* Fix the accuracy problem of allclose op when using float64 data type in static mode.

* Format the code style.

* Still has bugs.

* Fixed allclose_op bug, which cannot deal with some cases of fp64 inputs.

* improved CUDA kernel performance.

* Changed CUDA code.

* Fixed a bug in cuda kernel which cannot deal with large dimension input, and added an unittest for it.

* Add a test case for float32 input.

* Add allclose Op, and its function is analogous to numpy.allclose. It returns True if two tensors are elementwise equal within a tolerance.