microsoft / microsoft/winml-cli
P1-TEST-003: ISV Model Analysis Report Validation
Open
@DingmaomaoBJTU is already working on this.
Since Mar 31, 2026.
P1
testing
triaged
- Dominant language
- Python
- Stars
- 40
- Forks
- 11
- Avg merge
- 1d 8h
- Merged PRs (30d)
- 50
Description
Summary
Validate the ISV model analysis report (P1-MODEL-003) by cross-checking analysis results against runtime behavior and ensuring all documented issues are reproducible.
Context
The ISV model analysis report is only valuable if its findings are accurate. This validation step ensures that architecture findings, export failure root causes, and EP compatibility assessments are reproducible and correct.
From plans/release/0501_release_plan/P0_CHECKLIST.md (P1-TEST-003).
Current State
- No ISV analysis report yet (depends on P1-MODEL-003)
- No validation framework for ISV report accuracy
Desired State
- Each ISV report finding validated against runtime behavior
- Export failures reproducible in a clean environment
- EP compatibility assessments validated against runtime results (where hardware available)
Acceptance Criteria
- All ISV export failure root causes verified as reproducible in clean environment
- EP compatibility claims validated against static analyzer results
- At least 3 ISV models successfully exported + run E2E on at least 1 EP
- Analysis report accuracy: ≥ 90% of findings confirmed correct
- Discrepancies between report and runtime documented
Technical Notes
- Depends on P1-MODEL-003 (ISV analysis report)
- Validation environment: clean Python venv +
uv(no manual installations) - Hardware validation: best-effort on available P0 EPs
Related Files
plans/release/0501_release_plan/P0_CHECKLIST.md— P1-TEST-003
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.