Darknet: Open Source Neural Networks in C - Train a Classifier on CIFAR-10
阿新 • • 發佈：2018-11-08
Darknet: Open Source Neural Networks in C - Train a Classifier on CIFAR-10
Darknet is an open source neural network framework written in C and CUDA. It is fast, easy to install, and supports CPU and GPU computation. You can find the source on GitHub or you can read more about what Darknet can do right here:
https://github.com/pjreddie/darknet
Train a Classifier on CIFAR-10
https://pjreddie.com/darknet/train-cifar/
Learn how to train a classifier from scratch in Darknet.
This post will teach you how to train a classifier from scratch in Darknet. We’ll play with the CIFAR-10 dataset, a 10 class dataset of small images. Let’s get started!
https://www.cs.toronto.edu/~kriz/cifar.html
Install Darknet
If you don’t have installed already, do it:
git clone https://github.com/pjreddie/darknet
cd darknet
make
If it all worked, great! If something went wrong… um… try to fix it?
Get The Data
We’ll use my mirror of the CIFAR data because we want the pictures in image format. The original dataset comes in a binary format but I want this tutorial to generalize to whatever dataset you want to work with, so we’ll do it with images instead.
https://pjreddie.com/projects/cifar-10-dataset-mirror/
Let’s put the data in the data/ folder. To do that run:
cd data
wget https://pjreddie.com/media/files/cifar.tgz
tar xzf cifar.tgz
[email protected]:~/eclipse-darknet/darknet_181107/data$ wget https://pjreddie.com/media/files/cifar.tgz
--2018-11-07 19:54:24-- https://pjreddie.com/media/files/cifar.tgz
Resolving pjreddie.com (pjreddie.com)... 128.208.3.39
Connecting to pjreddie.com (pjreddie.com)|128.208.3.39|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 168584360 (161M) [application/octet-stream]
Saving to: ‘cifar.tgz’
cifar.tgz 100%[==========================================================>] 160.77M 3.45MB/s in 62s
2018-11-07 19:55:27 (2.60 MB/s) - ‘cifar.tgz’ saved [168584360/168584360]
[email protected]:~/eclipse-darknet/darknet_181107/data$ tar xzf cifar.tgz
[email protected]:~/eclipse-darknet/darknet_181107/data$
Now let’s look at what we have:
ls cifar
lists two directories with our data, train and test, and a file with the labels, labels.txt. You can look at labels.txt if you want and see what kinds of classes we will learn:
cat cifar/labels.txt
We also need to generate our paths files. These files will hold all the paths to the training and validation (or in this case testing) data. To do that, we’ll cd into our cifar directory, find all of the images, and write them to a file, then return to our base darknet directory.
cd cifar
find `pwd`/train -name \*.png > train.list
find `pwd`/test -name \*.png > test.list
cd ../..
[email protected]:~/eclipse-darknet/darknet_181107/data$ ls cifar
labels.txt test train
[email protected]:~/eclipse-darknet/darknet_181107/data$
[email protected]:~/eclipse-darknet/darknet_181107/data$ cat cifar/labels.txt
airplane
automobile
bird
cat
deer
dog
frog
horse
ship
truck
[email protected]:~/eclipse-darknet/darknet_181107/data$
[email protected]:~/eclipse-darknet/darknet_181107/data$ cd cifar/
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$ ll
total 2176
drwxr-xr-x 4 strong strong 4096 Nov 19 2016 ./
drwxrwxr-x 4 strong strong 4096 Nov 8 10:02 ../
-rw-r--r-- 1 strong strong 60 Nov 19 2016 labels.txt
drwxr-xr-x 2 strong strong 315392 Nov 19 2016 test/
drwxr-xr-x 2 strong strong 1892352 Nov 19 2016 train/
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$ find `pwd`/train -name \*.png > train.list
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$ find `pwd`/test -name \*.png > test.list
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$ cd ../..
[email protected]:~/eclipse-darknet/darknet_181107$
[email protected]:~/eclipse-darknet/darknet_181107$
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$ ll
total 6660
drwxr-xr-x 4 strong strong 4096 Nov 8 10:05 ./
drwxrwxr-x 4 strong strong 4096 Nov 8 10:02 ../
-rw-r--r-- 1 strong strong 60 Nov 19 2016 labels.txt
drwxr-xr-x 2 strong strong 315392 Nov 19 2016 test/
-rw-rw-r-- 1 strong strong 748890 Nov 8 10:05 test.list
drwxr-xr-x 2 strong strong 1892352 Nov 19 2016 train/
-rw-rw-r-- 1 strong strong 3838890 Nov 8 10:05 train.list
[email protected]:~/eclipse-darknet/darknet_181107/data/cifar$
Make A Dataset Config File
We have to give Darknet some metadata about CIFAR-10. Using your favorite editor, open up a new file in the cfg/ directory called cfg/cifar.data. In it you should have something like this:
classes=10
train = data/cifar/train.list
valid = data/cifar/test.list
labels = data/cifar/labels.txt
backup = backup/
top=2
- classes=10: the dataset has 10 different classes
- train = …: where to find the list of training files
- valid = …: where to find the list of validation files
- labels = …: where to find the list of possible classes
- backup = …: where to save backup weight files during training
- top = 2: calculate top-n accuracy at test time (in addition to top-1)
Make A Network Config File!
We need a network to train. In your cfg directory make another file called cfg/cifar_small.cfg. In it put this network:
[net]
batch=128
subdivisions=1
height=28
width=28
channels=3
max_crop=32
min_crop=32
hue=.1
saturation=.75
exposure=.75
learning_rate=0.1
policy=poly
power=4
max_batches = 5000
momentum=0.9
decay=0.0005
[convolutional]
batch_normalize=1
filters=32
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=16
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=64
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=32
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=64
size=1
stride=1
pad=1
activation=leaky
[convolutional]
filters=10
size=1
stride=1
pad=1
activation=leaky
[avgpool]
[softmax]
It’s a really small network so it won’t work very well but it’s good for this example. The network is just 4 convolutional layers and 2 maxpooling layers.
The last convolutional layer has 10 filters because we have 10 classes. It outputs a 7 x 7 x 10 image. We just want 10 predictions total so we use an average pooling layer to take the average across the image for each channel. This will give us our 10 predictions. We use a softmax to convert the predictions into a probability distribution. This layer also calculates our error as cross-entropy loss.
https://en.wikipedia.org/wiki/Softmax_function
Train The Model
Now we just have to run the training code!
terminal
./darknet classifier train cfg/cifar.data cfg/cifar_small.cfg
argument
classifier train ./train_cfg/cifar.data ./train_cfg/cifar_small.cfg
And watch it go!
You are just telling Darknet you want to train a classifier using the following data and network cfg files. On a CPU training may take an hour or more, even for this small network. If you have a GPU you should enable GPU training by following these instructions.
https://pjreddie.com/darknet/install/
Restarting Training
If you stop training you can always restart it using one of the model checkpoints it saves along the way:
terminal
./darknet classifier train cfg/cifar.data cfg/cifar_small.cfg backup/cifar_small.backup
argument
classifier train ./train_cfg/cifar.data ./train_cfg/cifar_small.cfg ./backup/cifar_small.backup
Validate The Model
Now we have to see how well our model is doing. We can calculate top-1 and top-2 validation accuracy using the valid command. We can run validation on a backup, the final weights file, or any saved epoch weight file:
terminal
./darknet classifier valid cfg/cifar.data cfg/cifar_small.cfg backup/cifar_small.backup
argument
classifier valid ./train_cfg/cifar.data ./train_cfg/cifar_small.cfg ./backup/cifar_small.backup
You will see a bunch of scrolling numbers that tell you your accuracy.
1. Makefile
GPU=1
CUDNN=1
OPENCV=0
OPENMP=0
DEBUG=0
2. Program Arguments
right-click on the darknet_181107 -> Properties -> Run/Debug Settings -> New -> C/C++ Application -> OK
Arguments
classifier train ./train_cfg/cifar.data ./train_cfg/cifar_small.cfg
3. Run/Debug Settings
Run darknet_181107-cifar_small
/home/strong/eclipse-darknet/darknet_181107/train_cfg/cifar_small.cfg
[net]
batch=128
subdivisions=1
height=28
width=28
channels=3
max_crop=32
min_crop=32
hue=.1
saturation=.75
exposure=.75
learning_rate=0.1
policy=poly
power=4
max_batches = 5000
momentum=0.9
decay=0.0005
[convolutional]
batch_normalize=1
filters=32
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=16
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=64
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[convolutional]
batch_normalize=1
filters=32
size=1
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=64
size=1
stride=1
pad=1
activation=leaky
[convolutional]
filters=10
size=1
stride=1
pad=1
activation=leaky
[avgpool]
[softmax]
/home/strong/eclipse-darknet/darknet_181107/train_cfg/cifar.cfg
[net]
batch=128
subdivisions=1
height=28
width=28
channels=3
max_crop=32
min_crop=32
hue=.1
saturation=.75
exposure=.75
learning_rate=0.4
policy=poly
power=4
max_batches = 5000
momentum=0.9
decay=0.0005
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=128
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[dropout]
probability=.5
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=256
size=3
stride=1
pad=1
activation=leaky
[maxpool]
size=2
stride=2
[dropout]
probability=.5
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[convolutional]
batch_normalize=1
filters=512
size=3
stride=1
pad=1
activation=leaky
[dropout]
probability=.5
[convolutional]
filters=10
size=1
stride=1
pad=1
activation=leaky
[avgpool]
[softmax]
groups=1
訓練 log
59985, 153.562: 0.345311, 0.304563 avg, 0.000000 rate, 0.005881 seconds, 7678080 images
Loaded: 0.000016 seconds
59986, 153.564: 0.222410, 0.296348 avg, 0.000000 rate, 0.006304 seconds, 7678208 images
Loaded: 0.000022 seconds
59987, 153.567: 0.188696, 0.285583 avg, 0.000000 rate, 0.005702 seconds, 7678336 images
Loaded: 0.000014 seconds
59988, 153.569: 0.232374, 0.280262 avg, 0.000000 rate, 0.006430 seconds, 7678464 images
Loaded: 0.000019 seconds
59989, 153.572: 0.247903, 0.277026 avg, 0.000000 rate, 0.005785 seconds, 7678592 images
Loaded: 0.000047 seconds
59990, 153.574: 0.291492, 0.278473 avg, 0.000000 rate, 0.005631 seconds, 7678720 images
Loaded: 0.000014 seconds
59991, 153.577: 0.250590, 0.275684 avg, 0.000000 rate, 0.005977 seconds, 7678848 images
Loaded: 0.000013 seconds
59992, 153.580: 0.168354, 0.264951 avg, 0.000000 rate, 0.005793 seconds, 7678976 images
Loaded: 0.000013 seconds
59993, 153.582: 0.331874, 0.271644 avg, 0.000000 rate, 0.005759 seconds, 7679104 images
Loaded: 0.000014 seconds
59994, 153.585: 0.235558, 0.268035 avg, 0.000000 rate, 0.005812 seconds, 7679232 images
Loaded: 0.000023 seconds
59995, 153.587: 0.288601, 0.270092 avg, 0.000000 rate, 0.006358 seconds, 7679360 images
Loaded: 0.000020 seconds
59996, 153.590: 0.398588, 0.282941 avg, 0.000000 rate, 0.005774 seconds, 7679488 images
Loaded: 0.000015 seconds
59997, 153.592: 0.318306, 0.286478 avg, 0.000000 rate, 0.006111 seconds, 7679616 images
Loaded: 0.000027 seconds
59998, 153.595: 0.299280, 0.287758 avg, 0.000000 rate, 0.008606 seconds, 7679744 images
Loaded: 0.000025 seconds
59999, 153.597: 0.335652, 0.292547 avg, 0.000000 rate, 0.006158 seconds, 7679872 images
Loaded: 0.000016 seconds
60000, 153.600: 0.277184, 0.291011 avg, 0.000000 rate, 0.005788 seconds, 7680000 images
Saving weights to backup//cifar_small.backup
Saving weights to backup//cifar_small.weights
訓練模型
4. Validate The Model
./darknet classifier valid cfg/cifar.data cfg/cifar_small.cfg backup/cifar_small.weights
/home/strong/eclipse-darknet/darknet_181107/cfg/cifar_small.cfg
[net]
batch=1
subdivisions=1
height=28
width=28
channels=3
max_crop=32
min_crop=32
hue=.1
saturation=.75
exposure=.75
learning_rate=0.1
policy=poly
power=4
max_batches = 5000
momentum=0.9
decay=0.0005
./darknet classifier valid cfg/cifar.data cfg/cifar_small.cfg backup/cifar_small.weights
[email protected]:~/eclipse-darknet/darknet_181107$ ./darknet classifier valid cfg/cifar.data cfg/cifar_small.cfg backup/cifar_small.weights
layer filters size input output
0 conv 32 3 x 3 / 1 28 x 28 x 3 -> 28 x 28 x 32 0.001 BFLOPs
1 max 2 x 2 / 2 28 x 28 x 32 -> 14 x 14 x 32
2 conv 16 1 x 1 / 1 14 x 14 x 32 -> 14 x 14 x 16 0.000 BFLOPs
3 conv 64 3 x 3 / 1 14 x 14 x 16 -> 14 x 14 x 64 0.004 BFLOPs
4 max 2 x 2 / 2 14 x 14 x 64 -> 7 x 7 x 64
5 conv 32 1 x 1 / 1 7 x 7 x 64 -> 7 x 7 x 32 0.000 BFLOPs
6 conv 128 3 x 3 / 1 7 x 7 x 32 -> 7 x 7 x 128 0.004 BFLOPs
7 conv 64 1 x 1 / 1 7 x 7 x 128 -> 7 x 7 x 64 0.001 BFLOPs
8 conv 10 1 x 1 / 1 7 x 7 x 64 -> 7 x 7 x 10 0.000 BFLOPs
9 avg 7 x 7 x 10 -> 10
10 softmax 10
Loading weights from backup/cifar_small.weights...Done!
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6251_automobile.png, 1, 0.279962, 0.092464,
0: top 1: 0.000000, top 2: 0.000000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9798_dog.png, 5, 0.000000, 0.000000,
1: top 1: 0.500000, top 2: 0.500000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3942_automobile.png, 1, 0.000000, 0.999985,
2: top 1: 0.666667, top 2: 0.666667
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6197_deer.png, 4, 0.006338, 0.000819,
3: top 1: 0.500000, top 2: 0.750000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1828_horse.png, 7, 0.005472, 0.000199,
4: top 1: 0.600000, top 2: 0.800000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6073_bird.png, 2, 0.002952, 0.000015,
5: top 1: 0.666667, top 2: 0.833333
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/609_truck.png, 9, 0.000000, 0.000000,
6: top 1: 0.714286, top 2: 0.857143
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/692_bird.png, 2, 0.727276, 0.000020,
7: top 1: 0.625000, top 2: 0.875000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5737_truck.png, 9, 0.000014, 0.024781,
8: top 1: 0.666667, top 2: 0.888889
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/325_automobile.png, 1, 0.000013, 0.995964,
9: top 1: 0.700000, top 2: 0.900000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4526_cat.png, 3, 0.000021, 0.000007,
10: top 1: 0.636364, top 2: 0.909091
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9568_ship.png, 8, 0.000452, 0.000001,
11: top 1: 0.666667, top 2: 0.916667
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/297_airplane.png, 0, 0.945728, 0.000021,
12: top 1: 0.692308, top 2: 0.923077
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6734_ship.png, 8, 0.001546, 0.000000,
13: top 1: 0.714286, top 2: 0.928571
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/311_frog.png, 6, 0.000000, 0.000000,
14: top 1: 0.733333, top 2: 0.933333
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9366_bird.png, 2, 0.017653, 0.001454,
15: top 1: 0.687500, top 2: 0.937500
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1527_dog.png, 5, 0.008312, 0.001587,
16: top 1: 0.647059, top 2: 0.882353
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5227_deer.png, 4, 0.000062, 0.000021,
17: top 1: 0.611111, top 2: 0.833333
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4738_dog.png, 5, 0.000267, 0.000009,
18: top 1: 0.631579, top 2: 0.842105
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4120_deer.png, 4, 0.000007, 0.000001,
19: top 1: 0.650000, top 2: 0.850000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6726_cat.png, 3, 0.000016, 0.000003,
20: top 1: 0.619048, top 2: 0.809524
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8411_ship.png, 8, 0.002263, 0.000005,
21: top 1: 0.636364, top 2: 0.818182
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7018_airplane.png, 0, 0.814647, 0.000041,
22: top 1: 0.652174, top 2: 0.826087
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/288_horse.png, 7, 0.001681, 0.000030,
23: top 1: 0.666667, top 2: 0.833333
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4415_airplane.png, 0, 0.973400, 0.000043,
24: top 1: 0.680000, top 2: 0.840000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3077_automobile.png, 1, 0.000000, 1.000000,
25: top 1: 0.692308, top 2: 0.846154
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3676_truck.png, 9, 0.403864, 0.039076,
26: top 1: 0.666667, top 2: 0.851852
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2555_horse.png, 7, 0.000005, 0.000000,
27: top 1: 0.678571, top 2: 0.857143
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7922_ship.png, 8, 0.000743, 0.000008,
28: top 1: 0.689655, top 2: 0.862069
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7965_horse.png, 7, 0.000000, 0.000000,
29: top 1: 0.700000, top 2: 0.866667
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2582_dog.png, 5, 0.000000, 0.000000,
30: top 1: 0.709677, top 2: 0.870968
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9110_automobile.png, 1, 0.000431, 0.910856,
31: top 1: 0.718750, top 2: 0.875000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1951_horse.png, 7, 0.000002, 0.000001,
32: top 1: 0.727273, top 2: 0.878788
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7101_frog.png, 6, 0.000000, 0.000000,
33: top 1: 0.735294, top 2: 0.882353
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5935_frog.png, 6, 0.000000, 0.000000,
34: top 1: 0.742857, top 2: 0.885714
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1453_dog.png, 5, 0.001009, 0.000870,
35: top 1: 0.750000, top 2: 0.888889
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5025_dog.png, 5, 0.000004, 0.000000,
36: top 1: 0.756757, top 2: 0.891892
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6214_dog.png, 5, 0.000004, 0.000001,
37: top 1: 0.763158, top 2: 0.894737
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3472_horse.png, 7, 0.000044, 0.000032,
38: top 1: 0.769231, top 2: 0.897436
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/712_ship.png, 8, 0.000019, 0.000000,
39: top 1: 0.775000, top 2: 0.900000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2885_dog.png, 5, 0.000000, 0.000000,
40: top 1: 0.780488, top 2: 0.902439
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3638_truck.png, 9, 0.000001, 0.055196,
41: top 1: 0.785714, top 2: 0.904762
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5073_ship.png, 8, 0.000233, 0.000084,
42: top 1: 0.790698, top 2: 0.906977
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2860_dog.png, 5, 0.000002, 0.000000,
43: top 1: 0.795455, top 2: 0.909091
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5266_deer.png, 4, 0.012169, 0.000558,
44: top 1: 0.800000, top 2: 0.911111
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5749_dog.png, 5, 0.000067, 0.000000,
45: top 1: 0.804348, top 2: 0.913043
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/382_airplane.png, 0, 0.999153, 0.000001,
46: top 1: 0.808511, top 2: 0.914894
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5852_cat.png, 3, 0.002267, 0.000229,
47: top 1: 0.812500, top 2: 0.916667
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6395_ship.png, 8, 0.036328, 0.000519,
48: top 1: 0.816327, top 2: 0.918367
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5418_ship.png, 8, 0.001279, 0.000325,
49: top 1: 0.820000, top 2: 0.920000
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4289_horse.png, 7, 0.002492, 0.000020,
50: top 1: 0.823529, top 2: 0.921569
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3581_deer.png, 4, 0.000031, 0.000012,
51: top 1: 0.826923, top 2: 0.923077
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2821_airplane.png, 0, 0.877437, 0.001708,
52: top 1: 0.830189, top 2: 0.924528
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4999_automobile.png, 1, 0.000042, 0.999080,
53: top 1: 0.833333, top 2: 0.925926
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5490_cat.png, 3, 0.000421, 0.000187,
54: top 1: 0.836364, top 2: 0.927273
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2575_dog.png, 5, 0.000000, 0.000000,
55: top 1: 0.839286, top 2: 0.928571
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/831_cat.png, 3, 0.000073, 0.001551,
56: top 1: 0.842105, top 2: 0.929825
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3752_cat.png, 3, 0.000514, 0.072850,
57: top 1: 0.827586, top 2: 0.913793
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6493_automobile.png, 1, 0.000303, 0.998797,
58: top 1: 0.830508, top 2: 0.915254
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9041_dog.png, 5, 0.000791, 0.001027,
59: top 1: 0.816667, top 2: 0.916667
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8301_dog.png, 5, 0.000008, 0.000323,
60: top 1: 0.819672, top 2: 0.918033
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8361_ship.png, 8, 0.003861, 0.044749,
61: top 1: 0.822581, top 2: 0.919355
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6716_ship.png, 8, 0.000835, 0.000001,
62: top 1: 0.825397, top 2: 0.920635
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5455_truck.png, 9, 0.006370, 0.930129,
63: top 1: 0.812500, top 2: 0.906250
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/171_truck.png, 9, 0.637036, 0.004199,
64: top 1: 0.800000, top 2: 0.907692
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/946_airplane.png, 0, 0.906047, 0.000008,
65: top 1: 0.803030, top 2: 0.909091
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1642_dog.png, 5, 0.000929, 0.000092,
66: top 1: 0.805970, top 2: 0.910448
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4680_truck.png, 9, 0.000000, 0.000120,
67: top 1: 0.808824, top 2: 0.911765
......
......
......
9877: top 1: 0.852399, top 2: 0.943713
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6753_deer.png, 4, 0.005128, 0.001055,
9878: top 1: 0.852313, top 2: 0.943618
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4169_cat.png, 3, 0.001398, 0.001112,
9879: top 1: 0.852328, top 2: 0.943623
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4067_automobile.png, 1, 0.000006, 0.999239,
9880: top 1: 0.852343, top 2: 0.943629
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/896_cat.png, 3, 0.000003, 0.000010,
9881: top 1: 0.852358, top 2: 0.943635
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9817_truck.png, 9, 0.000234, 0.571874,
9882: top 1: 0.852272, top 2: 0.943641
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7410_deer.png, 4, 0.022779, 0.002281,
9883: top 1: 0.852287, top 2: 0.943646
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/659_automobile.png, 1, 0.000000, 0.999293,
9884: top 1: 0.852301, top 2: 0.943652
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4979_truck.png, 9, 0.000000, 0.034305,
9885: top 1: 0.852316, top 2: 0.943658
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/70_bird.png, 2, 0.053408, 0.000952,
9886: top 1: 0.852331, top 2: 0.943663
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7079_horse.png, 7, 0.000030, 0.000000,
9887: top 1: 0.852346, top 2: 0.943669
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3926_truck.png, 9, 0.000000, 0.000002,
9888: top 1: 0.852361, top 2: 0.943675
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2539_ship.png, 8, 0.001803, 0.000901,
9889: top 1: 0.852376, top 2: 0.943680
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3609_truck.png, 9, 0.000017, 0.000133,
9890: top 1: 0.852391, top 2: 0.943686
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8332_truck.png, 9, 0.000022, 0.000261,
9891: top 1: 0.852406, top 2: 0.943692
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7186_bird.png, 2, 0.002604, 0.000044,
9892: top 1: 0.852421, top 2: 0.943698
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6172_automobile.png, 1, 0.001470, 0.980129,
9893: top 1: 0.852436, top 2: 0.943703
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3510_deer.png, 4, 0.000070, 0.000099,
9894: top 1: 0.852451, top 2: 0.943709
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3763_ship.png, 8, 0.000571, 0.000159,
9895: top 1: 0.852466, top 2: 0.943715
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7680_cat.png, 3, 0.000096, 0.000024,
9896: top 1: 0.852380, top 2: 0.943619
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6402_ship.png, 8, 0.057762, 0.000408,
9897: top 1: 0.852394, top 2: 0.943625
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9894_automobile.png, 1, 0.576769, 0.399480,
9898: top 1: 0.852308, top 2: 0.943631
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1170_horse.png, 7, 0.000000, 0.000000,
9899: top 1: 0.852323, top 2: 0.943636
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3703_frog.png, 6, 0.000015, 0.000006,
9900: top 1: 0.852338, top 2: 0.943642
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6038_deer.png, 4, 0.001107, 0.001192,
9901: top 1: 0.852353, top 2: 0.943648
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9639_bird.png, 2, 0.007043, 0.000009,
9902: top 1: 0.852368, top 2: 0.943653
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8104_horse.png, 7, 0.000078, 0.000010,
9903: top 1: 0.852383, top 2: 0.943659
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4725_truck.png, 9, 0.000004, 0.439126,
9904: top 1: 0.852398, top 2: 0.943665
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9302_cat.png, 3, 0.003163, 0.115956,
9905: top 1: 0.852413, top 2: 0.943671
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3495_bird.png, 2, 0.001203, 0.000024,
9906: top 1: 0.852327, top 2: 0.943575
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/973_automobile.png, 1, 0.000003, 0.994684,
9907: top 1: 0.852342, top 2: 0.943581
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9884_truck.png, 9, 0.000012, 0.118405,
9908: top 1: 0.852356, top 2: 0.943587
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5923_airplane.png, 0, 0.542622, 0.414726,
9909: top 1: 0.852371, top 2: 0.943592
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/864_airplane.png, 0, 0.999562, 0.000009,
9910: top 1: 0.852386, top 2: 0.943598
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/596_horse.png, 7, 0.000303, 0.000000,
9911: top 1: 0.852401, top 2: 0.943604
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9811_automobile.png, 1, 0.000004, 0.992264,
9912: top 1: 0.852416, top 2: 0.943609
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2414_dog.png, 5, 0.000097, 0.000005,
9913: top 1: 0.852431, top 2: 0.943615
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8742_truck.png, 9, 0.029223, 0.007781,
9914: top 1: 0.852446, top 2: 0.943621
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1778_ship.png, 8, 0.000715, 0.000025,
9915: top 1: 0.852461, top 2: 0.943626
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4413_frog.png, 6, 0.000008, 0.000005,
9916: top 1: 0.852476, top 2: 0.943632
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1373_cat.png, 3, 0.000083, 0.000003,
9917: top 1: 0.852490, top 2: 0.943638
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/447_airplane.png, 0, 0.799862, 0.000049,
9918: top 1: 0.852505, top 2: 0.943644
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8937_horse.png, 7, 0.000443, 0.000000,
9919: top 1: 0.852520, top 2: 0.943649
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9632_bird.png, 2, 0.000005, 0.000000,
9920: top 1: 0.852535, top 2: 0.943655
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8726_truck.png, 9, 0.000037, 0.000092,
9921: top 1: 0.852550, top 2: 0.943661
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3102_airplane.png, 0, 0.961033, 0.000016,
9922: top 1: 0.852565, top 2: 0.943666
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4246_dog.png, 5, 0.000006, 0.000002,
9923: top 1: 0.852580, top 2: 0.943672
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/873_bird.png, 2, 0.000028, 0.000000,
9924: top 1: 0.852594, top 2: 0.943678
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3012_automobile.png, 1, 0.047959, 0.891379,
9925: top 1: 0.852609, top 2: 0.943683
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6137_bird.png, 2, 0.011992, 0.000195,
9926: top 1: 0.852624, top 2: 0.943689
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7441_ship.png, 8, 0.000643, 0.000208,
9927: top 1: 0.852639, top 2: 0.943695
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4468_truck.png, 9, 0.004604, 0.012142,
9928: top 1: 0.852553, top 2: 0.943700
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1141_automobile.png, 1, 0.000000, 0.999708,
9929: top 1: 0.852568, top 2: 0.943706
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9511_deer.png, 4, 0.000000, 0.000007,
9930: top 1: 0.852583, top 2: 0.943712
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9987_airplane.png, 0, 0.992385, 0.000015,
9931: top 1: 0.852598, top 2: 0.943717
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5781_airplane.png, 0, 0.577285, 0.162381,
9932: top 1: 0.852612, top 2: 0.943723
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8861_deer.png, 4, 0.032639, 0.000428,
9933: top 1: 0.852627, top 2: 0.943729
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7323_cat.png, 3, 0.000069, 0.000009,
9934: top 1: 0.852642, top 2: 0.943734
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4489_horse.png, 7, 0.000004, 0.000000,
9935: top 1: 0.852657, top 2: 0.943740
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6826_frog.png, 6, 0.002698, 0.000597,
9936: top 1: 0.852672, top 2: 0.943746
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1823_airplane.png, 0, 0.171337, 0.001239,
9937: top 1: 0.852586, top 2: 0.943651
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4767_frog.png, 6, 0.000177, 0.000003,
9938: top 1: 0.852601, top 2: 0.943656
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2154_dog.png, 5, 0.001858, 0.004266,
9939: top 1: 0.852515, top 2: 0.943561
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5877_truck.png, 9, 0.000000, 0.000170,
9940: top 1: 0.852530, top 2: 0.943567
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5241_deer.png, 4, 0.000248, 0.000038,
9941: top 1: 0.852545, top 2: 0.943573
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5568_horse.png, 7, 0.000000, 0.000000,
9942: top 1: 0.852560, top 2: 0.943578
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8933_frog.png, 6, 0.000390, 0.000618,
9943: top 1: 0.852574, top 2: 0.943584
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3610_dog.png, 5, 0.000016, 0.000054,
9944: top 1: 0.852589, top 2: 0.943590
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5827_airplane.png, 0, 0.991940, 0.000001,
9945: top 1: 0.852604, top 2: 0.943595
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8920_frog.png, 6, 0.000000, 0.000000,
9946: top 1: 0.852619, top 2: 0.943601
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/178_dog.png, 5, 0.000002, 0.000002,
9947: top 1: 0.852533, top 2: 0.943607
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2585_cat.png, 3, 0.000264, 0.000004,
9948: top 1: 0.852548, top 2: 0.943612
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1492_truck.png, 9, 0.000088, 0.151010,
9949: top 1: 0.852462, top 2: 0.943518
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5876_frog.png, 6, 0.000485, 0.001064,
9950: top 1: 0.852477, top 2: 0.943523
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/231_automobile.png, 1, 0.000000, 0.999544,
9951: top 1: 0.852492, top 2: 0.943529
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5497_cat.png, 3, 0.000096, 0.000000,
9952: top 1: 0.852507, top 2: 0.943535
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6061_airplane.png, 0, 0.994684, 0.000114,
9953: top 1: 0.852522, top 2: 0.943540
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1471_ship.png, 8, 0.021756, 0.006553,
9954: top 1: 0.852536, top 2: 0.943546
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1329_dog.png, 5, 0.000205, 0.000068,
9955: top 1: 0.852551, top 2: 0.943552
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3146_bird.png, 2, 0.259927, 0.000991,
9956: top 1: 0.852466, top 2: 0.943457
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5212_dog.png, 5, 0.010953, 0.001141,
9957: top 1: 0.852480, top 2: 0.943463
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5734_frog.png, 6, 0.000047, 0.000043,
9958: top 1: 0.852495, top 2: 0.943468
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7503_ship.png, 8, 0.644954, 0.051710,
9959: top 1: 0.852410, top 2: 0.943474
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3483_airplane.png, 0, 0.999005, 0.000000,
9960: top 1: 0.852424, top 2: 0.943480
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4884_deer.png, 4, 0.000045, 0.000005,
9961: top 1: 0.852439, top 2: 0.943485
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6352_bird.png, 2, 0.000019, 0.000018,
9962: top 1: 0.852454, top 2: 0.943491
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4914_cat.png, 3, 0.000194, 0.000002,
9963: top 1: 0.852469, top 2: 0.943497
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1563_truck.png, 9, 0.000069, 0.000049,
9964: top 1: 0.852484, top 2: 0.943502
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4660_horse.png, 7, 0.003685, 0.000231,
9965: top 1: 0.852398, top 2: 0.943408
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1728_airplane.png, 0, 0.839480, 0.000329,
9966: top 1: 0.852413, top 2: 0.943413
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/8087_bird.png, 2, 0.002026, 0.000059,
9967: top 1: 0.852428, top 2: 0.943419
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2724_airplane.png, 0, 0.999640, 0.000009,
9968: top 1: 0.852443, top 2: 0.943425
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2247_dog.png, 5, 0.000000, 0.000000,
9969: top 1: 0.852457, top 2: 0.943430
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2481_horse.png, 7, 0.002196, 0.000149,
9970: top 1: 0.852472, top 2: 0.943436
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9943_horse.png, 7, 0.114012, 0.002700,
9971: top 1: 0.852487, top 2: 0.943442
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5295_airplane.png, 0, 0.993299, 0.000114,
9972: top 1: 0.852502, top 2: 0.943447
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4492_ship.png, 8, 0.000560, 0.000430,
9973: top 1: 0.852517, top 2: 0.943453
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/958_horse.png, 7, 0.000137, 0.000001,
9974: top 1: 0.852531, top 2: 0.943459
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/4827_truck.png, 9, 0.000005, 0.000009,
9975: top 1: 0.852546, top 2: 0.943464
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2760_bird.png, 2, 0.000104, 0.000025,
9976: top 1: 0.852461, top 2: 0.943470
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/521_horse.png, 7, 0.000084, 0.000000,
9977: top 1: 0.852475, top 2: 0.943476
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5437_airplane.png, 0, 0.996226, 0.000003,
9978: top 1: 0.852490, top 2: 0.943481
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1937_automobile.png, 1, 0.086177, 0.036568,
9979: top 1: 0.852405, top 2: 0.943387
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9683_frog.png, 6, 0.000055, 0.000663,
9980: top 1: 0.852420, top 2: 0.943392
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3345_horse.png, 7, 0.000000, 0.000000,
9981: top 1: 0.852434, top 2: 0.943398
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5817_cat.png, 3, 0.000059, 0.000001,
9982: top 1: 0.852449, top 2: 0.943404
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/6230_bird.png, 2, 0.000002, 0.000000,
9983: top 1: 0.852464, top 2: 0.943409
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2591_ship.png, 8, 0.000404, 0.000002,
9984: top 1: 0.852479, top 2: 0.943415
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9026_truck.png, 9, 0.086814, 0.005917,
9985: top 1: 0.852493, top 2: 0.943421
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9699_airplane.png, 0, 0.929331, 0.001589,
9986: top 1: 0.852508, top 2: 0.943426
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2863_frog.png, 6, 0.000001, 0.000007,
9987: top 1: 0.852523, top 2: 0.943432
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3409_truck.png, 9, 0.000000, 0.000005,
9988: top 1: 0.852538, top 2: 0.943438
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/3948_automobile.png, 1, 0.000570, 0.677844,
9989: top 1: 0.852553, top 2: 0.943443
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/5545_frog.png, 6, 0.000001, 0.000003,
9990: top 1: 0.852567, top 2: 0.943449
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/548_bird.png, 2, 0.000315, 0.000128,
9991: top 1: 0.852582, top 2: 0.943455
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9409_truck.png, 9, 0.000016, 0.000042,
9992: top 1: 0.852597, top 2: 0.943460
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/2836_cat.png, 3, 0.000004, 0.000082,
9993: top 1: 0.852612, top 2: 0.943466
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7377_cat.png, 3, 0.000009, 0.000002,
9994: top 1: 0.852626, top 2: 0.943472
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/7341_deer.png, 4, 0.000048, 0.000129,
9995: top 1: 0.852641, top 2: 0.943477
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9840_deer.png, 4, 0.000131, 0.000013,
9996: top 1: 0.852556, top 2: 0.943483
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9584_cat.png, 3, 0.000000, 0.000000,
9997: top 1: 0.852571, top 2: 0.943489
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/1751_automobile.png, 1, 0.000000, 0.999160,
9998: top 1: 0.852585, top 2: 0.943494
/home/strong/eclipse-darknet/darknet_181107/data/cifar/test/9359_frog.png, 6, 0.000131, 0.000000,
9999: top 1: 0.852600, top 2: 0.943500
[email protected]:~/eclipse-darknet/darknet_181107$
Wordbook
you only look once，YOLO
Visual Object Classes，VOC
Pattern Analysis, Statistical Modelling and Computational Learning，PASCAL
mean Average Precision，mAP：平均精度均值
floating point operations per second，FLOPS
frame rate or frame frequency, frames per second，FPS
hertz，Hz
billion，Bn
operations，Ops
configuration，cfg
ImageNet Large Scale Visual Recognition Challenge，ILSVRC
Microsoft Common Objects in Context，MS COCO
metadata ['metədeɪtə]：n. 元資料
probability distribution：概率分佈