reduce inference ci time

Disable one test in test_quantize_transpiler.

clean up channel

disable this two unittests to accelerate Mac unit test

test=develop

test=develop

Adding fused_embedding_fc_lstm op

fix random fail on test_api_impl

test=develop

* refine reduce by cub
* optimize KernelDepthwiseConvFilterGrad
* optimize depthwise conv and reduce mean and reduce sum
* fix bug: dilation
* cuda arch and cuda 8 compatible

test=develop

test=develop

test=develop

disable dist se resnet

test=develop

Current scope needs to be thread-safe for training

Hide out params in elementwise layers and scale layer

* Yuyang fix global gradient clip

* Share LoDs

* Revert unnecessary changes

* Fix bug in sequence_slice_op

* flags

* "follow comment"

Fix pip install in mac

test=develop

Fix kwargs in clip and logical API

scope's API modifies its internal state. And scope's
API can be called from multiple threads during traing.
Hence, we need locks to protect the scope's internal
states.

We can optimize it in the future. But the current
solution is buggy.

test=develop

test=develop

* show detail error log on ci

* test

* fix memopt and dist

* update apispec

* will fix different batch issue test=develop

- add naive executor
- fix concurrency performance issue

Fixed-point quantize transpiler.

test=develop

test=develop

Add ci job for Mac

clean v2 usage

* add GraphNum

test=develop

* add graph number check in parallelExecutor

test=develop

* fix transformer_model bug

test=develop

* fix graph num

  extended test_text_classification ot use new op