* amp.decorate optimizers set to None is ok

* refine unittest

* add unittest and refine example code

* refine unittest

* fix data parallel when VOCAB var in program

* run dy2stat pure fp16 in Linear model

* no use self._pure_fp16_inputs

* add test and fix Adam error in dy2stat pure fp16 training

* use paddle.optimizer.Adam

* run test in gpu

* change test time for CI

* enlarge atol for test_resnet_pure_fp16

* refine code and enlarge atol

* make custom_white_list and custom_black_list take effect for AMP and pure fp16

* check tracer is not None

* use default atol

* change filter_size

* change atol and add some NOTE

* Added HCCL backend support in dynamic graph mode

* fix segmentation fault

* add ut

[Dy2stat]Allow users to switch eval/train mode when using @to_static to decorate a function (#37383)

* Allow users to switch eval/train mode when using @to_static to decorate a function

* refine code for train() and eval()

* support `for i in [1,2,3]` statements in dy2stat

* add test case

* fix ci

* remove wrong code

* Expose func for varbase

* Expose func for varbase and enhance varbase init func

* Change func name and add test case for _CopyGradientWith

* Rename func

* Add test cases to increase coverage

* Refine the logic of _to func

* Replace numel() with _numel(), Add test code

* refine layer to

* delete comment

* refine logic

* refine code

* refine pure_fp16_init

* refine comment

* Support Py3 type annotations in @to_static

* support type hint for args in func

* support type hint assign

* if annotation and value(Constant) are diffent type, we use value type

* polish type_from_annotation()

* code format

* code format

* remove useless commentary

* fix review
Co-authored-by: NAurelius84 <zhangliujie@baidu.com>

* Layer.to reutrn self

* add device required

* remove additional warnning in layer.to

* remove additional warnning in layer.to

* remove additional warnning in layer.to

* remove additional warnning in layer.to

* remove additional warnning in layer.to

* [hapi] support dygrapg amp O2

* fix problem of static pure fp16 in hapi

* fix bug

* fix format

* fix ut

* follow comments

* update ut

* update amp save/load

* fix ut

* refine code format

Add Tokenizer related functionalities for Transformer model in order that the process of training and predicting is consistent.

* support the text string as an input Tensor
* support the "VOCAB"unordered_map<wstring, int> as an input Tensor to lookup tokens
* Tokenizer used for BERT. This tokenizer applies an end-to-end, text string to wordpiece tokenization.
* It first applies basic tokenization, followed by wordpiece tokenization.

* remove no_value using var.name

* fix unit test for CI

* fix unit test

* add test case

* fix test case

* add more test case

* catch the generatorfunction and intercept it.

* add test generator

* add test case

* refine the testcase

* fix no_grad context error in train mode when using save/load

* change net to train mode in test case

* add no_sync for parameters sync

* add pipeline for moe

* add attr is_distributed

* refine code

* refine black/white list for pure fp16

* refine amp level

* fix typo

* update tracer._amp_level

* modify format

* modify format

* change time to remove static tempfile

* delete remove_static_file() function

* init functional jacobian api

* finish test with dtype float32

* add float64 test case

* polish code

* use atol=1e-5 with dtype float64

* fix for ci

* set timeout for test_jacobian

* polish API docstring

* modify docstring

* split minimize() to step() + update()

* add unscale and step for grad_scaler

* add unittest

* refine code in minimize

* delete step in loss_scaler

* fix example bug

* refine comment

* refine unittest

* add unittest

* add pure fp16 major function in auto_cast & tracer

* support master weight in dygraph for pure fp16

* check mix dtype of fp16&fp32 for check_finite_and_unscale op

* change pure fp16 funtion name

* refine some bug in auto_cast

* refine auto_cast interface logic

* add param _casted_by_pure_fp16 for class Layer

* support state_dict hook for save model by user appointed dtype in pure_fp16_decorator

* refine pure_fp16_decorator as decorator

* add unittest

* add comment

* add comment

* support recompute

* add comment for auto_cast and decorator

* support to_static_state_dict for paddle.jit.save

* unlimite models num and optimizers num

* add lookup_table in black_list

* fix momentum and layer state_dict

* fix bug in layer state_dict

* fix bug in layer state_dict_helper

* refine unittest

* refine test_momentun_op

* refine interface and some code

* refine amp_decorator interface

* refine pure fp16 interface

* refine master weight interface

* fix no_grad context error in dy2stat

* remove useless comments

* fix error by drop_kids in python

* add test and fix review

* clip op extra information when export model,test=ocr

* rename clip_extra parameter to kwargs in save_inference_model, test=ocr

* support numpy.ndarray index.

* polish code.

* [NPU] fix depthwise_conv2d_grad, test=develop

* remove debug files, test=develop

* Add solutions to PyLayer which is unsupported in DataParallel

* modify note format for parallel.py

* modify docs of dataparallel

* add docs of dp with pylayer

* modify docs format

* modify example format

* change example of dp with pylayer

* add unittest for dp with pylayer

* modify ut

* merge latest codes

* update

* modify for CI-Coverage

* modify text-indent

* dy2stat_error: add revise suggestion for two error cases

* fix test_error

* fix review

This reverts commit ae93d9c2.