CMSIS-NN  
CMSIS NN Software Library
 
Loading...
Searching...
No Matches
Data Structures | Enumerations
Structure Types
Public

Enums and Data Structures used in public API. More...

Data Structures

struct  cmsis_nn_tile
 
struct  cmsis_nn_context
 
struct  cmsis_nn_bias_data
 
struct  cmsis_nn_dims
 
struct  cmsis_nn_lstm_dims
 
struct  cmsis_nn_per_channel_quant_params
 
struct  cmsis_nn_per_tensor_quant_params
 
struct  cmsis_nn_quant_params
 
struct  cmsis_nn_activation
 
struct  cmsis_nn_conv_params
 
struct  cmsis_nn_transpose_conv_params
 
struct  cmsis_nn_dw_conv_params
 
struct  cmsis_nn_pool_params
 
struct  cmsis_nn_fc_params
 
struct  cmsis_nn_bmm_params
 
struct  cmsis_nn_svdf_params
 
struct  cmsis_nn_softmax_lut_s16
 
struct  cmsis_nn_scaling
 
struct  cmsis_nn_lstm_gate
 
struct  cmsis_nn_lstm_params
 
struct  cmsis_nn_lstm_context
 

Enumerations

enum  arm_nn_activation_type {
  ARM_SIGMOID = 0 ,
  ARM_TANH = 1
}
 
enum  arm_cmsis_nn_status {
  ARM_CMSIS_NN_SUCCESS = 0 ,
  ARM_CMSIS_NN_ARG_ERROR = -1 ,
  ARM_CMSIS_NN_NO_IMPL_ERROR = -2 ,
  ARM_CMSIS_NN_FAILURE = -3
}
 

Description

Enums and Data Structures used in public API.

Data Structure Documentation

◆ cmsis_nn_tile

struct cmsis_nn_tile

CMSIS-NN object to contain the width and height of a tile

Data Fields
int32_t h

Height

int32_t w

Width

◆ cmsis_nn_context

struct cmsis_nn_context

CMSIS-NN object used for the function context.

Data Fields
void * buf

Pointer to a buffer needed for the optimization

int32_t size

Buffer size

◆ cmsis_nn_bias_data

struct cmsis_nn_bias_data

CMSIS-NN object used to hold bias data for int16 variants.

Data Fields
const void * data

Pointer to bias data

const bool is_int32_bias

Indicate type of bias data. True means int32 else int64

◆ cmsis_nn_dims

struct cmsis_nn_dims

CMSIS-NN object to contain the dimensions of the tensors

Data Fields
int32_t c

Input channels

int32_t h

Height

int32_t n

Generic dimension to contain either the batch size or output channels. Please refer to the function documentation for more information

int32_t w

Width

◆ cmsis_nn_lstm_dims

struct cmsis_nn_lstm_dims

CMSIS-NN object to contain LSTM specific input parameters related to dimensions

Data Fields
int32_t max_time
int32_t num_batches
int32_t num_inputs
int32_t num_outputs

◆ cmsis_nn_per_channel_quant_params

struct cmsis_nn_per_channel_quant_params

CMSIS-NN object for the per-channel quantization parameters

Data Fields
int32_t * multiplier

Multiplier values

int32_t * shift

Shift values

◆ cmsis_nn_per_tensor_quant_params

struct cmsis_nn_per_tensor_quant_params

CMSIS-NN object for the per-tensor quantization parameters

Data Fields
int32_t multiplier

Multiplier value

int32_t shift

Shift value

◆ cmsis_nn_quant_params

struct cmsis_nn_quant_params

CMSIS-NN object for quantization parameters. This struct supports both per-tensor and per-channels requantization and is recommended for new operators.

Data Fields
int32_t is_per_channel
int32_t * multiplier

Multiplier values

int32_t * shift

Shift values

◆ cmsis_nn_activation

struct cmsis_nn_activation

CMSIS-NN object for the quantized Relu activation

Data Fields
int32_t max

Max value used to clamp the result

int32_t min

Min value used to clamp the result

◆ cmsis_nn_conv_params

struct cmsis_nn_conv_params

CMSIS-NN object for the convolution layer parameters

Data Fields
cmsis_nn_activation activation
cmsis_nn_tile dilation
int32_t input_offset

The negative of the zero value for the input tensor

int32_t output_offset

The negative of the zero value for the output tensor

cmsis_nn_tile padding
cmsis_nn_tile stride

◆ cmsis_nn_transpose_conv_params

struct cmsis_nn_transpose_conv_params

CMSIS-NN object for the transpose convolution layer parameters

Data Fields
cmsis_nn_activation activation
cmsis_nn_tile dilation
int32_t input_offset

The negative of the zero value for the input tensor

int32_t output_offset

The negative of the zero value for the output tensor

cmsis_nn_tile padding
cmsis_nn_tile padding_offsets
cmsis_nn_tile stride

◆ cmsis_nn_dw_conv_params

struct cmsis_nn_dw_conv_params

CMSIS-NN object for the depthwise convolution layer parameters

Data Fields
cmsis_nn_activation activation
int32_t ch_mult

Channel Multiplier. ch_mult * in_ch = out_ch

cmsis_nn_tile dilation
int32_t input_offset

The negative of the zero value for the input tensor

int32_t output_offset

The negative of the zero value for the output tensor

cmsis_nn_tile padding
cmsis_nn_tile stride

◆ cmsis_nn_pool_params

struct cmsis_nn_pool_params

CMSIS-NN object for pooling layer parameters

Data Fields
cmsis_nn_activation activation
cmsis_nn_tile padding
cmsis_nn_tile stride

◆ cmsis_nn_fc_params

struct cmsis_nn_fc_params

CMSIS-NN object for Fully Connected layer parameters

Data Fields
cmsis_nn_activation activation
int32_t filter_offset

The negative of the zero value for the filter tensor

int32_t input_offset

The negative of the zero value for the input tensor

int32_t output_offset

The negative of the zero value for the output tensor

◆ cmsis_nn_bmm_params

struct cmsis_nn_bmm_params

CMSIS-NN object for Batch Matmul layer parameters

Data Fields
const bool adj_x
const bool adj_y
cmsis_nn_fc_params fc_params

◆ cmsis_nn_svdf_params

struct cmsis_nn_svdf_params

CMSIS-NN object for SVDF layer parameters

Data Fields
cmsis_nn_activation input_activation
int32_t input_offset

The negative of the zero value for the input tensor

cmsis_nn_activation output_activation
int32_t output_offset

The negative of the zero value for the output tensor

int32_t rank

◆ cmsis_nn_softmax_lut_s16

struct cmsis_nn_softmax_lut_s16

CMSIS-NN object for Softmax s16 layer parameters

Data Fields
const int16_t * exp_lut
const int16_t * one_by_one_lut

◆ cmsis_nn_scaling

struct cmsis_nn_scaling

CMSIS-NN object for quantization parameters

Data Fields
int32_t multiplier

Multiplier value

int32_t shift

Shift value

◆ cmsis_nn_lstm_gate

struct cmsis_nn_lstm_gate

CMSIS-NN object for LSTM gate parameters

Data Fields
arm_nn_activation_type activation_type
const void * bias
const void * hidden_effective_bias

Precomputed kernel_sum * lhs_offset

int32_t hidden_multiplier
int32_t hidden_shift
const void * hidden_weights
const void * input_effective_bias

Bias added with precomputed kernel_sum * lhs_offset

int32_t input_multiplier
int32_t input_shift
const void * input_weights

◆ cmsis_nn_lstm_params

struct cmsis_nn_lstm_params

CMSIS-NN object for LSTM parameters

Data Fields
int32_t batch_size
int32_t cell_clip

Min/max value of cell output

cmsis_nn_lstm_gate cell_gate
int32_t cell_scale_power
cmsis_nn_lstm_gate forget_gate
int32_t forget_to_cell_multiplier
int32_t forget_to_cell_shift
int32_t hidden_size

Size of output from the LSTM cell, used as output and recursively into the next time step

cmsis_nn_lstm_gate input_gate
int32_t input_offset
int32_t input_size

Size of new data input into the LSTM cell

int32_t input_to_cell_multiplier
int32_t input_to_cell_shift
cmsis_nn_lstm_gate output_gate
int32_t output_multiplier
int32_t output_offset
int32_t output_shift
int32_t time_major

0 if first dimension is batch, else first dimension is time

int32_t time_steps

◆ cmsis_nn_lstm_context

struct cmsis_nn_lstm_context

CMSIS-NN object for LSTM scratch buffers

Data Fields
void * cell_state
void * temp1
void * temp2

Enumeration Type Documentation

◆ arm_cmsis_nn_status

enum arm_cmsis_nn_status

Function return codes

Enumerator
ARM_CMSIS_NN_SUCCESS 

No error

ARM_CMSIS_NN_ARG_ERROR 

One or more arguments are incorrect

ARM_CMSIS_NN_NO_IMPL_ERROR 

No implementation available

ARM_CMSIS_NN_FAILURE 

Logical error

◆ arm_nn_activation_type

enum arm_nn_activation_type

Enum for specifying activation function types

Enumerator
ARM_SIGMOID 

Sigmoid activation function

ARM_TANH 

Tanh activation function