TF Keras NLP Demo Notebook (#81)

* messing around with tfhub to remind myself

* more tries for getting on the inside a statefulpartitionedcall

* keras module sub for tfhub keras models

* combined eager gradients

* compatbility with tfhub, nested, and eager models

* formatting fix

* fix layer naming in nested models

* use tfhub and tf_models as optional deps

* remove testing nb

* fix dependency names

* _layers is now a mapping. use values for layer obj

* fix unittest protobuf errors

* pr feedback and tf2 compatibility

* remove testing nb for now

* downgrade notebooks protobuf

* use default input/output tensor in subclass/fn tf models

* fmt fix

* code clean up with recursive nesting, use model nodes

* rename node attrs for older keras compatbility

* add nested model tests

* add TF Keras NLP demo notebook

* add new notebook dependencies to tests

* fix nb tests and minor compatibility changes

* sync with colab

* tf2 wrapper changes for nested/tfhub keras models

* fmt

Co-authored-by: piotrm <[email protected]>

README.md

@@ -16,6 +16,7 @@ To quickly play around with the TruLens library, check out the following CoLab n
 * PyTorch: [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1n77IGrPDO2XpeIVo_LQW0gY78enV-tY9?usp=sharing)
 * Tensorflow 2 / Keras: [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1f-ETsdlppODJGQCdMXG-jmGmfyWyW2VD?usp=sharing)
 * NLP with PyTorch: [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/18GcjsYMkRbxPDDS3J6BEbKnb7AY-1-Wa?usp=sharing)
+* NLP with Tensorflow 2 / Keras: [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1K09IvN7cMTkzsnb-uAeA0YQNfDU7Ibhs?usp=sharing)
 
 
 # Installation