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

#include "dark_cuda.hpp"
#include "tree.hpp"
#include <cblas-openblas64.h>

Include dependency graph for blas.hpp:

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

Macros
#define	axpy_cpu cblas_saxpy
	SAXPY constant times a vector plus a vector. Uses unrolled loops for increments equal to one.

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

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

#define	scal_cpu cblas_sscal
	Scales a vector by a constant. Uses unrolled loops for increment equal to 1.

Functions
void	activate_and_mult (float a1, float a2, size_t size, ACTIVATION a, float *dst)

void	adam_gpu (int n, float x, float m, float *v, float B1, float B2, float rate, float eps, int t)

void	adam_update_gpu (float w, float d, float m, float v, float B1, float B2, float eps, float decay, float rate, int n, int batch, int t)

void	add_3_arrays_activate (float a1, float a2, float a3, size_t size, ACTIVATION a, float dst)

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

void	add_bias_gpu (float output, float biases, int batch, int n, int size)

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

void	axpy_ongpu_offset (int N, float ALPHA, float X, int OFFX, int INCX, float Y, int OFFY, int INCY)

void	backward_bias_gpu (float bias_updates, float delta, int batch, int n, int size)

void	backward_implicit_gpu (int batch, int nweights, float weight_updates_gpu, float delta_gpu)

void	backward_sam_gpu (float in_w_h_c_delta, int size, int channel_size, float in_scales_c, float out_from_delta, float in_from_output, float *out_state_delta)

void	backward_scale_channels_gpu (float in_w_h_c_delta, int size, int channel_size, int batch_size, int scale_wh, float in_scales_c, float out_from_delta, float in_from_output, float *out_state_delta)

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

void	backward_scale_gpu (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)

void	backward_shortcut_multilayer_gpu (int src_outputs, int batch, int n, int outputs_of_layers_gpu, float layers_delta_gpu, 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	const_ongpu (int N, float ALPHA, float *X, int INCX)

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	constrain_weight_updates_ongpu (int N, float coef, float weights_gpu, float weight_updates_gpu)

void	coord_conv_gpu (float *dst, int size, int w, int h, int chan, int b, int type)

void	copy_ongpu (int N, float X, int INCX, float Y, int INCY)

void	copy_ongpu_offset (int N, float X, int OFFX, int INCX, float Y, int OFFY, int INCY)

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

void	expand_array_gpu (const float src_gpu, float dst_gpu, int size, int groups)

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

void	fast_mean_gpu (float x, int batch, int filters, int spatial, float mean)

void	fast_v_cbn_gpu (const float x, float mean, int batch, int filters, int spatial, int minibatch_index, int max_minibatch_index, float m_avg, float v_avg, float variance, const float alpha, float rolling_mean_gpu, float *rolling_variance_gpu, int inverse_variance, float epsilon)

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

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

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

void	fill_ongpu (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 (float *input, size_t 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	flatten_ongpu (float x, int spatial, int layers, int batch, int forward, float out)

void	forward_implicit_gpu (int batch, int nweights, float weight_gpu, float output_gpu)

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	gradient_centralization_gpu (int w, int h, int c, int f, float *in)

void	input_shortcut_gpu (float in, int batch, int w1, int h1, int c1, float add, int w2, int h2, int c2, float *out)

void	inverse_variance_ongpu (int size, float src, float dst, float epsilon)

int	is_nan_or_inf (float *input, size_t size)

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

void	l2_gpu (int n, float pred, float truth, float delta, float error)

void	mask_gpu_new_api (int N, float X, float mask_num, float mask, float val)

void	mask_ongpu (int N, float X, float mask_num, float mask)

float	math_vector_length (float *A, unsigned int feature_size)

void	mean_array_gpu (float src, int size, float alpha, float avg)

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	mean_gpu (float x, int batch, int filters, int spatial, float mean)

void	memcpy_ongpu (void dst, void src, int size_bytes)

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

void	mul_ongpu (int N, float X, int INCX, float Y, int INCY)

void	mult_add_into_gpu (int num, float a, float b, float *c)

void	mult_inverse_array_gpu (const float src_gpu, float dst_gpu, int size, float eps, float divider, float clip, float abs_add)

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)

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

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

void	normalize_scale_bias_gpu (float x, float mean, float variance, float scales, float *biases, int batch, int filters, int spatial, int inverse_variance, float epsilon)

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)

void	P_constrastive_f_det_gpu (int labels, unsigned int feature_size, float temperature, contrastive_params contrast_p, const int contrast_p_size)

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

float *	random_matrix (int rows, int cols)

void	reduce_and_expand_array_gpu (const float src_gpu, float dst_gpu, int size, int groups)

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

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

void	reset_nan_and_inf (float *input, size_t size)

void	rotate_weights_gpu (const float src_weight_gpu, float weight_deform_gpu, int nweights, int n, int size, int reverse)

void	sam_gpu (float in_w_h_c, int size, int channel_size, float scales_c, float *out)

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

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

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

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

void	scale_bias_gpu (float output, float biases, int batch, int n, int size)

void	scale_channels_gpu (float in_w_h_c, int size, int channel_size, int batch_size, int scale_wh, float scales_c, float *out)

void	shortcut_gpu (int batch, int w1, int h1, int c1, float add, int w2, int h2, int c2, float out)

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	shortcut_multilayer_gpu (int src_outputs, int batch, int n, int outputs_of_layers_gpu, float layers_output_gpu, float out, float in, float weights_gpu, int nweights, WEIGHTS_NORMALIZATION_T weights_normalization)

void	simple_copy_ongpu (int size, float src, float dst)

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

void	smooth_l1_gpu (int n, float pred, float truth, float delta, float error)

void	smooth_rotate_weights_gpu (const float src_weight_gpu, float weight_deform_gpu, int nweights, int n, int size, int angle, int reverse)

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_gpu (float input, int n, int offset, int groups, float temp, float output)

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

void	softmax_tree_gpu (float input, int spatial, int batch, int stride, float temp, float output, Darknet::Tree hier)

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

void	softmax_x_ent_gpu (int n, float pred, float truth, float delta, float error)

void	stretch_sway_flip_weights_gpu (const float src_weight_gpu, float weight_deform_gpu, int nweights, int n, int size, int angle, int reverse)

void	stretch_weights_gpu (const float src_weight_gpu, float weight_deform_gpu, int nweights, int n, int size, float scale, int reverse)

void	sum_of_mults (float a1, float a2, float b1, float b2, size_t size, float *dst)

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

void	sway_and_flip_weights_gpu (const float src_weight_gpu, float weight_deform_gpu, int nweights, int n, int size, int angle, int reverse)

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	upsample_gpu (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	variance_gpu (float x, float mean, int batch, int filters, int spatial, float *variance)

void	weighted_delta_gpu (float a, float b, float s, float da, float db, float ds, int num, float *dc)

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

void	weighted_sum_gpu (float a, float b, float s, int num, float c)

Detailed Description

BLAS: "Basic Linear Algebra Subprograms".

Macro Definition Documentation

◆ axpy_cpu

#define axpy_cpu cblas_saxpy

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

◆ copy_cpu

#define copy_cpu cblas_scopy

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

◆ dot_cpu

#define dot_cpu cblas_sdot

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

◆ scal_cpu

#define scal_cpu cblas_sscal

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

Function Documentation

◆ activate_and_mult()

void activate_and_mult	(	float *	a1,
		float *	a2,
		size_t	size,
		ACTIVATION	a,
		float *	dst
	)

Here is the call graph for this function:

◆ adam_gpu()

void adam_gpu	(	int	n,
		float *	x,
		float *	m,
		float *	v,
		float	B1,
		float	B2,
		float	rate,
		float	eps,
		int	t
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ adam_update_gpu()

void adam_update_gpu	(	float *	w,
		float *	d,
		float *	m,
		float *	v,
		float	B1,
		float	B2,
		float	eps,
		float	decay,
		float	rate,
		int	n,
		int	batch,
		int	t
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ add_3_arrays_activate()

void add_3_arrays_activate	(	float *	a1,
		float *	a2,
		float *	a3,
		size_t	size,
		ACTIVATION	a,
		float *	dst
	)

Here is the call graph for this function:

◆ add_bias()

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

Here is the caller graph for this function:

◆ add_bias_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ axpy_ongpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ axpy_ongpu_offset()

void axpy_ongpu_offset	(	int	N,
		float	ALPHA,
		float *	X,
		int	OFFX,
		int	INCX,
		float *	Y,
		int	OFFY,
		int	INCY
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ backward_bias_gpu()

void backward_bias_gpu	(	float *	bias_updates,
		float *	delta,
		int	batch,
		int	n,
		int	size
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ backward_implicit_gpu()

void backward_implicit_gpu	(	int	batch,
		int	nweights,
		float *	weight_updates_gpu,
		float *	delta_gpu
	)

Here is the call graph for this function:

◆ backward_sam_gpu()

void backward_sam_gpu	(	float *	in_w_h_c_delta,
		int	size,
		int	channel_size,
		float *	in_scales_c,
		float *	out_from_delta,
		float *	in_from_output,
		float *	out_state_delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ backward_scale_channels_gpu()

void backward_scale_channels_gpu	(	float *	in_w_h_c_delta,
		int	size,
		int	channel_size,
		int	batch_size,
		int	scale_wh,
		float *	in_scales_c,
		float *	out_from_delta,
		float *	in_from_output,
		float *	out_state_delta
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ 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_scale_gpu()

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

Here is the call graph for this function:

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:

◆ backward_shortcut_multilayer_gpu()

void backward_shortcut_multilayer_gpu	(	int	src_outputs,
		int	batch,
		int	n,
		int *	outputs_of_layers_gpu,
		float **	layers_delta_gpu,
		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
	)

◆ const_ongpu()

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

Here is the call graph for this function:

◆ 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:

◆ constrain_weight_updates_ongpu()

void constrain_weight_updates_ongpu	(	int	N,
		float	coef,
		float *	weights_gpu,
		float *	weight_updates_gpu
	)

Here is the call graph for this function:

◆ coord_conv_gpu()

void coord_conv_gpu	(	float *	dst,
		int	size,
		int	w,
		int	h,
		int	chan,
		int	b,
		int	type
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ copy_ongpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ copy_ongpu_offset()

void copy_ongpu_offset	(	int	N,
		float *	X,
		int	OFFX,
		int	INCX,
		float *	Y,
		int	OFFY,
		int	INCY
	)

Here is the call graph for this function:

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:

◆ expand_array_gpu()

void expand_array_gpu	(	const float *	src_gpu,
		float *	dst_gpu,
		int	size,
		int	groups
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fast_mean_delta_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fast_mean_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fast_v_cbn_gpu()

void fast_v_cbn_gpu	(	const float *	x,
		float *	mean,
		int	batch,
		int	filters,
		int	spatial,
		int	minibatch_index,
		int	max_minibatch_index,
		float *	m_avg,
		float *	v_avg,
		float *	variance,
		const float	alpha,
		float *	rolling_mean_gpu,
		float *	rolling_variance_gpu,
		int	inverse_variance,
		float	epsilon
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fast_variance_delta_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fast_variance_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ fill_cpu()

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

Here is the caller graph for this function:

◆ fill_ongpu()

void fill_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:

◆ 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()

void fix_nan_and_inf	(	float *	input,
		size_t	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:

◆ flatten_ongpu()

void flatten_ongpu	(	float *	x,
		int	spatial,
		int	layers,
		int	batch,
		int	forward,
		float *	out
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ forward_implicit_gpu()

void forward_implicit_gpu	(	int	batch,
		int	nweights,
		float *	weight_gpu,
		float *	output_gpu
	)

Here is the call 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:

◆ gradient_centralization_gpu()

void gradient_centralization_gpu	(	int	w,
		int	h,
		int	c,
		int	f,
		float *	in
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ input_shortcut_gpu()

void input_shortcut_gpu	(	float *	in,
		int	batch,
		int	w1,
		int	h1,
		int	c1,
		float *	add,
		int	w2,
		int	h2,
		int	c2,
		float *	out
	)

Here is the call graph for this function:

◆ inverse_variance_ongpu()

void inverse_variance_ongpu	(	int	size,
		float *	src,
		float *	dst,
		float	epsilon
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ is_nan_or_inf()

int is_nan_or_inf	(	float *	input,
		size_t	size
	)

Here is the call 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:

◆ l2_gpu()

void l2_gpu	(	int	n,
		float *	pred,
		float *	truth,
		float *	delta,
		float *	error
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ mask_gpu_new_api()

void mask_gpu_new_api	(	int	N,
		float *	X,
		float	mask_num,
		float *	mask,
		float	val
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ mask_ongpu()

void mask_ongpu	(	int	N,
		float *	X,
		float	mask_num,
		float *	mask
	)

Here is the call graph for this function:

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_array_gpu()

void mean_array_gpu	(	float *	src,
		int	size,
		float	alpha,
		float *	avg
	)

Here is the call graph for this function:

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:

◆ mean_gpu()

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

Here is the call graph for this function:

◆ memcpy_ongpu()

void memcpy_ongpu	(	void *	dst,
		void *	src,
		int	size_bytes
	)

Here is the call graph for this function:

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:

◆ mul_ongpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ mult_add_into_gpu()

void mult_add_into_gpu	(	int	num,
		float *	a,
		float *	b,
		float *	c
	)

Here is the call graph for this function:

◆ mult_inverse_array_gpu()

void mult_inverse_array_gpu	(	const float *	src_gpu,
		float *	dst_gpu,
		int	size,
		float	eps,
		float	divider,
		float	clip,
		float	abs_add
	)

Here is the call graph for this function:

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:

◆ normalize_delta_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ normalize_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ normalize_scale_bias_gpu()

void normalize_scale_bias_gpu	(	float *	x,
		float *	mean,
		float *	variance,
		float *	scales,
		float *	biases,
		int	batch,
		int	filters,
		int	spatial,
		int	inverse_variance,
		float	epsilon
	)

Here is the call graph for this function:

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:

◆ P_constrastive_f_det_gpu()

void P_constrastive_f_det_gpu	(	int *	labels,
		unsigned int	feature_size,
		float	temperature,
		contrastive_params *	contrast_p,
		const int	contrast_p_size
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ pow_ongpu()

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

Here is the call graph for this function:

◆ random_matrix()

float * random_matrix	(	int	rows,
		int	cols
	)

Here is the caller graph for this function:

◆ reduce_and_expand_array_gpu()

void reduce_and_expand_array_gpu	(	const float *	src_gpu,
		float *	dst_gpu,
		int	size,
		int	groups
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ 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:

◆ reorg_ongpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ reset_nan_and_inf()

void reset_nan_and_inf	(	float *	input,
		size_t	size
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ rotate_weights_gpu()

void rotate_weights_gpu	(	const float *	src_weight_gpu,
		float *	weight_deform_gpu,
		int	nweights,
		int	n,
		int	size,
		int	reverse
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ sam_gpu()

void sam_gpu	(	float *	in_w_h_c,
		int	size,
		int	channel_size,
		float *	scales_c,
		float *	out
	)

Here is the call graph for this function:

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_add_ongpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ scal_ongpu()

void scal_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:

◆ scale_bias()

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

Here is the caller graph for this function:

◆ scale_bias_gpu()

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

Here is the call graph for this function:

Here is the caller graph for this function:

◆ scale_channels_gpu()

void scale_channels_gpu	(	float *	in_w_h_c,
		int	size,
		int	channel_size,
		int	batch_size,
		int	scale_wh,
		float *	scales_c,
		float *	out
	)

Here is the call graph for this function:

Here is the caller graph for this function:

◆ shortcut_gpu()

void shortcut_gpu	(	int	batch,
		int	w1,
		int	h1,
		int	c1,
		float *	add,
		int	w2,
		int	h2,
		int	c2,
		float *	out
	)

Here is the call 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:

◆ shortcut_multilayer_gpu()

void shortcut_multilayer_gpu	(	int	src_outputs,
		int	batch,
		int	n,
		int *	outputs_of_layers_gpu,
		float **	layers_output_gpu,
		float *	out,
		float *	in,
		float *	weights_gpu,
		int	nweights,
		WEIGHTS_NORMALIZATION_T	weights_normalization
	)