Update dien.py

add a global variable initializer op in model def

fix

Update feature config generation method & Fix bugs

- Adopted a new feature config generation method
- Simplified input mechanism for multi-valued features.
- Fix bug: in some cases the order of model input placeholder and actual model input list does not match.
- Fix bug: Divide by zero overflow in `SequencePoolingLayer` when sequence length is 0.

Refactor & Add sequence input support

* Refactor Input&Embedding

* Support sequence(multi-value) input for AFM,AutoInt,DCN,DeepFM,FNN,NFM,PNN,xDeepFM models

Update for v0.2.1

* Add AutoInt & InteractingLayer

Release v0.2.0

Merge from dev (shenweichen#15)

* Fix bugs when using tensorflow version higher than 1.6.0.
* Now support tf version from 1.4.0 - 1.12.0 except for 1.7.* and 1.8.*
* Update docs

minor change

1. Modified output shape of BiInteractionPooling, InnerProductLayer to make
sure that the dimensions of the output and input of a layer are same if possible

2. Minimize the nesting of other layers in the custom layer, because I found that the statistics of the parameters number of model.summary()  are incorrect when using other layers.

refactor&update docs

Use a more space efficient implementation in CrossNet
Add custom_objects in utils which will be used in load_model function
Add FAQ in docs