ModuleNotFoundError: No module named 'tensorflow.keras.layers.experimental'Bug:

[veiro]

Bug Description

when import autokeras

import autokeras as ak

Bug Reproduction

Code for reproducing the bug:

https://colab.research.google.com/github/keras-team/autokeras/blob/master/docs/ipynb/structured_data_classification.ipynb

Setup Details

Include the details about the versions of:

• OS type and version:
• Python: 3
• autokeras:
• keras-tuner:
• scikit-learn:
• numpy:
• pandas:
• tensorflow:

Additional context

in <cell line: 4>()
2 import tensorflow as tf
3
----> 4 import autokeras as ak

7 frames
/usr/local/lib/python3.10/dist-packages/autokeras/keras_layers.py in
18 from tensorflow import nest
19 from tensorflow.keras import layers
---> 20 from tensorflow.keras.layers.experimental import preprocessing
21
22 from autokeras.utils import data_utils

ModuleNotFoundError: No module named 'tensorflow.keras.layers.experimental

[Aya121298]

I have the same problem

[sorowar-cse]

SaMEEEEEEEE

[JingxinLee]

same

[2abet]

The experimental bit has been rendered redundant. Use tf.keras.layers import RandomFlip instead.
https://www.tensorflow.org/api_docs/python/tf/keras/layers/RandomFlip

If you need the full preprocessing.. you can use from tensorflow.keras.layers import *

That should work: didnt try it tho.

[odeyale2016]

from tensorflow.keras.layers import *
The above code works for me

[luckyabsoluter]

It works on Windows.

python 3.9.20
tensorflow==2.10.1
autokeras==1.1.0
numpy==1.26.4
keras-nlp==0.10.0

or
tensorflow2.10 has a bug: See: https://stackoverflow.com/questions/73770146/warningtensorflowusing-a-while-loop-for-converting-cause-there-is-no-registere

python 3.9.20
tensorflow==2.8.4
autokeras==1.0.20
numpy==1.26.4
keras-nlp==0.10.0