tensorwatch

tensorwatch copied to clipboard

Kudos to the team for this great library. I was wondering if it is possible to integrate tensorwatch with other libraries as mentioned. I am asking this out of my curiosity derived while using interpret as it allows for crazy integrations.

Any example/hint would be great so that I can prepare a notebook and share.

Thank you! TensorWatch currently works well with PyTorch and it should work well with TensorFlow eager mode tensors. We are still working on Keras integration however. BTW, do you mean this library when you mention interpret? If you are looking for any specific integrations, we would love to know as well!

Yes, I mean that library.

TensorWatch currently works well with PyTorch and it should work well with TensorFlow eager mode tensors.

Examples would be great.

Also, since keras has been merged into TensorFlow (from its 1.13 release) I think that advantage can be taken. For example, I used interpret to generate explanations from a model created using keras from the tensorflow variant by wrapping it into a sklearn compatible version (see this notebook).

Hey, any updates on it?

Hi @sayakpaul, I've put keras support in our backlog, hope to be done sometime next week. For TensorFlow, we already support eager tensors so no changes there. I would keep this issue open until we are done with keras support.

Okay @sytelus. Could you please point me to an example which uses tensorwatch and tensorflow simultaneously or anything similar?

@sytelus a gentle reminder regarding the examples on TF.