You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Based on the Global Settings and Imports in the example notebook for multinomial_vae, it requires tensorflow==2.2.0-rc1 and keras==2.3.1. After I installed these two libraries and ran the code, I got AttributeError: module 'tensorflow.keras.initializers' has no attribute 'truncated_normal'\n when running multinomial_vae.py. According to Tensorflow 2.2 documentation, there isn't a class called truncated_normal but there is one named TruncatedNormal.
How do we replicate the issue?
Running the notebook would probably encounter this error.
Expected behavior (i.e. solution)
Change truncated_normal to TruncatedNormal in here and here should resolve this problem.
The text was updated successfully, but these errors were encountered:
Description
Based on the Global Settings and Imports in the example notebook for multinomial_vae, it requires
tensorflow==2.2.0-rc1
andkeras==2.3.1
. After I installed these two libraries and ran the code, I gotAttributeError: module 'tensorflow.keras.initializers' has no attribute 'truncated_normal'\n
when running multinomial_vae.py. According to Tensorflow 2.2 documentation, there isn't a class calledtruncated_normal
but there is one namedTruncatedNormal
.How do we replicate the issue?
Running the notebook would probably encounter this error.
Expected behavior (i.e. solution)
Change
truncated_normal
toTruncatedNormal
in here and here should resolve this problem.The text was updated successfully, but these errors were encountered: