In those ops, some api has execute run error, fix those op, and format some sample code.

Those ops include，scale， sum， sums，unique_with_counts，unique，
wherre， add error message and test case

add addmm op

* Add the matmul, elementwise_euqal, elementwise_sum ops to API2.0
* Fix the import meesage in common_ops_import
* Update the test cast for mm

* Add the sum op to API 2.0, test=develop
* Fix the import meesage in common_ops_import

*  add new op paddle.tensor.mul(x, y, x_num_col_dims=1, y_num_col_dims=1, out=None, name=None)
* add new op paddle.tensor.pow(input, exponent, out=None, name=None)
* add the aboved two new ops unittest (test api param out and name)

* add new op paddle.tensor.div(x, y, out=None, name=None) 
* add gpu and dygraph unittests.
* Performance optimization: scale op is not called when alpha=1. 
* op error message optimization.

* sin sqrt tanh atan add out, test=develop

* optimize doc, test=develop

* add dygraph test, test=develop