Add full mlx blockwise support#2233
Open
jessegrabowski wants to merge 4 commits into
Open
Conversation
jessegrabowski
requested review from
cetagostini,
drbenvincent,
ricardoV94 and
williambdean
June 16, 2026 02:44
ricardoV94
reviewed
Jun 16, 2026
ricardoV94
approved these changes
Jun 20, 2026
Member
|
failing test |
jessegrabowski
force-pushed
the
mlx-blockwise-upgrade
branch
6 times, most recently
from
June 22, 2026 01:50
d53a158 to
3f11bb2
Compare
Member
|
that ci looks like fun |
jessegrabowski
force-pushed
the
mlx-blockwise-upgrade
branch
from
July 24, 2026 06:40
3f11bb2 to
28ba533
Compare
ricardoV94
reviewed
Jul 24, 2026
| return [a, b], op(a, b), values | ||
|
|
||
|
|
||
| def _join_graph(rng, batch): |
Member
There was a problem hiding this comment.
just inline this code in the test it's 3 lines
| return [a, b], op(a, b), values | ||
|
|
||
|
|
||
| def _scalar_from_tensor_graph(rng, batch): |
| ) | ||
|
|
||
|
|
||
| def _makevector_graph(rng, batch): |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
MLX blockwise operations fail in non-trivial cases because our implementation isn't a full vectorize, it's just a single naive application of
mlx.vmap. We need to align and broadcast all input shapes, then applyvmaponce per non-core dimension. This is whatjnp.vectorizedoes under the hood. I just went ahead and re-implementedjnp.vectorizein MLX, giving us full vectorize support.Closes #2092 , and also address additional unreported cases (e.g. when data has batch dim).