Softmax Functions

Functions
arm_cmsis_nn_status	arm_softmax_f32 (const float32_t *input, int32_t num_rows, int32_t row_size, float32_t *output)
	Softmax using the float-native API signature.
arm_cmsis_nn_status	arm_softmax_f16 (const float16_t *input, int32_t num_rows, int32_t row_size, float16_t *output)
	Softmax using the float-native API signature.
void	arm_nn_softmax_1x2_f16 (const float16_t in[2], float16_t out[2])
	Specialized softmax helper for a single float32 row of length 2.
void	arm_nn_softmax_1x2_f32 (const float32_t in[2], float32_t out[2])
	Specialized softmax helper for a single float32 row of length 2.
arm_cmsis_nn_status	arm_softmax_s16 (const int16_t *input, const int32_t num_rows, const int32_t row_size, const int32_t mult, const int32_t shift, const cmsis_nn_softmax_lut_s16 *softmax_params, int16_t *output)
	S16 softmax function.
void	arm_softmax_s8 (const int8_t *input, const int32_t num_rows, const int32_t row_size, const int32_t mult, const int32_t shift, const int32_t diff_min, int8_t *output)
	S8 softmax function.
void	arm_softmax_s8_s16 (const int8_t *input, const int32_t num_rows, const int32_t row_size, const int32_t mult, const int32_t shift, const int32_t diff_min, int16_t *output)
	S8 to s16 softmax function.
void	arm_softmax_u8 (const uint8_t *input, const int32_t num_rows, const int32_t row_size, const int32_t mult, const int32_t shift, const int32_t diff_min, uint8_t *output)
	U8 softmax function.

Description

Function Documentation

arm_nn_softmax_1x2_f16()

void arm_nn_softmax_1x2_f16	(	const float16_t	in[2],
		float16_t	out[2]
	)

Specialized softmax helper for a single float32 row of length 2.

arm_nn_softmax_1x2_f32()

void arm_nn_softmax_1x2_f32	(	const float32_t	in[2],
		float32_t	out[2]
	)

Specialized softmax helper for a single float32 row of length 2.

arm_softmax_f16()

arm_cmsis_nn_status arm_softmax_f16	(	const float16_t *	input,
		int32_t	num_rows,
		int32_t	row_size,
		float16_t *	output
	)

Softmax using the float-native API signature.

Parameters

[in]	input	Pointer to the input matrix stored as num_rows rows of row_size values.
[in]	num_rows	Number of rows in the input matrix.
[in]	row_size	Number of columns per row.
[out]	output	Pointer to the output matrix.

Returns

ARM_CMSIS_NN_SUCCESS on success or ARM_CMSIS_NN_ARG_ERROR on invalid arguments.

arm_softmax_f32()

arm_cmsis_nn_status arm_softmax_f32	(	const float32_t *	input,
		int32_t	num_rows,
		int32_t	row_size,
		float32_t *	output
	)

Softmax using the float-native API signature.

Parameters

[in]	input	Pointer to the input matrix stored as num_rows rows of row_size values.
[in]	num_rows	Number of rows in the input matrix.
[in]	row_size	Number of columns per row.
[out]	output	Pointer to the output matrix.

Returns

ARM_CMSIS_NN_SUCCESS on success or ARM_CMSIS_NN_ARG_ERROR on invalid arguments.

arm_softmax_s16()

arm_cmsis_nn_status arm_softmax_s16	(	const int16_t *	input,
		const int32_t	num_rows,
		const int32_t	row_size,
		const int32_t	mult,
		const int32_t	shift,
		const cmsis_nn_softmax_lut_s16 *	softmax_params,
		int16_t *	output
	)

S16 softmax function.

Parameters

[in]	input	Pointer to the input tensor
[in]	num_rows	Number of rows in the input tensor
[in]	row_size	Number of elements in each input row
[in]	mult	Input quantization multiplier
[in]	shift	Input quantization shift within the range [0, 31]
[in]	softmax_params	Softmax s16 layer parameters with two pointers to LUTs speficied below. For indexing the high 9 bits are used and 7 remaining for interpolation. That means 512 entries for the 9-bit indexing and 1 extra for interpolation, i.e. 513 values for each LUT. Lookup table for exp(x), where x uniform distributed between [-10.0 , 0.0] Lookup table for 1 / (1 + x), where x uniform distributed between [0.0 , 1.0]
[out]	output	Pointer to the output tensor

Returns

The function returns ARM_CMSIS_NN_ARG_ERROR Argument error check failed ARM_CMSIS_NN_SUCCESS - Successful operation

Note

Supported framework: TensorFlow Lite micro (bit-accurate)

arm_softmax_s8()

void arm_softmax_s8	(	const int8_t *	input,
		const int32_t	num_rows,
		const int32_t	row_size,
		const int32_t	mult,
		const int32_t	shift,
		const int32_t	diff_min,
		int8_t *	output
	)

S8 softmax function.

Parameters

[in]	input	Pointer to the input tensor
[in]	num_rows	Number of rows in the input tensor
[in]	row_size	Number of elements in each input row
[in]	mult	Input quantization multiplier
[in]	shift	Input quantization shift within the range [0, 31]
[in]	diff_min	Minimum difference with max in row. Used to check if the quantized exponential operation can be performed
[out]	output	Pointer to the output tensor

Note

Supported framework: TensorFlow Lite micro (bit-accurate)

arm_softmax_s8_s16()

void arm_softmax_s8_s16	(	const int8_t *	input,
		const int32_t	num_rows,
		const int32_t	row_size,
		const int32_t	mult,
		const int32_t	shift,
		const int32_t	diff_min,
		int16_t *	output
	)

S8 to s16 softmax function.

Parameters

[in]	input	Pointer to the input tensor
[in]	num_rows	Number of rows in the input tensor
[in]	row_size	Number of elements in each input row
[in]	mult	Input quantization multiplier
[in]	shift	Input quantization shift within the range [0, 31]
[in]	diff_min	Minimum difference with max in row. Used to check if the quantized exponential operation can be performed
[out]	output	Pointer to the output tensor

Note

Supported framework: TensorFlow Lite micro (bit-accurate)

arm_softmax_u8()

void arm_softmax_u8	(	const uint8_t *	input,
		const int32_t	num_rows,
		const int32_t	row_size,
		const int32_t	mult,
		const int32_t	shift,
		const int32_t	diff_min,
		uint8_t *	output
	)

U8 softmax function.

Parameters

[in]	input	Pointer to the input tensor
[in]	num_rows	Number of rows in the input tensor
[in]	row_size	Number of elements in each input row
[in]	mult	Input quantization multiplier
[in]	shift	Input quantization shift within the range [0, 31]
[in]	diff_min	Minimum difference with max in row. Used to check if the quantized exponential operation can be performed
[out]	output	Pointer to the output tensor

Note

Supported framework: TensorFlow Lite micro (bit-accurate)