Skip to content

Fixes for ONNX and FP4 in 7.2 #4464

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
wants to merge 19 commits into
base: release/rocm-rel-7.2
Choose a base branch
from
Open

Fixes for ONNX and FP4 in 7.2 #4464

wants to merge 19 commits into from

Conversation

@causten
Copy link
Collaborator

@causten causten commented Nov 24, 2025

Motivation

  • Improve MXFP4 by pulling in updated rocMLIR
  • Minor fixes to docs

Technical Details

Changelog Category

    • Added: New functionality.
    • Changed: Changes to existing functionality.
    • Removed: Functionality or support that has been removed. (Compared to a previous release)
    • Optimized: Component performance that has been optimized or improved.
    • Resolved Issues: Known issues from a previous version that have been resolved.
    • Not Applicable: This PR is not to be included in the changelog.

github-actions bot and others added 19 commits November 17, 2025 10:41
@github-actions
Update onnxruntime main 8fe48049c089ec50f6f915ca97c37969b7194a96 (#4441)
892348d
@TedThemistokleous
Add right pad masking mode (#4392)
78f96b1 
Generate a raw mask from each mask_index number used for right padding.
@eddieliao
[AIMIGRAPHX-273] Fix Jenkins Parallel Stage Displaying Incorrect Time…
0d81fd4 
… Elapsed (#4424)

The parallel stage view on BlueOcean is not showing the correct total time after the whole parallel step is complete
@turneram
Refactor GroupQueryAttention (#4396)
d80331a 
Remove the GroupQueryAttention ref op and use equivalent ops in its place.
@bdevorem
Flash decoding round 1; AIMIGRAPHX-242 (#4393)
66b6e36 
rocMLIR has implemented split kv and GQA, which enables us to implement flash decoding.
@pfultz2
Show attributes in onnx trace (#4445)
8254ea9
@aarushjain29
change opset version for resnet50 example (#4446)
ce3870f
@pfultz2
[AIMIGRAPHX-326] Fix "reduce_sum: axes: value out of range" error in …
71d23dd 
…simplify_reshapes (#4443)
@shivadbhavsar
disable matching for dynamic shapes (#4362)
72ed531
@pfultz2
Add initial unit tests for jit gpu kernels (#4278)
074e4d8
@amd-hsivasun
[Ex CI] Enable PipelineRepoRef parameter (#4455)
4444596
@aarushjain29
clamping the scale (#4410)
1235768 
Updates since PR 4190 for clamping
@lakhinderwalia
quieten the verbose console output of compiler warnings and errors, i…
bdbf814 
…n tests that explicitly invoke them. (#4453)
@eddieliao
Use local variable in place of DOCKER_OPTS (#4454)
f9effd4
@dependabot
Bump rocm-docs-core from 1.29.0 to 1.30.0 in /docs/sphinx (#4449)
52785bb
@pfultz2
Algo min element if tests (#4447)
8a1e49f
@harkgill-amd
Docs: Change images to .png to resolve lint warnings (#4457)
d1a1b43
@pfultz2
Improve layout propagation in poinwise fusion when using broadcasted …
80fa637 
…inputs (#4432)
@github-actions
Update onnxruntime main 4665804592d1ae341094b93c0717b1c1c6656559 (#4458)
26f7e37
@causten causten requested a review from a team as a code owner November 24, 2025 16:28
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.