* init

* move CUDAStream to phi

* fix compilation

* merge develop

* add stream_owned_ member

* split cuda_stream.h

* fix cpu compile

* fix constructor

* fix bug

* fix windows compile

* fix inference test_levit

* fix windows tests

* Add kernel declarations

* Copy kernel implementation code

* Transfer implementation code

* Register new kernels

* Remove old kernels

* Fix code style

* Fix bugs

* mutable_data->HostAlloc

* Transfer infermeta

* Add yaml and update python api

* Add PADDLE_WITH_HIP check

* Update unittests

* Fix bugs

* Fix bugs

* Optimize directory structure

* Add output checks

* lu_impl.h->lu_kernel_impl.h
Co-authored-by: NBobholamovic <linmanhui@baidu.com>

* [Phi] Add yaml for assign_value

* [Phi] Fix the bug of the assign api and modify the unittest

* [Phi] Fix the bug when the tensor does not have the backend info

* [Phi] Replace the functional-style cast init by the brace-init

* [Phi] Cast the data explicitly

* depthwise_fp16

* depthwise_fp16

* depthwise_fp16

* depthwise_fp16

* test_activation_op unitest error, yaml & activation.py in_dygraph_mode incomplete

* fix test_activation_op unitest error, add yaml and dygraph test

* fix code style with pre-commit

* try to fix namespace error of abs in activation_functor.h

* fix namespace error of abs

* migrate dirichlet op kernel to phi

* fix dirichlet sample memory leak

* support auto fallback to  cpu kernel for cusom device

* fix some problem of kernel fallback

* Move frame OP to phi、add frame OP yaml config and supplement single test

* add Header file of in_dygraph_mode

* Modify variable name and FrameGradInferMeta multiplex UnchangedInferMeta

* move seq2col to phi

* Move api(lgamma) from legacy_api.yaml to api.yaml

* Move api(lgamma) from legacy_api.yaml to api.yaml

* Move api(lgamma) from legacy_api.yaml to api.yaml

* modify code style

* add x to X mapping

* add definition of lgamma

* delete redundant lgamma definitions

* Modify code comments

* Modify ops.py code format

* add lgamma  single test and lgamma api in fluid

* Optimized lgamma unittest

* Add kernel declarations

* Copy kernel implementation code

* Transfer implementation code

* Fix: Move out_grad to first

* Register new kernels

* Remove old kernels

* Move out_grad to last

* Fix bugs

* Transfer infermeta

* Add yaml files

* Add blank line

* Fix code style

* Optimize directory structure
Co-authored-by: NBobholamovic <linmanhui@baidu.com>

* fix typos in template for codegen of operators
* change the way to set attributes for grad op maker

* move crop_tensor from fluid to phi

* delete fluid header files

* fix crop_tensor_op dygraph_mode bug

* modify header files, add out tensor check

* move average_accumulates op to phi kernel

Add yaml and unittest for bmm op

* fix behavior of device_id=None in Tensor.cuda

* fix CI

* add reverse yaml

* add complex for einsum grad kernel

* pass the ci

* transfer the svd infer into phi infermeta

* remove the svd.h

* modify svd api

* fix svd error by insert optional

* multi stream support handle lazy init.

* support eigen lazy init

* update

* fix ci problem

* Move slogdeterminant op to phi

* Add yaml and unit test for slogdeterminant

* (modified) fc support fp16

* __CUDA_ARCH__ version

* delete half

* delete half

* add half support

* add half support

* add half support

* add_ymal_utest for phi grid_sampler op

* (modified) fc support fp16

* __CUDA_ARCH__ version

* delete half

* delete half