显示 Keras 模型结构和参数

废话少说，直接看代码

from keras.models import Sequential
from keras.layers import Dense
import numpy as np

np.random.seed(7)

dataset = np.loadtxt('pima-indians-diabetes.csv', delimiter = ',')
x = dataset[:, 0 : 8]
Y = dataset[:, 8]

model = Sequential()
model.add(Dense(10, input_dim=8, activation='sigmoid'))
model.add(Dense(5, activation='sigmoid'))
model.add(Dense(1, activation='sigmoid'))

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

model.fit(x=x, y=Y, epochs=100, batch_size=10, validation_split=0.2)


weights = np.array(model.get_weights())
print(weights.shape)    #(6,),三层结构，6个数组（每层一个全连接数组，一个偏置数组）

print(weights[0].shape) #(8, 10)，全连接
print(weights[0])       #[[...]]， 

print(weights[1].shape) #(10,)，偏置
print(weights[1])       #[...]

print(weights[2].shape) #(10, 5),全连接
print(weights[2])       #[[...]]

print(weights[3].shape) #(5,)，偏置
print(weights[3])       #[...]

print(weights[4].shape) #(5, 1)，全连接
print(weights[4])       #[[...]]

print(weights[5].shape) #(1,)，偏置
print(weights[5])       #[...]

运行后结果如下，

... ...
614/614 [==============================] - 0s 102us/step - loss: 0.5342 - accuracy: 0.7394 - val_loss: 0.5898 - val_accuracy: 0.7143
Epoch 98/100
614/614 [==============================] - 0s 92us/step - loss: 0.5343 - accuracy: 0.7345 - val_loss: 0.5963 - val_accuracy: 0.6883
Epoch 99/100
614/614 [==============================] - 0s 91us/step - loss: 0.5346 - accuracy: 0.7459 - val_loss: 0.5950 - val_accuracy: 0.7143
Epoch 100/100
614/614 [==============================] - 0s 94us/step - loss: 0.5343 - accuracy: 0.7378 - val_loss: 0.5992 - val_accuracy: 0.6883

(6,)
(8, 10)
[[-3.16984117e-01  7.68889785e-01 -6.09298885e-01  2.47692540e-01
  -4.28811222e-01 -3.27906728e-01 -9.68379498e-01  5.13341837e-02
  -1.25264615e-01 -4.91730124e-01]
 [ 2.11044028e-03 -4.49306726e-01  1.98752776e-01 -2.66629070e-01
  -1.40203267e-01  4.65806782e-01  1.77532241e-01 -7.84069952e-03
  -3.89660805e-01  9.84918177e-02]
 [ 2.31141478e-01 -1.12549305e-01  2.68504262e-01 -2.70382732e-01
   3.41944665e-01  2.30201110e-01 -1.10623777e-01 -1.36407129e-02
  -9.43713263e-03  1.61676109e-01]
 [ 9.12439078e-02  1.37153640e-01 -1.14635983e-03  1.55549422e-01
  -9.64731127e-02 -2.13079065e-01 -7.52463162e-01  1.12140961e-01
  -2.36525431e-01 -3.01530827e-02]
 [ 3.61231089e-01 -3.56254607e-01 -2.21178934e-01  3.97225380e-01
   1.04284659e-02 -5.62568188e-01 -4.30800825e-01  1.00965045e-01
   3.91700178e-01 -3.33490968e-01]
 [-3.33913565e-01 -1.10131040e-01 -3.23349953e-01  9.21180472e-02
   1.06020682e-01 -4.13157195e-02 -1.33292452e-01  3.75863463e-01
   1.40875010e-02  8.16177130e-02]
 [-7.82935143e-01 -1.88346773e-01 -1.04280390e-01  4.78980929e-01
  -1.18376267e+00  7.10901558e-01 -5.79004109e-01 -9.32134613e-02
   7.95687854e-01  2.71425750e-02]
 [-2.17972279e-01  2.39495710e-01 -1.78288132e-01  1.90169722e-01
  -3.07604492e-01 -3.15494061e-01 -2.90811323e-02 -2.84363925e-01
   3.06198806e-01 -4.56228644e-01]]
(10,)
[ 0.72589767 -0.28218535  0.44881618 -0.41012928  0.8835287   0.44842014
 -0.418379   -0.42579165 -0.31845236 -0.5013525 ]
(10, 5)
[[ 0.5809633   0.41399166  0.79127985  0.42651224 -0.54863864]
 [-0.64285654 -0.6585386  -1.0624979  -1.3291585   1.414629  ]
 [ 0.39915696  0.68231463  0.21638893 -0.14376499 -0.5282974 ]
 [-0.2700348  -1.2455482  -0.8031258  -0.64963794  0.7466293 ]
 [ 0.5946755   1.6657637   1.7360553   1.9690804  -2.1626413 ]
 [ 0.26416108 -0.30562374  0.27448004  0.6426197   0.51702225]
 [-0.6067534   0.02540117 -0.6464479  -0.07267946  0.2550761 ]
 [-0.08417924 -0.32010064 -0.20277596 -0.699236    0.1021447 ]
 [-1.1629095  -0.10892586 -1.0560298  -0.9182803   1.0299007 ]
 [-1.0873649  -1.3365886  -1.3386427  -1.4494324   1.0108947 ]]
(5,)
[-0.27321738 -0.218769   -0.17955495  0.05500561  0.37478566]
(5, 1)
[[-0.4963149 ]
 [-1.0566001 ]
 [-0.9775477 ]
 [-0.91774327]
 [ 1.0470334 ]]
(1,)
[0.26617393]