blas.hpp File Reference

BLAS: "Basic Linear Algebra Subprograms". More...

This graph shows which files directly or indirectly include this file:

Functions
void	add_bias (float *output, float *biases, int batch, int n, int size)
void	axpy_cpu (int N, float ALPHA, float *X, int INCX, float *Y, int INCY)
	SAXPY constant times a vector plus a vector. Uses unrolled loops for increments equal to one.
void	backward_scale_cpu (float *x_norm, float *delta, int batch, int n, int size, float *scale_updates)
void	backward_shortcut_multilayer_cpu (int size, int src_outputs, int batch, int n, int *outputs_of_layers, float **layers_delta, float *delta_out, float *delta_in, float *weights, float *weight_updates, int nweights, float *in, float **layers_output, WEIGHTS_NORMALIZATION_T weights_normalization)
int	check_sim (size_t i, size_t j, contrastive_params *contrast_p, int contrast_p_size)
void	constrain_cpu (int size, float ALPHA, float *X)
void	constrain_min_max_ongpu (int N, float MIN, float MAX, float *X, int INCX)
void	constrain_ongpu (int N, float ALPHA, float *X, int INCX)
void	copy_cpu (int N, float *X, int INCX, float *Y, int INCY)
	SCOPY copies a vector, x, to a vector, y. Uses unrolled loops for increments equal to 1.
float	cosine_similarity (float *A, float *B, unsigned int feature_size)
float	dot_cpu (int N, float *X, int INCX, float *Y, int INCY)
	SDOT forms the dot product of two vectors. Uses unrolled loops for increments equal to one.
void	fill_cpu (int N, float ALPHA, float *X, int INCX)
float	find_P_constrastive (size_t i, size_t j, contrastive_params *contrast_p, int contrast_p_size)
float	find_sim (size_t i, size_t j, contrastive_params *contrast_p, int contrast_p_size)
void	fix_nan_and_inf_cpu (float *input, size_t size)
void	flatten (float *x, int size, int layers, int batch, int forward)
void	get_embedding (float *src, int src_w, int src_h, int src_c, int embedding_size, int cur_w, int cur_h, int cur_n, int cur_b, float *dst)
void	grad_contrastive_loss_negative (size_t i, int *labels, size_t num_of_samples, float **z, unsigned int feature_size, float temperature, float *cos_sim, float *p_constrastive, float *delta, int wh)
void	grad_contrastive_loss_negative_f (size_t i, int *class_ids, int *labels, size_t num_of_samples, float **z, unsigned int feature_size, float temperature, float *delta, int wh, contrastive_params *contrast_p, int contrast_p_size, int neg_max)
void	grad_contrastive_loss_positive (size_t i, int *labels, size_t num_of_samples, float **z, unsigned int feature_size, float temperature, float *cos_sim, float *p_constrastive, float *delta, int wh)
void	grad_contrastive_loss_positive_f (size_t i, int *class_ids, int *labels, size_t num_of_samples, float **z, unsigned int feature_size, float temperature, float *delta, int wh, contrastive_params *contrast_p, int contrast_p_size)
void	l2_cpu (int n, float *pred, float *truth, float *delta, float *error)
float	math_vector_length (float *A, unsigned int feature_size)
void	mean_cpu (float *x, int batch, int filters, int spatial, float *mean)
void	mean_delta_cpu (float *delta, float *variance, int batch, int filters, int spatial, float *mean_delta)
void	mul_cpu (int N, float *X, int INCX, float *Y, int INCY)
void	normalize_cpu (float *x, float *mean, float *variance, int batch, int filters, int spatial)
void	normalize_delta_cpu (float *x, float *mean, float *variance, float *mean_delta, float *variance_delta, int batch, int filters, int spatial, float *delta)
float	P_constrastive (size_t i, size_t l, int *labels, size_t num_of_samples, float **z, unsigned int feature_size, float temperature, float *cos_sim, float *exp_cos_sim)
float	P_constrastive_f (size_t i, size_t l, int *labels, float **z, unsigned int feature_size, float temperature, contrastive_params *contrast_p, int contrast_p_size)
float	P_constrastive_f_det (size_t il, int *labels, float **z, unsigned int feature_size, float temperature, contrastive_params *contrast_p, int contrast_p_size)
float *	random_matrix (int rows, int cols)
void	reorg_cpu (float *x, int w, int h, int c, int batch, int stride, int forward, float *out)
void	scal_add_cpu (int N, float ALPHA, float BETA, float *X, int INCX)
void	scal_cpu (int N, float ALPHA, float *X, int INCX)
	Scales a vector by a constant. Uses unrolled loops for increment equal to 1.
void	scale_bias (float *output, float *scales, int batch, int n, int size)
void	shortcut_multilayer_cpu (int size, int src_outputs, int batch, int n, int *outputs_of_layers, float **layers_output, float *out, float *in, float *weights, int nweights, WEIGHTS_NORMALIZATION_T weights_normalization)
void	smooth_l1_cpu (int n, float *pred, float *truth, float *delta, float *error)
void	softmax (float *input, int n, float temp, float *output, int stride)
void	softmax_cpu (float *input, int n, int batch, int batch_offset, int groups, int group_offset, int stride, float temp, float *output)
void	softmax_x_ent_cpu (int n, float *pred, float *truth, float *delta, float *error)
void	test_blas ()
int	test_gpu_blas ()
void	upsample_cpu (float *in, int w, int h, int c, int batch, int stride, int forward, float scale, float *out)
void	variance_cpu (float *x, float *mean, int batch, int filters, int spatial, float *variance)
void	variance_delta_cpu (float *x, float *delta, float *mean, float *variance, int batch, int filters, int spatial, float *variance_delta)
void	weighted_sum_cpu (float *a, float *b, float *s, int num, float *c)

Detailed Description

BLAS: "Basic Linear Algebra Subprograms".

Function Documentation

add_bias()

void add_bias	(	float *	output,
		float *	biases,
		int	batch,
		int	n,
		int	size
	)

Here is the caller graph for this function:

axpy_cpu()

void axpy_cpu	(	int	N,
		float	ALPHA,
		float *	X,
		int	INCX,
		float *	Y,
		int	INCY
	)

SAXPY constant times a vector plus a vector. Uses unrolled loops for increments equal to one.

backward_scale_cpu()

void backward_scale_cpu	(	float *	x_norm,
		float *	delta,
		int	batch,
		int	n,
		int	size,
		float *	scale_updates
	)

Here is the caller graph for this function:

backward_shortcut_multilayer_cpu()

void backward_shortcut_multilayer_cpu	(	int	size,
		int	src_outputs,
		int	batch,
		int	n,
		int *	outputs_of_layers,
		float **	layers_delta,
		float *	delta_out,
		float *	delta_in,
		float *	weights,
		float *	weight_updates,
		int	nweights,
		float *	in,
		float **	layers_output,
		WEIGHTS_NORMALIZATION_T	weights_normalization
	)

Here is the call graph for this function:

Here is the caller graph for this function:

check_sim()

int check_sim	(	size_t	i,
		size_t	j,
		contrastive_params *	contrast_p,
		int	contrast_p_size
	)

constrain_cpu()

void constrain_cpu	(	int	size,
		float	ALPHA,
		float *	X
	)

constrain_min_max_ongpu()

void constrain_min_max_ongpu	(	int	N,
		float	MIN,
		float	MAX,
		float *	X,
		int	INCX
	)

Here is the call graph for this function:

Here is the caller graph for this function:

constrain_ongpu()

void constrain_ongpu	(	int	N,
		float	ALPHA,
		float *	X,
		int	INCX
	)

Here is the call graph for this function:

Here is the caller graph for this function:

copy_cpu()

void copy_cpu	(	int	N,
		float *	X,
		int	INCX,
		float *	Y,
		int	INCY
	)

SCOPY copies a vector, x, to a vector, y. Uses unrolled loops for increments equal to 1.

Here is the caller graph for this function:

cosine_similarity()

float cosine_similarity	(	float *	A,
		float *	B,
		unsigned int	feature_size
	)

Here is the caller graph for this function:

dot_cpu()

float dot_cpu	(	int	N,
		float *	X,
		int	INCX,
		float *	Y,
		int	INCY
	)

SDOT forms the dot product of two vectors. Uses unrolled loops for increments equal to one.

fill_cpu()

void fill_cpu	(	int	N,
		float	ALPHA,
		float *	X,
		int	INCX
	)

Here is the caller graph for this function:

find_P_constrastive()

float find_P_constrastive	(	size_t	i,
		size_t	j,
		contrastive_params *	contrast_p,
		int	contrast_p_size
	)

Here is the call graph for this function:

find_sim()

float find_sim	(	size_t	i,
		size_t	j,
		contrastive_params *	contrast_p,
		int	contrast_p_size
	)

Here is the call graph for this function:

Here is the caller graph for this function:

fix_nan_and_inf_cpu()

void fix_nan_and_inf_cpu	(	float *	input,
		size_t	size
	)

flatten()

void flatten	(	float *	x,
		int	size,
		int	layers,
		int	batch,
		int	forward
	)

Here is the caller graph for this function:

get_embedding()

void get_embedding	(	float *	src,
		int	src_w,
		int	src_h,
		int	src_c,
		int	embedding_size,
		int	cur_w,
		int	cur_h,
		int	cur_n,
		int	cur_b,
		float *	dst
	)

Here is the caller graph for this function:

grad_contrastive_loss_negative()

void grad_contrastive_loss_negative	(	size_t	i,
		int *	labels,
		size_t	num_of_samples,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		float *	cos_sim,
		float *	p_constrastive,
		float *	delta,
		int	wh
	)

Here is the call graph for this function:

Here is the caller graph for this function:

grad_contrastive_loss_negative_f()

void grad_contrastive_loss_negative_f	(	size_t	i,
		int *	class_ids,
		int *	labels,
		size_t	num_of_samples,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		float *	delta,
		int	wh,
		contrastive_params *	contrast_p,
		int	contrast_p_size,
		int	neg_max
	)

Here is the call graph for this function:

Here is the caller graph for this function:

grad_contrastive_loss_positive()

void grad_contrastive_loss_positive	(	size_t	i,
		int *	labels,
		size_t	num_of_samples,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		float *	cos_sim,
		float *	p_constrastive,
		float *	delta,
		int	wh
	)

Here is the call graph for this function:

Here is the caller graph for this function:

grad_contrastive_loss_positive_f()

void grad_contrastive_loss_positive_f	(	size_t	i,
		int *	class_ids,
		int *	labels,
		size_t	num_of_samples,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		float *	delta,
		int	wh,
		contrastive_params *	contrast_p,
		int	contrast_p_size
	)

Here is the call graph for this function:

Here is the caller graph for this function:

l2_cpu()

void l2_cpu	(	int	n,
		float *	pred,
		float *	truth,
		float *	delta,
		float *	error
	)

Here is the caller graph for this function:

math_vector_length()

float math_vector_length	(	float *	A,
		unsigned int	feature_size
	)

Here is the caller graph for this function:

mean_cpu()

void mean_cpu	(	float *	x,
		int	batch,
		int	filters,
		int	spatial,
		float *	mean
	)

Here is the caller graph for this function:

mean_delta_cpu()

void mean_delta_cpu	(	float *	delta,
		float *	variance,
		int	batch,
		int	filters,
		int	spatial,
		float *	mean_delta
	)

Here is the caller graph for this function:

mul_cpu()

void mul_cpu	(	int	N,
		float *	X,
		int	INCX,
		float *	Y,
		int	INCY
	)

Here is the caller graph for this function:

normalize_cpu()

void normalize_cpu	(	float *	x,
		float *	mean,
		float *	variance,
		int	batch,
		int	filters,
		int	spatial
	)

Here is the caller graph for this function:

normalize_delta_cpu()

void normalize_delta_cpu	(	float *	x,
		float *	mean,
		float *	variance,
		float *	mean_delta,
		float *	variance_delta,
		int	batch,
		int	filters,
		int	spatial,
		float *	delta
	)

Here is the caller graph for this function:

P_constrastive()

float P_constrastive	(	size_t	i,
		size_t	l,
		int *	labels,
		size_t	num_of_samples,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		float *	cos_sim,
		float *	exp_cos_sim
	)

Here is the call graph for this function:

Here is the caller graph for this function:

P_constrastive_f()

float P_constrastive_f	(	size_t	i,
		size_t	l,
		int *	labels,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		contrastive_params *	contrast_p,
		int	contrast_p_size
	)

Here is the call graph for this function:

Here is the caller graph for this function:

P_constrastive_f_det()

float P_constrastive_f_det	(	size_t	il,
		int *	labels,
		float **	z,
		unsigned int	feature_size,
		float	temperature,
		contrastive_params *	contrast_p,
		int	contrast_p_size
	)

Here is the caller graph for this function:

random_matrix()

float * random_matrix	(	int	rows,
		int	cols
	)

reorg_cpu()

void reorg_cpu	(	float *	x,
		int	w,
		int	h,
		int	c,
		int	batch,
		int	stride,
		int	forward,
		float *	out
	)

Here is the caller graph for this function:

scal_add_cpu()

void scal_add_cpu	(	int	N,
		float	ALPHA,
		float	BETA,
		float *	X,
		int	INCX
	)

Here is the caller graph for this function:

scal_cpu()

void scal_cpu	(	int	N,
		float	ALPHA,
		float *	X,
		int	INCX
	)

Scales a vector by a constant. Uses unrolled loops for increment equal to 1.

Here is the caller graph for this function:

scale_bias()

void scale_bias	(	float *	output,
		float *	scales,
		int	batch,
		int	n,
		int	size
	)

Here is the caller graph for this function:

shortcut_multilayer_cpu()

void shortcut_multilayer_cpu	(	int	size,
		int	src_outputs,
		int	batch,
		int	n,
		int *	outputs_of_layers,
		float **	layers_output,
		float *	out,
		float *	in,
		float *	weights,
		int	nweights,
		WEIGHTS_NORMALIZATION_T	weights_normalization
	)

Here is the call graph for this function:

Here is the caller graph for this function:

smooth_l1_cpu()

void smooth_l1_cpu	(	int	n,
		float *	pred,
		float *	truth,
		float *	delta,
		float *	error
	)

Here is the caller graph for this function:

softmax()

void softmax	(	float *	input,
		int	n,
		float	temp,
		float *	output,
		int	stride
	)

Here is the caller graph for this function:

softmax_cpu()

void softmax_cpu	(	float *	input,
		int	n,
		int	batch,
		int	batch_offset,
		int	groups,
		int	group_offset,
		int	stride,
		float	temp,
		float *	output
	)

Here is the call graph for this function:

Here is the caller graph for this function:

softmax_x_ent_cpu()

void softmax_x_ent_cpu	(	int	n,
		float *	pred,
		float *	truth,
		float *	delta,
		float *	error
	)

Here is the caller graph for this function:

test_blas()

void test_blas

(

)

test_gpu_blas()

int test_gpu_blas

(

)

upsample_cpu()

void upsample_cpu	(	float *	in,
		int	w,
		int	h,
		int	c,
		int	batch,
		int	stride,
		int	forward,
		float	scale,
		float *	out
	)

Here is the caller graph for this function:

variance_cpu()

void variance_cpu	(	float *	x,
		float *	mean,
		int	batch,
		int	filters,
		int	spatial,
		float *	variance
	)

Here is the caller graph for this function:

variance_delta_cpu()

void variance_delta_cpu	(	float *	x,
		float *	delta,
		float *	mean,
		float *	variance,
		int	batch,
		int	filters,
		int	spatial,
		float *	variance_delta
	)

Here is the caller graph for this function:

weighted_sum_cpu()

void weighted_sum_cpu	(	float *	a,
		float *	b,
		float *	s,
		int	num,
		float *	c
	)