* refine tensor.dtype for bloat16

* refine test

* revert

* refine bfloat16 print

* Add BaseTransformer for dy2st error message

* Fix return_transformer error

* Polish dy2st error info in runtime

* Fix UT error

* Polish runtime error code

* fix init()

* delete test_device_context

* replace CPUDeviceContext with CPUContext

* fix test_scalar

* remove dot_op.cc

* fix compile

* convert graph to program to let sSandaloneExecutor supporrt CompiledProgram

* skip case that compiled_program._program is None

* execute CompiledProgram._compile to apply build_strategy

* 2nd part of transpose update

* add switch_auto_tune option.

* add some changes according to Ci

* refine the structure of auto_tune_base.

* merge develop changes

* reset the switch_set_range and change unittest of transpose auto-tune

* change the kernel auto-tune logits

* [Dy2Stat]Polish break/continue statement transformer logic

* re-write the unit tests for compare xpu op

*test=kunlun

* re-write the unit tests for compare xpu op

*test=kunlun
Co-authored-by: Nrunzhech <runzh_chen@sjtu.edu.cn>

* [Dy2Stat]Enhance nonlocal machanism while returning single var

* [Dy2Stat]Enhance nonlocal machanism while returning single var

* move standard apis to api.yaml

* revert erfinv

* delete dot_op.h

* fix dot

* rerun ci

* save multi program into one param and seperate model

* export class property

* add code

* add unit test

* support scope_guard

* fix test

* merge and add base support for non-local for

* for and while non-local support

* fix ci errors: v1

* fix bug

* fix

* fix code

* fix

* fix

* fix

* add relu6 kernel and yaml

* format files

* format code and fix bug

* fix build failed

* support complexd selected_rows kernel in yaml

* support configuring optimizer api in yaml

* fix data transform bug

* add comparisons trt converter

* skip xpu conv2d fp16 unitest
*test=kunlun

* minor
*test=kunlun

* [Dy2Stat]Polish all API name of _jst

* add unittest for PR43688

1. test_parallel_executor_seresnext_base_gpu failed on 2 P100 GPUs with `470.82` driver.
```
======================================================================
FAIL: test_seresnext_with_learning_rate_decay (test_parallel_executor_seresnext_base_gpu.TestResnetGPU)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/opt/paddle/paddle/build/python/paddle/fluid/tests/unittests/test_parallel_executor_seresnext_base_gpu.py", line 32, in test_seresnext_with_learning_rate_decay
    self._compare_result_with_origin_model(
  File "/opt/paddle/paddle/build/python/paddle/fluid/tests/unittests/seresnext_test_base.py", line 56, in _compare_result_with_origin_model
    self.assertAlmostEquals(
AssertionError: 6.8825445 != 6.882531 within 1e-05 delta (1.335144e-05 difference)
----------------------------------------------------------------------
```
2. To be more accuracte on evaluating loss convergence, we proposed to apply IOU as metric, instead of comparing first and last loss values.
3. As offline discussion, we also evaluated convergence on P100 and A100 in 1000 interations to make sure this UT have the same convergence property on both devices. The curves are showed below.
![A100-Single, P100-Single and Diff (1)](https://user-images.githubusercontent.com/13541238/175461920-25df6101-6dd8-4387-862c-d1c8e9299c57.png)