float train_network_datum(network net)
{
#ifdef GPU
if(gpu_index >= 0) return train_network_datum_gpu(net);
#endif
*net.seen += net.batch;
net.train = 1;
forward_network(net);
backward_network(net);
float error = *net.cost;
if(((*net.seen)/net.batch)%net.subdivisions == 0) update_network(net);
return error;
}
float train_network_datum(network net, float *x, float *y)
{
#ifdef GPU
if(gpu_index >= 0) return train_network_datum_gpu(net, x, y);
#endif
network_state state;
state.input = x;
state.truth = y;
state.train = 1;
forward_network(net, state);
backward_network(net, state);
float error = get_network_cost(net);
if((net.seen/net.batch)%net.subdivisions == 0) update_network(net);
return error;
}
void reconstruct_picture(network net, float *features, image recon, image update, float rate, float momentum, float lambda, int smooth_size, int iters)
{
int iter = 0;
for (iter = 0; iter < iters; ++iter) {
image delta = make_image(recon.w, recon.h, recon.c);
NETWORK_STATE(state);
#ifdef GPU
state.input = cuda_make_array(recon.data, recon.w*recon.h*recon.c);
state.delta = cuda_make_array(delta.data, delta.w*delta.h*delta.c);
state.truth = cuda_make_array(features, get_network_output_size(net));
forward_network_gpu(net, state);
backward_network_gpu(net, state);
cuda_pull_array(state.delta, delta.data, delta.w*delta.h*delta.c);
cuda_free(state.input);
cuda_free(state.delta);
cuda_free(state.truth);
#else
state.input = recon.data;
state.delta = delta.data;
state.truth = features;
forward_network(net, state);
backward_network(net, state);
#endif
fltadd(update.data, delta.data, recon.w * recon.h * recon.c);
smooth(recon, update, lambda, smooth_size);
fltaddmul(recon.data, update.data, recon.w * recon.h * recon.c, rate);
scal_cpu(recon.w*recon.h*recon.c, momentum, update.data, 1);
//float mag = mag_array(recon.data, recon.w*recon.h*recon.c);
//scal_cpu(recon.w*recon.h*recon.c, 600/mag, recon.data, 1);
constrain_image(recon);
free_image(delta);
}
}
float train_network_batch(network net, data d, int n)
{
int i,j;
network_state state;
state.train = 1;
float sum = 0;
int batch = 2;
for(i = 0; i < n; ++i){
for(j = 0; j < batch; ++j){
int index = rand()%d.X.rows;
state.input = d.X.vals[index];
state.truth = d.y.vals[index];
forward_network(net, state);
backward_network(net, state);
sum += get_network_cost(net);
}
update_network(net);
}
return (float)sum/(n*batch);
}
void reconstruct_picture(network net, float *features, image recon, image update, float rate, float momentum, float lambda, int smooth_size)
{
scale_image(recon, 2);
translate_image(recon, -1);
image delta = make_image(recon.w, recon.h, recon.c);
network_state state = {0};
#ifdef GPU
state.input = cuda_make_array(recon.data, recon.w*recon.h*recon.c);
state.delta = cuda_make_array(delta.data, delta.w*delta.h*delta.c);
state.truth = cuda_make_array(features, get_network_output_size(net));
forward_network_gpu(net, state);
backward_network_gpu(net, state);
cuda_pull_array(state.delta, delta.data, delta.w*delta.h*delta.c);
cuda_free(state.input);
cuda_free(state.delta);
cuda_free(state.truth);
#else
state.input = recon.data;
state.delta = delta.data;
state.truth = features;
forward_network(net, state);
backward_network(net, state);
#endif
axpy_cpu(recon.w*recon.h*recon.c, 1, delta.data, 1, update.data, 1);
smooth(recon, update, lambda, smooth_size);
axpy_cpu(recon.w*recon.h*recon.c, rate, update.data, 1, recon.data, 1);
scal_cpu(recon.w*recon.h*recon.c, momentum, update.data, 1);
translate_image(recon, 1);
scale_image(recon, .5);
constrain_image(recon);
free_image(delta);
}
void optimize_picture(network *net, image orig, int max_layer, float scale, float rate, float thresh, int norm)
{
//scale_image(orig, 2);
//translate_image(orig, -1);
net->n = max_layer + 1;
int dx = rand()%16 - 8;
int dy = rand()%16 - 8;
int flip = rand()%2;
image crop = crop_image(orig, dx, dy, orig.w, orig.h);
image im = resize_image(crop, (int)(orig.w * scale), (int)(orig.h * scale));
if(flip) flip_image(im);
resize_network(net, im.w, im.h);
layer_t last = net->layers[net->n-1];
//net->layers[net->n - 1].activation = LINEAR;
image delta = make_image(im.w, im.h, im.c);
NETWORK_STATE(state);
#ifdef GPU
state.input = cuda_make_array(im.data, im.w*im.h*im.c);
state.delta = cuda_make_array(im.data, im.w*im.h*im.c);
forward_network_gpu(*net, state);
copy_ongpu(last.outputs, last.output_gpu, 1, last.delta_gpu, 1);
cuda_pull_array(last.delta_gpu, last.delta, last.outputs);
calculate_loss(last.delta, last.delta, last.outputs, thresh);
cuda_push_array(last.delta_gpu, last.delta, last.outputs);
backward_network_gpu(*net, state);
cuda_pull_array(state.delta, delta.data, im.w*im.h*im.c);
cuda_free(state.input);
cuda_free(state.delta);
#else
state.input = im.data;
state.delta = delta.data;
forward_network(*net, state);
fltcpy(last.delta, last.output, last.outputs);
calculate_loss(last.output, last.delta, last.outputs, thresh);
backward_network(*net, state);
#endif
if(flip) flip_image(delta);
//normalize_array(delta.data, delta.w*delta.h*delta.c);
image resized = resize_image(delta, orig.w, orig.h);
image out = crop_image(resized, -dx, -dy, orig.w, orig.h);
/*
image g = grayscale_image(out);
free_image(out);
out = g;
*/
//rate = rate / abs_mean(out.data, out.w*out.h*out.c);
if(norm) normalize_array(out.data, out.w*out.h*out.c);
fltaddmul(orig.data, out.data, orig.w * orig.h * orig.c, rate);
/*
normalize_array(orig.data, orig.w*orig.h*orig.c);
scale_image(orig, sqrt(var));
translate_image(orig, mean);
*/
//translate_image(orig, 1);
//scale_image(orig, .5);
//normalize_image(orig);
constrain_image(orig);
free_image(crop);
free_image(im);
free_image(delta);
free_image(resized);
free_image(out);
}
void train_dcgan(char *cfg, char *weight, char *acfg, char *aweight, int clear, int display, char *train_images, int maxbatch)
{
#ifdef GPU
char *backup_directory = "/home/kunle12/backup/";
srand(time(0));
char *base = basecfg(cfg);
char *abase = basecfg(acfg);
printf("%s\n", base);
network *gnet = load_network(cfg, weight, clear);
network *anet = load_network(acfg, aweight, clear);
//float orig_rate = anet->learning_rate;
int i, j, k;
layer imlayer = {0};
for (i = 0; i < gnet->n; ++i) {
if (gnet->layers[i].out_c == 3) {
imlayer = gnet->layers[i];
break;
}
}
printf("Learning Rate: %g, Momentum: %g, Decay: %g\n", gnet->learning_rate, gnet->momentum, gnet->decay);
int imgs = gnet->batch*gnet->subdivisions;
i = *gnet->seen/imgs;
data train, buffer;
list *plist = get_paths(train_images);
//int N = plist->size;
char **paths = (char **)list_to_array(plist);
load_args args= get_base_args(anet);
args.paths = paths;
args.n = imgs;
args.m = plist->size;
args.d = &buffer;
args.type = CLASSIFICATION_DATA;
args.threads=16;
args.classes = 1;
char *ls[2] = {"imagenet", "zzzzzzzz"};
args.labels = ls;
pthread_t load_thread = load_data_in_thread(args);
clock_t time;
gnet->train = 1;
anet->train = 1;
int x_size = gnet->inputs*gnet->batch;
int y_size = gnet->truths*gnet->batch;
float *imerror = cuda_make_array(0, y_size);
//int ay_size = anet->truths*anet->batch;
float aloss_avg = -1;
//data generated = copy_data(train);
if (maxbatch == 0) maxbatch = gnet->max_batches;
while (get_current_batch(gnet) < maxbatch) {
i += 1;
time=clock();
pthread_join(load_thread, 0);
train = buffer;
//translate_data_rows(train, -.5);
//scale_data_rows(train, 2);
load_thread = load_data_in_thread(args);
printf("Loaded: %lf seconds\n", sec(clock()-time));
data gen = copy_data(train);
for (j = 0; j < imgs; ++j) {
train.y.vals[j][0] = 1;
gen.y.vals[j][0] = 0;
}
time=clock();
for(j = 0; j < gnet->subdivisions; ++j){
get_next_batch(train, gnet->batch, j*gnet->batch, gnet->truth, 0);
int z;
for(z = 0; z < x_size; ++z){
gnet->input[z] = rand_normal();
}
for(z = 0; z < gnet->batch; ++z){
float mag = mag_array(gnet->input + z*gnet->inputs, gnet->inputs);
scale_array(gnet->input + z*gnet->inputs, gnet->inputs, 1./mag);
}
/*
for(z = 0; z < 100; ++z){
printf("%f, ", gnet->input[z]);
}
printf("\n");
printf("input: %f %f\n", mean_array(gnet->input, x_size), variance_array(gnet->input, x_size));
*/
//cuda_push_array(gnet->input_gpu, gnet->input, x_size);
//cuda_push_array(gnet->truth_gpu, gnet->truth, y_size);
*gnet->seen += gnet->batch;
forward_network(gnet);
fill_gpu(imlayer.outputs*imlayer.batch, 0, imerror, 1);
fill_cpu(anet->truths*anet->batch, 1, anet->truth, 1);
copy_cpu(anet->inputs*anet->batch, imlayer.output, 1, anet->input, 1);
anet->delta_gpu = imerror;
forward_network(anet);
backward_network(anet);
//float genaloss = *anet->cost / anet->batch;
//printf("%f\n", genaloss);
scal_gpu(imlayer.outputs*imlayer.batch, 1, imerror, 1);
scal_gpu(imlayer.outputs*imlayer.batch, 0, gnet->layers[gnet->n-1].delta_gpu, 1);
//printf("realness %f\n", cuda_mag_array(imerror, imlayer.outputs*imlayer.batch));
//printf("features %f\n", cuda_mag_array(gnet->layers[gnet->n-1].delta_gpu, imlayer.outputs*imlayer.batch));
axpy_gpu(imlayer.outputs*imlayer.batch, 1, imerror, 1, gnet->layers[gnet->n-1].delta_gpu, 1);
backward_network(gnet);
/*
for(k = 0; k < gnet->n; ++k){
layer l = gnet->layers[k];
cuda_pull_array(l.output_gpu, l.output, l.outputs*l.batch);
printf("%d: %f %f\n", k, mean_array(l.output, l.outputs*l.batch), variance_array(l.output, l.outputs*l.batch));
}
*/
for(k = 0; k < gnet->batch; ++k){
int index = j*gnet->batch + k;
copy_cpu(gnet->outputs, gnet->output + k*gnet->outputs, 1, gen.X.vals[index], 1);
}
}
harmless_update_network_gpu(anet);
data merge = concat_data(train, gen);
//randomize_data(merge);
float aloss = train_network(anet, merge);
//translate_image(im, 1);
//scale_image(im, .5);
//translate_image(im2, 1);
//scale_image(im2, .5);
#ifdef OPENCV
if(display){
image im = float_to_image(anet->w, anet->h, anet->c, gen.X.vals[0]);
image im2 = float_to_image(anet->w, anet->h, anet->c, train.X.vals[0]);
show_image(im, "gen", 1);
show_image(im2, "train", 1);
save_image(im, "gen");
save_image(im2, "train");
}
#endif
/*
if(aloss < .1){
anet->learning_rate = 0;
} else if (aloss > .3){
anet->learning_rate = orig_rate;
}
*/
update_network_gpu(gnet);
free_data(merge);
free_data(train);
free_data(gen);
if (aloss_avg < 0) aloss_avg = aloss;
aloss_avg = aloss_avg*.9 + aloss*.1;
printf("%d: adv: %f | adv_avg: %f, %f rate, %lf seconds, %d images\n", i, aloss, aloss_avg, get_current_rate(gnet), sec(clock()-time), i*imgs);
if(i%10000==0){
char buff[256];
sprintf(buff, "%s/%s_%d.weights", backup_directory, base, i);
save_weights(gnet, buff);
sprintf(buff, "%s/%s_%d.weights", backup_directory, abase, i);
save_weights(anet, buff);
}
if(i%1000==0){
char buff[256];
sprintf(buff, "%s/%s.backup", backup_directory, base);
save_weights(gnet, buff);
sprintf(buff, "%s/%s.backup", backup_directory, abase);
save_weights(anet, buff);
}
}
char buff[256];
sprintf(buff, "%s/%s_final.weights", backup_directory, base);
save_weights(gnet, buff);
#endif
free_network(gnet);
free_network(anet);
}
void train_prog(char *cfg, char *weight, char *acfg, char *aweight, int clear, int display, char *train_images, int maxbatch)
{
#ifdef GPU
char *backup_directory = "/home/kunle12/backup/";
srand(time(0));
char *base = basecfg(cfg);
char *abase = basecfg(acfg);
printf("%s\n", base);
network *gnet = load_network(cfg, weight, clear);
network *anet = load_network(acfg, aweight, clear);
int i, j, k;
layer imlayer = gnet->layers[gnet->n-1];
printf("Learning Rate: %g, Momentum: %g, Decay: %g\n", gnet->learning_rate, gnet->momentum, gnet->decay);
int imgs = gnet->batch*gnet->subdivisions;
i = *gnet->seen/imgs;
data train, buffer;
list *plist = get_paths(train_images);
char **paths = (char **)list_to_array(plist);
load_args args= get_base_args(anet);
args.paths = paths;
args.n = imgs;
args.m = plist->size;
args.d = &buffer;
args.type = CLASSIFICATION_DATA;
args.threads=16;
args.classes = 1;
char *ls[2] = {"imagenet", "zzzzzzzz"};
args.labels = ls;
pthread_t load_thread = load_data_in_thread(args);
clock_t time;
gnet->train = 1;
anet->train = 1;
int x_size = gnet->inputs*gnet->batch;
int y_size = gnet->truths*gnet->batch;
float *imerror = cuda_make_array(0, y_size);
float aloss_avg = -1;
if (maxbatch == 0) maxbatch = gnet->max_batches;
while (get_current_batch(gnet) < maxbatch) {
{
int cb = get_current_batch(gnet);
float alpha = (float) cb / (maxbatch/2);
if(alpha > 1) alpha = 1;
float beta = 1 - alpha;
printf("%f %f\n", alpha, beta);
set_network_alpha_beta(gnet, alpha, beta);
set_network_alpha_beta(anet, beta, alpha);
}
i += 1;
time=clock();
pthread_join(load_thread, 0);
train = buffer;
load_thread = load_data_in_thread(args);
printf("Loaded: %lf seconds\n", sec(clock()-time));
data gen = copy_data(train);
for (j = 0; j < imgs; ++j) {
train.y.vals[j][0] = 1;
gen.y.vals[j][0] = 0;
}
time=clock();
for (j = 0; j < gnet->subdivisions; ++j) {
get_next_batch(train, gnet->batch, j*gnet->batch, gnet->truth, 0);
int z;
for(z = 0; z < x_size; ++z){
gnet->input[z] = rand_normal();
}
/*
for(z = 0; z < gnet->batch; ++z){
float mag = mag_array(gnet->input + z*gnet->inputs, gnet->inputs);
scale_array(gnet->input + z*gnet->inputs, gnet->inputs, 1./mag);
}
*/
*gnet->seen += gnet->batch;
forward_network(gnet);
fill_gpu(imlayer.outputs*imlayer.batch, 0, imerror, 1);
fill_cpu(anet->truths*anet->batch, 1, anet->truth, 1);
copy_cpu(anet->inputs*anet->batch, imlayer.output, 1, anet->input, 1);
anet->delta_gpu = imerror;
forward_network(anet);
backward_network(anet);
//float genaloss = *anet->cost / anet->batch;
scal_gpu(imlayer.outputs*imlayer.batch, 1, imerror, 1);
scal_gpu(imlayer.outputs*imlayer.batch, 0, gnet->layers[gnet->n-1].delta_gpu, 1);
axpy_gpu(imlayer.outputs*imlayer.batch, 1, imerror, 1, gnet->layers[gnet->n-1].delta_gpu, 1);
backward_network(gnet);
for(k = 0; k < gnet->batch; ++k){
int index = j*gnet->batch + k;
copy_cpu(gnet->outputs, gnet->output + k*gnet->outputs, 1, gen.X.vals[index], 1);
}
}
harmless_update_network_gpu(anet);
data merge = concat_data(train, gen);
float aloss = train_network(anet, merge);
#ifdef OPENCV
if(display){
image im = float_to_image(anet->w, anet->h, anet->c, gen.X.vals[0]);
image im2 = float_to_image(anet->w, anet->h, anet->c, train.X.vals[0]);
show_image(im, "gen", 1);
show_image(im2, "train", 1);
save_image(im, "gen");
save_image(im2, "train");
}
#endif
update_network_gpu(gnet);
free_data(merge);
free_data(train);
free_data(gen);
if (aloss_avg < 0) aloss_avg = aloss;
aloss_avg = aloss_avg*.9 + aloss*.1;
printf("%d: adv: %f | adv_avg: %f, %f rate, %lf seconds, %d images\n", i, aloss, aloss_avg, get_current_rate(gnet), sec(clock()-time), i*imgs);
if(i%10000==0){
char buff[256];
sprintf(buff, "%s/%s_%d.weights", backup_directory, base, i);
save_weights(gnet, buff);
sprintf(buff, "%s/%s_%d.weights", backup_directory, abase, i);
save_weights(anet, buff);
}
if(i%1000==0){
char buff[256];
sprintf(buff, "%s/%s.backup", backup_directory, base);
save_weights(gnet, buff);
sprintf(buff, "%s/%s.backup", backup_directory, abase);
save_weights(anet, buff);
}
}
char buff[256];
sprintf(buff, "%s/%s_final.weights", backup_directory, base);
save_weights(gnet, buff);
#endif
free_network( gnet );
free_network( anet );
}