InnerProductLayer functions

Functions to acceleratе InnerProductLayer in Synet Framework. More...

Functions
SIMD_API void *	SimdSynetInnerProduct32fInit (size_t batch, size_t input, size_t output, SimdBool transpose, SimdConvolutionActivationType activation)
	Initilizes FP32 inner product algorithm. More...
SIMD_API size_t	SimdSynetInnerProduct32fInternalBufferSize (const void *context)
	Gets size of internal buffer used inside FP32 inner product algorithm. More...
SIMD_API void	SimdSynetInnerProduct32fSetParams (void *context, const float *weight, SimdBool *internal, const float *bias, const float *params)
	Sets weights, beases and parameters of activation function required for FP32 inner product algorithm. More...
SIMD_API void	SimdSynetInnerProduct32fForward (void *context, const float *src, float *dst)
	Performs forward propagation of FP32 inner product algorithm. More...
SIMD_API void	SimdSynetInnerProductLayerForward (const float *src, const float *weight, const float *bias, size_t count, size_t size, float *dst)
	This function is used for forward propagation of InnerProductLayer. More...
SIMD_API void	SimdSynetInnerProduct8i (size_t M, size_t N, size_t K, const uint8_t *src, const int8_t *weight, int32_t *dst, SimdSynetCompatibilityType compatibility)
	This function is used for INT8 forward propagation of InnerProductLayer. More...

Detailed Description

Functions to acceleratе InnerProductLayer in Synet Framework.

Function Documentation

SimdSynetInnerProduct32fInit()

void * SimdSynetInnerProduct32fInit	(	size_t	batch,
		size_t	input,
		size_t	output,
		SimdBool	transpose,
		SimdConvolutionActivationType	activation
	)

Initilizes FP32 inner product algorithm.

Parameters

[in]	batch	- a batch size.
[in]	input	- a input vector size.
[in]	output	- a output vector size.
[in]	transpose	- a flag of transposing of weight matrix.
[in]	activation	- an activation function type used after inner product.

Returns

a pointer to FP32 inner product context. On error it returns NULL. It must be released with using of function SimdRelease. This pointer is used in functions SimdSynetInnerProduct32fInternalBufferSize, :SimdSynetInnerProduct32fSetParams and SimdSynetInnerProduct32fForward.

SimdSynetInnerProduct32fInternalBufferSize()

size_t SimdSynetInnerProduct32fInternalBufferSize

(

const void *

context

)

Gets size of internal buffer used inside FP32 inner product algorithm.

Parameters

[in]

context

- a pointer to FP32 inner product context. It must be created by function SimdSynetInnerProduct32fInit and released by function SimdRelease.

Returns

size of internal buffer used inside FP32 inner product algorithm.

SimdSynetInnerProduct32fSetParams()

void SimdSynetInnerProduct32fSetParams	(	void *	context,
		const float *	weight,
		SimdBool *	internal,
		const float *	bias,
		const float *	params
	)

Sets weights, beases and parameters of activation function required for FP32 inner product algorithm.

Parameters

[in,out]	context	- a pointer to FP32 inner product context. It must be created by function SimdSynetInnerProduct32fInit and released by function SimdRelease.
[in]	weight	- a pointer to inner product weights.
[out]	internal	- a flag signalized that weight is stored in the internal buffer. Can be NULL.
[in]	bias	- a pointer to bias. Can be NULL.
[in]	params	- a pointer to parameters of activation functions (see SimdConvolutionActivationType). Can be NULL.

SimdSynetInnerProduct32fForward()

void SimdSynetInnerProduct32fForward	(	void *	context,
		const float *	src,
		float *	dst
	)

Performs forward propagation of FP32 inner product algorithm.

Parameters

[in]	context	- a pointer to FP32 inner product context. It must be created by function SimdSynetInnerProduct32fInit and released by function SimdRelease.
[in]	src	- a pointer to input tensor.
[out]	dst	- a pointer to output tensor.

SimdSynetInnerProductLayerForward()

void SimdSynetInnerProductLayerForward	(	const float *	src,
		const float *	weight,
		const float *	bias,
		size_t	count,
		size_t	size,
		float *	dst
	)

This function is used for forward propagation of InnerProductLayer.

Algorithm's details:

for(i = 0; i < count; ++i)
{
    dst[i] = (bias ? bias[i] : 0);
    for(j = 0; j < size; ++j)
       dst[i] += src[j]*weight[i*size + j];
}

Note

This function is used in Synet Framework.

Parameters

[in]	src	- a pointer to the input 32-bit float array. The size of the array must be equal to size.
[in]	weight	- a pointer to the 32-bit float array with weight coefficients. The size of the array must be equal to count*size.
[in]	bias	- a pointer to the 32-bit float array with bias coefficients. The size of the array must be equal to count. Can be NULL.
[in]	count	- a size of output array.
[in]	size	- a size of input array.
[out]	dst	- a pointer to the output 32-bit float array. The size of the array must be equal to count.

SimdSynetInnerProduct8i()

void SimdSynetInnerProduct8i	(	size_t	M,
		size_t	N,
		size_t	K,
		const uint8_t *	src,
		const int8_t *	weight,
		int32_t *	dst,
		SimdSynetCompatibilityType	compatibility
	)

This function is used for INT8 forward propagation of InnerProductLayer.

Algorithm's details:

for (i = 0; i < M; ++i)
{
    for (j = 0; j < N; ++j)
    {
        sum = 0;
        for (k = 0; k < K; ++k)
            sum += src[i * K + k] * weight[j * K + k];
        dst[i*N + j] = sum;
    }
}

Note

This function is used in Synet Framework.

Parameters

[in]	M	- a batch size.
[in]	N	- an output size.
[in]	K	- an input size.
[in]	src	- a pointer to the input 8-bit unsigned integer array. The size of the array must be equal to M*K.
[in]	weight	- a pointer to the 8-bit signed integer array with weight. The size of the array must be equal to N*K.
[out]	dst	- a pointer to the output 32-bit integer array. The size of the array must be equal to M*N.
[in]	compatibility	- a flags of calculation compatibility.