Skip to content
This repository has been archived by the owner on Aug 28, 2023. It is now read-only.

Latest commit

 

History

History
94 lines (77 loc) · 4.54 KB

pull_request_template.md

File metadata and controls

94 lines (77 loc) · 4.54 KB

OpenVINO Deep Learning Workbench Pull Request

All Submissions:

  • Create a meaningful name of the PR with reference to JIRA. Example: [43714] [NIGHTLY] Take correct tag for snyk scans
  • JIRA tickets:
    • 44177 - Model analyzer fails with topology type guessing
  • Set assignee to the PR.
  • Add corresponding labels to the PR.
  • Set appropriate milestone to the PR.
  • Added tests for new functionality (unit / e2e)
  • Followed code style guidelines:
    • Code covered with typings
    • No debug printing in production code
    • No commented blocks of code in production code

Submissions with migration changes:

REMOVE SECTION IF NOT APPLICABLE

  • Check that workbench still works after the update from the previous version

Submissions with dependencies changes:

REMOVE SECTION IF NOT APPLICABLE

  • Communicated to IP plan owner with the list of new production dependencies
  • Ensure all dependencies versions are frozen (including development dependencies)
  • Updated dependencies in dependencies/edge_node_setup.py. Ask artyomtugaryov or akashchi for more details.

Submissions with Netron-related changes:

REMOVE SECTION IF NOT APPLICABLE

  • Run npm run init-netron from the client directory to check copying Netron sources
  • Checked integration with latest Netron version on Ubuntu:
    • Delete node_modules directory
    • Install node modules (run npm i)
    • Run script for copying Netron sources (run npm run init-netron from the client directory)
  • Checked integration with latest Netron version on Windows
  • Check topologies visualization (both original and runtime graphs) with latest Netron in WB UI:
    • squeezenet
    • mobilenet-ssd
    • semantic-segmentation-adas

Submissions that contain model-analyzer changes:

REMOVE SECTION IF NOT APPLICABLE

  • PR in model-analyzer repository: openvinotoolkit/model_analyzer# .

Submissions that contain update of the OpenVINO package and image:

REMOVE SECTION IF NOT APPLICABLE

  • PR name should be special.
    Template: [Issue Number for Transition] Transition to the new package (<package number>, <WW>.<day number>`<YY>[, RC#(number of release candidate]).
    Example: [CI] Transition to the new package (#170, ww49.5`20, RC#2)
  • Updated openvino_image in openvino_version.yml in the private CI repository
  • Updated dependencies in dependencies/edge_node_setup.py. Ask artyomtugaryov or akashchi for more details.
  • Status of pre-commit pipeline: FILL HERE.
    Example: GREEN.
  • Status of custom deployment pipeline: FILL HERE.
    Example: GREEN. Custom deployment does not happen automatically.
    Therefore, there is a need to trigger it manually. Ask SalnikovIgor or akashchi for more details.
  • Link to the matching PR in model-analyzer repository: openvinotoolkit/model_analyzer# .
  • Status of model-analyzer pipeline: FILL HERE.
    Example: GREEN.

NOTES:

  • To mark the item as done, add x to the empty [ ], so that you get [x] for items that you have done.
    Ignore items that are not relevant to your PR.

  • To add link, add [Describing text](http://some-link).

  • Fill the checklist when creating the PR. Pull Requests that are not properly formatted will not be reviewed.

  • PRs with WIP label are not reviewed by default unless explicitly requested.

  • To learn more about GitHub markdown rules, refer to documentation

Glossary

Assignee - Assignee is the single person whose review you require. Feel free as many reviewers to the PR. Reviewer is a person who will benefit from being aware of the proposed changes or whose opinion you want to know. Rule: keep single assignee.

JIRA ticket. At least one ticket that current PR addresses. Do not include project code. For example: 12345.

Available commands in pull request comments

  1. /bump:client - bump to latest master branch in client submodule
  2. /bump:client:releases/2021/4 - bump to latest releases/2021/4 branch in client submodule
  3. /bump:client:yatarkan:feature/yt/feature-name - bump to latest feature/yt/feature-name branch from yatarkan remote fork in client submodule
  4. /update - merge base (target) branch into target branch and push changes

NOTE: For more details, please visit the DEVELOPER.md