Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tensorflow Addons #73

Open
yu-iskw opened this issue Sep 12, 2019 · 10 comments
Open

Tensorflow Addons #73

yu-iskw opened this issue Sep 12, 2019 · 10 comments
Labels

Comments

@yu-iskw
Copy link

yu-iskw commented Sep 12, 2019

@philipperemy

I would like to expand the code with tf.keras. Can I modify the code to release it on pypi?
I know the license is MIT, but I would like to confirm that just in case.

@philipperemy philipperemy changed the title Can I modify the code with tf.kears to release it on pypi? Can I modify the code with tf.keras to release it on pypi? Sep 13, 2019
@philipperemy
Copy link
Owner

@yu-iskw hey! I'm still not sure exactly at the moment what would be the impact of this change. I know Keras is now somehow part of Tensorflow but can still be imported with import keras instead of import tf.keras. So that would force users running on theanos/CNTK to move to tensorflow? I'm not sure how many users still use those frameworks with Keras. Or maybe we could have two different wheels. One for pure Keras and one for TF+Keras during the transition? What do you think?

@yu-iskw
Copy link
Author

yu-iskw commented Sep 13, 2019

Hi, thank you for replying the message. I am not sure how many people the change will impacts. As for me, one of the benefit to modify it with tf.keras is I am able to use tf.data.Dataset API. Especially, TFRecord is very convenient to use a large data set.

So what I want to do is to support the package to both of pure keras and tf.keras while the transition from keras to tf.keras is going. I don’t know the future plans of migrating between them. However, it would be worthwhile supporting both at the moment.

@philipperemy
Copy link
Owner

@yu-iskw sure if we can support both at the same time it could be nice! I guess that would be the best solution.

@philipperemy
Copy link
Owner

As long as nobody is impacted by the change I'm fine with that. If we have a way to import tf.keras instead of keras when tensorflow.keras is present, without breaking anything, then let's do that.

@evanatyourservice
Copy link

evanatyourservice commented Oct 20, 2019

I’ll be testing this later today using tensorflow 2.0 in a colab notebook, I’ll report back if I have to make any changes to make it run smoothly.

Edit: never mind I now see that pull req!

@philipperemy
Copy link
Owner

👍

@shun-lin
Copy link

shun-lin commented Nov 3, 2019

Hi!

I recently came across TCN and wanted to try it out with tf 2.0 and tf.keras. @philipperemy thank you so much for the keras version implementation :). If we also want to support tf.keras, I think another great place we can put it on for more visibility is tensorflow addons. @philipperemy may I help out with porting the tcn layer portion of the code to tensorflow.addons.layers?

@philipperemy
Copy link
Owner

@shun-lin if you can help out porting the TCN layer portion to tensorflow addons layers, it would be amazing!!

@philipperemy
Copy link
Owner

The TF2.0 and tf.keras is on the way :)

@shun-lin
Copy link

shun-lin commented Nov 4, 2019

@philipperemy thanks :)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

4 participants