Skip to content

[NPU] Added NPU compile tests support to model_hub_tests#36771

Draft
asukhobo wants to merge 22 commits into
openvinotoolkit:masterfrom
asukhobo:npu-model-hub-tests
Draft

[NPU] Added NPU compile tests support to model_hub_tests#36771
asukhobo wants to merge 22 commits into
openvinotoolkit:masterfrom
asukhobo:npu-model-hub-tests

Conversation

@asukhobo

@asukhobo asukhobo commented Jul 8, 2026

Copy link
Copy Markdown

Details:

Added NPU compile tests support to model_hub_tests. Added new tmp workflow for precommit testing for NPU compilation, will be merged into windows release/debug flow after.

Tickets:

  • E-104135

AI Assistance:

  • AI assistance used: yes
  • *github workflow + review and suggestions for current model_hub_tests suite. All changes were reviewed and tested locally(except github actions workflow) *

asukhobo and others added 6 commits July 8, 2026 13:59
Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
NPU has no runtime for dynamic shapes and the PyTorch frontend keeps input dims dynamic even when traced with a concrete example. Declare static input shapes (derived from the traced example) via convert_model(input=...) for the NPU device so the compiler-in-plugin can compile. CPU/GPU behavior is unchanged.

Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
…supported on NPU due to test fauilures

Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
@github-actions github-actions Bot added category: CI OpenVINO public CI category: TF FE OpenVINO TensorFlow FrontEnd github_actions Pull requests that update GitHub Actions code category: PyTorch FE OpenVINO PyTorch Frontend category: JAX FE OpenVINO JAX FrontEnd labels Jul 8, 2026
@sys-openvino-ci sys-openvino-ci added the ExternalIntelPR External contributor from Intel label Jul 8, 2026

@akashchi akashchi left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please re-request the review once the changes are in the main windows workflow.

Comment thread .github/workflows/job_npu_tests.yml Outdated
Comment thread .github/workflows/windows_vs2022_npu.yml Outdated
Comment on lines +6 to +9
runner:
description: 'Machine on which the tests would run'
type: string
required: true

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I can see that some of the commands are Windows-specific (expand-archive, ^, etc.). Are you planning on enabling this job for Linux as well? If so, I think the commands should be revised. If not, the image input can be removed.

Copy link
Copy Markdown
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sure, but right now Windows is the major platform for NPU, so it goes with higher priority to enable. Later we will integrate our NPU checks for Linux as well.

Signed-off-by: Sukhobochenko, Anton <anton.sukhobochenko@intel.com>
@asukhobo asukhobo force-pushed the npu-model-hub-tests branch from 7b72d7b to cda8074 Compare July 13, 2026 14:13
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

category: CI OpenVINO public CI category: JAX FE OpenVINO JAX FrontEnd category: PyTorch FE OpenVINO PyTorch Frontend category: TF FE OpenVINO TensorFlow FrontEnd ExternalIntelPR External contributor from Intel github_actions Pull requests that update GitHub Actions code

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants