Python深度学习之分类模型示例，MNIST数据集手写数字识别

最新推荐文章于 2024-05-14 15:42:33 发布
最新推荐文章于 2024-05-14 15:42:33 发布
阅读量1.7k 收藏 6
点赞数
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/97492699
版权

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。

我们把60000个训练样本分成两部分，前5000个为验证样本，后55000为训练样本。代码基本与Tensorflow官方一致，完整代码如下：（TensorFlow版本: 1.14.0，其它版本或许有差异）

#!/usr/bin/env python
import os
import tensorflow as tf

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
tf.compat.v1.enable_eager_execution()

print("TensorFlow Version:\t", tf.__version__)

mnist = tf.keras.datasets.mnist

(x_train_all, y_train_all), (x_test, y_test) = mnist.load_data()
x_train_all, x_test = x_train_all / 255.0, x_test / 255.0

x_valid, x_train = x_train_all[:5000], x_train_all[5000:]
y_valid, y_train = y_train_all[:5000], y_train_all[5000:]

model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Flatten(input_shape=(28, 28)))
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'))

model.compile(optimizer='adam',
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'],)

model.fit(x_train, y_train, epochs=5, validation_data=(x_valid, y_valid))
model.evaluate(x_test, y_test)

输出结果如下：

"C:\Program Files\Python\Python37\python.exe" "D:/Pycharm Projects/MLDemo/MLDemo.py"
TensorFlow Version:	 1.14.0
Train on 55000 samples, validate on 5000 samples
Epoch 1/5

   32/55000 [..............................] - ETA: 2:58 - loss: 2.4444 - acc: 0.1250
 1056/55000 [..............................] - ETA: 7s - loss: 1.6515 - acc: 0.4848  
 2080/55000 [>.............................] - ETA: 5s - loss: 1.2406 - acc: 0.6192
 3104/55000 [>.............................] - ETA: 4s - loss: 1.0329 - acc: 0.6811
 4192/55000 [=>............................] - ETA: 3s - loss: 0.9097 - acc: 0.7240
 5216/55000 [=>............................] - ETA: 3s - loss: 0.8194 - acc: 0.7510
 6240/55000 [==>...........................] - ETA: 3s - loss: 0.7554 - acc: 0.7708
 7168/55000 [==>...........................] - ETA: 3s - loss: 0.7100 - acc: 0.7846
 8288/55000 [===>..........................] - ETA: 2s - loss: 0.6664 - acc: 0.7986
 9376/55000 [====>.........................] - ETA: 2s - loss: 0.6278 - acc: 0.8104
10528/55000 [====>.........................] - ETA: 2s - loss: 0.6004 - acc: 0.8181
11584/55000 [=====>........................] - ETA: 2s - loss: 0.5733 - acc: 0.8266
12672/55000 [=====>........................] - ETA: 2s - loss: 0.5568 - acc: 0.8320
13792/55000 [======>.......................] - ETA: 2s - loss: 0.5393 - acc: 0.8381
14912/55000 [=======>......................] - ETA: 2s - loss: 0.5257 - acc: 0.8425
15904/55000 [=======>......................] - ETA: 2s - loss: 0.5144 - acc: 0.8462
16928/55000 [========>.....................] - ETA: 2s - loss: 0.5006 - acc: 0.8506
18048/55000 [========>.....................] - ETA: 1s - loss: 0.4886 - acc: 0.8543
19200/55000 [=========>....................] - ETA: 1s - loss: 0.4773 - acc: 0.8573
20256/55000 [==========>...................] - ETA: 1s - loss: 0.4660 - acc: 0.8609
21344/55000 [==========>...................] - ETA: 1s - loss: 0.4550 - acc: 0.8639
22464/55000 [===========>..................] - ETA: 1s - loss: 0.4442 - acc: 0.8674
23584/55000 [===========>..................] - ETA: 1s - loss: 0.4363 - acc: 0.8698
24608/55000 [============>.................] - ETA: 1s - loss: 0.4298 - acc: 0.8717
25696/55000 [=============>................] - ETA: 1s - loss: 0.4223 - acc: 0.8742
26816/55000 [=============>................] - ETA: 1s - loss: 0.4147 - acc: 0.8764
27968/55000 [==============>...............] - ETA: 1s - loss: 0.4070 - acc: 0.8784
29120/55000 [==============>...............] - ETA: 1s - loss: 0.4006 - acc: 0.8805
30240/55000 [===============>..............] - ETA: 1s - loss: 0.3949 - acc: 0.8824
31360/55000 [================>.............] - ETA: 1s - loss: 0.3903 - acc: 0.8838
32512/55000 [================>.............] - ETA: 1s - loss: 0.3842 - acc: 0.8858
33696/55000 [=================>............] - ETA: 1s - loss: 0.3797 - acc: 0.8872
34720/55000 [=================>............] - ETA: 1s - loss: 0.3743 - acc: 0.8886
35776/55000 [==================>...........] - ETA: 0s - loss: 0.3699 - acc: 0.8900
36864/55000 [===================>..........] - ETA: 0s - loss: 0.3651 - acc: 0.8914
38016/55000 [===================>..........] - ETA: 0s - loss: 0.3604 - acc: 0.8928
39136/55000 [====================>.........] - ETA: 0s - loss: 0.3563 - acc: 0.8940
40192/55000 [====================>.........] - ETA: 0s - loss: 0.3526 - acc: 0.8951
41248/55000 [=====================>........] - ETA: 0s - loss: 0.3492 - acc: 0.8962
42400/55000 [======================>.......] - ETA: 0s - loss: 0.3460 - acc: 0.8973
43488/55000 [======================>.......] - ETA: 0s - loss: 0.3423 - acc: 0.8984
44512/55000 [=======================>......] - ETA: 0s - loss: 0.3395 - acc: 0.8993
45440/55000 [=======================>......] - ETA: 0s - loss: 0.3365 - acc: 0.9002
46528/55000 [========================>.....] - ETA: 0s - loss: 0.3335 - acc: 0.9008
47648/55000 [========================>.....] - ETA: 0s - loss: 0.3307 - acc: 0.9017
48736/55000 [=========================>....] - ETA: 0s - loss: 0.3278 - acc: 0.9027
49856/55000 [==========================>...] - ETA: 0s - loss: 0.3252 - acc: 0.9035
50944/55000 [==========================>...] - ETA: 0s - loss: 0.3221 - acc: 0.9045
51968/55000 [===========================>..] - ETA: 0s - loss: 0.3199 - acc: 0.9051
53088/55000 [===========================>..] - ETA: 0s - loss: 0.3176 - acc: 0.9059
54112/55000 [============================>.] - ETA: 0s - loss: 0.3152 - acc: 0.9066
55000/55000 [==============================] - 3s 55us/sample - loss: 0.3129 - acc: 0.9072 - val_loss: 0.1477 - val_acc: 0.9608
Epoch 2/5

   32/55000 [..............................] - ETA: 8s - loss: 0.3416 - acc: 0.8750
 1088/55000 [..............................] - ETA: 2s - loss: 0.2193 - acc: 0.9366
 2144/55000 [>.............................] - ETA: 2s - loss: 0.1951 - acc: 0.9459
 3200/55000 [>.............................] - ETA: 2s - loss: 0.1856 - acc: 0.9491
 4320/55000 [=>............................] - ETA: 2s - loss: 0.1806 - acc: 0.9488
 5376/55000 [=>............................] - ETA: 2s - loss: 0.1830 - acc: 0.9488
 6432/55000 [==>...........................] - ETA: 2s - loss: 0.1749 - acc: 0.9496
 7584/55000 [===>..........................] - ETA: 2s - loss: 0.1707 - acc: 0.9499
 8736/55000 [===>..........................] - ETA: 2s - loss: 0.1692 - acc: 0.9497
 9792/55000 [====>.........................] - ETA: 2s - loss: 0.1684 - acc: 0.9504
10944/55000 [====>.........................] - ETA: 2s - loss: 0.1671 - acc: 0.9501
12096/55000 [=====>........................] - ETA: 2s - loss: 0.1670 - acc: 0.9506
13248/55000 [======>.......................] - ETA: 1s - loss: 0.1670 - acc: 0.9505
14432/55000 [======>.......................] - ETA: 1s - loss: 0.1661 - acc: 0.9504
15552/55000 [=======>......................] - ETA: 1s - loss: 0.1682 - acc: 0.9500
16672/55000 [========>.....................] - ETA: 1s - loss: 0.1671 - acc: 0.9507
17760/55000 [========>.....................] - ETA: 1s - loss: 0.1675 - acc: 0.9505
18912/55000 [=========>....................] - ETA: 1s - loss: 0.1660 - acc: 0.9510
19968/55000 [=========>....................] - ETA: 1s - loss: 0.1662 - acc: 0.9510
21120/55000 [==========>...................] - ETA: 1s - loss: 0.1653 - acc: 0.9510
22272/55000 [===========>..................] - ETA: 1s - loss: 0.1640 - acc: 0.9517
23424/55000 [===========>..................] - ETA: 1s - loss: 0.1633 - acc: 0.9523
24544/55000 [============>.................] - ETA: 1s - loss: 0.1625 - acc: 0.9527
25664/55000 [============>.................] - ETA: 1s - loss: 0.1620 - acc: 0.9530
26688/55000 [=============>................] - ETA: 1s - loss: 0.1619 - acc: 0.9530
27840/55000 [==============>...............] - ETA: 1s - loss: 0.1611 - acc: 0.9531
28960/55000 [==============>...............] - ETA: 1s - loss: 0.1620 - acc: 0.9529
30080/55000 [===============>..............] - ETA: 1s - loss: 0.1619 - acc: 0.9527
31232/55000 [================>.............] - ETA: 1s - loss: 0.1610 - acc: 0.9529
32384/55000 [================>.............] - ETA: 1s - loss: 0.1605 - acc: 0.9528
33472/55000 [=================>............] - ETA: 0s - loss: 0.1595 - acc: 0.9532
34592/55000 [=================>............] - ETA: 0s - loss: 0.1589 - acc: 0.9533
35744/55000 [==================>...........] - ETA: 0s - loss: 0.1588 - acc: 0.9534
36896/55000 [===================>..........] - ETA: 0s - loss: 0.1587 - acc: 0.9534
38016/55000 [===================>..........] - ETA: 0s - loss: 0.1575 - acc: 0.9537
39104/55000 [====================>.........] - ETA: 0s - loss: 0.1566 - acc: 0.9539
40256/55000 [====================>.........] - ETA: 0s - loss: 0.1561 - acc: 0.9541
41376/55000 [=====================>........] - ETA: 0s - loss: 0.1562 - acc: 0.9541
42496/55000 [======================>.......] - ETA: 0s - loss: 0.1555 - acc: 0.9542
43648/55000 [======================>.......] - ETA: 0s - loss: 0.1545 - acc: 0.9543
44800/55000 [=======================>......] - ETA: 0s - loss: 0.1539 - acc: 0.9543
45952/55000 [========================>.....] - ETA: 0s - loss: 0.1537 - acc: 0.9544
47072/55000 [========================>.....] - ETA: 0s - loss: 0.1532 - acc: 0.9545
48192/55000 [=========================>....] - ETA: 0s - loss: 0.1526 - acc: 0.9548
49184/55000 [=========================>....] - ETA: 0s - loss: 0.1528 - acc: 0.9548
50336/55000 [==========================>...] - ETA: 0s - loss: 0.1522 - acc: 0.9550
51424/55000 [===========================>..] - ETA: 0s - loss: 0.1521 - acc: 0.9551
52544/55000 [===========================>..] - ETA: 0s - loss: 0.1521 - acc: 0.9549
53696/55000 [============================>.] - ETA: 0s - loss: 0.1516 - acc: 0.9550
54848/55000 [============================>.] - ETA: 0s - loss: 0.1510 - acc: 0.9551
55000/55000 [==============================] - 3s 48us/sample - loss: 0.1512 - acc: 0.9551 - val_loss: 0.1084 - val_acc: 0.9680
Epoch 3/5

   32/55000 [..............................] - ETA: 6s - loss: 0.0403 - acc: 1.0000
 1184/55000 [..............................] - ETA: 2s - loss: 0.1262 - acc: 0.9654
 2304/55000 [>.............................] - ETA: 2s - loss: 0.1259 - acc: 0.9661
 3392/55000 [>.............................] - ETA: 2s - loss: 0.1216 - acc: 0.9652
 4544/55000 [=>............................] - ETA: 2s - loss: 0.1202 - acc: 0.9650
 5696/55000 [==>...........................] - ETA: 2s - loss: 0.1192 - acc: 0.9631
 6880/55000 [==>...........................] - ETA: 2s - loss: 0.1125 - acc: 0.9658
 8000/55000 [===>..........................] - ETA: 2s - loss: 0.1099 - acc: 0.9665
 9152/55000 [===>..........................] - ETA: 2s - loss: 0.1088 - acc: 0.9667
10304/55000 [====>.........................] - ETA: 2s - loss: 0.1080 - acc: 0.9664
11488/55000 [=====>........................] - ETA: 1s - loss: 0.1066 - acc: 0.9672
12576/55000 [=====>........................] - ETA: 1s - loss: 0.1069 - acc: 0.9676
13696/55000 [======>.......................] - ETA: 1s - loss: 0.1061 - acc: 0.9678
14848/55000 [=======>......................] - ETA: 1s - loss: 0.1069 - acc: 0.9670
16000/55000 [=======>......................] - ETA: 1s - loss: 0.1075 - acc: 0.9668
17120/55000 [========>.....................] - ETA: 1s - loss: 0.1084 - acc: 0.9661
18240/55000 [========>.....................] - ETA: 1s - loss: 0.1079 - acc: 0.9663
19392/55000 [=========>....................] - ETA: 1s - loss: 0.1081 - acc: 0.9663
20544/55000 [==========>...................] - ETA: 1s - loss: 0.1076 - acc: 0.9667
21696/55000 [==========>...................] - ETA: 1s - loss: 0.1069 - acc: 0.9669
22816/55000 [===========>..................] - ETA: 1s - loss: 0.1070 - acc: 0.9669
24000/55000 [============>.................] - ETA: 1s - loss: 0.1079 - acc: 0.9667
25152/55000 [============>.................] - ETA: 1s - loss: 0.1079 - acc: 0.9671
26272/55000 [=============>................] - ETA: 1s - loss: 0.1083 - acc: 0.9670
27392/55000 [=============>................] - ETA: 1s - loss: 0.1093 - acc: 0.9667
28576/55000 [==============>...............] - ETA: 1s - loss: 0.1100 - acc: 0.9667
29760/55000 [===============>..............] - ETA: 1s - loss: 0.1096 - acc: 0.9668
30880/55000 [===============>..............] - ETA: 1s - loss: 0.1096 - acc: 0.9669
32000/55000 [================>.............] - ETA: 1s - loss: 0.1102 - acc: 0.9667
33120/55000 [=================>............] - ETA: 0s - loss: 0.1100 - acc: 0.9668
34240/55000 [=================>............] - ETA: 0s - loss: 0.1101 - acc: 0.9669
35200/55000 [==================>...........] - ETA: 0s - loss: 0.1100 - acc: 0.9668
36352/55000 [==================>...........] - ETA: 0s - loss: 0.1103 - acc: 0.9667
37472/55000 [===================>..........] - ETA: 0s - loss: 0.1104 - acc: 0.9666
38624/55000 [====================>.........] - ETA: 0s - loss: 0.1104 - acc: 0.9668
39776/55000 [====================>.........] - ETA: 0s - loss: 0.1105 - acc: 0.9668
40896/55000 [=====================>........] - ETA: 0s - loss: 0.1108 - acc: 0.9668
42016/55000 [=====================>........] - ETA: 0s - loss: 0.1106 - acc: 0.9670
43168/55000 [======================>.......] - ETA: 0s - loss: 0.1098 - acc: 0.9672
44320/55000 [=======================>......] - ETA: 0s - loss: 0.1104 - acc: 0.9671
45440/55000 [=======================>......] - ETA: 0s - loss: 0.1109 - acc: 0.9669
46496/55000 [========================>.....] - ETA: 0s - loss: 0.1109 - acc: 0.9669
47648/55000 [========================>.....] - ETA: 0s - loss: 0.1109 - acc: 0.9668
48800/55000 [=========================>....] - ETA: 0s - loss: 0.1109 - acc: 0.9668
49920/55000 [==========================>...] - ETA: 0s - loss: 0.1107 - acc: 0.9669
51040/55000 [==========================>...] - ETA: 0s - loss: 0.1111 - acc: 0.9667
52192/55000 [===========================>..] - ETA: 0s - loss: 0.1116 - acc: 0.9665
53344/55000 [============================>.] - ETA: 0s - loss: 0.1115 - acc: 0.9665
54464/55000 [============================>.] - ETA: 0s - loss: 0.1119 - acc: 0.9665
55000/55000 [==============================] - 3s 47us/sample - loss: 0.1119 - acc: 0.9664 - val_loss: 0.0966 - val_acc: 0.9714
Epoch 4/5

   32/55000 [..............................] - ETA: 5s - loss: 0.0525 - acc: 0.9688
 1184/55000 [..............................] - ETA: 2s - loss: 0.1083 - acc: 0.9645
 2240/55000 [>.............................] - ETA: 2s - loss: 0.0960 - acc: 0.9692
 3360/55000 [>.............................] - ETA: 2s - loss: 0.1016 - acc: 0.9685
 4512/55000 [=>............................] - ETA: 2s - loss: 0.1044 - acc: 0.9694
 5632/55000 [==>...........................] - ETA: 2s - loss: 0.0989 - acc: 0.9714
 6720/55000 [==>...........................] - ETA: 2s - loss: 0.0991 - acc: 0.9707
 7840/55000 [===>..........................] - ETA: 2s - loss: 0.1021 - acc: 0.9694
 8992/55000 [===>..........................] - ETA: 2s - loss: 0.1013 - acc: 0.9694
10144/55000 [====>.........................] - ETA: 2s - loss: 0.1003 - acc: 0.9688
11168/55000 [=====>........................] - ETA: 2s - loss: 0.1003 - acc: 0.9680
12288/55000 [=====>........................] - ETA: 1s - loss: 0.0979 - acc: 0.9690
13408/55000 [======>.......................] - ETA: 1s - loss: 0.0965 - acc: 0.9699
14592/55000 [======>.......................] - ETA: 1s - loss: 0.0944 - acc: 0.9706
15744/55000 [=======>......................] - ETA: 1s - loss: 0.0940 - acc: 0.9708
16864/55000 [========>.....................] - ETA: 1s - loss: 0.0930 - acc: 0.9709
17984/55000 [========>.....................] - ETA: 1s - loss: 0.0929 - acc: 0.9710
19136/55000 [=========>....................] - ETA: 1s - loss: 0.0930 - acc: 0.9712
20288/55000 [==========>...................] - ETA: 1s - loss: 0.0928 - acc: 0.9713
21376/55000 [==========>...................] - ETA: 1s - loss: 0.0935 - acc: 0.9710
22464/55000 [===========>..................] - ETA: 1s - loss: 0.0929 - acc: 0.9712
23520/55000 [===========>..................] - ETA: 1s - loss: 0.0948 - acc: 0.9708
24576/55000 [============>.................] - ETA: 1s - loss: 0.0941 - acc: 0.9709
25632/55000 [============>.................] - ETA: 1s - loss: 0.0930 - acc: 0.9712
26784/55000 [=============>................] - ETA: 1s - loss: 0.0927 - acc: 0.9712
27904/55000 [==============>...............] - ETA: 1s - loss: 0.0923 - acc: 0.9713
29056/55000 [==============>...............] - ETA: 1s - loss: 0.0935 - acc: 0.9711
30144/55000 [===============>..............] - ETA: 1s - loss: 0.0946 - acc: 0.9706
31232/55000 [================>.............] - ETA: 1s - loss: 0.0946 - acc: 0.9703
32352/55000 [================>.............] - ETA: 1s - loss: 0.0950 - acc: 0.9703
33504/55000 [=================>............] - ETA: 0s - loss: 0.0949 - acc: 0.9703
34624/55000 [=================>............] - ETA: 0s - loss: 0.0948 - acc: 0.9704
35744/55000 [==================>...........] - ETA: 0s - loss: 0.0941 - acc: 0.9705
36896/55000 [===================>..........] - ETA: 0s - loss: 0.0934 - acc: 0.9707
38016/55000 [===================>..........] - ETA: 0s - loss: 0.0930 - acc: 0.9710
39136/55000 [====================>.........] - ETA: 0s - loss: 0.0932 - acc: 0.9708
40256/55000 [====================>.........] - ETA: 0s - loss: 0.0932 - acc: 0.9710
41408/55000 [=====================>........] - ETA: 0s - loss: 0.0929 - acc: 0.9711
42560/55000 [======================>.......] - ETA: 0s - loss: 0.0927 - acc: 0.9712
43552/55000 [======================>.......] - ETA: 0s - loss: 0.0926 - acc: 0.9712
44672/55000 [=======================>......] - ETA: 0s - loss: 0.0930 - acc: 0.9711
45824/55000 [=======================>......] - ETA: 0s - loss: 0.0930 - acc: 0.9712
46944/55000 [========================>.....] - ETA: 0s - loss: 0.0928 - acc: 0.9712
48096/55000 [=========================>....] - ETA: 0s - loss: 0.0926 - acc: 0.9712
49216/55000 [=========================>....] - ETA: 0s - loss: 0.0923 - acc: 0.9714
50336/55000 [==========================>...] - ETA: 0s - loss: 0.0927 - acc: 0.9712
51392/55000 [===========================>..] - ETA: 0s - loss: 0.0925 - acc: 0.9712
52448/55000 [===========================>..] - ETA: 0s - loss: 0.0927 - acc: 0.9711
53536/55000 [============================>.] - ETA: 0s - loss: 0.0925 - acc: 0.9713
54688/55000 [============================>.] - ETA: 0s - loss: 0.0924 - acc: 0.9714
55000/55000 [==============================] - 3s 48us/sample - loss: 0.0924 - acc: 0.9714 - val_loss: 0.0799 - val_acc: 0.9770
Epoch 5/5

   32/55000 [..............................] - ETA: 5s - loss: 0.1389 - acc: 0.9375
 1088/55000 [..............................] - ETA: 2s - loss: 0.0770 - acc: 0.9715
 2144/55000 [>.............................] - ETA: 2s - loss: 0.0785 - acc: 0.9748
 3200/55000 [>.............................] - ETA: 2s - loss: 0.0706 - acc: 0.9775
 4256/55000 [=>............................] - ETA: 2s - loss: 0.0716 - acc: 0.9777
 5248/55000 [=>............................] - ETA: 2s - loss: 0.0713 - acc: 0.9777
 6240/55000 [==>...........................] - ETA: 2s - loss: 0.0735 - acc: 0.9771
 7296/55000 [==>...........................] - ETA: 2s - loss: 0.0732 - acc: 0.9775
 8352/55000 [===>..........................] - ETA: 2s - loss: 0.0762 - acc: 0.9769
 9408/55000 [====>.........................] - ETA: 2s - loss: 0.0772 - acc: 0.9760
10464/55000 [====>.........................] - ETA: 2s - loss: 0.0758 - acc: 0.9762
11552/55000 [=====>........................] - ETA: 2s - loss: 0.0755 - acc: 0.9759
12608/55000 [=====>........................] - ETA: 2s - loss: 0.0786 - acc: 0.9752
13664/55000 [======>.......................] - ETA: 2s - loss: 0.0776 - acc: 0.9755
14816/55000 [=======>......................] - ETA: 1s - loss: 0.0764 - acc: 0.9758
16000/55000 [=======>......................] - ETA: 1s - loss: 0.0765 - acc: 0.9756
17152/55000 [========>.....................] - ETA: 1s - loss: 0.0769 - acc: 0.9754
18240/55000 [========>.....................] - ETA: 1s - loss: 0.0776 - acc: 0.9753
19392/55000 [=========>....................] - ETA: 1s - loss: 0.0776 - acc: 0.9754
20576/55000 [==========>...................] - ETA: 1s - loss: 0.0779 - acc: 0.9753
21728/55000 [==========>...................] - ETA: 1s - loss: 0.0770 - acc: 0.9754
22784/55000 [===========>..................] - ETA: 1s - loss: 0.0773 - acc: 0.9754
23872/55000 [============>.................] - ETA: 1s - loss: 0.0773 - acc: 0.9755
24992/55000 [============>.................] - ETA: 1s - loss: 0.0775 - acc: 0.9755
26080/55000 [=============>................] - ETA: 1s - loss: 0.0773 - acc: 0.9757
27008/55000 [=============>................] - ETA: 1s - loss: 0.0773 - acc: 0.9755
27872/55000 [==============>...............] - ETA: 1s - loss: 0.0783 - acc: 0.9752
28960/55000 [==============>...............] - ETA: 1s - loss: 0.0789 - acc: 0.9750
29984/55000 [===============>..............] - ETA: 1s - loss: 0.0801 - acc: 0.9744
30592/55000 [===============>..............] - ETA: 1s - loss: 0.0797 - acc: 0.9745
31136/55000 [===============>..............] - ETA: 1s - loss: 0.0795 - acc: 0.9744
32032/55000 [================>.............] - ETA: 1s - loss: 0.0794 - acc: 0.9744
33024/55000 [=================>............] - ETA: 1s - loss: 0.0793 - acc: 0.9745
34080/55000 [=================>............] - ETA: 1s - loss: 0.0789 - acc: 0.9748
35200/55000 [==================>...........] - ETA: 0s - loss: 0.0785 - acc: 0.9749
36256/55000 [==================>...........] - ETA: 0s - loss: 0.0790 - acc: 0.9746
37440/55000 [===================>..........] - ETA: 0s - loss: 0.0785 - acc: 0.9749
38496/55000 [===================>..........] - ETA: 0s - loss: 0.0784 - acc: 0.9749
39552/55000 [====================>.........] - ETA: 0s - loss: 0.0785 - acc: 0.9749
40608/55000 [=====================>........] - ETA: 0s - loss: 0.0786 - acc: 0.9749
41696/55000 [=====================>........] - ETA: 0s - loss: 0.0785 - acc: 0.9749
42816/55000 [======================>.......] - ETA: 0s - loss: 0.0782 - acc: 0.9749
43904/55000 [======================>.......] - ETA: 0s - loss: 0.0776 - acc: 0.9750
44960/55000 [=======================>......] - ETA: 0s - loss: 0.0775 - acc: 0.9750
46048/55000 [========================>.....] - ETA: 0s - loss: 0.0771 - acc: 0.9752
47136/55000 [========================>.....] - ETA: 0s - loss: 0.0772 - acc: 0.9752
48224/55000 [=========================>....] - ETA: 0s - loss: 0.0770 - acc: 0.9753
49248/55000 [=========================>....] - ETA: 0s - loss: 0.0770 - acc: 0.9754
50272/55000 [==========================>...] - ETA: 0s - loss: 0.0769 - acc: 0.9754
51232/55000 [==========================>...] - ETA: 0s - loss: 0.0771 - acc: 0.9753
52192/55000 [===========================>..] - ETA: 0s - loss: 0.0770 - acc: 0.9754
52992/55000 [===========================>..] - ETA: 0s - loss: 0.0771 - acc: 0.9753
53568/55000 [============================>.] - ETA: 0s - loss: 0.0771 - acc: 0.9754
54496/55000 [============================>.] - ETA: 0s - loss: 0.0772 - acc: 0.9752
55000/55000 [==============================] - 3s 52us/sample - loss: 0.0773 - acc: 0.9752 - val_loss: 0.0723 - val_acc: 0.9788

   32/10000 [..............................] - ETA: 0s - loss: 0.0410 - acc: 0.9688
 2496/10000 [======>.......................] - ETA: 0s - loss: 0.1088 - acc: 0.9671
 5280/10000 [==============>...............] - ETA: 0s - loss: 0.0977 - acc: 0.9705
 7936/10000 [======================>.......] - ETA: 0s - loss: 0.0810 - acc: 0.9757
10000/10000 [==============================] - 0s 22us/sample - loss: 0.0756 - acc: 0.9776

Process finished with exit code 0

准确率达到97.76%，还是相当不错的吧。至此，深度学习之“Hello World”示例完毕~~~

ctrigger
关注
  • 0
    点赞
  • 踩
  • 6
    收藏
    觉得还不错? 一键收藏
  • 0
    评论
Tensorflow实现逻辑回归（使用MNIST数据集）
MrLittleDog的博客
09-02 5751
今天我们讲一下如何使用Tensorflow实现逻辑回归，代码中采用了MNIST数据集。 首先，我们去获取MNIST数据集（下载链接：http://yann.lecun.com/exdb/mnist/），并将其读取到程序中。代码实现如下所示： import tensorflow as tf from tensorflow.examples.tutorials.mnist import inpu...
机器学习之算法案例手写数字识别
qq_46556714的博客
05-21 1780
算法案例手写数字识别 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。 选择算法，并保存模型 import pickle from sklearn.linear_model import LogisticRegression from sklearn.metrics import accuracy_score from sklearn.ensemble import RandomForestC
Pytorch深度学习案例1：mnist手写数字识别
最新发布
2301_76606951的博客
05-14 844
的 Iterable。输入层为(1,28,28)，经过第一个卷积层（卷积核为3，stride为1，padding为0）变为（32，26，26），经过第一个池化层（池化核为2）变为（32，13，13），经过第二个卷积层（卷积核为3，stride为1，padding为0）变为（64，11，11），经过第二个池化层（池化核为2）变为（64，5，5）(pred.argmax(1) == y).type(torch.float).sum().item()表示计算预测正确的样本数量，并将其作为一个标量值返回。
[PyTorch]——MNIST数据集的简单深度神经网络实现
咭咭熊的博客
03-05 740
1.题目描述： 使用Pytorch实现一个简单的深度神经网络，试图用的数据集是MNIST,预期结果可以达到98%左右的准确率。 该神经网络由1个输入层、1个全连接层结构的隐含层和1个输出层构建。 2.数据集的下载 2.具体实现过程 ...
MNIST数据集介绍
傅里叶也头大
01-21 918
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。点此下载Minist数据集 该站点包含四个文件 直接下载下来的数据是无法通过解压或者应用程序打开的，因为这些文件不是任何标准的图像格式而是以字节的形式进行存储的，所以必须编写程序来打开它。 ...
手写数字识别及python实现
star_sky_sc的博客
05-10 1万+
目录 1、总体流程 2、代码实现 下载数据集 确定激活函数、损失函数、计算梯度函数等 神经网络的搭建 模型的训练与验证 测试模型的泛化能力 1、总体流程 step1：下载数据集、读取数据 step2：搭建神经网络（确定输出层、隐藏层（层数）、输出层的结构） step3：初始化偏置和权重 step4：设置损失函数、激活函数 step5：设置超参数 step6：神经网络训练数据（通过误差反向传播求导、学习） step7：测试验证数据集（确定Loss、精确度） step8...
MNIST_never8fu_深度学习python_MNIST数据集_MNIST识别_MNIST_
10-01
MNIST数据集是机器学习领域一个非常经典的数据集，主要用于训练和测试手写数字识别的算法。这个数据集由Yann LeCun等人在1998年创建，包含60,000个训练样本和10,000个测试样本，每个样本都是28x28像素的灰度图像，...
Python手写数字识别【Mnist数据集】
01-04
手写数字识别要求利用MNIST数据集里的70000张手写体数字的图像，建立神经网络模型，进行0到9的分类，并能够对其他来源的图片进行识别，识别准确率大于97%。图片示例如下。 该设计要求学生基于TensorFlow深度学习平台...
手写数字识别MNIST数据集及卷积神经网络代码（含注释）
04-09
MNIST数据集是机器学习领域一个经典且广泛使用的数据集，主要用于训练和测试手写数字识别的算法。这个数据集包含60,000个训练样本和10,000个测试样本，每个样本都是28x28像素的灰度图像，对应0到9的手写数字。在深度...
Python深度学习项目：CNN手写数字识别测试准确率
05-04
这个任务是基于经典的MNIST数据集，它包含大量的手写数字图片，是许多机器学习和深度学习教程的首选示例。 **MNIST数据集** MNIST数据集是一个广泛使用的数据库，包含60,000个训练样本和10,000个测试样本。每个样本...
Python tensorflow实现mnist手写数字识别示例【非卷积与卷积实现】
09-18
主要介绍了Python tensorflow实现mnist手写数字识别,结合实例形式分析了基于tensorflow模块使用非卷积与卷积算法实现手写数字识别的具体操作技巧,需要的朋友可以参考下
CNN网络实现手写数字（MNIST）识别 代码分析
Sylvia136655的博客
07-19 5620
CNN网络实现手写数字（MNIST）识别 代码分析(自学用) Github代码源文件 本文是学习了使用Pytorch框架的CNN网络实现手写数字（MNIST）识别 #导入需要的包 import numpy as np //第三方库，用于进行科学计算 import torch from torch import nn from PIL import Image // Python Image Library,python第三方图像处理库 import matplotlib.pyplot as plt
用PyTorch实现MNIST手写数字识别（最新，非常详细）
热门推荐
qq_45588019的博客
01-20 6万+
1 MNIST数据集(提取码: zm7q )
Python代码实现简单的MNIST手写数字识别（适合初学者看）
zugexiaodui的博客
08-12 5万+
补充：由于很多同学找我要原数据集和代码，所以我上传到了资源里，https://download.csdn.net/download/zugexiaodui/10913834 初学机器学习，第一步是做一个简单的手写数字识别，我选用的是MNIST数据集（用其他数据集也可以，原理都差不多），算法是KNN（下载库直接调用函数，算法的具体实现没有过多关心）。在网上也看到过MNIST数据集的Python代码...
python 简单使用MNIST数据集+卷积神经网络实现手写数字识别
做减求空
06-03 2267
import tensorflow as tf import tensorflow.examples.tutorials.mnist.input_data as input_data #1 读取数据 mnist = input_data.read_data_sets("MNIST_data/", one_hot=True) #2 建立模型 使用卷积神经网络 #2.1 输入图像与标签 x = tf.placeholder("float", shape = [None, 28,28,1]) y_ = .
mnist手写数字识别python_基于Tensorflow的MNIST手写数字识别分类
weixin_39926943的博客
12-03 192
本文实例为大家分享了基于Tensorflow的MNIST手写数字识别分类的具体实现代码，供大家参考，具体内容如下代码如下：import tensorflow as tfimport numpy as npfrom tensorflow.examples.tutorials.mnist import input_datafrom tensorflow.contrib.tensorboard.plug...
神经网络的学习-搭建神经网络实现mnist数据集分类
泥签的博客
02-06 5482
从零开始，不用框架，搭建一个俩层人工神经网络实现mnist手写数据集图像分类。
深度学习视觉领域常用数据集汇总
wonengguwozai的博客
12-08 883
原文地址：http://www.dlworld.cn/YeJieDongTai/1550.html [导读]“大数据时代”，数据为王!无论是数据挖掘还是目前大热的深度学习领域都离不开“大数据”。大公司们一般会有自己的数据，但对于创业公司或是高校老师、学生来说，“Where can I get large datasets open to the public?”是不得不面对的一个问题。
使用MNIST数据集训练手写数字识别模型python
05-27
好的，我可以为您提供一个基于Python的手写数字识别模型的训练代码示例，使用的是MNIST数据集。 首先，需要安装必要的库，包括TensorFlow和Keras。您可以使用以下命令来安装它们： ``` pip install tensorflow ...
评论
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