Add reduce_merge files for reduce op #32697
Merged
Conversation
This file contains hidden or 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! |
xingfeng01
reviewed
May 7, 2021
paddle/fluid/operators/reduce_ops/reduce_merge/reduce_functor_op.h
Outdated
Show resolved
Hide resolved
Contributor
|
建议文件放在reduce_ops下即可，后缀改为.cuh |
xingfeng01
reviewed
May 7, 2021
xingfeng01
reviewed
May 7, 2021
xingfeng01
reviewed
May 7, 2021
xingfeng01
reviewed
May 7, 2021
xingfeng01
reviewed
May 7, 2021
Contributor
|
注意clang-format |
|
|
AnnaTrainingG
force-pushed
the
reduce_merge
branch
from
07f509e to
8378a5e
Compare
AnnaTrainingG
force-pushed
the
reduce_merge
branch
from
5a68611 to
1c44bae
Compare
…a malloc and writed more notes
AnnaTrainingG
force-pushed
the
reduce_merge
branch
from
1c44bae to
f83712c
Compare
ZzSean
reviewed
May 21, 2021
ZzSean
reviewed
May 24, 2021
Merged
Contributor
|
LGTM |
Xreki
reviewed
May 25, 2021
| #include "paddle/fluid/operators/amp/fp16_type_traits.h" | ||
| #include "paddle/fluid/platform/device_context.h" | ||
| #include "paddle/fluid/platform/hostdevice.h" | ||
| #include "paddle/fluid/platform/macros.h" |
Contributor
There was a problem hiding this comment.
极小化include的头文件，看起来以上头文件都没有实际依赖。
| } | ||
|
|
||
| template <typename T, size_t ElementCount, typename VectorLikeType> | ||
| static inline paddle::framework::Array<T, ElementCount> from( |
Contributor
There was a problem hiding this comment.
这个函数是从什么转到什么？单个的From函数一般是用作类的成员函数，如果是单独的函数，最好命名更清楚一些。
| vec.size(), ElementCount)); | ||
| size_t n = static_cast<size_t>(vec.size()); | ||
| paddle::framework::Array<T, ElementCount> ret; | ||
| for (size_t i = 0; i < n; ++i) ret[i] = vec[i]; |
|
|
||
| template <typename Tx, typename Ty, int BlockDim, typename ReduceOp, | ||
| typename TransformOp, int kRank, int kReduceRank> | ||
| static void launchKernel(const Tx* x_data, Ty* y_data, |
Contributor
There was a problem hiding this comment.
launchKernel -> LaunchKernel，首字母大写。
|
|
||
| template <typename Tx, typename Ty, int BlockDim, typename ReduceOp, | ||
| typename TransformOp> | ||
| static void launchReduceKernel(const Tx* x_data, Ty* y_data, |
Contributor
There was a problem hiding this comment.
launchReduceKernel -> LaunchReduceKernel
Xreki
approved these changes
May 25, 2021
Contributor
Xreki
left a comment
Xreki
left a comment
There was a problem hiding this comment.
LGTM. 先合入这个pr，下个pr尽快接入算子进行ci验证。
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
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
Performance optimization
PR changes
APIs
Describe
add reduce_op.h and functor.h in reduce_merge for reduce op
性能测试数据：