OVEP: Bug Fixes, Refactoring, and Contrib Ops Update#23742
Merged
jywu-msft merged 7 commits intomicrosoft:mainfrom Feb 21, 2025
Merged
OVEP: Bug Fixes, Refactoring, and Contrib Ops Update#23742jywu-msft merged 7 commits intomicrosoft:mainfrom
jywu-msft merged 7 commits intomicrosoft:mainfrom
Conversation
* update: Update MSFT Contrib Ops from OV * modified data_ops.cc to remove unsupported ops * disabled tests for EmbedLayerNormalisation and MatMulNBits --------- Co-authored-by: n1harika <[email protected]>
* Internal ci for PTL 1.1 (#523) * update: Update MSFT Contrib Ops in OVEP (#521) * update: Update MSFT Contrib Ops from OV * modified data_ops.cc to remove unsupported ops * disabled tests for EmbedLayerNormalisation and MatMulNBits --------- Co-authored-by: n1harika <[email protected]> * Add Max op to no_dimension_supported list --------- Co-authored-by: jatinwadhwa921 <[email protected]> Co-authored-by: Ankit Maheshkar <[email protected]> Co-authored-by: n1harika <[email protected]>
Python bindings don't call the provider factory shutdown method. We relied on this to avoid destruction order issues with statically scoped ov::Core objects. Refactor ov core and shared context lifetimes such that we don't need to rely on shutdown calls to manage life times and we avoid a static lifetime of the core. Co-authored-by: Eric Crawford <[email protected]>
…582) This change addresses an issue where a Pad op in quantized models fails due to an unsupported dimension input. The fix adds logic to detect if a Pad op is part of a quantized model by checking for a DequantizeLinear input. If found, the op is marked as quantized and the unsupported dimension check is bypassed, ensuring that the pad_value remains constant as required by the VPUX compiler. Related-to: EISW-152222 Co-authored-by: Surendar Rama Sitaraman <[email protected]>
Contributor
Author
Member
|
/azp run Linux OpenVINO CI Pipeline |
|
Azure Pipelines successfully started running 1 pipeline(s). |
Contributor
|
/azp run Linux CPU CI Pipeline, Linux CPU Minimal Build E2E CI Pipeline, Linux GPU CI Pipeline, Linux GPU TensorRT CI Pipeline, MacOS CI Pipeline, ONNX Runtime Web CI Pipeline, onnxruntime-binary-size-checks-ci-pipeline, Linux QNN CI Pipeline |
|
Azure Pipelines successfully started running 8 pipeline(s). |
Contributor
|
/azp run Windows CPU CI Pipeline, Windows GPU CI Pipeline, Windows GPU TensorRT CI Pipeline, Windows ARM64 QNN CI Pipeline, Windows x64 QNN CI Pipeline, Big Models |
|
Azure Pipelines successfully started running 5 pipeline(s). |
Contributor
|
/azp run Linux Android Emulator QNN CI Pipeline, Windows GPU CUDA CI Pipeline, Windows GPU DML CI Pipeline, Windows GPU Doc Gen CI Pipeline |
|
Azure Pipelines successfully started running 4 pipeline(s). |
Contributor
|
/azp run Win TRT Minimal CUDA Test CI Pipeline |
|
No pipelines are associated with this pull request. |
Contributor
|
/azp run Windows TRT Minimal CUDA Test CI Pipeline |
|
No pipelines are associated with this pull request. |
Contributor
|
/azp run Windows TensorRT Minimal CUDA Test CI Pipeline |
|
No pipelines are associated with this pull request. |
Contributor
|
/azp run Win_TRT_Minimal_CUDA_Test_CI |
|
Azure Pipelines successfully started running 1 pipeline(s). |
Contributor
Could you list the contrib ops supported by this PR? |
Contributor
Author
|
These are contrib ops
|
Member
|
/azp run Linux GPU CI Pipeline |
|
Azure Pipelines successfully started running 1 pipeline(s). |
Contributor
@saurabhkale17 Please list them in the git commit message. |
yihonglyu
reviewed
Feb 20, 2025
Contributor
yihonglyu
left a comment
yihonglyu
left a comment
There was a problem hiding this comment.
Could you enable tests for the contrib ops added in this PR?
yihonglyu
reviewed
Feb 20, 2025
yihonglyu
reviewed
Feb 20, 2025
jywu-msft
approved these changes
Feb 21, 2025
jatinwadhwa921
added a commit
to intel/onnxruntime
that referenced
this pull request
Feb 21, 2025
### Description This pull request combines multiple improvements, bug fixes for the OpenVINO Execution Provider (OVEP). The changes are summarized as follows: 1. Support for various contrib Ops in OVEP. 2. Dimension Check Fixes for Greater, Pad, and MAX Ops: Fixed dimension check failures for the Greater, Pad, and MAX ops in OVEP, ensuring they now pass validation for all supported models. 3. Refactor Core and Shared Context Lifetimes: Refactored the lifetimes of the OpenVINO core and shared context to remove dependency on shutdown calls. This change avoids relying on static lifetime management and improves stability and resource cleanup. 4. Fix for Duplicate DQ Node Removal: Addressed an issue where duplicate Dequantize (DQ) nodes that were initializers were incorrectly removed. Initializers should always be preserved, and this fix ensures that all duplicate DQ nodes that are initializers are retained. --------- Co-authored-by: Ankit Maheshkar <[email protected]> Co-authored-by: n1harika <[email protected]> Co-authored-by: rayngun <[email protected]> Co-authored-by: jatinwadhwa921 <[email protected]> Co-authored-by: Eric Crawford <[email protected]> Co-authored-by: Surendar Rama Sitaraman <[email protected]>
guschmue
pushed a commit
that referenced
this pull request
Mar 6, 2025
### Description This pull request combines multiple improvements, bug fixes for the OpenVINO Execution Provider (OVEP). The changes are summarized as follows: 1. Support for various contrib Ops in OVEP. 2. Dimension Check Fixes for Greater, Pad, and MAX Ops: Fixed dimension check failures for the Greater, Pad, and MAX ops in OVEP, ensuring they now pass validation for all supported models. 3. Refactor Core and Shared Context Lifetimes: Refactored the lifetimes of the OpenVINO core and shared context to remove dependency on shutdown calls. This change avoids relying on static lifetime management and improves stability and resource cleanup. 4. Fix for Duplicate DQ Node Removal: Addressed an issue where duplicate Dequantize (DQ) nodes that were initializers were incorrectly removed. Initializers should always be preserved, and this fix ensures that all duplicate DQ nodes that are initializers are retained. --------- Co-authored-by: Ankit Maheshkar <[email protected]> Co-authored-by: n1harika <[email protected]> Co-authored-by: rayngun <[email protected]> Co-authored-by: jatinwadhwa921 <[email protected]> Co-authored-by: Eric Crawford <[email protected]> Co-authored-by: Surendar Rama Sitaraman <[email protected]>
ashrit-ms
pushed a commit
that referenced
this pull request
Mar 17, 2025
### Description This pull request combines multiple improvements, bug fixes for the OpenVINO Execution Provider (OVEP). The changes are summarized as follows: 1. Support for various contrib Ops in OVEP. 2. Dimension Check Fixes for Greater, Pad, and MAX Ops: Fixed dimension check failures for the Greater, Pad, and MAX ops in OVEP, ensuring they now pass validation for all supported models. 3. Refactor Core and Shared Context Lifetimes: Refactored the lifetimes of the OpenVINO core and shared context to remove dependency on shutdown calls. This change avoids relying on static lifetime management and improves stability and resource cleanup. 4. Fix for Duplicate DQ Node Removal: Addressed an issue where duplicate Dequantize (DQ) nodes that were initializers were incorrectly removed. Initializers should always be preserved, and this fix ensures that all duplicate DQ nodes that are initializers are retained. --------- Co-authored-by: Ankit Maheshkar <[email protected]> Co-authored-by: n1harika <[email protected]> Co-authored-by: rayngun <[email protected]> Co-authored-by: jatinwadhwa921 <[email protected]> Co-authored-by: Eric Crawford <[email protected]> Co-authored-by: Surendar Rama Sitaraman <[email protected]>
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
6 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
This pull request combines multiple improvements, bug fixes for the OpenVINO Execution Provider (OVEP). The changes are summarized as follows:
Support for various contrib Ops in OVEP.
Dimension Check Fixes for Greater, Pad, and MAX Ops: Fixed dimension check failures for the Greater, Pad, and MAX ops in OVEP, ensuring they now pass validation for all supported models.
Refactor Core and Shared Context Lifetimes: Refactored the lifetimes of the OpenVINO core and shared context to remove dependency on shutdown calls. This change avoids relying on static lifetime management and improves stability and resource cleanup.
Fix for Duplicate DQ Node Removal: Addressed an issue where duplicate Dequantize (DQ) nodes that were initializers were incorrectly removed. Initializers should always be preserved, and this fix ensures that all duplicate DQ nodes that are initializers are retained.