[DTensor] Add torch symbol and prim for _grouped_mm #2503
+91
−1
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
kshitij12345
changed the title
This was referenced Sep 26, 2025
kshitij12345
force-pushed
the
dtensor-prims._grouped_mm
branch
from
7d785d7 to
9221690
Compare
kshitij12345
changed the title
kshitij12345
requested review from
mruberry,
lantiga,
t-vi and
KaelanDt
as code owners
|
We need to make the access to |
Have made the access conditional, thanks @t-vi |
|
I have pushed a couple of commits after changing PR status from draft to ready, but the Lit Job haven't been triggered. 
|
crcrpar
reviewed
Oct 2, 2025
Comment on lines
+255
to
+268
| "input_shardings", | ||
| [ | ||
| ( | ||
| [ | ||
| Shard( | ||
| -1, | ||
| ) | ||
| ], | ||
| [ | ||
| Shard(1), | ||
| ], | ||
| [Replicate()], | ||
| ), | ||
| ], |
There was a problem hiding this comment.
QQ: what's the type of input_shardings? Tuple of two list of Shard/Replicates?
There was a problem hiding this comment.
It is
[
([Shard(-1)], [Shard(1)], [Replicate()]),
]
NOTE: Each elements of the tuple is Sequence[Placement] as expected by distribute_tensor
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Outdated suggestions cannot be applied.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.
Related #2338
TODO:
Figure out why we seeNeed to set environment variables for nvFuser multi-device to work. See changes toError from segmentation group 1: The singleton Communicator isn't available. This is most likely because the instance wasn't successfully initialized due to lack of a multi-process running (e.g. mpirun or torchrun).only when running this primitive.helper.py