Skip to content

Conversation

@AakashKumarNain
Copy link
Contributor

No description provided.

Copy link
Contributor

@fchollet fchollet left a comment

Choose a reason for hiding this comment

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

Thank you for the PR! This will be a great new example.

@AakashKumarNain
Copy link
Contributor Author

Thanks @fchollet for the review. I will make the necessary changes

@AakashKumarNain
Copy link
Contributor Author

@fchollet I have made all the changes except for removing the keras.backend part. I need to check if I need to change anything if I am directly using tf.nn instead of keras.backend

Copy link
Contributor

@fchollet fchollet left a comment

Choose a reason for hiding this comment

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

Looks great! Thanks for the update.

@fchollet
Copy link
Contributor

Note that the motivation for using tf.nn APIs is that they're documented and better-maintained. The backend APIs are undocumented, have not been touched in some time, and will eventually be deprecated. But for now, this will do.

Copy link
Contributor

@fchollet fchollet left a comment

Choose a reason for hiding this comment

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

Looking good! Please address the last 2 comments and you can add the autogenerated files.

@AakashKumarNain
Copy link
Contributor Author

Thanks for the review. I have made all the changes.

Note that the motivation for using tf.nn APIs is that they're documented and better-maintained. The backend APIs are undocumented, have not been touched in some time, and will eventually be deprecated. But for now, this will do.

Totally agree with you. My only concern here is that for some APIs, we haven't provided clear cut code examples for their usage (for example CTC related functionalities). So, both things are required: removing the backend and improving the documentation in nn. package

Copy link
Contributor

@fchollet fchollet left a comment

Choose a reason for hiding this comment

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

LGTM, thank you for the great contribution!

@fchollet fchollet merged commit 4ce9f51 into keras-team:master Jun 26, 2020
@AakashKumarNain
Copy link
Contributor Author

Thanks for the review

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.

3 participants