SetBiasType

Function Usage

Sets the position, data format, data type, and transpose status of the bias matrix. These settings must be consistent with those in the kernel.

Prototype

1
int32_t SetBiasType(TPosition pos, CubeFormat type, DataType dataType)

Parameters

Table 1 Parameters

Parameter

Input/Output

Description

pos

Input

Position in the buffer where the bias matrix is located, which can be TPosition::GM or TPosition::VECOUT.

type

Input

Data format of the bias matrix, which can be CubeFormat::ND.

dataType

Input

Data type of the bias matrix, which can be DataType::DT_FLOAT, DataType::DT_FLOAT16, or DataType::DT_INT32.

The data type of Bias can be set to INT32 only when the data type of matrices A and B is INT8.

Returns

The value -1 indicates that the setting fails, and the value 0 indicates that the setting is successful.

Precautions

None

Example

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
auto ascendcPlatform = platform_ascendc::PlatformAscendC(context->GetPlatformInfo());
matmul_tiling::MatmulApiTiling tiling(ascendcPlatform); 
tiling.SetAType(matmul_tiling::TPosition::GM, matmul_tiling::CubeFormat::ND, matmul_tiling::DataType::DT_FLOAT16); 
tiling.SetBType(matmul_tiling::TPosition::GM, matmul_tiling::CubeFormat::ND, matmul_tiling::DataType::DT_FLOAT16);  
tiling.SetCType(matmul_tiling::TPosition::GM, matmul_tiling::CubeFormat::ND, matmul_tiling::DataType::DT_FLOAT);
tiling.SetBiasType(matmul_tiling::TPosition::GM, matmul_tiling::CubeFormat::ND, matmul_tiling::DataType::DT_FLOAT);  // Set the bias matrix.
tiling.SetShape(1024, 1024, 1024);   
tiling.SetOrgShape(1024, 1024, 1024);   
tiling.SetBias(true);   
tiling.SetBufferSpace(-1, -1, -1);
optiling::TCubeTiling tilingData;   
int ret = tiling.GetTiling(tilingData);