Fix a bug in save_inference_model and prune when the program is initailized by load_inference_model (#10011)

* Fix bug in save_inference_model and prune when the program is initialized by load_inference_program.

* Save the transpiled program instead.

Add neural transformer leanring rate decay function

* Add topk Python API.

* Add unit test.

* Remove the repeated API.

- EPS added to softmax mkldnn primitive outcome is limited to training
phase

Fixes after review

clang format fixes

clang format fixes

* init

* add some check

* add dist transpile logic

* add insert op for block

* init change get_pserver_program

* optimize code

* fix a bug

* can run now

* start to do table split

* start to process table gradient

* complete pserver part

* can send_vars now

* revert cpplint

* fix a bug

* optimize code

* move dist test to models

* revert the interface of distribute_transpiler.transpile

* fix prefetch_block

* optimize trainspiler code

* add comment to sum_op

* add warning log

* fix comment

* fix test_send_recv

* fix test_send_recv

* fix train with no distributed table

* optimize GetDims