accelerate memory optimize process

Imperative training network to the end

Pass graph to parallel executor instead of program

fix build issue on windows for sample prop op

* refile profiler
test=develop

* follow comment
test=develop

* Refine doc of uniform_random and fix dtype
* Update defaule value in the arguments

add per kernel config and remove const_cast.

* - Implemented draft of primitive desc keeping in Tensor

test=develop

- TransposeMKLDNNHandler::AcquireSrcMemory was reimplemented

- Added nchw and nc formats setting for sake of compatiblity

Fixed unit tests

- Worakaround to problem with 5D data in conv

- Added 3D and 1D MKL-DNN formats for name handles for tensor

test=develop

- Fix to UTs

test=develop

- Conv fp32 op was updated

Cosmetic fixes

test=develop

- tensor mkldnn cosmetics

test=develop

- Moved most of mkl-dnn specific code from Tensor to mkl-dnn utils

* - Lint fixes

test=develop

* - setting prim dec in Tensor , sets also layout to kMKLDNN

test=develop

* - Moved creation of prim desc totally out of Tensor

test=develop

* - Cosmetic fixes adter review

test=develop

fix pool3d doc. test=develop

test=develop

test=develop

test=develop

test=develop

fix bug for sampled softmax

fix cpu broadcast

test=develop

Change *(smart_ptr.get()) -> *smart_ptr

test=develop

* [Don't merge now]update_readme_to_1.3

* fix sth

test=develop

* update reademe_cn

test=develop

* fix en

test=develop

add sample_logits  and sampled_softmax_with_cross_entropy op

fix adaptive pool doc.test=develop

test=develop

test=develop

test=develop

polish parameter names

test=develop

test=develop

test=develop

reason: dereferencing smart pointer is the same as the underlying pointer
test=develop