& descriptor, - const WorkloadInfo& info) const -{ - return std::make_unique<NeonConvertBf16ToFp32Workload>(descriptor, info); -} - std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateConvertFp16ToFp32( const ConvertFp16ToFp32QueueDescriptor& descriptor, const WorkloadInfo& info) const @@ -669,13 +650,6 @@ std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateConvertFp16ToFp32( return std::make_unique<NeonConvertFp16ToFp32Workload>(descriptor, info); } -std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateConvertFp32ToBf16( - const ConvertFp32ToBf16QueueDescriptor& descriptor, - const WorkloadInfo& info) const -{ - return std::make_unique<NeonConvertFp32ToBf16Workload>(descriptor, info); -} - std::unique_ptr<IWorkload> NeonWorkloadFactory::CreateConvertFp32ToFp16( const ConvertFp32ToFp16QueueDescriptor& descriptor, const WorkloadInfo& info) const |