[Semi Auto] Entropy SPMD Rule #55394
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
… semi-auto/rule-base
…i-auto/embedding-rule
|
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. |
zhiqiu
reviewed
Jul 19, 2023
| input_specs_size, | ||
| 2, | ||
| phi::errors::InvalidArgument("The size of InputSpec of cross entropy " | ||
| "with softmax should be 1, but got [%d].", |
There was a problem hiding this comment.
Suggested change
| "with softmax should be 1, but got [%d].", | |
| "with softmax should be 2, but got [%d].", |
| from paddle.distributed.fleet import auto | ||
|
|
||
|
|
||
| class TestEmbeddingSPMDRule(unittest.TestCase): |
Comment on lines
+150
to
+151
| REGISTER_SPMD_RULE(cross_entropy_with_softmax, CrossEntropyWithSoftmaxSPMDRule); | ||
| REGISTER_SPMD_RULE(softmax_with_cross_entropy, CrossEntropyWithSoftmaxSPMDRule); |
There was a problem hiding this comment.
The two ops have same attrs?
There was a problem hiding this comment.
yes, here is to adopt for the difference between paddle static mode and dygraph mode, we will unify it in future.
cqulilujia
pushed a commit
to cqulilujia/Paddle
that referenced
this pull request
Jul 24, 2023
* base rule * add sharidng merge * add sharidng axis merge * define unified data class for inferencing dist_attr * test wrap DistTensorSpec in dygraph mode * matmul main logic done * shape int64 * common cc * define unified data class for inferencing dist_attr * test wrap DistTensorSpec in dygraph mode * define python api and wrap function in static mode for DistTensorSpec * revise syntax * map bugfix * broadcast func * compile 1 * add unitest * add registry * update unitest * bugfix * bugfix * add pybind * bugfix * bugfix macro gloabl name space * bugfix macro gloabl name space * pybind * pybind test * pybind bugfixed1 * pybind bugfixed2 * pybind unitest * merge dev * merge dev * merge dev * fixed cmake conflict * fixed cmake conflict * rename get method * revise inferforward output type * revise comment * replicated rule * replicated rule 2 * revert bug deps * add rule * add unitest * add rule * add unitest * move ut of auto_parallel * fix ut * bugfix * bugfix * bugfix * bugfix * bugfix * bugfix * bugfix * resolute input sharding conflict maybe * fixed comment * add rule * add unitest * fixed typoes --------- Co-authored-by: Yichen Zhang <zhangyichen03@baidu.com> Co-authored-by: zhiqiu <chenqiuliang@baidu.com>
wz1qqx
pushed a commit
to wz1qqx/Paddle
that referenced
this pull request
Jul 31, 2023
* base rule * add sharidng merge * add sharidng axis merge * define unified data class for inferencing dist_attr * test wrap DistTensorSpec in dygraph mode * matmul main logic done * shape int64 * common cc * define unified data class for inferencing dist_attr * test wrap DistTensorSpec in dygraph mode * define python api and wrap function in static mode for DistTensorSpec * revise syntax * map bugfix * broadcast func * compile 1 * add unitest * add registry * update unitest * bugfix * bugfix * add pybind * bugfix * bugfix macro gloabl name space * bugfix macro gloabl name space * pybind * pybind test * pybind bugfixed1 * pybind bugfixed2 * pybind unitest * merge dev * merge dev * merge dev * fixed cmake conflict * fixed cmake conflict * rename get method * revise inferforward output type * revise comment * replicated rule * replicated rule 2 * revert bug deps * add rule * add unitest * add rule * add unitest * move ut of auto_parallel * fix ut * bugfix * bugfix * bugfix * bugfix * bugfix * bugfix * bugfix * resolute input sharding conflict maybe * fixed comment * add rule * add unitest * fixed typoes --------- Co-authored-by: Yichen Zhang <zhangyichen03@baidu.com> Co-authored-by: zhiqiu <chenqiuliang@baidu.com>
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
Function optimization
PR changes
Others
Description
Pcard-70448
SPMD rule for entropy ops.
Support shard on broadcast axes and normalized axis.