/Python/When training= false still dropout

When training= false still dropout

from keras import layers
from keras import models
from keras import backend as K
import numpy as np

inp = layers.Input(shape=(10,))
out = layers.Dropout(0.5)(inp)

model = models.Model(inp, out)
model.layers[-1].trainable = False  # set dropout layer as non-trainable
model.compile(optimizer='adam', loss='mse') # IMPORTANT: we must always compile model after changing `trainable` attribute

# create a custom backend function so that we can control the learning phase
func = K.function(model.inputs + [K.learning_phase()], model.outputs)

x = np.ones((1,10))
# learning phase = 1, i.e. training mode
print(func([x, 1]))
# the output will be:
[array([[2., 2., 2., 0., 0., 2., 2., 2., 0., 0.]], dtype=float32)]
# as you can see some of the neurons have been dropped

# now set learning phase = 0, i.e test mode
print(func([x, 0]))
# the output will be:
[array([[1., 1., 1., 1., 1., 1., 1., 1., 1., 1.]], dtype=float32)]
# unsurprisingly, no neurons have been dropped in test phase

Source: stackoverflow.com

Similar codes

2 votes238 views1 answer
Frequency count of values in pandas dataframe
Python
0 vote306 views1 answer
How to calculate log 10 in python
Python
0 vote242 views1 answer
Binary table dataframe
Python
2 votes227 views1 answer
Python click buttons on websites
Python
0 vote236 views1 answer
Python log10
Python
1 vote246 views1 answer
Panda dataframe read csv change string to float
Python
6 votes263 views2 answers
Python pandas series to dataframe
Python
0 vote210 views1 answer
Python running standard deviation
Python
1 vote225 views1 answer
Aws lambda Unable to import module 'lambda_function': No module named 'requests'
Python
0 vote218 views1 answer
Jupyter notebook print all rows dataframe
Python