Expose a interface to create all variables and enable the test of not creating variables every time.

* Support testing during training by ParallelExecutor.

* Add unit test.

* Improve the interface.

* Follow comments.

Added support for dim equals 2 in activation functions

Rewrite memroy/detail

Fix debuger bugs

run prefetch prog on server

Fix seq tag model dist transpile

Update reader

Move filemanager to `edl` repo.

Update release doc

fix compiler error on `tensor_py.h`

fix missing core.so on mac