Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Restructure workflows steps #343

Merged
merged 150 commits into from
May 7, 2024

Conversation

PawelPeczek-Roboflow
Copy link
Collaborator

@PawelPeczek-Roboflow PawelPeczek-Roboflow commented Apr 9, 2024

Description

This is new version of Execution Engine for workflows. Shipped changes:

  • workflows blocks with state
  • workflows blocks as standalone modules with interface that makes it possible to fully describe the behaviour of block and manifest it externally via OpenAPI 3.0 and custom metadata that are drawn automatically from the code
  • EE compatible with new self-describing blocks
  • Blocks possible to control flow of workflow execution - hence Condition is no longer the only, hard-coded block with that capability
  • loading external blocks from workflows plugins
  • state of workflow block can be initialised automatically by EE
  • auto-generated docs with help of @SkalskiP
  • API to get extensive descriptions of loaded blocks dynamically (with information what is connectable to what)
  • API to get dynamic step outputs
  • API to validate workflow

Type of change

Please delete options that are not relevant.

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • This change requires a documentation update (shipped with PR)

How has this change been tested, please provide a testcase or example of how you tested the change?

  • automated unit tests for workflows running in CI for grid of Python versions
  • automated integration tests for workflows running in CI for grid of Python versions
  • integration tests for new endpoints in API
  • deployed and tested at staging env

Any specific deployment considerations

  • deployment to prod to be aligned with UI changes by @hansent

Docs

@PawelPeczek-Roboflow PawelPeczek-Roboflow changed the title Feature/restructure workflows steps Restructure workflows steps Apr 9, 2024
@grzegorz-roboflow grzegorz-roboflow linked an issue May 3, 2024 that may be closed by this pull request
2 tasks
"short_description": SHORT_DESCRIPTION,
"long_description": LONG_DESCRIPTION,
"license": "Apache-2.0",
"block_type": "fusion",
Copy link
Contributor

Choose a reason for hiding this comment

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

Also block_type may go out of sync with type


def _validate_used_kinds_uniqueness(declared_kinds: List[Kind]) -> None:
kinds_names_counter = Counter(k.name for k in declared_kinds)
non_unique_kinds = [k for k, v in kinds_names_counter.items() if v > 1]
Copy link
Contributor

Choose a reason for hiding this comment

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

could do this same way as you have done a few lines above

    kinds_names = [k.name for k in declared_kinds]
    if len(declared_kinds) != len(set(declared_kinds)):
        raise

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

yes, but my point was to deliver problematic elements in error message to provide more context

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Duplicate Inference with client.infer_on_stream() on Windows
5 participants