* change learning_rate_decay to learning_rate_scheduler
* optimize code
* change nn.cast to tensor.cast

Currently, our tests run with 2 GPUs, the init time is absurdly long:
about 4s for each process.  Currently, we run each OP test on
different processes. This PR:

1. create cmake function py_test_modules which will generate the
Makefile that runs a list of Python unittest module in a single Python
process.

2. move all "python unittest compatible" (e.g., used the unittest
package, not just a regular python file). from fluid/tests to
fluid/tests/unittests.

3. cmake now will run all OP tests in fluid/tests/unittests in a
single process, except the time-consuming tests, they are separated
into different processes to utilize parallelism. Please make sure to
use the unittest package if you put the python test file in
fluid/tests/unittests

4. remove all exit(0) from fluid/tests/unittests/*.py, exit(0) is used
to disable unittest, we can not do it when running all tests in a
single process since it will terminate the process without running the
other tests. Instead, the test is disabled in
fluid/tests/unittests/CMakeLists.txt. FIXME is added for each disabled
item. Please disable the unittest from
fluid/tests/unittests/CMakeLists.txt, instead of adding exit(0) to the
Python file, for all Python file in fluid/tests/unittests/.

5. add an option WITH_FAST_BUNDLE_TEST. When OFF, will run the unit
tests in separate process so that they can be tested individually.

* init polynomial_decay

* test polynomial_decay

* complete polynomial_decay

* fix conditional block op

* init scalar-switch-case-op

* switch op can compile

* complete forward switch_op

* add GetMatchCaseIndex

* add switch_grad_op

* init switch Python API

* add test_switch

* support set block list in python

* fix scope problem

* complete test

* optimize test

* optimize test

* rm backward part

* clear grad op

* polynomial_decay use switch op

* revert conditional_block_op and reshape_op

* add piecewise_decay and test

* fix piecewise_decay

* try to use condition op for switch

* can work

* clean old code

* revert

* rm switch_op.cc

* optimize code

* add attr is_scalar_condition for condition_block_op

* fix comment

* fix comment

* add export

* add basic interface for learning rate decay
* add exponential_decay
* add natural_exp_decay
* add inverse_time_decay