darknet/src/deconvolutional_kernels.cu

137 lines
4.5 KiB
Plaintext
Raw Normal View History

2015-11-16 06:51:26 +03:00
#include "cuda_runtime.h"
#include "curand.h"
#include "cublas_v2.h"
2015-02-11 06:41:03 +03:00
extern "C" {
#include "convolutional_layer.h"
#include "deconvolutional_layer.h"
#include "batchnorm_layer.h"
2015-02-11 06:41:03 +03:00
#include "gemm.h"
#include "blas.h"
#include "im2col.h"
#include "col2im.h"
#include "utils.h"
#include "cuda.h"
}
extern "C" void forward_deconvolutional_layer_gpu(layer l, network net)
2015-02-11 06:41:03 +03:00
{
int i;
int m = l.size*l.size*l.n;
int n = l.h*l.w;
int k = l.c;
2015-02-11 06:41:03 +03:00
fill_ongpu(l.outputs*l.batch, 0, l.output_gpu, 1);
2015-02-11 06:41:03 +03:00
for(i = 0; i < l.batch; ++i){
float *a = l.weights_gpu;
float *b = net.input_gpu + i*l.c*l.h*l.w;
float *c = net.workspace;
2015-02-11 06:41:03 +03:00
gemm_ongpu(1,0,m,n,k,1,a,m,b,n,0,c,n);
col2im_ongpu(net.workspace, l.out_c, l.out_h, l.out_w, l.size, l.stride, l.pad, l.output_gpu+i*l.outputs);
2015-02-11 06:41:03 +03:00
}
if (l.batch_normalize) {
forward_batchnorm_layer_gpu(l, net);
} else {
add_bias_gpu(l.output_gpu, l.biases_gpu, l.batch, l.n, l.out_w*l.out_h);
}
activate_array_ongpu(l.output_gpu, l.batch*l.n*l.out_w*l.out_h, l.activation);
2015-02-11 06:41:03 +03:00
}
extern "C" void backward_deconvolutional_layer_gpu(layer l, network net)
2015-02-11 06:41:03 +03:00
{
int i;
constrain_ongpu(l.outputs*l.batch, 1, l.delta_gpu, 1);
gradient_array_ongpu(l.output_gpu, l.outputs*l.batch, l.activation, l.delta_gpu);
if(l.batch_normalize){
backward_batchnorm_layer_gpu(l, net);
} else {
backward_bias_gpu(l.bias_updates_gpu, l.delta_gpu, l.batch, l.n, l.out_w*l.out_h);
}
2015-02-11 06:41:03 +03:00
//if(net.delta_gpu) memset(net.delta_gpu, 0, l.batch*l.h*l.w*l.c*sizeof(float));
2015-02-11 06:41:03 +03:00
for(i = 0; i < l.batch; ++i){
int m = l.c;
int n = l.size*l.size*l.n;
int k = l.h*l.w;
2015-02-11 06:41:03 +03:00
float *a = net.input_gpu + i*m*k;
float *b = net.workspace;
float *c = l.weight_updates_gpu;
2015-02-11 06:41:03 +03:00
im2col_ongpu(l.delta_gpu + i*l.outputs, l.out_c, l.out_h, l.out_w,
l.size, l.stride, l.pad, b);
gemm_ongpu(0,1,m,n,k,1,a,k,b,k,1,c,n);
2015-02-11 06:41:03 +03:00
if(net.delta_gpu){
int m = l.c;
int n = l.h*l.w;
int k = l.size*l.size*l.n;
2015-02-11 06:41:03 +03:00
float *a = l.weights_gpu;
float *b = net.workspace;
float *c = net.delta_gpu + i*n*m;
2015-02-11 06:41:03 +03:00
gemm_ongpu(0,0,m,n,k,1,a,k,b,n,1,c,n);
2015-02-11 06:41:03 +03:00
}
}
}
extern "C" void pull_deconvolutional_layer(layer l)
2015-02-11 06:41:03 +03:00
{
cuda_pull_array(l.weights_gpu, l.weights, l.c*l.n*l.size*l.size);
cuda_pull_array(l.biases_gpu, l.biases, l.n);
cuda_pull_array(l.weight_updates_gpu, l.weight_updates, l.c*l.n*l.size*l.size);
cuda_pull_array(l.bias_updates_gpu, l.bias_updates, l.n);
if (l.batch_normalize){
cuda_pull_array(l.scales_gpu, l.scales, l.n);
cuda_pull_array(l.rolling_mean_gpu, l.rolling_mean, l.n);
cuda_pull_array(l.rolling_variance_gpu, l.rolling_variance, l.n);
}
2015-02-11 06:41:03 +03:00
}
extern "C" void push_deconvolutional_layer(layer l)
2015-02-11 06:41:03 +03:00
{
cuda_push_array(l.weights_gpu, l.weights, l.c*l.n*l.size*l.size);
cuda_push_array(l.biases_gpu, l.biases, l.n);
cuda_push_array(l.weight_updates_gpu, l.weight_updates, l.c*l.n*l.size*l.size);
cuda_push_array(l.bias_updates_gpu, l.bias_updates, l.n);
if (l.batch_normalize){
cuda_push_array(l.scales_gpu, l.scales, l.n);
cuda_push_array(l.rolling_mean_gpu, l.rolling_mean, l.n);
cuda_push_array(l.rolling_variance_gpu, l.rolling_variance, l.n);
}
2015-02-11 06:41:03 +03:00
}
void update_deconvolutional_layer_gpu(layer l, int batch, float learning_rate, float momentum, float decay)
2015-02-11 06:41:03 +03:00
{
int size = l.size*l.size*l.c*l.n;
2015-02-11 06:41:03 +03:00
if(l.adam){
2017-04-13 00:22:53 +03:00
adam_update_gpu(l.weights_gpu, l.weight_updates_gpu, l.m_gpu, l.v_gpu, l.B1, l.B2, l.eps, decay, learning_rate, size, batch, l.t);
adam_update_gpu(l.biases_gpu, l.bias_updates_gpu, l.bias_m_gpu, l.bias_v_gpu, l.B1, l.B2, l.eps, decay, learning_rate, l.n, batch, l.t);
if(l.scales_gpu){
2017-04-13 00:22:53 +03:00
adam_update_gpu(l.scales_gpu, l.scale_updates_gpu, l.scale_m_gpu, l.scale_v_gpu, l.B1, l.B2, l.eps, decay, learning_rate, l.n, batch, l.t);
}
}else{
axpy_ongpu(size, -decay*batch, l.weights_gpu, 1, l.weight_updates_gpu, 1);
axpy_ongpu(size, learning_rate/batch, l.weight_updates_gpu, 1, l.weights_gpu, 1);
scal_ongpu(size, momentum, l.weight_updates_gpu, 1);
axpy_ongpu(l.n, learning_rate/batch, l.bias_updates_gpu, 1, l.biases_gpu, 1);
scal_ongpu(l.n, momentum, l.bias_updates_gpu, 1);
2015-02-11 06:41:03 +03:00
if(l.scales_gpu){
axpy_ongpu(l.n, learning_rate/batch, l.scale_updates_gpu, 1, l.scales_gpu, 1);
scal_ongpu(l.n, momentum, l.scale_updates_gpu, 1);
}
}
2015-02-11 06:41:03 +03:00
}