Skip to content

Commit

Permalink
fix bug for rotation
Browse files Browse the repository at this point in the history
  • Loading branch information
@zhxfl
zhxfl committed May 9, 2015
1 parent 71af185 commit 9045885
Show file tree
Hide file tree
Showing 6 changed files with 33 additions and 32 deletions.
4 changes: 2 additions & 2 deletions Config/Cifar10Config.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8,12 +8,12 @@ BATCH_SIZE = 128; #test image size should be divided with no remai
CHANNELS = 3; #1, 3, 4#
CROP = 8; #0<= crop <=imgSize#
SCALE = 0.0; #13% of ImgSize#
ROTATION = 10.0; #angle#
ROTATION = 0.0; #angle#
DISTORTION = 0.0; #just for mnist#
SHOWIMAGE = false; #show the images after transformation#
HORIZONTAL = true; #horizontal reflection#
TEST_EPOCH = 10; #the period to get the test dataset's error rate#
WHITE_NOISE = 0.01;
WHITE_NOISE = 0.0;


[
Expand Down
6 changes: 3 additions & 3 deletions dataAugmentation/cuTrasformation.cu
View file
Original file line number Diff line number Diff line change
Expand Up @@ -190,10 +190,10 @@ __global__ void g_generateDistortionMap(
__syncthreads();

float rand1 = rand[blockIdx.x];
float rand2 = rand[blockIdx.x];
float rand2 = rand[blockIdx.x + 1];
if(fabs(dMaxRotation) >= 0.01){
rand1 += 1.0;
rand2 += 1.0;
if(rand1 <= 0.0) rand1 = 0.0;
if(rand2 <= 0.0) rand2 = 0.0;
}

for(int is = 0; is < ImgSize2; is += blockDim.x)
Expand Down
Empty file added dataAugmentation/pca.cu
View file
Empty file.
Empty file added dataAugmentation/pca.cuh
View file
Empty file.
1 change: 1 addition & 0 deletions main.cpp
View file
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
#pragma warning (disable: 4819)
#include <stdlib.h>
#include <math.h>
#include <cuda_runtime.h>
Expand Down
54 changes: 27 additions & 27 deletions net.cu
View file
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,7 @@ std::vector<ConfigBase*>que;
void cuSaveConvNet()
{
FILE *pOut = fopen("Result/checkPoint.txt", "w");
for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = Layers::instance()->get(que[i]->m_name);
layer->save(pOut);
}
Expand All @@ -52,7 +52,7 @@ void cuReadConvNet(
{
FILE *pIn = fopen(path, "r");

for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = Layers::instance()->get(que[i]->m_name);
layer->initFromCheckpoint(pIn);
}
Expand All @@ -65,7 +65,7 @@ void buildNetWork(int trainLen, int testLen)
/*BFS*/
std::queue<ConfigBase*>qqq;
std::set<ConfigBase*> inque;
for(int i = 0; i < Config::instance()->getFirstLayers().size(); i++){
for(int i = 0; i < (int)Config::instance()->getFirstLayers().size(); i++){
qqq.push(Config::instance()->getFirstLayers()[i]);
inque.insert(Config::instance()->getFirstLayers()[i]);
}
Expand All @@ -90,7 +90,7 @@ void buildNetWork(int trainLen, int testLen)
}else if(top->m_type == std::string("COMBINELAYER")){
ConfigCombineLayer *bl = static_cast<ConfigCombineLayer*>(top);
bool flag = true;
for(int i = 0; i < bl->m_inputs.size(); i++){
for(int i = 0; i < (int)bl->m_inputs.size(); i++){
ConfigBase* cb = Config::instance()->getLayerByName(bl->m_inputs[i]);
if(finish.find(cb) == finish.end()){
qqq.push(top);
Expand Down Expand Up @@ -119,7 +119,7 @@ void buildNetWork(int trainLen, int testLen)

sprintf(logStr, "layer %15s:", top->m_name.c_str());
LOG(logStr, "Result/log.txt");
for(int n = 0; n < top->m_next.size(); n++){
for(int n = 0; n < (int)top->m_next.size(); n++){
if(inque.find(top->m_next[n]) == inque.end()){
qqq.push(top->m_next[n]);
inque.insert(top->m_next[n]);
Expand Down Expand Up @@ -150,7 +150,7 @@ void cuFreeCNNMemory(
void updataWB()
{
/*updateWb*/
for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = Layers::instance()->get(que[i]->m_name);
layer->updateWeight();
}
Expand All @@ -161,21 +161,21 @@ void updataWB()
void getNetworkCost(int* y)
{
/*feedforward*/
for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
if(que[i]->m_type == std::string("SOFTMAX")){
SoftMax* sm = (SoftMax*)Layers::instance()->get(que[i]->m_name);
sm->setPredict(y);
}
}


for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = Layers::instance()->get(que[i]->m_name);
layer->feedforward();
}

/*backpropagation*/
for(int i = que.size() - 1; i >=0; i--){
for(int i = (int)que.size() - 1; i >=0; i--){
ConfigBase* top = que[i];
LayerBase* layer = Layers::instance()->get(top->m_name);
layer->backpropagation();
Expand Down Expand Up @@ -211,12 +211,12 @@ void resultProdict(int* vote,int start)
{
/*feedforward*/

for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = Layers::instance()->get(que[i]->m_name);
layer->feedforward();
}

for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
if(que[i]->m_type == std::string("SOFTMAX")){
g_getCorrect<<<dim3(1), Config::instance()->getBatchSize()>>>(
Layers::instance()->get(que[i]->m_name)->getOutputs()->getDev(),
Expand Down Expand Up @@ -341,19 +341,19 @@ void predictTestDate(cuMatrixVector<float>&x,
for (int h = 0; h < hlen; h++) {
for (int c = 0; c < clen; c++) {
dl->getBatchImageWithStreams(testX, 0);
for (int p = 0; p < (testX.size() + batch - 1) / batch; p++) {
for (int p = 0; p < ((int)testX.size() + batch - 1) / batch; p++) {
dl->synchronize();
printf("test %2d%%", 100 * p / ((testX.size() + batch - 1) / batch));
int tstart = p * batch;
if(tstart + batch <= testX.size() - batch)
if(tstart + batch <= (int)testX.size() - batch)
dl->getBatchImageWithStreams(testX, tstart + batch);
else {
int start = testX.size() - batch;
dl->getBatchImageWithStreams(testX, start);
}
if(tstart + batch > testX.size()){
tstart = testX.size() - batch;
if(tstart + batch > (int)testX.size()){
tstart = (int)testX.size() - batch;
}
dl->testData(cropr[c], cropc[c], rotate[rid], scalex[sidx], scaley[sidy], h);
Expand Down Expand Up @@ -390,15 +390,15 @@ void getBatchImageWithStreams(cuMatrixVector<float>&x,
cuMatrixVector<float>&batchImg,
int start,
cudaStream_t stream1){
for(int i = 0; i < batchImg.size(); i++){
for(int i = 0; i < (int)batchImg.size(); i++){
memcpy(batchImg[i]->getHost(), x[i + start]->getHost(), sizeof(float) * batchImg[i]->getLen());
batchImg[i]->toGpu(stream1);
}
}
float getCost(){
float cost = 0.0;
for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = (LayerBase*)Layers::instance()->get(que[i]->m_name);
layer->calCost();
layer->printCost();
Expand Down Expand Up @@ -436,19 +436,19 @@ void cuTrainNetwork(cuMatrixVector<float>&x,
int epochs = 10000;
float lrate = 0.05;
float Momentum = 0.9;
float lrate = 0.05f;
float Momentum = 0.9f;
int id = 0;
for (int epo = 0; epo < epochs; epo++) {
if (id >= nlrate.size())
if (id >= (int)nlrate.size())
break;
lrate = nlrate[id];
Momentum = nMomentum[id];
Config::instance()->setLrate(lrate);
Config::instance()->setMomentum(Momentum);
float start, end;
start = clock();
start = (float)clock();
cuApplyRandom(batch, clock() + epo, ImgSize);
Config::instance()->setTraining(true);
Expand All @@ -459,19 +459,19 @@ void cuTrainNetwork(cuMatrixVector<float>&x,
DataLayer *dl = static_cast<DataLayer*>(Layers::instance()->get("data"));
dl->getBatchImageWithStreams(x, 0);
for (int k = 0; k < (x.size() + batch - 1) / batch; k ++) {
for (int k = 0; k < ((int)x.size() + batch - 1) / batch; k ++) {
dl->synchronize();
int start = k * batch;
printf("train %2d%%", 100 * start / ((x.size() + batch - 1)));
if(start + batch <= x.size() - batch)
if(start + batch <= (int)x.size() - batch)
dl->getBatchImageWithStreams(x, start + batch);
else{
int tstart = x.size() - batch;
dl->getBatchImageWithStreams(x, tstart);
}
if(start + batch > x.size()){
start = x.size() - batch;
if(start + batch > (int)x.size()){
start = (int)x.size() - batch;
}
dl->trainData();
Expand All @@ -482,7 +482,7 @@ void cuTrainNetwork(cuMatrixVector<float>&x,
float cost = getCost();
end = clock();
end = (float)clock();
sprintf(logStr, "epoch=%d time=%.03lfs cost=%f Momentum=%.06lf lrate=%.08lf\n",
epo, (float) (end - start) / CLOCKS_PER_SEC,
cost,
Expand All @@ -500,7 +500,7 @@ void cuTrainNetwork(cuMatrixVector<float>&x,
sprintf(logStr, "===================weight value================\n");
LOG(logStr, "Result/log.txt");
for(int i = 0; i < que.size(); i++){
for(int i = 0; i < (int)que.size(); i++){
LayerBase* layer = Layers::instance()->get(que[i]->m_name);
layer->printParameter();
}
Expand Down

0 comments on commit 9045885

Please sign in to comment.