* Update Transformer using Paddle-1.6 apis. (#3614)

* Update Transformer using Paddle-1.6 apis.

* Add check_version for Paddle-1.6 and reorganize utils including check, configuration, etc.

* mv neural_machine_translation to  PaddleMT (#3732)

* Paddle1.6 transformer fix (#3755)

* Revert "mv neural_machine_translation to  PaddleMT (#3732)"

This reverts commit a46365b6.

* Delete Transformer unremoved palm dir.

* Rename neural_machine_translation as PaddleMT.

* Update to PaddleMT in README.md

Fix the data file empty check.
Fix the utf8 file encoding.
Refine the README.

* Update transformer to be unified

* To support CE when unifing Transformer

* add multi process implementation for yolov3

To avoid the error when use_program_cache is True in Transformer infe…rence.

To avoid the data decode error.

* Shift up fluid models

* Update nlp submodules

* Add redirect guide in org readme

* Move HiNAS_models to AutoDL

* Delete file
* Rever the raw file

* Adjust the structure of whole repo

* Correct the readme for speec

* Update submodules LAC, Senta and SimNet

* Load Data --> 64 sec
* Shuffle/Batch --> 14 sec