convolutional_kernels.cu File Reference

Functions
void	assisted_activation2_gpu (float alpha, float *output, float *gt_gpu, float *a_avg_gpu, int size, int channels, int batches)
__global__ void	assisted_activation2_kernel (float alpha, float *output, float *gt_gpu, float *a_avg_gpu, int size, int channels, int batches)
void	assisted_activation_gpu (float alpha, float *output, float *gt_gpu, float *a_avg_gpu, int size, int channels, int batches)
__global__ void	assisted_activation_kernel (float alpha, float *output, float *gt_gpu, float *a_avg_gpu, int size, int channels, int batches)
void	assisted_excitation_forward_gpu (Darknet::Layer &l, Darknet::NetworkState state)
void	backward_convolutional_layer_gpu (Darknet::Layer &l, Darknet::NetworkState state)
void	binarize_gpu (float *x, int n, float *binary)
void	binarize_input_gpu (float *input, int n, int size, float *binary)
__global__ void	binarize_input_kernel (float *input, int n, int size, float *binary)
__global__ void	binarize_kernel (float *x, int n, float *binary)
void	binarize_weights_gpu (float *weights, int n, int size, float *binary)
__global__ void	binarize_weights_kernel (float *weights, int n, int size, float *binary)
__global__ void	binarize_weights_mean_kernel (float *weights, int n, int size, float *binary, float *mean_arr_gpu)
void	calc_avg_activation_gpu (float *src, float *dst, int size, int channels, int batches)
__global__ void	calc_avg_activation_kernel (float *src, float *dst, int size, int channels, int batches)
void	cuda_convert_f16_to_f32 (float *input_f16, size_t size, float *output_f32)
void	cuda_convert_f32_to_f16 (float *input_f32, size_t size, float *output_f16)
__global__ void	cuda_f16_to_f32 (half *input_f16, size_t size, float *output_f32)
__global__ void	cuda_f32_to_f16 (float *input_f32, size_t size, half *output_f16)
half *	cuda_make_f16_from_f32_array (float *src, size_t n)
void	fast_binarize_weights_gpu (float *weights, int n, int size, float *binary, float *mean_arr_gpu)
void	forward_convolutional_layer_gpu (Darknet::Layer &l, Darknet::NetworkState state)
void	pull_convolutional_layer (Darknet::Layer &l)
void	push_convolutional_layer (Darknet::Layer &l)
__global__ void	reduce_kernel (float *weights, int n, int size, float *mean_arr_gpu)
__global__ void	set_zero_kernel (float *src, int size)
void	update_convolutional_layer_gpu (Darknet::Layer &l, int batch, float learning_rate_init, float momentum, float decay, float loss_scale)
__inline__ __device__ float	warpAllReduceSum (float val)

Function Documentation

assisted_activation2_gpu()

void assisted_activation2_gpu	(	float	alpha,
		float *	output,
		float *	gt_gpu,
		float *	a_avg_gpu,
		int	size,
		int	channels,
		int	batches
	)

Here is the call graph for this function:

assisted_activation2_kernel()

__global__ void assisted_activation2_kernel	(	float	alpha,
		float *	output,
		float *	gt_gpu,
		float *	a_avg_gpu,
		int	size,
		int	channels,
		int	batches
	)

Here is the caller graph for this function:

assisted_activation_gpu()

void assisted_activation_gpu	(	float	alpha,
		float *	output,
		float *	gt_gpu,
		float *	a_avg_gpu,
		int	size,
		int	channels,
		int	batches
	)

Here is the call graph for this function:

Here is the caller graph for this function:

assisted_activation_kernel()

__global__ void assisted_activation_kernel	(	float	alpha,
		float *	output,
		float *	gt_gpu,
		float *	a_avg_gpu,
		int	size,
		int	channels,
		int	batches
	)

Here is the caller graph for this function:

assisted_excitation_forward_gpu()

void assisted_excitation_forward_gpu	(	Darknet::Layer &	l,
		Darknet::NetworkState	state
	)

Here is the call graph for this function:

Here is the caller graph for this function:

backward_convolutional_layer_gpu()

void backward_convolutional_layer_gpu	(	Darknet::Layer &	l,
		Darknet::NetworkState	state
	)

Here is the call graph for this function:

Here is the caller graph for this function:

binarize_gpu()

void binarize_gpu	(	float *	x,
		int	n,
		float *	binary
	)

Here is the call graph for this function:

Here is the caller graph for this function:

binarize_input_gpu()

void binarize_input_gpu	(	float *	input,
		int	n,
		int	size,
		float *	binary
	)

Here is the call graph for this function:

binarize_input_kernel()

__global__ void binarize_input_kernel	(	float *	input,
		int	n,
		int	size,
		float *	binary
	)

Here is the caller graph for this function:

binarize_kernel()

__global__ void binarize_kernel	(	float *	x,
		int	n,
		float *	binary
	)

Here is the caller graph for this function:

binarize_weights_gpu()

void binarize_weights_gpu	(	float *	weights,
		int	n,
		int	size,
		float *	binary
	)

Here is the call graph for this function:

Here is the caller graph for this function:

binarize_weights_kernel()

__global__ void binarize_weights_kernel	(	float *	weights,
		int	n,
		int	size,
		float *	binary
	)

Here is the caller graph for this function:

binarize_weights_mean_kernel()

__global__ void binarize_weights_mean_kernel	(	float *	weights,
		int	n,
		int	size,
		float *	binary,
		float *	mean_arr_gpu
	)

Here is the caller graph for this function:

calc_avg_activation_gpu()

void calc_avg_activation_gpu	(	float *	src,
		float *	dst,
		int	size,
		int	channels,
		int	batches
	)

Here is the call graph for this function:

Here is the caller graph for this function:

calc_avg_activation_kernel()

__global__ void calc_avg_activation_kernel	(	float *	src,
		float *	dst,
		int	size,
		int	channels,
		int	batches
	)

Here is the caller graph for this function:

cuda_convert_f16_to_f32()

void cuda_convert_f16_to_f32	(	float *	input_f16,
		size_t	size,
		float *	output_f32
	)

Here is the call graph for this function:

Here is the caller graph for this function:

cuda_convert_f32_to_f16()

void cuda_convert_f32_to_f16	(	float *	input_f32,
		size_t	size,
		float *	output_f16
	)

Here is the call graph for this function:

Here is the caller graph for this function:

cuda_f16_to_f32()

__global__ void cuda_f16_to_f32	(	half *	input_f16,
		size_t	size,
		float *	output_f32
	)

Here is the caller graph for this function:

cuda_f32_to_f16()

__global__ void cuda_f32_to_f16	(	float *	input_f32,
		size_t	size,
		half *	output_f16
	)

Here is the caller graph for this function:

cuda_make_f16_from_f32_array()

half * cuda_make_f16_from_f32_array	(	float *	src,
		size_t	n
	)

Here is the call graph for this function:

Here is the caller graph for this function:

fast_binarize_weights_gpu()

void fast_binarize_weights_gpu	(	float *	weights,
		int	n,
		int	size,
		float *	binary,
		float *	mean_arr_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

forward_convolutional_layer_gpu()

void forward_convolutional_layer_gpu	(	Darknet::Layer &	l,
		Darknet::NetworkState	state
	)

Here is the caller graph for this function:

pull_convolutional_layer()

void pull_convolutional_layer

(

Darknet::Layer &

l

)

Here is the call graph for this function:

Here is the caller graph for this function:

push_convolutional_layer()

void push_convolutional_layer

(

Darknet::Layer &

l

)

Here is the call graph for this function:

Here is the caller graph for this function:

reduce_kernel()

__global__ void reduce_kernel	(	float *	weights,
		int	n,
		int	size,
		float *	mean_arr_gpu
	)

Here is the call graph for this function:

Here is the caller graph for this function:

set_zero_kernel()

__global__ void set_zero_kernel	(	float *	src,
		int	size
	)

Here is the caller graph for this function:

update_convolutional_layer_gpu()

void update_convolutional_layer_gpu	(	Darknet::Layer &	l,
		int	batch,
		float	learning_rate_init,
		float	momentum,
		float	decay,
		float	loss_scale
	)

Here is the call graph for this function:

Here is the caller graph for this function:

warpAllReduceSum()

__inline__ __device__ float warpAllReduceSum

(

float

val

)

Here is the caller graph for this function: