enable cpu machine to run paddle model in gpu lib

* fix_conv_grad_grad

* fix_bug, test=develop

* support reshape doubel grad, test=develop

* fix reshape double grad, pass converage, test=develop

* fix review, test=develop

* fix_norm_model_average_bug

* test=develop

* refine comment test=develop

* refine comment test=develop

* refine eng doc for hard_sigmoid op
test=develop
test=document_fix

* refine the description of hard_sigmoid
 test=develop
test=document_fix

* update API.spec
test=document_fix

* Refine the decription of parameters of HardSigmoid op
test=develop, test=document_fix

* Update API.spec for hard_sigmoid op
test=develop, test=document_fix

* enhance embedding error message test=develop

* enforce .h error test=develop

* fix unittest code test=develop

* Fix fp16 dtype in embedding test=develop

* add import warnings test=develop

* Fix the error message of assign op
test=develop

* Refine input type checking for assign op
test=develop

* Refine unittest of assign op
test=develop

* add input type and dtype check, enhance shape error message for concat_op
test=develop

* enhance shape check
test=develop

* improve coverage

test=develop

* fix APIs: relu, relu6, hash

test=develop
test=document_fix

* fix relu6 doc

test=develop
test=document_fix

* fix API.spec

test=develop
test=document_fix

* add description link for hash

test=develop
test=document_fix

* improve data feeder related API

* Refine topk stack squeeze api en doc

test=develop
test=document_fix
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* Refine api en doc

test=develop
test=decument_fix
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* refine api.spec and returns

test=develop
test=document_fix
Signed-off-by: Nzhaoyuchen <zhaoyuchen01@baidu.com>

* add input type and dtype check for accuracy_op

* add input type and dtype check for accuracy_op

* modify python error on accuracy_op,add test=develop

* modify details on accuracy_op, test=develop

* test float16, test=develop

* add warning, test=develop

test=document_fix

* enhance mul_op input error message test=develop

fix sign op input error check
test=develop

* Fix api docs. test=document-fix

* Fix en docs. test=develop

* Fix the doc of dynamic_gru. test=document_fix

* Update API.spec. test=document_fix

* Fix codestyle test=develop, test=document_fix

* fix APIs: cast, concat, tensor_array_to_tensor, argmin, argmax, argsort

test=develop
test=document_fix

* fix API:cos, exp, ceil, elu, brelu English doc
test=develop
test=document_fix

* refine 13 en doc

* fix Print, is_empty, read_file test=document_fix

* fix the precise roi poop op test=develop
add roi backward implementation, fix the output-channel

* Refine the main comment of DynamicRNN.

* Refine the documentation of DynamicRNN's step_input function.

* Refine the documentation of DynamicRNN's static_input function.

* Refine the documentation of DynamicRNN's block function.

* Refine the documentation of DynamicRNN's memory function.

* Refine the documentation of DynamicRNN's update_memory and output function.

* Refine the code format and remove the method list.

* Refine the documentation of DynamicRNN's __call__ function.
test=develop
test=document_fix

* Minor modification.
test=develop
test=document_fix

* Fix some typo.

* Update API.spec.
test=develop
test=document_fix

* Refine the English according to the comments.

* Update API.spec.
test=develop
test=document_fix

* Fix some typo.

* Update API.spec.

* Add fp16 in input.dtype check test=develop

* Add warning of fp16 in CPU test=develop

* add unittest code for fp16 test=develop

* fix float16 list error test=develop

* Polish En doc for some APIs
* Update some comments and API.spec

* optimize the english description for APIs test=develop test=document_fix

* test=document_fix

* Update metrics.py

* test=document_fix

* test=document_fix

* test=document_fix

* test=document_fix

* fix GradientClipByNorm english doc
test=develop
test=document_fix

* modify english api test=develop test=document_fix

- leaky_relu
- less_than
- log
- logical_and
- logical_or
- logical_xor
- logical_not

* fix fuse_reduce_op quantization bug test=develop

* close fuse_all_reduce_ops in PaddleSlim, test=develop

* fix English Doc of API:layers.py_func/sum

fix APIs, to_variable、NCE、PRelu、softmax、rankloss for dygraph, test=document_fix, test=develop (#20142)

* fix some en doc test=develop test=document_fix

* fix new fluid.data API.spec test=develop test=document_fix

* fix typo test=develop, test=document_fix

* fix typo test=develop test=document_fix

* fix typo test=develop test=document_fix

* fix test=develop test=document_fix

* fix according review test=develop test=document_fix

* fix review test=develop test=document_fix

* fix according review advice test=develop test=document_fix

* fix the doc  test=develop test=document_fix

* fix API.spec test=develop test=document_fix