AttributeError: module 'tensorflow_core.python.keras.api._v2.keras.losses' has no attribute 'softmax_cross_entropy'

Question

I have an AttributeError: module 'tensorflow_core.python.keras.api._v2.keras.losses' has no attribute 'softmax_cross_entropy' error when using tf.losses.softmax_cross_entropy. Could someone help me?

Hey, Could you post a code example describing how you got this error ? — Bruce Swain, Mar 13 '20 at 13:07
Well, I think I should use this: tf.compat.v1.losses.softmax_cross_entropy — , Mar 13 '20 at 13:32

score 0 · Answer 1 · answered Mar 13 '20 at 19:29

0

The tf.losses now point to tf.keras.losses. You can get identical behavior by using tf.losses.categorical_crossentropy with from_logits set to True

answered Mar 13 '20 at 19:29

Srihari Humbarwadi

2,532
1
10
28

score 0 · Answer 2 · answered Dec 23 '21 at 07:48

Sometimes we encounter this error, especially when running on online binders like jupyter notebook. Instead of writing

tf.losses.softmax_cross_entropy

try

loss = 'softmax_cross_entropy'

or either of the below

tf.keras.losses.CategoricalCrossentropy()
loss = 'categorical_crossentropy'

You may also want to use from_logits=True as an argument - which shall look like

tf.keras.losses.CategoricalCrossentropy(from_logits=True)

while keep metrics something like

metrics=['accuracy']

score 0 · Answer 3 · answered May 26 '23 at 03:48

0

It might be loss=tf.keras.losses.BinaryCrossentropy()

answered May 26 '23 at 03:48

aoyilmaz

139
1
10

AttributeError: module 'tensorflow_core.python.keras.api._v2.keras.losses' has no attribute 'softmax_cross_entropy'

3 Answers3