Merge branch 'master' of github.com:pjreddie/darknet

Makefile

@@ -68,8 +68,8 @@ EXECOBJ = $(addprefix $(OBJDIR), $(EXECOBJA))
 OBJS = $(addprefix $(OBJDIR), $(OBJ))
 DEPS = $(wildcard src/*.h) Makefile include/darknet.h
 
-#all: obj backup results $(SLIB) $(ALIB) $(EXEC)
-all: obj  results $(SLIB) $(ALIB) $(EXEC)
+all: obj backup results $(SLIB) $(ALIB) $(EXEC)
+#all: obj  results $(SLIB) $(ALIB) $(EXEC)
 
 
 $(EXEC): $(EXECOBJ) $(ALIB)

cfg/alexnet.cfg

@@ -90,6 +90,3 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/cifar.cfg

@@ -119,6 +119,3 @@ activation=leaky
 
 [softmax]
 groups=1
-
-[cost]
-

cfg/cifar.test.cfg

@@ -115,5 +115,3 @@ activation=leaky
 groups=1
 temperature=3
 
-[cost]
-

cfg/darknet19.cfg

@@ -203,6 +203,3 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/darknet19_448.cfg

@@ -195,6 +195,3 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/densenet201.cfg

@@ -1949,6 +1949,3 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/extraction.cfg

@@ -201,6 +201,3 @@ activation=leaky
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/extraction22k.cfg

@@ -204,6 +204,3 @@ activation=leaky
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/go.cfg

@@ -130,6 +130,3 @@ stride=1
 
 [softmax]
 
-[cost]
-type=sse
-

cfg/go.test.cfg

@@ -129,6 +129,4 @@ stride=1
 
 [softmax]
 
-[cost]
-type=sse
 

cfg/gru.cfg

@@ -27,6 +27,4 @@ activation=linear
 
 [softmax]
 
-[cost]
-type=sse
 

cfg/resnet152.cfg

@@ -1458,6 +1458,3 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
-

cfg/resnet50.cfg

@@ -506,6 +506,4 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
 

cfg/rnn.cfg

@@ -35,6 +35,4 @@ activation=leaky
 
 [softmax]
 
-[cost]
-type=sse
 

cfg/rnn.train.cfg

@@ -35,6 +35,4 @@ activation=leaky
 
 [softmax]
 
-[cost]
-type=sse
 

cfg/strided.cfg

@@ -180,6 +180,3 @@ activation=ramp
 
 [softmax]
 
-[cost]
-type=sse
-

cfg/tiny.cfg

@@ -171,6 +171,4 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
 

cfg/vgg-16.cfg

@@ -148,6 +148,4 @@ activation=linear
 [softmax]
 groups=1
 
-[cost]
-type=sse
 

src/convolutional_kernels.cu

@@ -111,9 +111,13 @@ void forward_convolutional_layer_gpu(convolutional_layer l, network net)
             float *a = l.weights_gpu + j*l.nweights/l.groups;
             float *b = net.workspace;
             float *c = l.output_gpu + (i*l.groups + j)*n*m;
+            float *im = net.input_gpu + (i*l.groups + j)*l.c/l.groups*l.h*l.w;
 
-            im2col_gpu(net.input_gpu + (i*l.groups + j)*l.c/l.groups*l.h*l.w,
-                l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, b);
+            if (l.size == 1){
+                b = im;
+            } else {
+                im2col_gpu(im, l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, b);
+            }
             gemm_gpu(0,0,m,n,k,1,a,k,b,n,1,c,n);
         }
     }
@@ -237,9 +241,9 @@ void backward_convolutional_layer_gpu(convolutional_layer l, network net)
             float *c = l.weight_updates_gpu + j*l.nweights/l.groups;
 
             float *im  = net.input_gpu+(i*l.groups + j)*l.c/l.groups*l.h*l.w;
+            float *imd = net.delta_gpu+(i*l.groups + j)*l.c/l.groups*l.h*l.w;
 
-            im2col_gpu(im, l.c/l.groups, l.h, l.w,
-                    l.size, l.stride, l.pad, b);
+            im2col_gpu(im, l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, b);
             gemm_gpu(0,1,m,n,k,1,a,k,b,k,1,c,n);
 
             if (net.delta_gpu) {
@@ -247,11 +251,15 @@ void backward_convolutional_layer_gpu(convolutional_layer l, network net)
                 a = l.weights_gpu + j*l.nweights/l.groups;
                 b = l.delta_gpu + (i*l.groups + j)*m*k;
                 c = net.workspace;
+                if (l.size == 1) {
+                    c = imd;
+                }
 
                 gemm_gpu(1,0,n,k,m,1,a,n,b,k,0,c,k);
 
-                col2im_gpu(net.workspace, l.c/l.groups, l.h, l.w, l.size, l.stride, 
-                    l.pad, net.delta_gpu + (i*l.groups + j)*l.c/l.groups*l.h*l.w);
+                if (l.size != 1) {
+                    col2im_gpu(net.workspace, l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, imd);
+                }
                 if(l.binary || l.xnor) {
                     swap_binary(&l);
                 }

src/convolutional_layer.c

@@ -463,9 +463,13 @@ void forward_convolutional_layer(convolutional_layer l, network net)
             float *a = l.weights + j*l.nweights/l.groups;
             float *b = net.workspace;
             float *c = l.output + (i*l.groups + j)*n*m;
+            float *im =  net.input + (i*l.groups + j)*l.c/l.groups*l.h*l.w;
 
-            im2col_cpu(net.input + (i*l.groups + j)*l.c/l.groups*l.h*l.w,
-                l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, b);
+            if (l.size == 1) {
+                b = im;
+            } else {
+                im2col_cpu(im, l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, b);
+            }
             gemm(0,0,m,n,k,1,a,k,b,n,1,c,n);
         }
     }
@@ -502,20 +506,30 @@ void backward_convolutional_layer(convolutional_layer l, network net)
             float *c = l.weight_updates + j*l.nweights/l.groups;
 
             float *im  = net.input + (i*l.groups + j)*l.c/l.groups*l.h*l.w;
+            float *imd = net.delta + (i*l.groups + j)*l.c/l.groups*l.h*l.w;
 
+            if(l.size == 1){
+                b = im;
+            } else {
                 im2col_cpu(im, l.c/l.groups, l.h, l.w, 
                         l.size, l.stride, l.pad, b);
+            }
+
             gemm(0,1,m,n,k,1,a,k,b,k,1,c,n);
 
             if (net.delta) {
                 a = l.weights + j*l.nweights/l.groups;
                 b = l.delta + (i*l.groups + j)*m*k;
                 c = net.workspace;
+                if (l.size == 1) {
+                    c = imd;
+                }
 
                 gemm(1,0,n,k,m,1,a,n,b,k,0,c,k);
 
-                col2im_cpu(net.workspace, l.c/l.groups, l.h, l.w, l.size, l.stride, 
-                    l.pad, net.delta + (i*l.groups + j)*l.c/l.groups*l.h*l.w);
+                if (l.size != 1) {
+                    col2im_cpu(net.workspace, l.c/l.groups, l.h, l.w, l.size, l.stride, l.pad, imd);
+                }
             }
         }
     }

src/utils.c

@@ -7,6 +7,7 @@
 #include <float.h>
 #include <limits.h>
 #include <time.h>
+#include <sys/time.h>
 
 #include "utils.h"
 
@@ -25,9 +26,11 @@ double get_wall_time()
 
 double what_time_is_it_now()
 {
-    struct timespec now;
-    clock_gettime(CLOCK_REALTIME, &now);
-    return now.tv_sec + now.tv_nsec*1e-9;
+    struct timeval time;
+    if (gettimeofday(&time,NULL)){
+        return 0;
+    }
+    return (double)time.tv_sec + (double)time.tv_usec * .000001;
 }
 
 int *read_intlist(char *gpu_list, int *ngpus, int d)