activation_kernels.cu File Reference

Functions
__global__ void	activate_array_gelu_kernel (float *x, int n)
__global__ void	activate_array_hard_mish_kernel (float *x, int n, float *activation_input, float *output_gpu)
void	activate_array_hard_mish_ongpu (float *x, int n, float *activation_input_gpu, float *output_gpu)
__global__ void	activate_array_hardtan_kernel (float *x, int n)
__global__ void	activate_array_kernel (float *x, int n, ACTIVATION a)
__global__ void	activate_array_leaky_kernel (float *x, int n)
__global__ void	activate_array_logistic_kernel (float *x, int n)
__global__ void	activate_array_mish_kernel (float *x, int n, float *activation_input, float *output_gpu)
void	activate_array_mish_ongpu (float *x, int n, float *activation_input_gpu, float *output_gpu)
__global__ void	activate_array_normalize_channels_kernel (float *x, int size, int batch, int channels, int wh_step, float *output_gpu)
void	activate_array_normalize_channels_ongpu (float *x, int n, int batch, int channels, int wh_step, float *output_gpu)
__global__ void	activate_array_normalize_channels_softmax_kernel (float *x, int size, int batch, int channels, int wh_step, float *output_gpu, int use_max_val)
void	activate_array_normalize_channels_softmax_ongpu (float *x, int n, int batch, int channels, int wh_step, float *output_gpu, int use_max_val)
void	activate_array_ongpu (float *x, int n, ACTIVATION a)
__global__ void	activate_array_relu6_kernel (float *x, int n)
__global__ void	activate_array_relu_kernel (float *x, int n)
__global__ void	activate_array_selu_kernel (float *x, int n)
__global__ void	activate_array_swish_kernel (float *x, int n, float *output_sigmoid_gpu, float *output_gpu)
void	activate_array_swish_ongpu (float *x, int n, float *output_sigmoid_gpu, float *output_gpu)
__global__ void	activate_array_tanh_kernel (float *x, int n)
__device__ float	activate_kernel (float x, ACTIVATION a)
void	binary_activate_array_gpu (float *x, int n, int size, BINARY_ACTIVATION a, float *y)
__global__ void	binary_activate_array_kernel (float *x, int n, int s, BINARY_ACTIVATION a, float *y)
void	binary_gradient_array_gpu (float *x, float *dx, int n, int size, BINARY_ACTIVATION a, float *y)
__global__ void	binary_gradient_array_kernel (float *x, float *dy, int n, int s, BINARY_ACTIVATION a, float *dx)
__device__ float	device_max (const float a, const float b)
__device__ float	device_min (const float a, const float b)
__device__ float	elu_activate_kernel (float x)
__device__ float	elu_gradient_kernel (float x)
__device__ float	gelu_activate_kernel (float x)
__device__ float	gelu_gradient_kernel (float x)
__global__ void	gradient_array_gelu_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_hard_mish_kernel (int n, float *activation_input_gpu, float *delta)
void	gradient_array_hard_mish_ongpu (int n, float *activation_input_gpu, float *delta)
__global__ void	gradient_array_hardtan_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_kernel (float *x, int n, ACTIVATION a, float *delta)
__global__ void	gradient_array_leaky_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_logistic_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_mish_kernel (int n, float *activation_input_gpu, float *delta)
void	gradient_array_mish_ongpu (int n, float *activation_input_gpu, float *delta)
__global__ void	gradient_array_normalize_channels_kernel (float *x, int size, int batch, int channels, int wh_step, float *delta_gpu)
void	gradient_array_normalize_channels_ongpu (float *output_gpu, int n, int batch, int channels, int wh_step, float *delta_gpu)
__global__ void	gradient_array_normalize_channels_softmax_kernel (float *x, int size, int batch, int channels, int wh_step, float *delta_gpu)
void	gradient_array_normalize_channels_softmax_ongpu (float *output_gpu, int n, int batch, int channels, int wh_step, float *delta_gpu)
void	gradient_array_ongpu (float *x, int n, ACTIVATION a, float *delta)
__global__ void	gradient_array_relu6_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_relu_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_revleaky_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_selu_kernel (float *x, int n, float *delta)
__global__ void	gradient_array_swish_kernel (float *x, int n, float *sigmoid_gpu, float *delta)
void	gradient_array_swish_ongpu (float *x, int n, float *sigmoid_gpu, float *delta)
__global__ void	gradient_array_tanh_kernel (float *x, int n, float *delta)
__device__ float	gradient_kernel (float x, ACTIVATION a)
__device__ float	hard_mish_yashas (float x)
__device__ float	hard_mish_yashas_grad (float x)
__device__ float	hardtan_activate_kernel (float x)
__device__ float	hardtan_gradient_kernel (float x)
__device__ float	leaky_activate_kernel (float x)
__device__ float	leaky_gradient_kernel (float x)
__device__ float	lhtan_activate_kernel (float x)
__device__ float	lhtan_gradient_kernel (float x)
__device__ float	linear_activate_kernel (float x)
__device__ float	linear_gradient_kernel (float x)
__device__ float	loggy_activate_kernel (float x)
__device__ float	loggy_gradient_kernel (float x)
__device__ float	logistic_activate_kernel (float x)
__device__ float	logistic_gradient_kernel (float x)
__device__ float	mish_njuffa (float x)
__device__ float	mish_yashas (float x)
__device__ float	mish_yashas2 (float x)
__device__ float	plse_activate_kernel (float x)
__device__ float	plse_gradient_kernel (float x)
__device__ float	ramp_activate_kernel (float x)
__device__ float	ramp_gradient_kernel (float x)
__device__ float	relie_activate_kernel (float x)
__device__ float	relie_gradient_kernel (float x)
__device__ float	relu6_activate_kernel (float x)
__device__ float	relu6_gradient_kernel (float x)
__device__ float	relu_activate_kernel (float x)
__device__ float	relu_gradient_kernel (float x)
__device__ float	sech_gpu (float x)
__device__ float	selu_activate_kernel (float x)
__device__ float	selu_gradient_kernel (float x)
__device__ float	softplus_kernel (float x, float threshold=20)
__device__ float	stair_activate_kernel (float x)
__device__ float	stair_gradient_kernel (float x)
__device__ float	tanh_activate_kernel (float x)
__device__ float	tanh_gradient_kernel (float x)

Function Documentation

activate_array_gelu_kernel()

__global__ void activate_array_gelu_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_hard_mish_kernel()

__global__ void activate_array_hard_mish_kernel	(	float *	x,
		int	n,
		float *	activation_input,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_hard_mish_ongpu()

void activate_array_hard_mish_ongpu	(	float *	x,
		int	n,
		float *	activation_input_gpu,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_hardtan_kernel()

__global__ void activate_array_hardtan_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_kernel()

__global__ void activate_array_kernel	(	float *	x,
		int	n,
		ACTIVATION	a
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_leaky_kernel()

__global__ void activate_array_leaky_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_logistic_kernel()

__global__ void activate_array_logistic_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_mish_kernel()

__global__ void activate_array_mish_kernel	(	float *	x,
		int	n,
		float *	activation_input,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_mish_ongpu()

void activate_array_mish_ongpu	(	float *	x,
		int	n,
		float *	activation_input_gpu,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_normalize_channels_kernel()

__global__ void activate_array_normalize_channels_kernel	(	float *	x,
		int	size,
		int	batch,
		int	channels,
		int	wh_step,
		float *	output_gpu
	)

Here is the caller graph for this function:

activate_array_normalize_channels_ongpu()

void activate_array_normalize_channels_ongpu	(	float *	x,
		int	n,
		int	batch,
		int	channels,
		int	wh_step,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_normalize_channels_softmax_kernel()

__global__ void activate_array_normalize_channels_softmax_kernel	(	float *	x,
		int	size,
		int	batch,
		int	channels,
		int	wh_step,
		float *	output_gpu,
		int	use_max_val
	)

Here is the caller graph for this function:

activate_array_normalize_channels_softmax_ongpu()

void activate_array_normalize_channels_softmax_ongpu	(	float *	x,
		int	n,
		int	batch,
		int	channels,
		int	wh_step,
		float *	output_gpu,
		int	use_max_val
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_ongpu()

void activate_array_ongpu	(	float *	x,
		int	n,
		ACTIVATION	a
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_relu6_kernel()

__global__ void activate_array_relu6_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_relu_kernel()

__global__ void activate_array_relu_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_selu_kernel()

__global__ void activate_array_selu_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_swish_kernel()

__global__ void activate_array_swish_kernel	(	float *	x,
		int	n,
		float *	output_sigmoid_gpu,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_swish_ongpu()

void activate_array_swish_ongpu	(	float *	x,
		int	n,
		float *	output_sigmoid_gpu,
		float *	output_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_array_tanh_kernel()

__global__ void activate_array_tanh_kernel	(	float *	x,
		int	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

activate_kernel()

__device__ float activate_kernel	(	float	x,
		ACTIVATION	a
	)

Here is the call graph for this function:

Here is the caller graph for this function:

binary_activate_array_gpu()

void binary_activate_array_gpu	(	float *	x,
		int	n,
		int	size,
		BINARY_ACTIVATION	a,
		float *	y
	)

Here is the call graph for this function:

binary_activate_array_kernel()

__global__ void binary_activate_array_kernel	(	float *	x,
		int	n,
		int	s,
		BINARY_ACTIVATION	a,
		float *	y
	)

Here is the caller graph for this function:

binary_gradient_array_gpu()

void binary_gradient_array_gpu	(	float *	x,
		float *	dx,
		int	n,
		int	size,
		BINARY_ACTIVATION	a,
		float *	y
	)

Here is the call graph for this function:

binary_gradient_array_kernel()

__global__ void binary_gradient_array_kernel	(	float *	x,
		float *	dy,
		int	n,
		int	s,
		BINARY_ACTIVATION	a,
		float *	dx
	)

Here is the caller graph for this function:

device_max()

__device__ float device_max	(	const float	a,
		const float	b
	)

Here is the caller graph for this function:

device_min()

__device__ float device_min	(	const float	a,
		const float	b
	)

Here is the caller graph for this function:

elu_activate_kernel()

__device__ float elu_activate_kernel

(

float

x

)

Here is the caller graph for this function:

elu_gradient_kernel()

__device__ float elu_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

gelu_activate_kernel()

__device__ float gelu_activate_kernel

(

float

x

)

Here is the caller graph for this function:

gelu_gradient_kernel()

__device__ float gelu_gradient_kernel

(

float

x

)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_gelu_kernel()

__global__ void gradient_array_gelu_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_hard_mish_kernel()

__global__ void gradient_array_hard_mish_kernel	(	int	n,
		float *	activation_input_gpu,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_hard_mish_ongpu()

void gradient_array_hard_mish_ongpu	(	int	n,
		float *	activation_input_gpu,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_hardtan_kernel()

__global__ void gradient_array_hardtan_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_kernel()

__global__ void gradient_array_kernel	(	float *	x,
		int	n,
		ACTIVATION	a,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_leaky_kernel()

__global__ void gradient_array_leaky_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_logistic_kernel()

__global__ void gradient_array_logistic_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_mish_kernel()

__global__ void gradient_array_mish_kernel	(	int	n,
		float *	activation_input_gpu,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_mish_ongpu()

void gradient_array_mish_ongpu	(	int	n,
		float *	activation_input_gpu,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_normalize_channels_kernel()

__global__ void gradient_array_normalize_channels_kernel	(	float *	x,
		int	size,
		int	batch,
		int	channels,
		int	wh_step,
		float *	delta_gpu
	)

Here is the caller graph for this function:

gradient_array_normalize_channels_ongpu()

void gradient_array_normalize_channels_ongpu	(	float *	output_gpu,
		int	n,
		int	batch,
		int	channels,
		int	wh_step,
		float *	delta_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_normalize_channels_softmax_kernel()

__global__ void gradient_array_normalize_channels_softmax_kernel	(	float *	x,
		int	size,
		int	batch,
		int	channels,
		int	wh_step,
		float *	delta_gpu
	)

Here is the caller graph for this function:

gradient_array_normalize_channels_softmax_ongpu()

void gradient_array_normalize_channels_softmax_ongpu	(	float *	output_gpu,
		int	n,
		int	batch,
		int	channels,
		int	wh_step,
		float *	delta_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_ongpu()

void gradient_array_ongpu	(	float *	x,
		int	n,
		ACTIVATION	a,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_relu6_kernel()

__global__ void gradient_array_relu6_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_relu_kernel()

__global__ void gradient_array_relu_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_revleaky_kernel()

__global__ void gradient_array_revleaky_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_selu_kernel()

__global__ void gradient_array_selu_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_swish_kernel()

__global__ void gradient_array_swish_kernel	(	float *	x,
		int	n,
		float *	sigmoid_gpu,
		float *	delta
	)

Here is the caller graph for this function:

gradient_array_swish_ongpu()

void gradient_array_swish_ongpu	(	float *	x,
		int	n,
		float *	sigmoid_gpu,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_array_tanh_kernel()

__global__ void gradient_array_tanh_kernel	(	float *	x,
		int	n,
		float *	delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

gradient_kernel()

__device__ float gradient_kernel	(	float	x,
		ACTIVATION	a
	)

Here is the call graph for this function:

Here is the caller graph for this function:

hard_mish_yashas()

__device__ float hard_mish_yashas

(

float

x

)

Here is the caller graph for this function:

hard_mish_yashas_grad()

__device__ float hard_mish_yashas_grad

(

float

x

)

Here is the caller graph for this function:

hardtan_activate_kernel()

__device__ float hardtan_activate_kernel

(

float

x

)

Here is the caller graph for this function:

hardtan_gradient_kernel()

__device__ float hardtan_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

leaky_activate_kernel()

__device__ float leaky_activate_kernel

(

float

x

)

Here is the caller graph for this function:

leaky_gradient_kernel()

__device__ float leaky_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

lhtan_activate_kernel()

__device__ float lhtan_activate_kernel

(

float

x

)

Here is the caller graph for this function:

lhtan_gradient_kernel()

__device__ float lhtan_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

linear_activate_kernel()

__device__ float linear_activate_kernel

(

float

x

)

Here is the caller graph for this function:

linear_gradient_kernel()

__device__ float linear_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

loggy_activate_kernel()

__device__ float loggy_activate_kernel

(

float

x

)

Here is the caller graph for this function:

loggy_gradient_kernel()

__device__ float loggy_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

logistic_activate_kernel()

__device__ float logistic_activate_kernel

(

float

x

)

Here is the caller graph for this function:

logistic_gradient_kernel()

__device__ float logistic_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

mish_njuffa()

__device__ float mish_njuffa

(

float

x

)

mish_yashas()

__device__ float mish_yashas

(

float

x

)

mish_yashas2()

__device__ float mish_yashas2

(

float

x

)

Here is the caller graph for this function:

plse_activate_kernel()

__device__ float plse_activate_kernel

(

float

x

)

Here is the caller graph for this function:

plse_gradient_kernel()

__device__ float plse_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

ramp_activate_kernel()

__device__ float ramp_activate_kernel

(

float

x

)

Here is the caller graph for this function:

ramp_gradient_kernel()

__device__ float ramp_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

relie_activate_kernel()

__device__ float relie_activate_kernel

(

float

x

)

Here is the caller graph for this function:

relie_gradient_kernel()

__device__ float relie_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

relu6_activate_kernel()

__device__ float relu6_activate_kernel

(

float

x

)

Here is the call graph for this function:

Here is the caller graph for this function:

relu6_gradient_kernel()

__device__ float relu6_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

relu_activate_kernel()

__device__ float relu_activate_kernel

(

float

x

)

Here is the caller graph for this function:

relu_gradient_kernel()

__device__ float relu_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

sech_gpu()

__device__ float sech_gpu

(

float

x

)

Here is the caller graph for this function:

selu_activate_kernel()

__device__ float selu_activate_kernel

(

float

x

)

Here is the caller graph for this function:

selu_gradient_kernel()

__device__ float selu_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

softplus_kernel()

__device__ float softplus_kernel	(	float	x,
		float	threshold = 20
	)

Here is the caller graph for this function:

stair_activate_kernel()

__device__ float stair_activate_kernel

(

float

x

)

Here is the caller graph for this function:

stair_gradient_kernel()

__device__ float stair_gradient_kernel

(

float

x

)

Here is the caller graph for this function:

tanh_activate_kernel()

__device__ float tanh_activate_kernel

(

float

x

)

Here is the caller graph for this function:

tanh_gradient_kernel()

__device__ float tanh_gradient_kernel

(

float

x

)

Here is the caller graph for this function: