[AutoParallel] Support disttensor for Tensor.copy_ #58369
Merged
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
|
你的PR提交成功，感谢你对开源项目的贡献! |
|
❌ The PR is not created using PR's template. You can refer to this Demo. |
wanghuancoder
changed the title
chenwhql
reviewed
Oct 25, 2023
| @@ -128,6 +136,60 @@ void Tensor::copy_(const Tensor &src, | |||
| auto *dev_ctx = pool.GetMutable( | |||
| place.GetType() == target_place.GetType() ? target_place : place); | |||
|
|
|||
| #ifdef PADDLE_WITH_DISTRIBUTE | |||
There was a problem hiding this comment.
auto parallel可以放在DenseTensor kernel的分支，是DenseTensor Kernel的一个扩展机制
paddle/phi/api/lib/tensor_method.cc
Outdated
|
|
||
| // 1. InferSpmd (Infer DistAttr of Inputs&Outputs) | ||
| auto meta_dist_input_x = MakeDistMetaTensor(*src.impl()); | ||
| auto spmd_info = phi::distributed::ElementwiseUnaryInferSpmd( |
There was a problem hiding this comment.
这里咱们是不是直接创建一个输出，把输入的设置上去就行，不用走InferSpmd，这个elementwise的逻辑还挺多的
There was a problem hiding this comment.
这里不会出现：这个Tensor不需要在这个卡上计算的情况吗？
paddle/phi/api/lib/tensor_method.cc
Outdated
| // 4. Select Kernel | ||
|
|
||
| // 5. Reshard Input | ||
| auto dist_input_x = ReshardApiInputToKernelInput( |
There was a problem hiding this comment.
好像也不需要reshard，原先的值给输出copy过去就行
paddle/phi/api/lib/tensor_method.cc
Outdated
| phi::Copy(*dev_ctx, *input_x, target_place, blocking, dense_out); | ||
|
|
||
| // 9. Reshard Partial Output to Replicated (Temporary) | ||
| ReshardOutputPartialAxisToReplicated(dev_ctx, dist_out); |
LiYuRio
approved these changes
Oct 26, 2023
| phi::UnchangedInferMeta(MakeMetaTensor(*(src.impl_)), &meta_dist_out); | ||
|
|
||
| if (rank_is_in_current_mesh) { | ||
| // 4. Select Kernel |
LiYuRio
reviewed
Oct 26, 2023
| // API `copy_` does not need to set DistAttr for output. | ||
| return; | ||
| } | ||
| #endif |
danleifeng
pushed a commit
to danleifeng/Paddle
that referenced
this pull request
Nov 14, 2023
* support disttensor for tensor.copy_
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
Description
Tensor.copy_是典型的PHI api逻辑，会传入DistTensor。需要补充DistTensor分支。
PCard-70444