Manages the workflows for the ML pipelines for the workflow engine. Adds necessary information such as the execution informations (usertoken, computing resources, etc.) on the fly to them.
├── [+] oas/ # Mirror of the services/oas directory (needed for the docker build)
├── [+] src/ # Source code of the app
│ ├── api/ # Entry point of the rest api
│ ├── decorators/ # Shared decorators
│ ├── exceptions/ # Collection of exceptions
│ ├── parser/ # Implementation of different parsers
│ ├── repository/ # Implementation of different storage locations of the pipelines
│ └── services/ # Services the api exposes
├── [+] tests/ # Tests of the app
├── README.md # Inception
├── gunicorn_conf.py # Configuration of the webserver
└── pyproject.toml # Pyproject file
- Install the project with
poetry install
- Build the components and pipelines (see templates)
- Direct the
PIPELINE_TEMPLATES_DIR
variable to the directory where you built the pipelines - Run the server with
poetry run python src/main.py
Here there's a list of the most relevant scripts:
openapi.sh
: generate the openapi server and client code
On this project is implemented, at the moment, these kind of tests:
- ✅ Unit tests
- ✅ Integration tests
⁉️ E2E tests⁉️ Visual tests