Skip to content

Conversation

hereismari
Copy link
Contributor

@hereismari hereismari commented Jan 5, 2019

  • Use tf.keras
  • Use default initializer
  • Print target sequence in translation training python script
  • Change number of epochs from 100 to 20 (Model is overfitting)

Using the default regularizers the model is apparently saved correctly to tensorflow.js, I've tested with the translation example.


This change is Reviewable

Copy link
Collaborator

@caisq caisq left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the PR, @mari-linhares !

Can you update the PR description to say "initializer" instead of "regularizer"?

Reviewable status: 0 of 1 approvals obtained

Copy link
Collaborator

@caisq caisq left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Actually, I'll revise the PR description for you :)

@caisq caisq merged commit 3c6557b into tensorflow:master Jan 6, 2019
@hereismari
Copy link
Contributor Author

Sorry for the wrong description and thanks for fixing it!

@hereismari hereismari changed the title Use tf.keras + default regularizers. Use tf.keras + default initializers. Jan 6, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants