This is a multi-container project used to serve inference requests to ML models and register their predictions in the context of the research project SPIRA.
The source code is contained in the directory src.
The containers used in this project and their connections are configured in the file docker-compose.yml.
The tests are contained in the tests directory.
The file docker-compose.yml also specifies the necessary env files for the project, that should be contained in the envs directory.
The variables of each env file are the same as the attributes of the classes declared in src/settings.py. The main application uses pydantic to directly bind the environment variables to the respective attributes of the Settings classes. Additional environment files for the services other than the main one are specified by their respective documentations and the files should also be contained in in the envs directory.
Dependency management is done with poetry and they are configured in the file poetry.lock. Dependencies can be added and removed using the poetry cli.
Build the project:
docker compose build api
Run the project:
docker compose --profile production up
Run tests:
make all-tests
The Makefile divides the tests in sections. To run a specific test section:
make {test-section}