[Phi]Move kron kernel to phi #40427
Merged
[Phi]Move kron kernel to phi #40427
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thanks for your contribution! |
| // limitations under the License. | ||
|
|
||
| #include "paddle/phi/kernels/impl/kron_grad_kernel_impl.h" | ||
| #include "paddle/phi/kernels/kron_grad_kernel.h" |
| // limitations under the License. | ||
|
|
||
| #include "paddle/phi/kernels/impl/kron_kernel_impl.h" | ||
| #include "paddle/phi/kernels/kron_kernel.h" |
| namespace phi { | ||
|
|
||
| namespace ops = paddle::operators; | ||
| namespace plat = paddle::platform; |
There was a problem hiding this comment.
是否可以用phi::dtype 命名空间下的complex ？
zyfncg
reviewed
Mar 14, 2022
Comment on lines
21
to
22
| namespace ops = paddle::operators; | ||
| namespace plat = paddle::platform; |
There was a problem hiding this comment.
在phi下最好还是不使用paddle::xxx相关namespace的别名，会增加后续替换的难度
| const plat::complex<T>* dout_; | ||
| const plat::complex<T>* A_; | ||
| const plat::complex<T>* B_; | ||
| plat::complex<T>* dout_a_; |
| p_dout_y = dout_y.data<T>(); | ||
| } | ||
|
|
||
| plat::ForRange<Context> for_range(dev_ctx, numel); |
Comment on lines
239
to
244
| ops::TensorReduceImpl<T, T, kps::AddFunctor, kps::IdentityFunctor<T>>( | ||
| *ctx, dout_x, dx, kps::IdentityFunctor<T>(), {1}, stream); | ||
| } | ||
| if (dy) { | ||
| ops::TensorReduceImpl<T, T, kps::AddFunctor, kps::IdentityFunctor<T>>( | ||
| *ctx, dout_y, dy, kps::IdentityFunctor<T>(), {1}, stream); |
Comment on lines
20
to
21
| #include "paddle/fluid/framework/op_registry.h" | ||
| #include "paddle/fluid/platform/for_range.h" |
There was a problem hiding this comment.
op_registry.h这里应该不需要了
for_range.h使用phi下的
| #include "paddle/phi/kernels/funcs/eigen/common.h" | ||
| #include "paddle/phi/kernels/funcs/eigen/eigen_function.h" | ||
| #if defined(__NVCC__) || defined(__HIPCC__) | ||
| #include "paddle/fluid/operators/reduce_ops/reduce_op.cu.h" |
There was a problem hiding this comment.
reduce_op.cu.h可以使用paddle/phi/kernels/funcs/reduce_function.h代替
paddle/phi/ops/compat/kron_sig.cc
Outdated
Comment on lines
19
to
21
| KernelSignature KronOpArgumentMapping(const ArgumentMappingContext& ctx) { | ||
| return KernelSignature("kron", {"X", "Y"}, {}, {"Out"}); | ||
| } |
There was a problem hiding this comment.
这里的前向ArgumentMapping看上去没有特殊case，感觉可以不写，试试直接使用默认的参数映射能不能work？
| auto stream = dev_ctx.stream(); // it is a cuda device_context | ||
| auto* ctx = reinterpret_cast<const plat::CUDADeviceContext*>(&dev_ctx); | ||
| if (dx) { | ||
| ops::TensorReduceImpl<T, T, kps::AddFunctor, kps::IdentityFunctor<T>>( |
There was a problem hiding this comment.
ops::TensorReduceImpl 已迁移，这里可用 phi::funcs::ReduceKernel
| *ctx, dout_x, dx, kps::IdentityFunctor<T>(), {1}, stream); | ||
| } | ||
| if (dy) { | ||
| ops::TensorReduceImpl<T, T, kps::AddFunctor, kps::IdentityFunctor<T>>( |
| #include <algorithm> | ||
| #include <vector> | ||
|
|
||
| #include "paddle/fluid/framework/op_registry.h" |
There was a problem hiding this comment.
phi目录下用不到原来的op 注册头文件
| #include <vector> | ||
|
|
||
| #include "paddle/fluid/framework/op_registry.h" | ||
| #include "paddle/fluid/platform/for_range.h" |
There was a problem hiding this comment.
引用这个 paddle/phi/kernels/funcs/for_range.h
| #include "paddle/phi/kernels/funcs/eigen/common.h" | ||
| #include "paddle/phi/kernels/funcs/eigen/eigen_function.h" | ||
| #if defined(__NVCC__) || defined(__HIPCC__) | ||
| #include "paddle/fluid/operators/reduce_ops/reduce_op.cu.h" |
There was a problem hiding this comment.
ops::TensorReduceImpl 已迁移，这里可用 phi::funcs::ReduceKernel ， 此头文件可以不用了
YuanRisheng
reviewed
Mar 14, 2022
| p_shape_y = dim_y.Get(); | ||
| #endif | ||
|
|
||
| paddle::platform::ForRange<Context> for_range(dev_ctx, numel); |
zyfncg
approved these changes
Mar 14, 2022
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR types
Others
PR changes
Others
Describe
[Phi]Move kron kernel to phi