Add tp tests for yolo3512 and yolo3tiny512

Signed-off-by: Micaela Verucchi <micaelaverucchi@gmail.com>
This commit is contained in:
Micaela Verucchi
2020-03-26 15:06:49 +01:00
parent 3eda9b9219
commit 7c55dcb708
7 changed files with 1202 additions and 5 deletions
+7 -1
View File
@@ -41,7 +41,7 @@ find_package(yaml-cpp REQUIRED)
file(GLOB tkdnn_SRC "src/*.cpp")
set(tkdnn_LIBS kernels ${CUDA_LIBRARIES} ${CUDA_CUBLAS_LIBRARIES} ${CUDNN_LIBRARIES} ${OpenCV_LIBS} yaml-cpp)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -std=c++11")
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11")
include_directories(${CMAKE_CURRENT_SOURCE_DIR}/include ${CUDA_INCLUDE_DIRS} ${OPENCV_INCLUDE_DIRS} ${NVINFER_INCLUDES})
add_library(tkDNN SHARED ${tkdnn_SRC})
target_link_libraries(tkDNN ${tkdnn_LIBS} nvinfer_plugin)
@@ -88,12 +88,18 @@ target_link_libraries(test_yolo3 tkDNN)
add_executable(test_yolo3_512 tests/yolo3_512/yolo3_512.cpp)
target_link_libraries(test_yolo3_512 tkDNN)
add_executable(test_yolo3_512tp tests/yolo3_512tp/yolo3_512tp.cpp)
target_link_libraries(test_yolo3_512tp tkDNN)
add_executable(test_yolo3_tiny tests/yolo3_tiny/yolo3_tiny.cpp)
target_link_libraries(test_yolo3_tiny tkDNN)
add_executable(test_yolo3_tiny512 tests/yolo3_tiny512/yolo3_tiny512.cpp)
target_link_libraries(test_yolo3_tiny512 tkDNN)
add_executable(test_yolo3_tiny512tp tests/yolo3_tiny512tp/yolo3_tiny512tp.cpp)
target_link_libraries(test_yolo3_tiny512tp tkDNN)
add_executable(test_yolo3_berkeley tests/yolo3_berkeley/yolo3_berkeley.cpp)
target_link_libraries(test_yolo3_berkeley tkDNN)
+7
View File
@@ -0,0 +1,7 @@
classes : 3 #number of classes
map_points : 101 #number of recall points (0 for all, 101 for COCO, 11 PascalVOC)
map_levels : 10 #number of IoU step for the AP
map_step : 0.05 #step of IoU
IoU_thresh : 0.5 #starting IoU threshold
conf_thresh : 0.0 #threshold on the condifence of the bbox
verbose : false #print on screen information
+3 -4
View File
@@ -42,10 +42,6 @@ int main(int argc, char *argv[])
int classes, map_points, map_levels;
float map_step, IoU_thresh, conf_thresh;
//read mAP parameters
readParams( config_filename, classes, map_points, map_levels, map_step,
IoU_thresh, conf_thresh, verbose);
if(argc > 1)
net = argv[1];
if(argc > 2)
@@ -62,6 +58,9 @@ int main(int argc, char *argv[])
if(!fileExist(labels_path))
FatalError("Wrong labels file path.");
//read mAP parameters
readParams( config_filename, classes, map_points, map_levels, map_step,
IoU_thresh, conf_thresh, verbose);
std::ofstream times;
if(write_res_on_file)
+787
View File
@@ -0,0 +1,787 @@
[net]
# Testing
#batch=1
#subdivisions=1
# Training
batch=16
subdivisions=1
width=512
height=512
channels=3
momentum=0.9
decay=0.0005
angle=0
saturation = 1.5
exposure = 1.5
hue=.1
learning_rate=0.001
burn_in=1000
max_batches = 500200
policy=steps
steps=400000,450000
scales=.1,.1
[convolutional]
batch_normalize=1
filters=32
size=3
stride=1
pad=1
activation=leaky
# Downsample
[convolutional]
batch_normalize=1
filters=64
size=3
stride=2
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=32
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=64
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
# Downsample
[convolutional]
batch_normalize=1
filters=128
size=3
stride=2
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=64
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=64
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
# Downsample
[convolutional]
batch_normalize=1
filters=256
size=3
stride=2
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
# Downsample
[convolutional]
batch_normalize=1
filters=512
size=3
stride=2
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
# Downsample
[convolutional]
batch_normalize=1
filters=1024
size=3
stride=2
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky
[shortcut]
from=-3
activation=linear
######################
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=1024
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=1024
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=1024
activation=leaky
[convolutional]
size=1
stride=1
pad=1
filters=24
activation=linear
[yolo]
mask = 6,7,8
anchors = 10.256,16.494, 11.724,18.558, 17.678,16.437, 25.619,14.985, 46.845,79.02, 58.643,81.204, 23.646,208.56, 30.837,211.57, 37.921,211.16
classes=3
num=9
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1
[route]
layers = -4
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[upsample]
stride=2
[route]
layers = -1, 61
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=512
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=512
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=512
activation=leaky
[convolutional]
size=1
stride=1
pad=1
filters=24
activation=linear
[yolo]
mask = 3,4,5
anchors = 10.256,16.494, 11.724,18.558, 17.678,16.437, 25.619,14.985, 46.845,79.02, 58.643,81.204, 23.646,208.56, 30.837,211.57, 37.921,211.16
classes=3
num=9
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1
[route]
layers = -4
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[upsample]
stride=2
[route]
layers = -1, 36
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=256
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=256
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
size=3
stride=1
pad=1
filters=256
activation=leaky
[convolutional]
size=1
stride=1
pad=1
filters=24
activation=linear
[yolo]
mask = 0,1,2
anchors = 10.256,16.494, 11.724,18.558, 17.678,16.437, 25.619,14.985, 46.845,79.02, 58.643,81.204, 23.646,208.56, 30.837,211.57, 37.921,211.16
classes=3
num=9
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1
+93
View File
@@ -0,0 +1,93 @@
#include<iostream>
#include<vector>
#include "tkdnn.h"
int main() {
// Network layout
tk::dnn::dataDim_t dim(1, 3, 512, 512, 1);
tk::dnn::Network net(dim);
// create yolo3 model
std::string bin_path = "../tests/yolo3_512tp";
// downloadWeightsifDoNotExist("../tests/yolo3_512tp/layers/input.bin", bin_path, );
int classes = 3;
tk::dnn::Yolo *yolo [3];
#include "models/Yolo3.h"
// fill classes names
for(int i=0; i<3; i++) {
yolo[i]->classesNames = {"Dent", "Wrinkle", "UnsealedFlaps"};
}
// Load input
dnnType *data;
dnnType *input_h;
readBinaryFile(input_bin, dim.tot(), &input_h, &data);
//print network model
net.print();
//convert network to tensorRT
tk::dnn::NetworkRT netRT(&net, "yolo3_512tp.rt");
// the network have 3 outputs
tk::dnn::dataDim_t out_dim[3];
for(int i=0; i<3; i++) out_dim[i] = yolo[i]->output_dim;
dnnType *cudnn_out[3], *rt_out[3];
tk::dnn::dataDim_t dim1 = dim; //input dim
printCenteredTitle(" CUDNN inference ", '=', 30); {
dim1.print();
TIMER_START
net.infer(dim1, data);
TIMER_STOP
dim1.print();
}
for(int i=0; i<3; i++) cudnn_out[i] = yolo[i]->dstData;
printCenteredTitle(" compute detections ", '=', 30);
TIMER_START
int ndets = 0;
tk::dnn::Yolo::detection *dets = tk::dnn::Yolo::allocateDetections(tk::dnn::Yolo::MAX_DETECTIONS, classes);
for(int i=0; i<3; i++) yolo[i]->computeDetections(dets, ndets, net.input_dim.w, net.input_dim.h, 0.5);
tk::dnn::Yolo::mergeDetections(dets, ndets, classes);
for(int j=0; j<ndets; j++) {
tk::dnn::Yolo::box b = dets[j].bbox;
int x0 = (b.x-b.w/2.);
int x1 = (b.x+b.w/2.);
int y0 = (b.y-b.h/2.);
int y1 = (b.y+b.h/2.);
int cl = 0;
for(int c = 0; c < classes; ++c){
float prob = dets[j].prob[c];
if(prob > 0)
cl = c;
}
std::cout<<cl<<": "<<x0<<" "<<y0<<" "<<x1<<" "<<y1<<"\n";
}
TIMER_STOP
tk::dnn::dataDim_t dim2 = dim;
printCenteredTitle(" TENSORRT inference ", '=', 30); {
dim2.print();
TIMER_START
netRT.infer(dim2, data);
TIMER_STOP
dim2.print();
}
for(int i=0; i<3; i++) rt_out[i] = (dnnType*)netRT.buffersRT[i+1];
for(int i=0; i<3; i++) {
printCenteredTitle((std::string(" YOLO ") + std::to_string(i) + " CHECK RESULTS ").c_str(), '=', 30);
dnnType *out, *out_h;
int odim = out_dim[i].tot();
readBinaryFile(output_bins[i], odim, &out_h, &out);
std::cout<<"CUDNN vs correct"; checkResult(odim, cudnn_out[i], out);
std::cout<<"TRT vs correct"; checkResult(odim, rt_out[i], out);
std::cout<<"CUDNN vs TRT "; checkResult(odim, cudnn_out[i], rt_out[i]);
}
return 0;
}
+123
View File
@@ -0,0 +1,123 @@
#include<iostream>
#include "tkdnn.h"
const char *input_bin = "../tests/yolo3_tiny512tp/layers/input.bin";
const char *c0_bin = "../tests/yolo3_tiny512tp/layers/c0.bin";
const char *c2_bin = "../tests/yolo3_tiny512tp/layers/c2.bin";
const char *c4_bin = "../tests/yolo3_tiny512tp/layers/c4.bin";
const char *c6_bin = "../tests/yolo3_tiny512tp/layers/c6.bin";
const char *c8_bin = "../tests/yolo3_tiny512tp/layers/c8.bin";
const char *c10_bin = "../tests/yolo3_tiny512tp/layers/c10.bin";
const char *c12_bin = "../tests/yolo3_tiny512tp/layers/c12.bin";
const char *c13_bin = "../tests/yolo3_tiny512tp/layers/c13.bin";
const char *c14_bin = "../tests/yolo3_tiny512tp/layers/c14.bin";
const char *c15_bin = "../tests/yolo3_tiny512tp/layers/c15.bin";
const char *c18_bin = "../tests/yolo3_tiny512tp/layers/c18.bin";
const char *c21_bin = "../tests/yolo3_tiny512tp/layers/c21.bin";
const char *c22_bin = "../tests/yolo3_tiny512tp/layers/c22.bin";
const char *g16_bin = "../tests/yolo3_tiny512tp/layers/g16.bin";
const char *g23_bin = "../tests/yolo3_tiny512tp/layers/g23.bin";
// const char *output_bin = "../tests/yolo3_tiny512tp/layers/output.bin";
const char *output_bin = "../tests/yolo3_tiny512tp/debug/layer23_out.bin";
int main() {
int classes = 3;
// Network layout
tk::dnn::dataDim_t dim(1, 3, 512, 512, 1);
tk::dnn::Network net(dim);
tk::dnn::Conv2d c0 (&net, 16, 3, 3, 1, 1, 1, 1, c0_bin, true);
tk::dnn::Activation a0 (&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Pooling p1 (&net, 2, 2, 2, 2, tk::dnn::POOLING_MAX);
tk::dnn::Conv2d c2 (&net, 32, 3, 3, 1, 1, 1, 1, c2_bin, true);
tk::dnn::Activation a2 (&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Pooling p3 (&net, 2, 2, 2, 2, tk::dnn::POOLING_MAX);
tk::dnn::Conv2d c4 (&net, 64, 3, 3, 1, 1, 1, 1, c4_bin, true);
tk::dnn::Activation a4 (&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Pooling p5 (&net, 2, 2, 2, 2, tk::dnn::POOLING_MAX);
tk::dnn::Conv2d c6 (&net, 128, 3, 3, 1, 1, 1, 1, c6_bin, true);
tk::dnn::Activation a6 (&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Pooling p7(&net, 2, 2, 2, 2, tk::dnn::POOLING_MAX);
tk::dnn::Conv2d c8(&net, 256, 3, 3, 1, 1, 1, 1, c8_bin, true);
tk::dnn::Activation a8(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Pooling p9(&net, 2, 2, 2, 2, tk::dnn::POOLING_MAX);
tk::dnn::Conv2d c10(&net, 512, 3, 3, 1, 1, 1, 1, c10_bin, true);
tk::dnn::Activation a10(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Pooling p11(&net, 2, 2, 1, 1,0,0, tk::dnn::POOLING_MAX, false, true);
tk::dnn::Conv2d c12(&net, 1024, 3, 3, 1, 1, 1, 1, c12_bin, true);
tk::dnn::Activation a12(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Conv2d c13(&net, 256, 1, 1, 1, 1, 0, 0, c13_bin, true);
tk::dnn::Activation a13(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Conv2d c14(&net, 512, 3, 3, 1, 1, 1, 1, c14_bin, true);
tk::dnn::Activation a14(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Conv2d c15(&net, 24, 1, 1, 1, 1, 0, 0, c15_bin, false);
tk::dnn::Yolo yolo0 (&net, classes, 2, g16_bin);
tk::dnn::Layer *m17_layers[1] = { &a13 };
tk::dnn::Route m17 (&net, m17_layers, 1);
tk::dnn::Conv2d c18(&net, 128, 1, 1, 1, 1, 0, 0, c18_bin, true);
tk::dnn::Activation a18(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Upsample u19 (&net, 2);
tk::dnn::Layer *m20_layers[2] = { &u19, &a8 };
tk::dnn::Route m20 (&net, m20_layers, 2);
tk::dnn::Conv2d c21(&net, 256, 3, 3, 1, 1, 1, 1, c21_bin, true);
tk::dnn::Activation a21(&net, tk::dnn::ACTIVATION_LEAKY);
tk::dnn::Conv2d c22(&net, 24, 1, 1, 1, 1, 0, 0, c22_bin, false);
tk::dnn::Yolo yolo1 (&net, classes, 2, g23_bin);
// Load input
dnnType *data;
dnnType *input_h;
readBinaryFile(input_bin, dim.tot(), &input_h, &data);
//print network model
net.print();
// convert network to tensorRT
tk::dnn::NetworkRT netRT(&net, "yolo3_tiny512tp.rt");
dnnType *out_data, *out_data2; // cudnn output, tensorRT output
tk::dnn::dataDim_t dim1 = dim; //input dim
printCenteredTitle(" CUDNN inference ", '=', 30); {
dim1.print();
TIMER_START
out_data = net.infer(dim1, data);
TIMER_STOP
dim1.print();
}
tk::dnn::dataDim_t dim2 = dim;
printCenteredTitle(" TENSORRT inference ", '=', 30); {
dim2.print();
TIMER_START
out_data2 = netRT.infer(dim2, data);
TIMER_STOP
dim2.print();
}
printCenteredTitle(" CHECK RESULTS ", '=', 30);
dnnType *out, *out_h;
int out_dim = net.getOutputDim().tot();
readBinaryFile(output_bin, out_dim, &out_h, &out);
std::cout<<"CUDNN vs correct"; checkResult(out_dim, out_data, out);
std::cout<<"TRT vs correct"; checkResult(out_dim, out_data2, out);
std::cout<<"CUDNN vs TRT "; checkResult(out_dim, out_data, out_data2);
return 0;
}
+182
View File
@@ -0,0 +1,182 @@
[net]
# Testing
batch=1
subdivisions=1
# Training
# batch=64
# subdivisions=2
width=512
height=512
channels=3
momentum=0.9
decay=0.0005
angle=0
saturation = 1.5
exposure = 1.5
hue=.1
learning_rate=0.001
burn_in=1000
max_batches = 500200
policy=steps
steps=400000,450000
scales=.1,.1
[convolutional]
batch_normalize=1
filters=16
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=32
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=64
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=1
[convolutional]
batch_normalize=1
filters=1024
size=3
stride=1
pad=1
activation=leaky
###########
[convolutional]
batch_normalize=1
filters=256
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[convolutional]
size=1
stride=1
pad=1
filters=24
activation=linear
[yolo]
mask = 3,4,5
anchors = 10.638,16.801, 13.183,19.091, 24.568,12.24, 54.462,77.421, 29.199,210.49, 37.495,212.21
classes=3
num=6
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1
[route]
layers = -4
[convolutional]
batch_normalize=1
filters=128
size=1
stride=1
pad=1
activation=leaky
[upsample]
stride=2
[route]
layers = -1, 8
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[convolutional]
size=1
stride=1
pad=1
filters=24
activation=linear
[yolo]
mask = 0,1,2
anchors = 10.638,16.801, 13.183,19.091, 24.568,12.24, 54.462,77.421, 29.199,210.49, 37.495,212.21
classes=3
num=6
jitter=.3
ignore_thresh = .7
truth_thresh = 1
random=1