Repository navigation
Add Openvino support - #1461
Merged
Merged
Add Openvino support#1461
Conversation
Supreet Singh Palne (spalne)
marked this pull request as ready for review
October 7, 2026 21:03
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
xieofxie
reviewed
Oct 9, 2026
xieofxie
reviewed
Oct 9, 2026
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
xieofxie
reviewed
Oct 9, 2026
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
xieofxie
approved these changes
Oct 9, 2026
Qiong Wu (qiowu) (DingmaomaoBJTU)
added a commit
that referenced
this pull request
Oct 9, 2026
Documents Intel OpenVINO NPU prerequisites, Qwen3 bundle build and perf commands, pipeline stage placement, and optional runtime `--ep-options load_config=...` usage. The options are passed to GenaiSession at runtime; they are not embedded in the model-ID auto-build bundle. Includes the minimal type annotation fix in `src/winml/modelkit/export/value_range.py` required to pass the repository's Mypy lint check. Depends on #1461. --------- Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Contributor
There was a problem hiding this comment.
🟡 Changes recommended
Shared sidecar paths can select incorrect weights, and the advertised OpenVINO configuration option is not implemented.
2 open findings
What changed in this PR
Adds OpenVINO NPU/GPU support to Qwen3 GenAI bundles and performance reporting.
Changes:
- Adds OpenVINO routing, defaults, and provider-option overrides.
- Handles external weights for shared OpenVINO EPContexts.
- Includes generated responses in reports and JSON.
| File | Description |
|---|---|
src/winml/modelkit/session/genai_session.py |
Adds provider overrides and shared-weight handling. |
src/winml/modelkit/session/ep_device.py |
Catalogs OpenVINO defaults. |
src/winml/modelkit/models/winml/genai_bundle.py |
Forwards device and assembler options. |
src/winml/modelkit/models/hf/qwen3/genai.py |
Adds OpenVINO Qwen3 routing. |
src/winml/modelkit/commands/perf.py |
Forwards GenAI EP options. |
src/winml/modelkit/commands/_perf_genai.py |
Reports generated responses. |
src/winml/modelkit/export/value_range.py |
Improves typing annotations. |
tests/unit/session/test_genai_session.py |
Tests overrides and external weights. |
tests/unit/session/test_ep_device.py |
Tests OpenVINO defaults. |
tests/unit/models/winml/test_genai_bundle_registry.py |
Tests registered OpenVINO targets. |
tests/unit/models/winml/test_genai_bundle_orchestrator.py |
Tests device forwarding. |
tests/unit/models/qwen3/test_genai_config.py |
Tests OpenVINO pipeline configuration. |
tests/unit/commands/test_perf_genai.py |
Tests options and response reporting. |
tests/unit/commands/test_build_genai_bundle.py |
Tests OpenVINO GPU builds. |
🧠 Review effort: Balanced
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
# Conflicts: # src/winml/modelkit/export/value_range.py
xieofxie
approved these changes
Oct 9, 2026
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
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.
Adds OpenVINO NPU/GPU support for Qwen3 GenAI bundles and
winml perf --runtime ort-genai.--ep-options KEY=VALUEoption (for example,--ep-options load_config='{"NPU":{"NPU_TURBO":"NO"}}'). Explicit nested values override matching defaults while unspecified defaults remain.Verified on Intel NPU with Qwen3-0.6B: correct answers, ~1.5 s load, ~45–53 tokens/sec.