深度学习之利用TensorFlow实现简单的全连接层网络（MNIST数据集）

最新推荐文章于 2024-05-14 10:08:47 发布
最新推荐文章于 2024-05-14 10:08:47 发布
阅读量906 收藏 4
点赞数 1
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/ctrigger/article/details/100060474
版权

Tensorflow是一个基于数据流编程（Dataflow Programming）的符号数学系统，被广泛应用于各类机器学习（Machine Learning）算法的编程实现，其前身是谷歌的神经网络算法库DistBelief。MNIST数据集是机器学习领域中非常经典的一个数据集，由60000个训练样本和10000个测试样本组成，每个样本都是一张28*28像素的灰度手写数字图片。

完整代码（TensorFlow 2.0.0-RC0）如下：

#!/usr/bin/env python
import os
os.environ["TF_CPP_MIN_LOG_LEVEL"]='2'
 
import tensorflow as tf
 
# 打印TensorFlow版本号
print(tf.__name__, tf.__version__)
 
# 下载MNIST数据集
(x_train_all, y_train_all), (x_test, y_test) = tf.keras.datasets.mnist.load_data()
 
# 预处理数据 [0, 255] => [0.0, 1.0]
x_train_all, x_test = x_train_all / 255.0, x_test / 255.0
 
# 分割数据60000条数据集，前5000条为验证数据集，后55000条为训练数据集
x_valid, x_train = x_train_all[:5000], x_train_all[5000:]
y_valid, y_train = y_train_all[:5000], y_train_all[5000:]
 
# 构建全连接神经网络模型
model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Flatten(input_shape=(28, 28)))
model.add(tf.keras.layers.Dense(256, activation='relu'))
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dropout(0.2))
model.add(tf.keras.layers.Dense(10, activation='softmax'))
 
# 编译模型
model.compile(optimizer='adam',
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])
 
# 训练模型
model.fit(x_train, y_train, epochs=10, validation_data=(x_valid, y_valid))
 
# 评估模型
model.evaluate(x_test, y_test)

运行结果如下：

tensorflow 2.0.0-rc0
Train on 55000 samples, validate on 5000 samples
Epoch 1/10

   32/55000 [..............................] - ETA: 12:36 - loss: 2.3241 - accuracy: 0.0312
  736/55000 [..............................] - ETA: 36s - loss: 1.6744 - accuracy: 0.4837  
 1472/55000 [..............................] - ETA: 19s - loss: 1.2319 - accuracy: 0.6257
 2208/55000 [>.............................] - ETA: 14s - loss: 1.0136 - accuracy: 0.6893
 2944/55000 [>.............................] - ETA: 11s - loss: 0.8826 - accuracy: 0.7279
 3712/55000 [=>............................] - ETA: 9s - loss: 0.7942 - accuracy: 0.7548 
 4416/55000 [=>............................] - ETA: 8s - loss: 0.7364 - accuracy: 0.7720
 5152/55000 [=>............................] - ETA: 7s - loss: 0.6807 - accuracy: 0.7869
 5920/55000 [==>...........................] - ETA: 7s - loss: 0.6410 - accuracy: 0.8008
 6720/55000 [==>...........................] - ETA: 6s - loss: 0.6054 - accuracy: 0.8129
 7488/55000 [===>..........................] - ETA: 6s - loss: 0.5759 - accuracy: 0.8233
 8288/55000 [===>..........................] - ETA: 5s - loss: 0.5529 - accuracy: 0.8307
 9088/55000 [===>..........................] - ETA: 5s - loss: 0.5291 - accuracy: 0.8379
 9888/55000 [====>.........................] - ETA: 5s - loss: 0.5088 - accuracy: 0.8439
10688/55000 [====>.........................] - ETA: 4s - loss: 0.4923 - accuracy: 0.8502
11488/55000 [=====>........................] - ETA: 4s - loss: 0.4774 - accuracy: 0.8549
12288/55000 [=====>........................] - ETA: 4s - loss: 0.4635 - accuracy: 0.8587
13088/55000 [======>.......................] - ETA: 4s - loss: 0.4506 - accuracy: 0.8632
13856/55000 [======>.......................] - ETA: 4s - loss: 0.4418 - accuracy: 0.8662
14656/55000 [======>.......................] - ETA: 3s - loss: 0.4293 - accuracy: 0.8697
15456/55000 [=======>......................] - ETA: 3s - loss: 0.4201 - accuracy: 0.8727
16256/55000 [=======>......................] - ETA: 3s - loss: 0.4108 - accuracy: 0.8754
16992/55000 [========>.....................] - ETA: 3s - loss: 0.4016 - accuracy: 0.8784
17792/55000 [========>.....................] - ETA: 3s - loss: 0.3926 - accuracy: 0.8808
18592/55000 [=========>....................] - ETA: 3s - loss: 0.3851 - accuracy: 0.8832
19360/55000 [=========>....................] - ETA: 3s - loss: 0.3780 - accuracy: 0.8854
20128/55000 [=========>....................] - ETA: 3s - loss: 0.3704 - accuracy: 0.8875
20928/55000 [==========>...................] - ETA: 2s - loss: 0.3649 - accuracy: 0.8891
21696/55000 [==========>...................] - ETA: 2s - loss: 0.3605 - accuracy: 0.8906
22496/55000 [===========>..................] - ETA: 2s - loss: 0.3541 - accuracy: 0.8926
23264/55000 [===========>..................] - ETA: 2s - loss: 0.3497 - accuracy: 0.8940
24064/55000 [============>.................] - ETA: 2s - loss: 0.3437 - accuracy: 0.8957
24864/55000 [============>.................] - ETA: 2s - loss: 0.3372 - accuracy: 0.8978
25632/55000 [============>.................] - ETA: 2s - loss: 0.3329 - accuracy: 0.8992
26400/55000 [=============>................] - ETA: 2s - loss: 0.3290 - accuracy: 0.9002
27200/55000 [=============>................] - ETA: 2s - loss: 0.3248 - accuracy: 0.9015
28000/55000 [==============>...............] - ETA: 2s - loss: 0.3206 - accuracy: 0.9028
28768/55000 [==============>...............] - ETA: 2s - loss: 0.3176 - accuracy: 0.9035
29536/55000 [===============>..............] - ETA: 2s - loss: 0.3143 - accuracy: 0.9046
30336/55000 [===============>..............] - ETA: 1s - loss: 0.3102 - accuracy: 0.9060
31136/55000 [===============>..............] - ETA: 1s - loss: 0.3064 - accuracy: 0.9071
31936/55000 [================>.............] - ETA: 1s - loss: 0.3033 - accuracy: 0.9082
32736/55000 [================>.............] - ETA: 1s - loss: 0.3001 - accuracy: 0.9092
33504/55000 [=================>............] - ETA: 1s - loss: 0.2968 - accuracy: 0.9102
34176/55000 [=================>............] - ETA: 1s - loss: 0.2932 - accuracy: 0.9114
34848/55000 [==================>...........] - ETA: 1s - loss: 0.2902 - accuracy: 0.9122
35552/55000 [==================>...........] - ETA: 1s - loss: 0.2879 - accuracy: 0.9131
36288/55000 [==================>...........] - ETA: 1s - loss: 0.2854 - accuracy: 0.9139
37056/55000 [===================>..........] - ETA: 1s - loss: 0.2831 - accuracy: 0.9145
37856/55000 [===================>..........] - ETA: 1s - loss: 0.2804 - accuracy: 0.9152
38656/55000 [====================>.........] - ETA: 1s - loss: 0.2786 - accuracy: 0.9158
39360/55000 [====================>.........] - ETA: 1s - loss: 0.2771 - accuracy: 0.9161
40064/55000 [====================>.........] - ETA: 1s - loss: 0.2750 - accuracy: 0.9169
40800/55000 [=====================>........] - ETA: 1s - loss: 0.2725 - accuracy: 0.9177
41536/55000 [=====================>........] - ETA: 1s - loss: 0.2703 - accuracy: 0.9184
42304/55000 [======================>.......] - ETA: 0s - loss: 0.2678 - accuracy: 0.9192
43104/55000 [======================>.......] - ETA: 0s - loss: 0.2663 - accuracy: 0.9197
43872/55000 [======================>.......] - ETA: 0s - loss: 0.2652 - accuracy: 0.9202
44544/55000 [=======================>......] - ETA: 0s - loss: 0.2638 - accuracy: 0.9206
45248/55000 [=======================>......] - ETA: 0s - loss: 0.2626 - accuracy: 0.9210
45952/55000 [========================>.....] - ETA: 0s - loss: 0.2603 - accuracy: 0.9216
46688/55000 [========================>.....] - ETA: 0s - loss: 0.2584 - accuracy: 0.9222
47488/55000 [========================>.....] - ETA: 0s - loss: 0.2572 - accuracy: 0.9227
48288/55000 [=========================>....] - ETA: 0s - loss: 0.2558 - accuracy: 0.9231
49088/55000 [=========================>....] - ETA: 0s - loss: 0.2539 - accuracy: 0.9237
49856/55000 [==========================>...] - ETA: 0s - loss: 0.2515 - accuracy: 0.9244
50624/55000 [==========================>...] - ETA: 0s - loss: 0.2497 - accuracy: 0.9249
51392/55000 [===========================>..] - ETA: 0s - loss: 0.2480 - accuracy: 0.9254
52128/55000 [===========================>..] - ETA: 0s - loss: 0.2469 - accuracy: 0.9256
52832/55000 [===========================>..] - ETA: 0s - loss: 0.2451 - accuracy: 0.9261
53600/55000 [============================>.] - ETA: 0s - loss: 0.2437 - accuracy: 0.9266
54400/55000 [============================>.] - ETA: 0s - loss: 0.2421 - accuracy: 0.9271
55000/55000 [==============================] - 4s 80us/sample - loss: 0.2409 - accuracy: 0.9275 - val_loss: 0.1166 - val_accuracy: 0.9646
Epoch 2/10

   32/55000 [..............................] - ETA: 5s - loss: 0.1135 - accuracy: 0.9688
  736/55000 [..............................] - ETA: 3s - loss: 0.1157 - accuracy: 0.9701
 1440/55000 [..............................] - ETA: 3s - loss: 0.1066 - accuracy: 0.9701
 2144/55000 [>.............................] - ETA: 3s - loss: 0.1001 - accuracy: 0.9706
 2880/55000 [>.............................] - ETA: 3s - loss: 0.1073 - accuracy: 0.9674
 3648/55000 [>.............................] - ETA: 3s - loss: 0.1026 - accuracy: 0.9682
 4416/55000 [=>............................] - ETA: 3s - loss: 0.0989 - accuracy: 0.9692
 5152/55000 [=>............................] - ETA: 3s - loss: 0.0989 - accuracy: 0.9695
 5952/55000 [==>...........................] - ETA: 3s - loss: 0.1006 - accuracy: 0.9689
 6720/55000 [==>...........................] - ETA: 3s - loss: 0.1008 - accuracy: 0.9690
 7488/55000 [===>..........................] - ETA: 3s - loss: 0.0999 - accuracy: 0.9696
 8288/55000 [===>..........................] - ETA: 3s - loss: 0.1004 - accuracy: 0.9697
 9056/55000 [===>..........................] - ETA: 3s - loss: 0.0998 - accuracy: 0.9702
 9824/55000 [====>.........................] - ETA: 3s - loss: 0.0992 - accuracy: 0.9701
10592/55000 [====>.........................] - ETA: 3s - loss: 0.0986 - accuracy: 0.9700
11392/55000 [=====>........................] - ETA: 2s - loss: 0.0994 - accuracy: 0.9698
12128/55000 [=====>........................] - ETA: 2s - loss: 0.1004 - accuracy: 0.9694
12928/55000 [======>.......................] - ETA: 2s - loss: 0.1007 - accuracy: 0.9692
13728/55000 [======>.......................] - ETA: 2s - loss: 0.0990 - accuracy: 0.9698
14464/55000 [======>.......................] - ETA: 2s - loss: 0.0989 - accuracy: 0.9699
15264/55000 [=======>......................] - ETA: 2s - loss: 0.0980 - accuracy: 0.9703
16064/55000 [=======>......................] - ETA: 2s - loss: 0.0982 - accuracy: 0.9701
16864/55000 [========>.....................] - ETA: 2s - loss: 0.1002 - accuracy: 0.9689
17600/55000 [========>.....................] - ETA: 2s - loss: 0.1006 - accuracy: 0.9690
18400/55000 [=========>....................] - ETA: 2s - loss: 0.1021 - accuracy: 0.9687
19168/55000 [=========>....................] - ETA: 2s - loss: 0.1018 - accuracy: 0.9685
19936/55000 [=========>....................] - ETA: 2s - loss: 0.1013 - accuracy: 0.9688
20704/55000 [==========>...................] - ETA: 2s - loss: 0.1010 - accuracy: 0.9691
21472/55000 [==========>...................] - ETA: 2s - loss: 0.1013 - accuracy: 0.9689
22240/55000 [===========>..................] - ETA: 2s - loss: 0.1003 - accuracy: 0.9692
23008/55000 [===========>..................] - ETA: 2s - loss: 0.0994 - accuracy: 0.9695
23776/55000 [===========>..................] - ETA: 2s - loss: 0.0999 - accuracy: 0.9695
24576/55000 [============>.................] - ETA: 2s - loss: 0.1000 - accuracy: 0.9694
25376/55000 [============>.................] - ETA: 1s - loss: 0.1000 - accuracy: 0.9694
26176/55000 [=============>................] - ETA: 1s - loss: 0.1003 - accuracy: 0.9692
26976/55000 [=============>................] - ETA: 1s - loss: 0.1005 - accuracy: 0.9690
27776/55000 [==============>...............] - ETA: 1s - loss: 0.1003 - accuracy: 0.9690
28576/55000 [==============>...............] - ETA: 1s - loss: 0.1005 - accuracy: 0.9689
29376/55000 [===============>..............] - ETA: 1s - loss: 0.1008 - accuracy: 0.9687
30176/55000 [===============>..............] - ETA: 1s - loss: 0.1004 - accuracy: 0.9689
30944/55000 [===============>..............] - ETA: 1s - loss: 0.1010 - accuracy: 0.9687
31712/55000 [================>.............] - ETA: 1s - loss: 0.1007 - accuracy: 0.9689
32480/55000 [================>.............] - ETA: 1s - loss: 0.1001 - accuracy: 0.9691
33248/55000 [=================>............] - ETA: 1s - loss: 0.1000 - accuracy: 0.9690
34016/55000 [=================>............] - ETA: 1s - loss: 0.1006 - accuracy: 0.9690
34784/55000 [=================>............] - ETA: 1s - loss: 0.1004 - accuracy: 0.9691
35552/55000 [==================>...........] - ETA: 1s - loss: 0.0998 - accuracy: 0.9691
36320/55000 [==================>...........] - ETA: 1s - loss: 0.1001 - accuracy: 0.9690
37088/55000 [===================>..........] - ETA: 1s - loss: 0.1006 - accuracy: 0.9689
37856/55000 [===================>..........] - ETA: 1s - loss: 0.1009 - accuracy: 0.9687
38624/55000 [====================>.........] - ETA: 1s - loss: 0.1005 - accuracy: 0.9688
39392/55000 [====================>.........] - ETA: 1s - loss: 0.1004 - accuracy: 0.9689
40192/55000 [====================>.........] - ETA: 0s - loss: 0.0999 - accuracy: 0.9691
40992/55000 [=====================>........] - ETA: 0s - loss: 0.1003 - accuracy: 0.9690
41792/55000 [=====================>........] - ETA: 0s - loss: 0.0998 - accuracy: 0.9690
42592/55000 [======================>.......] - ETA: 0s - loss: 0.1002 - accuracy: 0.9689
43392/55000 [======================>.......] - ETA: 0s - loss: 0.1007 - accuracy: 0.9688
44192/55000 [=======================>......] - ETA: 0s - loss: 0.1007 - accuracy: 0.9689
44992/55000 [=======================>......] - ETA: 0s - loss: 0.1003 - accuracy: 0.9690
45792/55000 [=======================>......] - ETA: 0s - loss: 0.0999 - accuracy: 0.9691
46592/55000 [========================>.....] - ETA: 0s - loss: 0.1001 - accuracy: 0.9690
47392/55000 [========================>.....] - ETA: 0s - loss: 0.0997 - accuracy: 0.9692
48128/55000 [=========================>....] - ETA: 0s - loss: 0.0997 - accuracy: 0.9692
48896/55000 [=========================>....] - ETA: 0s - loss: 0.0996 - accuracy: 0.9692
49664/55000 [==========================>...] - ETA: 0s - loss: 0.0994 - accuracy: 0.9693
50432/55000 [==========================>...] - ETA: 0s - loss: 0.0996 - accuracy: 0.9693
51232/55000 [==========================>...] - ETA: 0s - loss: 0.0999 - accuracy: 0.9692
52032/55000 [===========================>..] - ETA: 0s - loss: 0.1004 - accuracy: 0.9691
52832/55000 [===========================>..] - ETA: 0s - loss: 0.1005 - accuracy: 0.9690
53600/55000 [============================>.] - ETA: 0s - loss: 0.1013 - accuracy: 0.9689
54400/55000 [============================>.] - ETA: 0s - loss: 0.1013 - accuracy: 0.9689
55000/55000 [==============================] - 4s 69us/sample - loss: 0.1013 - accuracy: 0.9688 - val_loss: 0.0875 - val_accuracy: 0.9730
Epoch 3/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0327 - accuracy: 1.0000
  800/55000 [..............................] - ETA: 3s - loss: 0.0693 - accuracy: 0.9800
 1600/55000 [..............................] - ETA: 3s - loss: 0.0682 - accuracy: 0.9775
 2304/55000 [>.............................] - ETA: 3s - loss: 0.0687 - accuracy: 0.9766
 3040/55000 [>.............................] - ETA: 3s - loss: 0.0646 - accuracy: 0.9780
 3744/55000 [=>............................] - ETA: 3s - loss: 0.0662 - accuracy: 0.9776
 4480/55000 [=>............................] - ETA: 3s - loss: 0.0658 - accuracy: 0.9777
 5184/55000 [=>............................] - ETA: 3s - loss: 0.0655 - accuracy: 0.9774
 5888/55000 [==>...........................] - ETA: 3s - loss: 0.0644 - accuracy: 0.9784
 6592/55000 [==>...........................] - ETA: 3s - loss: 0.0620 - accuracy: 0.9795
 7296/55000 [==>...........................] - ETA: 3s - loss: 0.0639 - accuracy: 0.9788
 8064/55000 [===>..........................] - ETA: 3s - loss: 0.0643 - accuracy: 0.9785
 8864/55000 [===>..........................] - ETA: 3s - loss: 0.0676 - accuracy: 0.9782
 9632/55000 [====>.........................] - ETA: 3s - loss: 0.0679 - accuracy: 0.9785
10400/55000 [====>.........................] - ETA: 3s - loss: 0.0676 - accuracy: 0.9783
11200/55000 [=====>........................] - ETA: 3s - loss: 0.0675 - accuracy: 0.9781
12000/55000 [=====>........................] - ETA: 2s - loss: 0.0672 - accuracy: 0.9785
12800/55000 [=====>........................] - ETA: 2s - loss: 0.0681 - accuracy: 0.9781
13600/55000 [======>.......................] - ETA: 2s - loss: 0.0672 - accuracy: 0.9784
14400/55000 [======>.......................] - ETA: 2s - loss: 0.0679 - accuracy: 0.9783
15200/55000 [=======>......................] - ETA: 2s - loss: 0.0684 - accuracy: 0.9780
16000/55000 [=======>......................] - ETA: 2s - loss: 0.0676 - accuracy: 0.9783
16800/55000 [========>.....................] - ETA: 2s - loss: 0.0668 - accuracy: 0.9785
17568/55000 [========>.....................] - ETA: 2s - loss: 0.0662 - accuracy: 0.9788
18336/55000 [=========>....................] - ETA: 2s - loss: 0.0664 - accuracy: 0.9790
19104/55000 [=========>....................] - ETA: 2s - loss: 0.0664 - accuracy: 0.9791
19840/55000 [=========>....................] - ETA: 2s - loss: 0.0672 - accuracy: 0.9789
20640/55000 [==========>...................] - ETA: 2s - loss: 0.0671 - accuracy: 0.9789
21408/55000 [==========>...................] - ETA: 2s - loss: 0.0674 - accuracy: 0.9788
22112/55000 [===========>..................] - ETA: 2s - loss: 0.0672 - accuracy: 0.9787
22816/55000 [===========>..................] - ETA: 2s - loss: 0.0684 - accuracy: 0.9785
23552/55000 [===========>..................] - ETA: 2s - loss: 0.0680 - accuracy: 0.9784
24288/55000 [============>.................] - ETA: 2s - loss: 0.0676 - accuracy: 0.9786
25056/55000 [============>.................] - ETA: 2s - loss: 0.0677 - accuracy: 0.9786
25856/55000 [=============>................] - ETA: 1s - loss: 0.0673 - accuracy: 0.9787
26656/55000 [=============>................] - ETA: 1s - loss: 0.0675 - accuracy: 0.9786
27456/55000 [=============>................] - ETA: 1s - loss: 0.0672 - accuracy: 0.9785
28256/55000 [==============>...............] - ETA: 1s - loss: 0.0678 - accuracy: 0.9783
29056/55000 [==============>...............] - ETA: 1s - loss: 0.0687 - accuracy: 0.9781
29856/55000 [===============>..............] - ETA: 1s - loss: 0.0686 - accuracy: 0.9782
30656/55000 [===============>..............] - ETA: 1s - loss: 0.0688 - accuracy: 0.9780
31424/55000 [================>.............] - ETA: 1s - loss: 0.0691 - accuracy: 0.9780
32224/55000 [================>.............] - ETA: 1s - loss: 0.0687 - accuracy: 0.9782
33024/55000 [=================>............] - ETA: 1s - loss: 0.0686 - accuracy: 0.9783
33824/55000 [=================>............] - ETA: 1s - loss: 0.0688 - accuracy: 0.9782
34624/55000 [=================>............] - ETA: 1s - loss: 0.0692 - accuracy: 0.9780
35424/55000 [==================>...........] - ETA: 1s - loss: 0.0689 - accuracy: 0.9782
36224/55000 [==================>...........] - ETA: 1s - loss: 0.0686 - accuracy: 0.9782
37024/55000 [===================>..........] - ETA: 1s - loss: 0.0686 - accuracy: 0.9783
37824/55000 [===================>..........] - ETA: 1s - loss: 0.0697 - accuracy: 0.9781
38592/55000 [====================>.........] - ETA: 1s - loss: 0.0701 - accuracy: 0.9779
39392/55000 [====================>.........] - ETA: 1s - loss: 0.0699 - accuracy: 0.9781
40192/55000 [====================>.........] - ETA: 0s - loss: 0.0696 - accuracy: 0.9780
40992/55000 [=====================>........] - ETA: 0s - loss: 0.0698 - accuracy: 0.9779
41792/55000 [=====================>........] - ETA: 0s - loss: 0.0702 - accuracy: 0.9778
42592/55000 [======================>.......] - ETA: 0s - loss: 0.0703 - accuracy: 0.9778
43392/55000 [======================>.......] - ETA: 0s - loss: 0.0702 - accuracy: 0.9779
44192/55000 [=======================>......] - ETA: 0s - loss: 0.0702 - accuracy: 0.9779
44992/55000 [=======================>......] - ETA: 0s - loss: 0.0708 - accuracy: 0.9777
45792/55000 [=======================>......] - ETA: 0s - loss: 0.0710 - accuracy: 0.9776
46592/55000 [========================>.....] - ETA: 0s - loss: 0.0710 - accuracy: 0.9775
47360/55000 [========================>.....] - ETA: 0s - loss: 0.0709 - accuracy: 0.9775
48160/55000 [=========================>....] - ETA: 0s - loss: 0.0710 - accuracy: 0.9775
48960/55000 [=========================>....] - ETA: 0s - loss: 0.0708 - accuracy: 0.9775
49760/55000 [==========================>...] - ETA: 0s - loss: 0.0706 - accuracy: 0.9776
50496/55000 [==========================>...] - ETA: 0s - loss: 0.0705 - accuracy: 0.9776
51264/55000 [==========================>...] - ETA: 0s - loss: 0.0705 - accuracy: 0.9776
52064/55000 [===========================>..] - ETA: 0s - loss: 0.0705 - accuracy: 0.9777
52864/55000 [===========================>..] - ETA: 0s - loss: 0.0710 - accuracy: 0.9777
53664/55000 [============================>.] - ETA: 0s - loss: 0.0711 - accuracy: 0.9776
54464/55000 [============================>.] - ETA: 0s - loss: 0.0709 - accuracy: 0.9777
55000/55000 [==============================] - 4s 69us/sample - loss: 0.0708 - accuracy: 0.9777 - val_loss: 0.0758 - val_accuracy: 0.9790
Epoch 4/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0532 - accuracy: 0.9688
  800/55000 [..............................] - ETA: 3s - loss: 0.0421 - accuracy: 0.9825
 1600/55000 [..............................] - ETA: 3s - loss: 0.0350 - accuracy: 0.9875
 2400/55000 [>.............................] - ETA: 3s - loss: 0.0384 - accuracy: 0.9862
 3200/55000 [>.............................] - ETA: 3s - loss: 0.0374 - accuracy: 0.9869
 4000/55000 [=>............................] - ETA: 3s - loss: 0.0381 - accuracy: 0.9875
 4800/55000 [=>............................] - ETA: 3s - loss: 0.0349 - accuracy: 0.9894
 5600/55000 [==>...........................] - ETA: 3s - loss: 0.0364 - accuracy: 0.9887
 6400/55000 [==>...........................] - ETA: 3s - loss: 0.0380 - accuracy: 0.9883
 7200/55000 [==>...........................] - ETA: 3s - loss: 0.0387 - accuracy: 0.9878
 8000/55000 [===>..........................] - ETA: 3s - loss: 0.0390 - accuracy: 0.9872
 8800/55000 [===>..........................] - ETA: 2s - loss: 0.0398 - accuracy: 0.9872
 9600/55000 [====>.........................] - ETA: 2s - loss: 0.0424 - accuracy: 0.9855
10400/55000 [====>.........................] - ETA: 2s - loss: 0.0432 - accuracy: 0.9856
11168/55000 [=====>........................] - ETA: 2s - loss: 0.0438 - accuracy: 0.9855
11968/55000 [=====>........................] - ETA: 2s - loss: 0.0439 - accuracy: 0.9852
12768/55000 [=====>........................] - ETA: 2s - loss: 0.0443 - accuracy: 0.9854
13568/55000 [======>.......................] - ETA: 2s - loss: 0.0452 - accuracy: 0.9850
14368/55000 [======>.......................] - ETA: 2s - loss: 0.0478 - accuracy: 0.9843
15168/55000 [=======>......................] - ETA: 2s - loss: 0.0482 - accuracy: 0.9842
15968/55000 [=======>......................] - ETA: 2s - loss: 0.0486 - accuracy: 0.9842
16736/55000 [========>.....................] - ETA: 2s - loss: 0.0500 - accuracy: 0.9839
17536/55000 [========>.....................] - ETA: 2s - loss: 0.0496 - accuracy: 0.9841
18336/55000 [=========>....................] - ETA: 2s - loss: 0.0504 - accuracy: 0.9840
19136/55000 [=========>....................] - ETA: 2s - loss: 0.0505 - accuracy: 0.9840
19936/55000 [=========>....................] - ETA: 2s - loss: 0.0501 - accuracy: 0.9840
20736/55000 [==========>...................] - ETA: 2s - loss: 0.0501 - accuracy: 0.9841
21536/55000 [==========>...................] - ETA: 2s - loss: 0.0497 - accuracy: 0.9843
22336/55000 [===========>..................] - ETA: 2s - loss: 0.0496 - accuracy: 0.9842
23136/55000 [===========>..................] - ETA: 2s - loss: 0.0494 - accuracy: 0.9844
23872/55000 [============>.................] - ETA: 2s - loss: 0.0499 - accuracy: 0.9842
24672/55000 [============>.................] - ETA: 1s - loss: 0.0495 - accuracy: 0.9844
25472/55000 [============>.................] - ETA: 1s - loss: 0.0500 - accuracy: 0.9840
26272/55000 [=============>................] - ETA: 1s - loss: 0.0499 - accuracy: 0.9840
27072/55000 [=============>................] - ETA: 1s - loss: 0.0498 - accuracy: 0.9839
27872/55000 [==============>...............] - ETA: 1s - loss: 0.0503 - accuracy: 0.9838
28672/55000 [==============>...............] - ETA: 1s - loss: 0.0504 - accuracy: 0.9838
29472/55000 [===============>..............] - ETA: 1s - loss: 0.0508 - accuracy: 0.9836
30272/55000 [===============>..............] - ETA: 1s - loss: 0.0507 - accuracy: 0.9835
31072/55000 [===============>..............] - ETA: 1s - loss: 0.0504 - accuracy: 0.9835
31840/55000 [================>.............] - ETA: 1s - loss: 0.0506 - accuracy: 0.9833
32640/55000 [================>.............] - ETA: 1s - loss: 0.0506 - accuracy: 0.9833
33440/55000 [=================>............] - ETA: 1s - loss: 0.0510 - accuracy: 0.9831
34240/55000 [=================>............] - ETA: 1s - loss: 0.0510 - accuracy: 0.9832
35008/55000 [==================>...........] - ETA: 1s - loss: 0.0508 - accuracy: 0.9834
35808/55000 [==================>...........] - ETA: 1s - loss: 0.0510 - accuracy: 0.9833
36608/55000 [==================>...........] - ETA: 1s - loss: 0.0511 - accuracy: 0.9833
37408/55000 [===================>..........] - ETA: 1s - loss: 0.0513 - accuracy: 0.9832
38208/55000 [===================>..........] - ETA: 1s - loss: 0.0512 - accuracy: 0.9832
39008/55000 [====================>.........] - ETA: 1s - loss: 0.0513 - accuracy: 0.9832
39808/55000 [====================>.........] - ETA: 0s - loss: 0.0516 - accuracy: 0.9831
40608/55000 [=====================>........] - ETA: 0s - loss: 0.0517 - accuracy: 0.9831
41408/55000 [=====================>........] - ETA: 0s - loss: 0.0518 - accuracy: 0.9830
42208/55000 [======================>.......] - ETA: 0s - loss: 0.0522 - accuracy: 0.9829
43008/55000 [======================>.......] - ETA: 0s - loss: 0.0520 - accuracy: 0.9830
43808/55000 [======================>.......] - ETA: 0s - loss: 0.0519 - accuracy: 0.9829
44576/55000 [=======================>......] - ETA: 0s - loss: 0.0521 - accuracy: 0.9828
45376/55000 [=======================>......] - ETA: 0s - loss: 0.0523 - accuracy: 0.9827
46176/55000 [========================>.....] - ETA: 0s - loss: 0.0522 - accuracy: 0.9826
46976/55000 [========================>.....] - ETA: 0s - loss: 0.0519 - accuracy: 0.9827
47776/55000 [=========================>....] - ETA: 0s - loss: 0.0519 - accuracy: 0.9827
48576/55000 [=========================>....] - ETA: 0s - loss: 0.0519 - accuracy: 0.9828
49376/55000 [=========================>....] - ETA: 0s - loss: 0.0522 - accuracy: 0.9827
50176/55000 [==========================>...] - ETA: 0s - loss: 0.0525 - accuracy: 0.9826
50976/55000 [==========================>...] - ETA: 0s - loss: 0.0530 - accuracy: 0.9825
51744/55000 [===========================>..] - ETA: 0s - loss: 0.0534 - accuracy: 0.9824
52544/55000 [===========================>..] - ETA: 0s - loss: 0.0535 - accuracy: 0.9824
53344/55000 [============================>.] - ETA: 0s - loss: 0.0538 - accuracy: 0.9824
54144/55000 [============================>.] - ETA: 0s - loss: 0.0537 - accuracy: 0.9823
54880/55000 [============================>.] - ETA: 0s - loss: 0.0539 - accuracy: 0.9823
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0538 - accuracy: 0.9823 - val_loss: 0.0793 - val_accuracy: 0.9792
Epoch 5/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0048 - accuracy: 1.0000
  832/55000 [..............................] - ETA: 3s - loss: 0.0287 - accuracy: 0.9892
 1632/55000 [..............................] - ETA: 3s - loss: 0.0304 - accuracy: 0.9902
 2432/55000 [>.............................] - ETA: 3s - loss: 0.0308 - accuracy: 0.9893
 3232/55000 [>.............................] - ETA: 3s - loss: 0.0311 - accuracy: 0.9901
 4032/55000 [=>............................] - ETA: 3s - loss: 0.0370 - accuracy: 0.9883
 4800/55000 [=>............................] - ETA: 3s - loss: 0.0380 - accuracy: 0.9879
 5600/55000 [==>...........................] - ETA: 3s - loss: 0.0363 - accuracy: 0.9882
 6400/55000 [==>...........................] - ETA: 3s - loss: 0.0360 - accuracy: 0.9884
 7200/55000 [==>...........................] - ETA: 3s - loss: 0.0377 - accuracy: 0.9875
 8000/55000 [===>..........................] - ETA: 3s - loss: 0.0391 - accuracy: 0.9870
 8768/55000 [===>..........................] - ETA: 2s - loss: 0.0384 - accuracy: 0.9872
 9536/55000 [====>.........................] - ETA: 2s - loss: 0.0370 - accuracy: 0.9878
10336/55000 [====>.........................] - ETA: 2s - loss: 0.0369 - accuracy: 0.9876
11136/55000 [=====>........................] - ETA: 2s - loss: 0.0368 - accuracy: 0.9878
11968/55000 [=====>........................] - ETA: 2s - loss: 0.0381 - accuracy: 0.9874
12736/55000 [=====>........................] - ETA: 2s - loss: 0.0377 - accuracy: 0.9876
13504/55000 [======>.......................] - ETA: 2s - loss: 0.0383 - accuracy: 0.9873
14304/55000 [======>.......................] - ETA: 2s - loss: 0.0376 - accuracy: 0.9875
15104/55000 [=======>......................] - ETA: 2s - loss: 0.0367 - accuracy: 0.9878
15872/55000 [=======>......................] - ETA: 2s - loss: 0.0356 - accuracy: 0.9882
16672/55000 [========>.....................] - ETA: 2s - loss: 0.0357 - accuracy: 0.9881
17472/55000 [========>.....................] - ETA: 2s - loss: 0.0369 - accuracy: 0.9878
18208/55000 [========>.....................] - ETA: 2s - loss: 0.0370 - accuracy: 0.9875
18976/55000 [=========>....................] - ETA: 2s - loss: 0.0373 - accuracy: 0.9874
19744/55000 [=========>....................] - ETA: 2s - loss: 0.0378 - accuracy: 0.9873
20512/55000 [==========>...................] - ETA: 2s - loss: 0.0384 - accuracy: 0.9872
21312/55000 [==========>...................] - ETA: 2s - loss: 0.0388 - accuracy: 0.9869
22112/55000 [===========>..................] - ETA: 2s - loss: 0.0395 - accuracy: 0.9866
22880/55000 [===========>..................] - ETA: 2s - loss: 0.0398 - accuracy: 0.9863
23680/55000 [===========>..................] - ETA: 2s - loss: 0.0402 - accuracy: 0.9863
24480/55000 [============>.................] - ETA: 1s - loss: 0.0404 - accuracy: 0.9862
25280/55000 [============>.................] - ETA: 1s - loss: 0.0404 - accuracy: 0.9862
26080/55000 [=============>................] - ETA: 1s - loss: 0.0407 - accuracy: 0.9862
26880/55000 [=============>................] - ETA: 1s - loss: 0.0404 - accuracy: 0.9862
27584/55000 [==============>...............] - ETA: 1s - loss: 0.0402 - accuracy: 0.9862
28384/55000 [==============>...............] - ETA: 1s - loss: 0.0403 - accuracy: 0.9862
29184/55000 [==============>...............] - ETA: 1s - loss: 0.0405 - accuracy: 0.9861
29984/55000 [===============>..............] - ETA: 1s - loss: 0.0402 - accuracy: 0.9862
30784/55000 [===============>..............] - ETA: 1s - loss: 0.0398 - accuracy: 0.9862
31584/55000 [================>.............] - ETA: 1s - loss: 0.0399 - accuracy: 0.9862
32384/55000 [================>.............] - ETA: 1s - loss: 0.0395 - accuracy: 0.9863
33152/55000 [=================>............] - ETA: 1s - loss: 0.0398 - accuracy: 0.9863
33952/55000 [=================>............] - ETA: 1s - loss: 0.0396 - accuracy: 0.9863
34752/55000 [=================>............] - ETA: 1s - loss: 0.0399 - accuracy: 0.9863
35552/55000 [==================>...........] - ETA: 1s - loss: 0.0398 - accuracy: 0.9863
36352/55000 [==================>...........] - ETA: 1s - loss: 0.0395 - accuracy: 0.9864
37152/55000 [===================>..........] - ETA: 1s - loss: 0.0396 - accuracy: 0.9863
37952/55000 [===================>..........] - ETA: 1s - loss: 0.0399 - accuracy: 0.9864
38752/55000 [====================>.........] - ETA: 1s - loss: 0.0401 - accuracy: 0.9864
39552/55000 [====================>.........] - ETA: 1s - loss: 0.0397 - accuracy: 0.9865
40352/55000 [=====================>........] - ETA: 0s - loss: 0.0398 - accuracy: 0.9866
41152/55000 [=====================>........] - ETA: 0s - loss: 0.0404 - accuracy: 0.9863
41952/55000 [=====================>........] - ETA: 0s - loss: 0.0401 - accuracy: 0.9864
42752/55000 [======================>.......] - ETA: 0s - loss: 0.0402 - accuracy: 0.9865
43552/55000 [======================>.......] - ETA: 0s - loss: 0.0402 - accuracy: 0.9864
44352/55000 [=======================>......] - ETA: 0s - loss: 0.0403 - accuracy: 0.9864
45152/55000 [=======================>......] - ETA: 0s - loss: 0.0410 - accuracy: 0.9862
45920/55000 [========================>.....] - ETA: 0s - loss: 0.0411 - accuracy: 0.9861
46688/55000 [========================>.....] - ETA: 0s - loss: 0.0411 - accuracy: 0.9861
47488/55000 [========================>.....] - ETA: 0s - loss: 0.0411 - accuracy: 0.9860
48288/55000 [=========================>....] - ETA: 0s - loss: 0.0416 - accuracy: 0.9859
49088/55000 [=========================>....] - ETA: 0s - loss: 0.0420 - accuracy: 0.9859
49888/55000 [==========================>...] - ETA: 0s - loss: 0.0421 - accuracy: 0.9858
50688/55000 [==========================>...] - ETA: 0s - loss: 0.0423 - accuracy: 0.9857
51488/55000 [===========================>..] - ETA: 0s - loss: 0.0421 - accuracy: 0.9857
52288/55000 [===========================>..] - ETA: 0s - loss: 0.0422 - accuracy: 0.9857
53088/55000 [===========================>..] - ETA: 0s - loss: 0.0422 - accuracy: 0.9856
53920/55000 [============================>.] - ETA: 0s - loss: 0.0423 - accuracy: 0.9856
54720/55000 [============================>.] - ETA: 0s - loss: 0.0423 - accuracy: 0.9856
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0423 - accuracy: 0.9856 - val_loss: 0.0757 - val_accuracy: 0.9772
Epoch 6/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0050 - accuracy: 1.0000
  768/55000 [..............................] - ETA: 4s - loss: 0.0242 - accuracy: 0.9922
 1568/55000 [..............................] - ETA: 3s - loss: 0.0278 - accuracy: 0.9911
 2368/55000 [>.............................] - ETA: 3s - loss: 0.0281 - accuracy: 0.9907
 3168/55000 [>.............................] - ETA: 3s - loss: 0.0265 - accuracy: 0.9905
 3936/55000 [=>............................] - ETA: 3s - loss: 0.0256 - accuracy: 0.9901
 4736/55000 [=>............................] - ETA: 3s - loss: 0.0249 - accuracy: 0.9903
 5536/55000 [==>...........................] - ETA: 3s - loss: 0.0241 - accuracy: 0.9911
 6336/55000 [==>...........................] - ETA: 3s - loss: 0.0253 - accuracy: 0.9910
 7136/55000 [==>...........................] - ETA: 3s - loss: 0.0265 - accuracy: 0.9910
 7936/55000 [===>..........................] - ETA: 3s - loss: 0.0268 - accuracy: 0.9913
 8736/55000 [===>..........................] - ETA: 3s - loss: 0.0260 - accuracy: 0.9916
 9536/55000 [====>.........................] - ETA: 2s - loss: 0.0259 - accuracy: 0.9915
10336/55000 [====>.........................] - ETA: 2s - loss: 0.0252 - accuracy: 0.9917
11136/55000 [=====>........................] - ETA: 2s - loss: 0.0267 - accuracy: 0.9914
11936/55000 [=====>........................] - ETA: 2s - loss: 0.0278 - accuracy: 0.9910
12736/55000 [=====>........................] - ETA: 2s - loss: 0.0288 - accuracy: 0.9906
13504/55000 [======>.......................] - ETA: 2s - loss: 0.0296 - accuracy: 0.9904
14304/55000 [======>.......................] - ETA: 2s - loss: 0.0316 - accuracy: 0.9899
15104/55000 [=======>......................] - ETA: 2s - loss: 0.0311 - accuracy: 0.9901
15904/55000 [=======>......................] - ETA: 2s - loss: 0.0314 - accuracy: 0.9899
16704/55000 [========>.....................] - ETA: 2s - loss: 0.0312 - accuracy: 0.9899
17504/55000 [========>.....................] - ETA: 2s - loss: 0.0312 - accuracy: 0.9899
18304/55000 [========>.....................] - ETA: 2s - loss: 0.0309 - accuracy: 0.9900
19104/55000 [=========>....................] - ETA: 2s - loss: 0.0310 - accuracy: 0.9897
19904/55000 [=========>....................] - ETA: 2s - loss: 0.0317 - accuracy: 0.9894
20704/55000 [==========>...................] - ETA: 2s - loss: 0.0322 - accuracy: 0.9894
21504/55000 [==========>...................] - ETA: 2s - loss: 0.0328 - accuracy: 0.9894
22304/55000 [===========>..................] - ETA: 2s - loss: 0.0329 - accuracy: 0.9894
23008/55000 [===========>..................] - ETA: 2s - loss: 0.0329 - accuracy: 0.9893
23744/55000 [===========>..................] - ETA: 2s - loss: 0.0328 - accuracy: 0.9894
24512/55000 [============>.................] - ETA: 1s - loss: 0.0339 - accuracy: 0.9891
25312/55000 [============>.................] - ETA: 1s - loss: 0.0359 - accuracy: 0.9886
26112/55000 [=============>................] - ETA: 1s - loss: 0.0361 - accuracy: 0.9885
26912/55000 [=============>................] - ETA: 1s - loss: 0.0363 - accuracy: 0.9885
27712/55000 [==============>...............] - ETA: 1s - loss: 0.0357 - accuracy: 0.9887
28448/55000 [==============>...............] - ETA: 1s - loss: 0.0362 - accuracy: 0.9885
29152/55000 [==============>...............] - ETA: 1s - loss: 0.0358 - accuracy: 0.9886
29888/55000 [===============>..............] - ETA: 1s - loss: 0.0357 - accuracy: 0.9886
30688/55000 [===============>..............] - ETA: 1s - loss: 0.0359 - accuracy: 0.9887
31424/55000 [================>.............] - ETA: 1s - loss: 0.0359 - accuracy: 0.9888
32224/55000 [================>.............] - ETA: 1s - loss: 0.0361 - accuracy: 0.9887
33024/55000 [=================>............] - ETA: 1s - loss: 0.0360 - accuracy: 0.9887
33824/55000 [=================>............] - ETA: 1s - loss: 0.0362 - accuracy: 0.9886
34592/55000 [=================>............] - ETA: 1s - loss: 0.0363 - accuracy: 0.9885
35392/55000 [==================>...........] - ETA: 1s - loss: 0.0362 - accuracy: 0.9885
36192/55000 [==================>...........] - ETA: 1s - loss: 0.0361 - accuracy: 0.9885
36992/55000 [===================>..........] - ETA: 1s - loss: 0.0360 - accuracy: 0.9886
37792/55000 [===================>..........] - ETA: 1s - loss: 0.0360 - accuracy: 0.9886
38592/55000 [====================>.........] - ETA: 1s - loss: 0.0360 - accuracy: 0.9886
39392/55000 [====================>.........] - ETA: 1s - loss: 0.0361 - accuracy: 0.9886
40192/55000 [====================>.........] - ETA: 0s - loss: 0.0357 - accuracy: 0.9887
40992/55000 [=====================>........] - ETA: 0s - loss: 0.0354 - accuracy: 0.9888
41792/55000 [=====================>........] - ETA: 0s - loss: 0.0353 - accuracy: 0.9888
42592/55000 [======================>.......] - ETA: 0s - loss: 0.0352 - accuracy: 0.9888
43392/55000 [======================>.......] - ETA: 0s - loss: 0.0349 - accuracy: 0.9889
44192/55000 [=======================>......] - ETA: 0s - loss: 0.0349 - accuracy: 0.9889
44992/55000 [=======================>......] - ETA: 0s - loss: 0.0350 - accuracy: 0.9889
45792/55000 [=======================>......] - ETA: 0s - loss: 0.0353 - accuracy: 0.9888
46560/55000 [========================>.....] - ETA: 0s - loss: 0.0355 - accuracy: 0.9887
47360/55000 [========================>.....] - ETA: 0s - loss: 0.0354 - accuracy: 0.9887
48160/55000 [=========================>....] - ETA: 0s - loss: 0.0354 - accuracy: 0.9887
48960/55000 [=========================>....] - ETA: 0s - loss: 0.0352 - accuracy: 0.9888
49760/55000 [==========================>...] - ETA: 0s - loss: 0.0350 - accuracy: 0.9889
50528/55000 [==========================>...] - ETA: 0s - loss: 0.0352 - accuracy: 0.9888
51328/55000 [==========================>...] - ETA: 0s - loss: 0.0353 - accuracy: 0.9888
52128/55000 [===========================>..] - ETA: 0s - loss: 0.0353 - accuracy: 0.9887
52928/55000 [===========================>..] - ETA: 0s - loss: 0.0351 - accuracy: 0.9888
53728/55000 [============================>.] - ETA: 0s - loss: 0.0350 - accuracy: 0.9887
54528/55000 [============================>.] - ETA: 0s - loss: 0.0350 - accuracy: 0.9887
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0348 - accuracy: 0.9888 - val_loss: 0.0853 - val_accuracy: 0.9792
Epoch 7/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0057 - accuracy: 1.0000
  832/55000 [..............................] - ETA: 3s - loss: 0.0387 - accuracy: 0.9868
 1632/55000 [..............................] - ETA: 3s - loss: 0.0307 - accuracy: 0.9902
 2432/55000 [>.............................] - ETA: 3s - loss: 0.0271 - accuracy: 0.9918
 3232/55000 [>.............................] - ETA: 3s - loss: 0.0252 - accuracy: 0.9929
 4032/55000 [=>............................] - ETA: 3s - loss: 0.0240 - accuracy: 0.9931
 4768/55000 [=>............................] - ETA: 3s - loss: 0.0228 - accuracy: 0.9937
 5568/55000 [==>...........................] - ETA: 3s - loss: 0.0220 - accuracy: 0.9941
 6368/55000 [==>...........................] - ETA: 3s - loss: 0.0234 - accuracy: 0.9939
 7168/55000 [==>...........................] - ETA: 3s - loss: 0.0242 - accuracy: 0.9937
 7936/55000 [===>..........................] - ETA: 3s - loss: 0.0260 - accuracy: 0.9929
 8736/55000 [===>..........................] - ETA: 2s - loss: 0.0256 - accuracy: 0.9930
 9536/55000 [====>.........................] - ETA: 2s - loss: 0.0254 - accuracy: 0.9926
10304/55000 [====>.........................] - ETA: 2s - loss: 0.0260 - accuracy: 0.9921
11104/55000 [=====>........................] - ETA: 2s - loss: 0.0253 - accuracy: 0.9923
11904/55000 [=====>........................] - ETA: 2s - loss: 0.0249 - accuracy: 0.9922
12704/55000 [=====>........................] - ETA: 2s - loss: 0.0245 - accuracy: 0.9921
13472/55000 [======>.......................] - ETA: 2s - loss: 0.0242 - accuracy: 0.9924
14272/55000 [======>.......................] - ETA: 2s - loss: 0.0245 - accuracy: 0.9922
15072/55000 [=======>......................] - ETA: 2s - loss: 0.0241 - accuracy: 0.9923
15872/55000 [=======>......................] - ETA: 2s - loss: 0.0254 - accuracy: 0.9920
16672/55000 [========>.....................] - ETA: 2s - loss: 0.0255 - accuracy: 0.9918
17440/55000 [========>.....................] - ETA: 2s - loss: 0.0255 - accuracy: 0.9916
18240/55000 [========>.....................] - ETA: 2s - loss: 0.0256 - accuracy: 0.9916
19040/55000 [=========>....................] - ETA: 2s - loss: 0.0253 - accuracy: 0.9916
19840/55000 [=========>....................] - ETA: 2s - loss: 0.0258 - accuracy: 0.9914
20640/55000 [==========>...................] - ETA: 2s - loss: 0.0256 - accuracy: 0.9916
21440/55000 [==========>...................] - ETA: 2s - loss: 0.0258 - accuracy: 0.9915
22240/55000 [===========>..................] - ETA: 2s - loss: 0.0260 - accuracy: 0.9914
23040/55000 [===========>..................] - ETA: 2s - loss: 0.0263 - accuracy: 0.9912
23808/55000 [===========>..................] - ETA: 2s - loss: 0.0266 - accuracy: 0.9912
24608/55000 [============>.................] - ETA: 1s - loss: 0.0266 - accuracy: 0.9912
25408/55000 [============>.................] - ETA: 1s - loss: 0.0267 - accuracy: 0.9911
26208/55000 [=============>................] - ETA: 1s - loss: 0.0268 - accuracy: 0.9911
27008/55000 [=============>................] - ETA: 1s - loss: 0.0282 - accuracy: 0.9907
27808/55000 [==============>...............] - ETA: 1s - loss: 0.0284 - accuracy: 0.9907
28608/55000 [==============>...............] - ETA: 1s - loss: 0.0292 - accuracy: 0.9905
29408/55000 [===============>..............] - ETA: 1s - loss: 0.0293 - accuracy: 0.9905
30208/55000 [===============>..............] - ETA: 1s - loss: 0.0292 - accuracy: 0.9906
31008/55000 [===============>..............] - ETA: 1s - loss: 0.0289 - accuracy: 0.9907
31808/55000 [================>.............] - ETA: 1s - loss: 0.0292 - accuracy: 0.9906
32608/55000 [================>.............] - ETA: 1s - loss: 0.0293 - accuracy: 0.9906
33376/55000 [=================>............] - ETA: 1s - loss: 0.0293 - accuracy: 0.9905
34176/55000 [=================>............] - ETA: 1s - loss: 0.0295 - accuracy: 0.9904
34976/55000 [==================>...........] - ETA: 1s - loss: 0.0299 - accuracy: 0.9903
35744/55000 [==================>...........] - ETA: 1s - loss: 0.0302 - accuracy: 0.9903
36480/55000 [==================>...........] - ETA: 1s - loss: 0.0303 - accuracy: 0.9903
37280/55000 [===================>..........] - ETA: 1s - loss: 0.0304 - accuracy: 0.9903
38080/55000 [===================>..........] - ETA: 1s - loss: 0.0307 - accuracy: 0.9902
38848/55000 [====================>.........] - ETA: 1s - loss: 0.0305 - accuracy: 0.9902
39648/55000 [====================>.........] - ETA: 0s - loss: 0.0303 - accuracy: 0.9903
40480/55000 [=====================>........] - ETA: 0s - loss: 0.0302 - accuracy: 0.9904
41280/55000 [=====================>........] - ETA: 0s - loss: 0.0302 - accuracy: 0.9904
42080/55000 [=====================>........] - ETA: 0s - loss: 0.0302 - accuracy: 0.9904
42880/55000 [======================>.......] - ETA: 0s - loss: 0.0305 - accuracy: 0.9903
43648/55000 [======================>.......] - ETA: 0s - loss: 0.0309 - accuracy: 0.9902
44448/55000 [=======================>......] - ETA: 0s - loss: 0.0309 - accuracy: 0.9901
45248/55000 [=======================>......] - ETA: 0s - loss: 0.0307 - accuracy: 0.9901
46048/55000 [========================>.....] - ETA: 0s - loss: 0.0306 - accuracy: 0.9901
46816/55000 [========================>.....] - ETA: 0s - loss: 0.0309 - accuracy: 0.9900
47648/55000 [========================>.....] - ETA: 0s - loss: 0.0309 - accuracy: 0.9900
48448/55000 [=========================>....] - ETA: 0s - loss: 0.0310 - accuracy: 0.9899
49248/55000 [=========================>....] - ETA: 0s - loss: 0.0310 - accuracy: 0.9899
50048/55000 [==========================>...] - ETA: 0s - loss: 0.0312 - accuracy: 0.9898
50848/55000 [==========================>...] - ETA: 0s - loss: 0.0310 - accuracy: 0.9898
51648/55000 [===========================>..] - ETA: 0s - loss: 0.0314 - accuracy: 0.9898
52448/55000 [===========================>..] - ETA: 0s - loss: 0.0312 - accuracy: 0.9899
53248/55000 [============================>.] - ETA: 0s - loss: 0.0313 - accuracy: 0.9898
54016/55000 [============================>.] - ETA: 0s - loss: 0.0314 - accuracy: 0.9897
54784/55000 [============================>.] - ETA: 0s - loss: 0.0315 - accuracy: 0.9897
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0315 - accuracy: 0.9897 - val_loss: 0.0797 - val_accuracy: 0.9802
Epoch 8/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0419 - accuracy: 1.0000
  832/55000 [..............................] - ETA: 3s - loss: 0.0173 - accuracy: 0.9952
 1632/55000 [..............................] - ETA: 3s - loss: 0.0157 - accuracy: 0.9951
 2432/55000 [>.............................] - ETA: 3s - loss: 0.0135 - accuracy: 0.9955
 3232/55000 [>.............................] - ETA: 3s - loss: 0.0151 - accuracy: 0.9950
 4032/55000 [=>............................] - ETA: 3s - loss: 0.0151 - accuracy: 0.9950
 4832/55000 [=>............................] - ETA: 3s - loss: 0.0171 - accuracy: 0.9944
 5632/55000 [==>...........................] - ETA: 3s - loss: 0.0176 - accuracy: 0.9941
 6432/55000 [==>...........................] - ETA: 3s - loss: 0.0164 - accuracy: 0.9944
 7232/55000 [==>...........................] - ETA: 3s - loss: 0.0177 - accuracy: 0.9935
 8032/55000 [===>..........................] - ETA: 3s - loss: 0.0195 - accuracy: 0.9924
 8832/55000 [===>..........................] - ETA: 2s - loss: 0.0202 - accuracy: 0.9921
 9568/55000 [====>.........................] - ETA: 2s - loss: 0.0214 - accuracy: 0.9920
10368/55000 [====>.........................] - ETA: 2s - loss: 0.0219 - accuracy: 0.9917
11168/55000 [=====>........................] - ETA: 2s - loss: 0.0217 - accuracy: 0.9919
11936/55000 [=====>........................] - ETA: 2s - loss: 0.0210 - accuracy: 0.9922
12736/55000 [=====>........................] - ETA: 2s - loss: 0.0204 - accuracy: 0.9925
13536/55000 [======>.......................] - ETA: 2s - loss: 0.0207 - accuracy: 0.9925
14336/55000 [======>.......................] - ETA: 2s - loss: 0.0208 - accuracy: 0.9925
15136/55000 [=======>......................] - ETA: 2s - loss: 0.0207 - accuracy: 0.9926
15936/55000 [=======>......................] - ETA: 2s - loss: 0.0202 - accuracy: 0.9928
16736/55000 [========>.....................] - ETA: 2s - loss: 0.0201 - accuracy: 0.9928
17536/55000 [========>.....................] - ETA: 2s - loss: 0.0201 - accuracy: 0.9929
18336/55000 [=========>....................] - ETA: 2s - loss: 0.0199 - accuracy: 0.9929
19136/55000 [=========>....................] - ETA: 2s - loss: 0.0197 - accuracy: 0.9930
19936/55000 [=========>....................] - ETA: 2s - loss: 0.0195 - accuracy: 0.9931
20736/55000 [==========>...................] - ETA: 2s - loss: 0.0196 - accuracy: 0.9933
21536/55000 [==========>...................] - ETA: 2s - loss: 0.0194 - accuracy: 0.9934
22336/55000 [===========>..................] - ETA: 2s - loss: 0.0195 - accuracy: 0.9933
23136/55000 [===========>..................] - ETA: 2s - loss: 0.0198 - accuracy: 0.9933
23936/55000 [============>.................] - ETA: 2s - loss: 0.0203 - accuracy: 0.9932
24736/55000 [============>.................] - ETA: 1s - loss: 0.0203 - accuracy: 0.9932
25536/55000 [============>.................] - ETA: 1s - loss: 0.0200 - accuracy: 0.9933
26336/55000 [=============>................] - ETA: 1s - loss: 0.0199 - accuracy: 0.9934
27136/55000 [=============>................] - ETA: 1s - loss: 0.0198 - accuracy: 0.9934
27904/55000 [==============>...............] - ETA: 1s - loss: 0.0197 - accuracy: 0.9934
28704/55000 [==============>...............] - ETA: 1s - loss: 0.0202 - accuracy: 0.9933
29504/55000 [===============>..............] - ETA: 1s - loss: 0.0203 - accuracy: 0.9933
30304/55000 [===============>..............] - ETA: 1s - loss: 0.0207 - accuracy: 0.9932
31104/55000 [===============>..............] - ETA: 1s - loss: 0.0206 - accuracy: 0.9933
31904/55000 [================>.............] - ETA: 1s - loss: 0.0205 - accuracy: 0.9933
32704/55000 [================>.............] - ETA: 1s - loss: 0.0208 - accuracy: 0.9933
33504/55000 [=================>............] - ETA: 1s - loss: 0.0210 - accuracy: 0.9933
34304/55000 [=================>............] - ETA: 1s - loss: 0.0209 - accuracy: 0.9933
35104/55000 [==================>...........] - ETA: 1s - loss: 0.0209 - accuracy: 0.9932
35872/55000 [==================>...........] - ETA: 1s - loss: 0.0207 - accuracy: 0.9934
36672/55000 [===================>..........] - ETA: 1s - loss: 0.0207 - accuracy: 0.9933
37472/55000 [===================>..........] - ETA: 1s - loss: 0.0210 - accuracy: 0.9933
38272/55000 [===================>..........] - ETA: 1s - loss: 0.0210 - accuracy: 0.9933
39072/55000 [====================>.........] - ETA: 1s - loss: 0.0209 - accuracy: 0.9933
39872/55000 [====================>.........] - ETA: 0s - loss: 0.0209 - accuracy: 0.9932
40608/55000 [=====================>........] - ETA: 0s - loss: 0.0211 - accuracy: 0.9931
41408/55000 [=====================>........] - ETA: 0s - loss: 0.0211 - accuracy: 0.9931
42208/55000 [======================>.......] - ETA: 0s - loss: 0.0215 - accuracy: 0.9930
42976/55000 [======================>.......] - ETA: 0s - loss: 0.0214 - accuracy: 0.9930
43744/55000 [======================>.......] - ETA: 0s - loss: 0.0217 - accuracy: 0.9929
44544/55000 [=======================>......] - ETA: 0s - loss: 0.0219 - accuracy: 0.9928
45344/55000 [=======================>......] - ETA: 0s - loss: 0.0220 - accuracy: 0.9928
46144/55000 [========================>.....] - ETA: 0s - loss: 0.0221 - accuracy: 0.9927
46944/55000 [========================>.....] - ETA: 0s - loss: 0.0222 - accuracy: 0.9927
47712/55000 [=========================>....] - ETA: 0s - loss: 0.0221 - accuracy: 0.9927
48512/55000 [=========================>....] - ETA: 0s - loss: 0.0220 - accuracy: 0.9927
49312/55000 [=========================>....] - ETA: 0s - loss: 0.0221 - accuracy: 0.9927
50112/55000 [==========================>...] - ETA: 0s - loss: 0.0223 - accuracy: 0.9927
50880/55000 [==========================>...] - ETA: 0s - loss: 0.0222 - accuracy: 0.9927
51680/55000 [===========================>..] - ETA: 0s - loss: 0.0223 - accuracy: 0.9926
52480/55000 [===========================>..] - ETA: 0s - loss: 0.0225 - accuracy: 0.9926
53248/55000 [============================>.] - ETA: 0s - loss: 0.0229 - accuracy: 0.9925
54048/55000 [============================>.] - ETA: 0s - loss: 0.0229 - accuracy: 0.9926
54848/55000 [============================>.] - ETA: 0s - loss: 0.0230 - accuracy: 0.9925
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0230 - accuracy: 0.9925 - val_loss: 0.0864 - val_accuracy: 0.9802
Epoch 9/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0439 - accuracy: 1.0000
  800/55000 [..............................] - ETA: 3s - loss: 0.0231 - accuracy: 0.9937
 1600/55000 [..............................] - ETA: 3s - loss: 0.0326 - accuracy: 0.9919
 2400/55000 [>.............................] - ETA: 3s - loss: 0.0319 - accuracy: 0.9917
 3200/55000 [>.............................] - ETA: 3s - loss: 0.0299 - accuracy: 0.9916
 4000/55000 [=>............................] - ETA: 3s - loss: 0.0319 - accuracy: 0.9905
 4800/55000 [=>............................] - ETA: 3s - loss: 0.0294 - accuracy: 0.9910
 5600/55000 [==>...........................] - ETA: 3s - loss: 0.0281 - accuracy: 0.9912
 6400/55000 [==>...........................] - ETA: 3s - loss: 0.0276 - accuracy: 0.9917
 7200/55000 [==>...........................] - ETA: 3s - loss: 0.0277 - accuracy: 0.9914
 8000/55000 [===>..........................] - ETA: 3s - loss: 0.0282 - accuracy: 0.9912
 8800/55000 [===>..........................] - ETA: 2s - loss: 0.0287 - accuracy: 0.9911
 9600/55000 [====>.........................] - ETA: 2s - loss: 0.0282 - accuracy: 0.9916
10400/55000 [====>.........................] - ETA: 2s - loss: 0.0271 - accuracy: 0.9919
11200/55000 [=====>........................] - ETA: 2s - loss: 0.0263 - accuracy: 0.9921
12000/55000 [=====>........................] - ETA: 2s - loss: 0.0267 - accuracy: 0.9921
12800/55000 [=====>........................] - ETA: 2s - loss: 0.0255 - accuracy: 0.9924
13536/55000 [======>.......................] - ETA: 2s - loss: 0.0249 - accuracy: 0.9928
14336/55000 [======>.......................] - ETA: 2s - loss: 0.0237 - accuracy: 0.9932
15136/55000 [=======>......................] - ETA: 2s - loss: 0.0238 - accuracy: 0.9933
15936/55000 [=======>......................] - ETA: 2s - loss: 0.0232 - accuracy: 0.9933
16704/55000 [========>.....................] - ETA: 2s - loss: 0.0233 - accuracy: 0.9933
17504/55000 [========>.....................] - ETA: 2s - loss: 0.0236 - accuracy: 0.9934
18304/55000 [========>.....................] - ETA: 2s - loss: 0.0238 - accuracy: 0.9932
19104/55000 [=========>....................] - ETA: 2s - loss: 0.0243 - accuracy: 0.9929
19904/55000 [=========>....................] - ETA: 2s - loss: 0.0240 - accuracy: 0.9930
20736/55000 [==========>...................] - ETA: 2s - loss: 0.0236 - accuracy: 0.9931
21536/55000 [==========>...................] - ETA: 2s - loss: 0.0235 - accuracy: 0.9931
22336/55000 [===========>..................] - ETA: 2s - loss: 0.0238 - accuracy: 0.9929
23136/55000 [===========>..................] - ETA: 2s - loss: 0.0243 - accuracy: 0.9927
23936/55000 [============>.................] - ETA: 2s - loss: 0.0245 - accuracy: 0.9926
24736/55000 [============>.................] - ETA: 1s - loss: 0.0248 - accuracy: 0.9925
25536/55000 [============>.................] - ETA: 1s - loss: 0.0248 - accuracy: 0.9925
26336/55000 [=============>................] - ETA: 1s - loss: 0.0250 - accuracy: 0.9925
27136/55000 [=============>................] - ETA: 1s - loss: 0.0250 - accuracy: 0.9924
27936/55000 [==============>...............] - ETA: 1s - loss: 0.0247 - accuracy: 0.9925
28736/55000 [==============>...............] - ETA: 1s - loss: 0.0246 - accuracy: 0.9925
29536/55000 [===============>..............] - ETA: 1s - loss: 0.0249 - accuracy: 0.9923
30336/55000 [===============>..............] - ETA: 1s - loss: 0.0249 - accuracy: 0.9924
31136/55000 [===============>..............] - ETA: 1s - loss: 0.0244 - accuracy: 0.9925
31936/55000 [================>.............] - ETA: 1s - loss: 0.0242 - accuracy: 0.9926
32736/55000 [================>.............] - ETA: 1s - loss: 0.0242 - accuracy: 0.9926
33536/55000 [=================>............] - ETA: 1s - loss: 0.0239 - accuracy: 0.9926
34336/55000 [=================>............] - ETA: 1s - loss: 0.0247 - accuracy: 0.9923
35136/55000 [==================>...........] - ETA: 1s - loss: 0.0245 - accuracy: 0.9923
35936/55000 [==================>...........] - ETA: 1s - loss: 0.0243 - accuracy: 0.9924
36704/55000 [===================>..........] - ETA: 1s - loss: 0.0241 - accuracy: 0.9925
37472/55000 [===================>..........] - ETA: 1s - loss: 0.0238 - accuracy: 0.9926
38272/55000 [===================>..........] - ETA: 1s - loss: 0.0237 - accuracy: 0.9926
39072/55000 [====================>.........] - ETA: 1s - loss: 0.0234 - accuracy: 0.9926
39872/55000 [====================>.........] - ETA: 0s - loss: 0.0236 - accuracy: 0.9925
40672/55000 [=====================>........] - ETA: 0s - loss: 0.0242 - accuracy: 0.9924
41472/55000 [=====================>........] - ETA: 0s - loss: 0.0241 - accuracy: 0.9924
42272/55000 [======================>.......] - ETA: 0s - loss: 0.0244 - accuracy: 0.9922
43072/55000 [======================>.......] - ETA: 0s - loss: 0.0241 - accuracy: 0.9923
43872/55000 [======================>.......] - ETA: 0s - loss: 0.0243 - accuracy: 0.9923
44608/55000 [=======================>......] - ETA: 0s - loss: 0.0243 - accuracy: 0.9922
45408/55000 [=======================>......] - ETA: 0s - loss: 0.0244 - accuracy: 0.9921
46208/55000 [========================>.....] - ETA: 0s - loss: 0.0245 - accuracy: 0.9921
47008/55000 [========================>.....] - ETA: 0s - loss: 0.0245 - accuracy: 0.9920
47808/55000 [=========================>....] - ETA: 0s - loss: 0.0243 - accuracy: 0.9921
48576/55000 [=========================>....] - ETA: 0s - loss: 0.0245 - accuracy: 0.9921
49376/55000 [=========================>....] - ETA: 0s - loss: 0.0244 - accuracy: 0.9921
50176/55000 [==========================>...] - ETA: 0s - loss: 0.0244 - accuracy: 0.9920
50976/55000 [==========================>...] - ETA: 0s - loss: 0.0242 - accuracy: 0.9921
51776/55000 [===========================>..] - ETA: 0s - loss: 0.0241 - accuracy: 0.9921
52576/55000 [===========================>..] - ETA: 0s - loss: 0.0239 - accuracy: 0.9922
53376/55000 [============================>.] - ETA: 0s - loss: 0.0237 - accuracy: 0.9923
54176/55000 [============================>.] - ETA: 0s - loss: 0.0238 - accuracy: 0.9923
54976/55000 [============================>.] - ETA: 0s - loss: 0.0240 - accuracy: 0.9923
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0240 - accuracy: 0.9923 - val_loss: 0.1497 - val_accuracy: 0.9720
Epoch 10/10

   32/55000 [..............................] - ETA: 5s - loss: 0.0166 - accuracy: 1.0000
  832/55000 [..............................] - ETA: 3s - loss: 0.0260 - accuracy: 0.9904
 1632/55000 [..............................] - ETA: 3s - loss: 0.0246 - accuracy: 0.9902
 2432/55000 [>.............................] - ETA: 3s - loss: 0.0186 - accuracy: 0.9926
 3232/55000 [>.............................] - ETA: 3s - loss: 0.0182 - accuracy: 0.9935
 4000/55000 [=>............................] - ETA: 3s - loss: 0.0191 - accuracy: 0.9930
 4768/55000 [=>............................] - ETA: 3s - loss: 0.0233 - accuracy: 0.9922
 5440/55000 [=>............................] - ETA: 3s - loss: 0.0247 - accuracy: 0.9917
 6240/55000 [==>...........................] - ETA: 3s - loss: 0.0249 - accuracy: 0.9912
 7040/55000 [==>...........................] - ETA: 3s - loss: 0.0235 - accuracy: 0.9916
 7840/55000 [===>..........................] - ETA: 3s - loss: 0.0241 - accuracy: 0.9916
 8640/55000 [===>..........................] - ETA: 3s - loss: 0.0251 - accuracy: 0.9916
 9440/55000 [====>.........................] - ETA: 2s - loss: 0.0263 - accuracy: 0.9913
10208/55000 [====>.........................] - ETA: 2s - loss: 0.0265 - accuracy: 0.9911
11008/55000 [=====>........................] - ETA: 2s - loss: 0.0275 - accuracy: 0.9910
11808/55000 [=====>........................] - ETA: 2s - loss: 0.0279 - accuracy: 0.9909
12608/55000 [=====>........................] - ETA: 2s - loss: 0.0268 - accuracy: 0.9911
13408/55000 [======>.......................] - ETA: 2s - loss: 0.0280 - accuracy: 0.9908
14208/55000 [======>.......................] - ETA: 2s - loss: 0.0272 - accuracy: 0.9911
14976/55000 [=======>......................] - ETA: 2s - loss: 0.0264 - accuracy: 0.9914
15776/55000 [=======>......................] - ETA: 2s - loss: 0.0257 - accuracy: 0.9916
16576/55000 [========>.....................] - ETA: 2s - loss: 0.0254 - accuracy: 0.9916
17344/55000 [========>.....................] - ETA: 2s - loss: 0.0253 - accuracy: 0.9915
18112/55000 [========>.....................] - ETA: 2s - loss: 0.0248 - accuracy: 0.9918
18880/55000 [=========>....................] - ETA: 2s - loss: 0.0243 - accuracy: 0.9919
19680/55000 [=========>....................] - ETA: 2s - loss: 0.0237 - accuracy: 0.9921
20448/55000 [==========>...................] - ETA: 2s - loss: 0.0234 - accuracy: 0.9923
21248/55000 [==========>...................] - ETA: 2s - loss: 0.0234 - accuracy: 0.9923
22048/55000 [===========>..................] - ETA: 2s - loss: 0.0238 - accuracy: 0.9922
22848/55000 [===========>..................] - ETA: 2s - loss: 0.0236 - accuracy: 0.9923
23648/55000 [===========>..................] - ETA: 2s - loss: 0.0238 - accuracy: 0.9922
24448/55000 [============>.................] - ETA: 1s - loss: 0.0235 - accuracy: 0.9923
25248/55000 [============>.................] - ETA: 1s - loss: 0.0235 - accuracy: 0.9924
26048/55000 [=============>................] - ETA: 1s - loss: 0.0232 - accuracy: 0.9925
26848/55000 [=============>................] - ETA: 1s - loss: 0.0229 - accuracy: 0.9926
27648/55000 [==============>...............] - ETA: 1s - loss: 0.0226 - accuracy: 0.9927
28448/55000 [==============>...............] - ETA: 1s - loss: 0.0224 - accuracy: 0.9928
29280/55000 [==============>...............] - ETA: 1s - loss: 0.0223 - accuracy: 0.9929
30080/55000 [===============>..............] - ETA: 1s - loss: 0.0219 - accuracy: 0.9930
30880/55000 [===============>..............] - ETA: 1s - loss: 0.0218 - accuracy: 0.9930
31680/55000 [================>.............] - ETA: 1s - loss: 0.0218 - accuracy: 0.9931
32480/55000 [================>.............] - ETA: 1s - loss: 0.0218 - accuracy: 0.9931
33280/55000 [=================>............] - ETA: 1s - loss: 0.0217 - accuracy: 0.9931
34080/55000 [=================>............] - ETA: 1s - loss: 0.0217 - accuracy: 0.9930
34880/55000 [==================>...........] - ETA: 1s - loss: 0.0218 - accuracy: 0.9929
35648/55000 [==================>...........] - ETA: 1s - loss: 0.0223 - accuracy: 0.9927
36416/55000 [==================>...........] - ETA: 1s - loss: 0.0225 - accuracy: 0.9927
37216/55000 [===================>..........] - ETA: 1s - loss: 0.0225 - accuracy: 0.9927
38016/55000 [===================>..........] - ETA: 1s - loss: 0.0230 - accuracy: 0.9926
38848/55000 [====================>.........] - ETA: 1s - loss: 0.0231 - accuracy: 0.9926
39648/55000 [====================>.........] - ETA: 0s - loss: 0.0229 - accuracy: 0.9926
40480/55000 [=====================>........] - ETA: 0s - loss: 0.0228 - accuracy: 0.9927
41280/55000 [=====================>........] - ETA: 0s - loss: 0.0226 - accuracy: 0.9928
42112/55000 [=====================>........] - ETA: 0s - loss: 0.0225 - accuracy: 0.9928
42912/55000 [======================>.......] - ETA: 0s - loss: 0.0224 - accuracy: 0.9928
43712/55000 [======================>.......] - ETA: 0s - loss: 0.0224 - accuracy: 0.9928
44480/55000 [=======================>......] - ETA: 0s - loss: 0.0225 - accuracy: 0.9927
45248/55000 [=======================>......] - ETA: 0s - loss: 0.0225 - accuracy: 0.9927
46048/55000 [========================>.....] - ETA: 0s - loss: 0.0226 - accuracy: 0.9927
46848/55000 [========================>.....] - ETA: 0s - loss: 0.0229 - accuracy: 0.9926
47648/55000 [========================>.....] - ETA: 0s - loss: 0.0229 - accuracy: 0.9926
48384/55000 [=========================>....] - ETA: 0s - loss: 0.0229 - accuracy: 0.9925
49184/55000 [=========================>....] - ETA: 0s - loss: 0.0228 - accuracy: 0.9925
49984/55000 [==========================>...] - ETA: 0s - loss: 0.0226 - accuracy: 0.9926
50784/55000 [==========================>...] - ETA: 0s - loss: 0.0224 - accuracy: 0.9927
51552/55000 [===========================>..] - ETA: 0s - loss: 0.0224 - accuracy: 0.9927
52352/55000 [===========================>..] - ETA: 0s - loss: 0.0225 - accuracy: 0.9927
53152/55000 [===========================>..] - ETA: 0s - loss: 0.0224 - accuracy: 0.9928
53952/55000 [============================>.] - ETA: 0s - loss: 0.0225 - accuracy: 0.9928
54752/55000 [============================>.] - ETA: 0s - loss: 0.0223 - accuracy: 0.9928
55000/55000 [==============================] - 4s 68us/sample - loss: 0.0223 - accuracy: 0.9928 - val_loss: 0.0845 - val_accuracy: 0.9786

10000/10000 [==============================] - 0s 31us/sample - loss: 0.0463 - accuracy: 0.9801
[Finished in 42.1s]

 

ctrigger
关注
  • 1
    点赞
  • 踩
  • 4
    收藏
    觉得还不错? 一键收藏
  • 1
    评论
用tensorflow实现简单的全连接层网络
weixin_42729843的博客
11-09 5880
参考网上给的例程使用tensorflow实现了一个简单的全连接层网络的搭建。训练一个神经网络匹配二元函数。 步骤主要如下： 导入相应的包 定义添加层 准备训练数据 构建神经网络层 定义损失函数和训练优化使损失函数最小化 初始化变量，激活结构 迭代训练 1、首先导入相应的包，有tensorflow、numpy import tensorflow as tf import nump...
tensorflow实现残差网络方式(mnist数据集)
09-16
在本文中，我们将深入探讨如何使用TensorFlow框架实现残差网络（ResNet）来处理MNIST数据集。残差网络是深度学习领域的一个重要突破，由何凯明等人提出，它解决了深度神经网络中梯度消失和训练难度增大的问题。尽管...
TensorFlow：全连接层的实现
南淮北安的博客
12-30 8123
文章目录一、张量方式实现二、层方式实现 一、张量方式实现 在 TensorFlow 中，要实现全连接层，只需要定义好权值张量W 和偏置张量b，并利用 TensorFlow 提供的批量矩阵相乘函数tf.matmul()即可完成网络层的计算。 如下代码创建输入X 矩阵为???? = 2个样本，每个样本的输入特征长度为???????????? = 784，输出节点数为???????????????? =256，故定义权值矩阵W 的shape 为[78...
2024年最新TensorFlow2 一小时学会全连接层_tensorflow2 keras全连接，如何系统全面性学习大数据开发语言
最新发布
2401_84615650的博客
05-14 273
Squential (序列模型) 是各层次之间依次顺序的线性关系. 模型结构通过一个列表来制定.可以帮助我们实现全连接.
Tensorflow（十六） —— 全连接层
Cyrus_May的博客
04-16 2054
Tensorflow（十六） —— 全连接层1. 全连接层2. Multi Layers 1. 全连接层 import tensorflow as tf from tensorflow import keras # **************** fully connected layers x = tf.random.normal([1000,784]) net = keras.layers.Dense(512) out = net(x) print("out:",out.shape) prin
用tensorflow搭建mnist全连接神经网络
Mr.zwX
01-26 1152
#避免不必要的警告 import warnings warnings.filterwarnings("ignore") #初步导包 import numpy as np import matplotlib.pyplot as plt import matplotlib.cm as cm %matplotlib inline #将绘图嵌入jupyter notebook #从keras库导入mnis...
tensorflow mnist数据集全连接神经网络 python简单实现
井底小蛙的博客
05-10 304
#coding:utf-8 import tensorflow as tf from tensorflow import keras from keras import layers """ 基于tensorflow框架/mnist数据集,建一个三层全连接神经网络的10分类模型; python代码的简单实现 参考文献： Tensorflow文档 函数式API部分 https://tensorflow.google.cn/guide/keras/functional 详细解释：Epoch、batc.
network_深度学习_全连接网络_MNIST数据集_
09-30
标题中的“network_深度学习_全连接网络_MNIST数据集_”暗示了我们要讨论的是一个基于深度学习的全连接神经网络模型，该模型应用于MNIST数据集进行手写数字识别。MNIST数据集是机器学习领域的一个经典案例，包含了60...
使用TensorFlow框架演示了卷积神经网络在MNIST数据集上的应用
01-20
总结来说，卷积神经网络利用TensorFlow框架在MNIST数据集上的应用，展示了其在图像识别领域的强大能力。通过卷积层学习特征，池化层降维，ReLU层增加非线性，以及全连接层进行分类，CNN能够高效地识别手写数字。这种...
使用 TensorFlow 实现深度神经网络对 mnist 数据集进行训练和测试.zip
01-13
在本项目中，我们将探讨如何使用 TensorFlow 这一强大的开源库来实现深度神经网络，并针对经典的手写数字识别数据集 MNIST 进行训练和测试。MNIST 数据集包含60,000个训练样本和10,000个测试样本，是初学者学习深度...
使用tensorflow实现VGG网络,训练mnist数据集方式
09-16
总的来说，使用TensorFlow实现VGG网络并训练MNIST数据集是一个典型的深度学习任务，涉及到模型架构的理解、数据处理技巧以及训练策略的选择。通过这个过程，可以深入理解深度学习模型的工作原理，同时提升在实际项目...
(16)tensorflow全连接的层实现
小蜗Leon的博客
09-04 405
全连接的层实现 功能 函数代码 层实现方式 layers.Dense(units, activation) 获取 Dense 类的权值矩阵 fc.kernel 获取 Dense 类的偏置向量 fc.bias 返回待优化参数列表 fc.trainable_variables 层实现 layers.Dense(units, activation) units指定层输出节点数 activation指定激活函数 ayers.Dense在调用时会根据输入数据自动生成输入节点数 im
深度学习 之四 【TensorFlow 两层神经网络】
VictorZhang
06-05 2068
在上一篇博客中，我们实现了一层神经网络，现在，我们将通过TensorFlow实现两层神经网络；添加一个隐藏层到网络，让它建模更复杂的功能。而且隐藏层上使用非线性激活函数可以模拟非线性函数。 使用Relu在隐藏层作为激活函数，形成多层网络，如下图： 那么，这个隐藏层激活函数，我们就使用 ReLu ，Relu是非线性激活函数，又称作修正线性单元，Relu对于所有的负的输入，返回0，大于等于0的...
tensorflow 2.0 实战-mnist 手写全连接
ZHE
04-17 2292
3.6 张量前向传播向前传播计算误差计算梯度更新梯度完整代码及注解 我们已经学到： tensor 创建 索引和切片 reshape and broadcasting 数学操作 实战前向传播： out=relu{relu{relu{X@W1+b1}@W2+b2}@W3+b+3}out = relu\{relu\{relu\{X@W_1+b_1 \}@W_2+b_2 \}@W_3+b+3\}ou...
TensorFlow2可视化网络层数（keras中的plot_model模块）
Alex抱着爆米花的博客
03-22 603
今天在看论文的时候发现有keras中的plot_model模块进行可视化，效果还不错 安装说明 先装Graphviz，再装pydot_ng Graphviz 1、到官网下载graphviz的安装包 https://graphviz.org/download/ 记得勾选这个选项 2、测试： 在cmd中输入 dot -version并回车，若显示出graphviz的相关版本信息，则安装配置成功。 dot -version 3、在终端中使用命令行安装graphviz： pip install graph
Tensorflow <一> 一层全连接网络实现XOR
Pilgrim
05-14 2533
注释已经很详细了 # -*- coding: utf-8 -*- # @author: Zafedom # @date: 17/5/14 下午9:17 ''' 练习tf 使用一层的网络实现xor 这个代码很拼运气，不见得结果好 文档参考：https://yjango.gitbooks.io/superorganism/content/dai_ma_yan_shi.html 代码参考：
tensorflow全连接层降维
weixin_34284188的博客
08-03 1191
tf.layers.dense( inputs = inputs,##输入## units = units,##输出的维度## activation = activation, use_bias = use_bias,##True or False## trainable = sel...
【TensorFlow&PyTorch】深度学习常用神经网络层学习笔记
m0_63235356的博客
07-09 420
TensorFlow:Module: tf.nn | TensorFlow Core v2.9.1PyTorch:torch.nn — PyTorch 1.12 documentationPyTorch:torch.nn - PyTorch中文文档以上为文档链接，TensorFlow的很详细了，PyTorch的建议两个合着看。以下仅选取本人觉得常用的接口以及参数（即不全，想看详细内容的自行通过链接查看文档）卷积输出的宽=((输入的宽-卷积核宽+2*填充的层数)/步距)+1.输出的宽高都同理,池化层输出大
tensorflow如何查看建立的神经网络的层次信息
小白的博客
06-19 1113
想要查看你建立的网络层次信息，输入model.summary()即可，model是你的模型名字。
TensorFlow实战：深度解析MNIST数据集的三层BP神经网络
本篇文章主要探讨了在TensorFlow中操作MNIST数据集，实现神经网络进行图像分类预测的过程。文章的目的是通过实践学习神经网络的基础理论，特别是softmax函数的应用，以及如何利用TensorFlow这个强大的深度学习框架来...
评论 1
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值