[webgpu] Fix GatherBlockQuantized on Intel ADL/TGL platforms - #26526
Merged
Guenther Schmuelling (guschmue) merged 3 commits intoNov 13, 2025
Conversation
Contributor
Author
|
The first commit is only the early draft to demonstrate the fixing. |
Contributor
Author
Yulong Wang (fs-eire)
previously approved these changes
Nov 10, 2025
Contributor
|
Thanks for the fix! LGTM |
Contributor
|
/azp run Linux QNN CI Pipeline, Win_TRT_Minimal_CUDA_Test_CI, Windows ARM64 QNN CI Pipeline, Windows GPU Doc Gen CI Pipeline |
Yulong Wang (fs-eire)
approved these changes
Nov 12, 2025
|
Azure Pipelines successfully started running 4 pipeline(s). |
Guenther Schmuelling (guschmue)
approved these changes
Nov 12, 2025
Contributor
|
tested on tgl with qwen3-0.6b and int4 embeddings - works! |
Contributor
Author
|
Not sure why CI failed. Probably kick off a retry. |
Contributor
|
/azp run Linux QNN CI Pipeline,Win_TRT_Minimal_CUDA_Test_CI,Windows ARM64 QNN CI Pipeline,Windows GPU Doc Gen CI Pipeline |
|
Azure Pipelines successfully started running 4 pipeline(s). |
Contributor
|
/azp run web_Release / build_onnxruntime_web,web_Debug / build_onnxruntime_web,Test Linux TensorRT x64 Release,Test Linux CUDA x64 Release |
|
No pipelines are associated with this pull request. |
Guenther Schmuelling (guschmue)
merged commit Nov 13, 2025
d6219b6
into
microsoft:main
157 of 184 checks passed
Rohanjames1997
pushed a commit
to Rohanjames1997/onnxruntime
that referenced
this pull request
Dec 4, 2025
…soft#26526) ### Description The `GatherBlockQuantized` operation was using incorrect `data_indices` during execution on Intel Alder Lake (ADL) and Tiger Lake (TGL) platforms. This change sets the proper `data_indices`, resolving correctness issues encountered with the Phi-4-mini model on these architectures. ### Motivation and Context See above.
Sumit Kumar (Sumit2318)
pushed a commit
that referenced
this pull request
Jan 6, 2026
### Description The `GatherBlockQuantized` operation was using incorrect `data_indices` during execution on Intel Alder Lake (ADL) and Tiger Lake (TGL) platforms. This change sets the proper `data_indices`, resolving correctness issues encountered with the Phi-4-mini model on these architectures. ### Motivation and Context See above.
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
3 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.
Description
The
GatherBlockQuantizedoperation was using incorrectdata_indicesduring execution on Intel Alder Lake (ADL) and Tiger Lake (TGL) platforms.This change sets the proper
data_indices, resolving correctness issues encountered with the Phi-4-mini model on these architectures.Motivation and Context
See above.