[AutoParallel] Support operators have mixed inputs. #57774
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! |
chenwhql
reviewed
Sep 26, 2023
| local_t_2 = paddle.to_tensor(np_array, dtype='float16') | ||
| elif np_array.dtype == np.int32: | ||
| local_t_1 = paddle.to_tensor(np_array, dtype='int32') | ||
| # local_t_2 = paddle.to_tensor(np_array, dtype='float16') |
There was a problem hiding this comment.
注释忘记删除了，已经恢复
| x = np.random.random(size=[4, 4]).astype("float32") | ||
| y = np.random.random(size=[4, 4]).astype("float32") | ||
| local_x, dist_x = self.create_local_and_dist_tensor_pair(x) | ||
| local_y, dist_y = self.create_two_local_tensor_pair(y) |
There was a problem hiding this comment.
local_y, dist_y -> local_y1, local_y2？
| @@ -72,6 +72,13 @@ def FindParsingFunctionFromAttributeType(atype): | |||
| " auto {} = {}(\"{}\", \"{}\", args, {}, {});\n" | |||
| ) | |||
|
|
|||
| CONVERT_INPUT_TENSORS_TO_DIST_TENSOR_TEMPLATE = """ | |||
| const phi::distributed::ProcessMesh* mesh = nullptr; | |||
| auto mesh_ptr = &mesh; | |||
There was a problem hiding this comment.
这里是不是直接传入&mesh参数即可，不用新增一行赋值
| @@ -325,7 +332,9 @@ def GeneratePythonCFunction(self): | |||
| inplace_returns_pos_map = {} | |||
| # Generate Python-C Tensors Parsing Logic | |||
| get_eager_tensor_str = "" | |||
| input_names = "mesh_ptr, " | |||
There was a problem hiding this comment.
这个可以放到上面模板里吗？这样看模板能知道mesh传进去了，inputs则只包含api的输入，语义更准确一些
Frida-a
pushed a commit
to Frida-a/Paddle
that referenced
this pull request
Oct 14, 2023
* [AutoParallel] Support operators have mixed inputs like DenseTensor and DistTensor. * Polish code with review comments.
jiahy0825
pushed a commit
to jiahy0825/Paddle
that referenced
this pull request
Oct 16, 2023
* [AutoParallel] Support operators have mixed inputs like DenseTensor and DistTensor. * Polish code with review comments.
danleifeng
pushed a commit
to danleifeng/Paddle
that referenced
this pull request
Nov 14, 2023
* [AutoParallel] Support operators have mixed inputs like DenseTensor and DistTensor. * Polish code with review comments.
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
Pcard-73145
Support operators have mixed inputs like DenseTensor and DistTensor. DenseTensor will be sharded to replicate DistTensor automatically.
Former implementation is deployed in
phiAPIs: PR 57684. However, inputs are allconstreference, which means converting inputDenseTensortoDistTensoris an unsafe operation. Meanwhile, there is modification of inputs inAutoGradlevel like convert them to continuous Tensors. As a result, modification ofphiinput likex_tmpinAutoGradlevel will not change the real inputx, but backward node relies on the real inputxinformation as Tensor metas. It's more properly to implement this sharding strategy inPython-Clevel.