[TOPI] Implement Einsum with reduction axes #12913
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
github-actions
bot
requested review from
MasterJH5574,
junrushao,
masahi and
spectrometerHBH
include/tvm/topi/einsum.h
Outdated
| // Neglecting oshape assign check temporally | ||
| return oshape; | ||
| } | ||
| Array<PrimExpr> NumpyEinsumShape(const std::string subscripts, |
Contributor
There was a problem hiding this comment.
Just curious: what does the “Numpy” prefix stand for?
Member
Author
There was a problem hiding this comment.
this is the name used previously, indeed there's no reason using numpy prefix, we can probably find a better name
Contributor
There was a problem hiding this comment.
Sounds good. The function is just doing pure shape inference... Just some random thought, what about “InferEinsumShape”?
Comment on lines
-176
to
-179
| TVM_REGISTER_GLOBAL("topi.einsum").set_body([](TVMArgs args, TVMRetValue* rv) { | ||
| *rv = einsum(args[0], args[1]); | ||
| }); | ||
|
|
Contributor
There was a problem hiding this comment.
Is this no longer used?
Member
Author
There was a problem hiding this comment.
it's moved to einsum.cc file
MasterJH5574
approved these changes
Sep 27, 2022
Contributor
MasterJH5574
left a comment
MasterJH5574
left a comment
There was a problem hiding this comment.
Thanks @vinx13 for the fix!!
xinetzone
pushed a commit
to daobook/tvm
that referenced
this pull request
Nov 25, 2022
* [TOPI] Implement Einsum with reduction axes * address comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.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.
The current Einsum doesn't utilize
ReduceOpfor reduction. It instead unrolls the whole reduction when defining the computation. It makes the generated result not suitable for analysis as it doesn't contain reduction information. It's also broken for large reductions. This PR provided a new implementation that generates the reduction axes correctly.Graph level optimizations might be needed for efficient Einsum. This PR only provides compute definition for further analysis and optimizations.
cc @spectrometerHBH @MasterJH5574 @junrushao @masahi