cogment 1.3.0

Cogment python SDK

cogment-py-sdk

Cogment is an innovative open source AI platform designed to leverage the advent of AI to benefit humankind through human-AI collaboration developed by AI Redefined. Cogment enables AI researchers and engineers to build, train and operate AI agents in simulated or real environments shared with humans. For the full user documentation visit https://docs.cogment.ai

This module, cogment-py-sdk, is the Python SDK for making use of Cogment when working with Python. It's full documentation can be consulted at https://docs.cogment.ai/cogment/cogment-api-reference/python/.

Developers

Local setup

Make sure you have the following installed:

• Python (any version >=3.7 should work),
• Poetry.

Install the dependencies, including downloading and building the cogment protobuf API, by navigating to the python SDK directory and run the following

poetry install

Define used Cogment protobuf API

The version of the used cogment protobuf API is defined in the .cogment-api.yaml file at the root of the repository. The following can be defined:

• cogment_api_version: "latest", is the default, it retrieves the latest build of the cogment-api develop,
• cogment_api_version: "vMAJOR.MINOR.PATCH[-PRERELEASE]", retrieves an official release of cogment-api.
• cogment_api_path: "../path/to/cogment-api", retrieves a local version of cogment-api found at the given path ; if set, this overrides cogment_api_version.

⚠️ when building a docker image, cogment_api_path needs to exists in the docker file system. In practice it means it should be a subdirectory of the current directory.

Tests

To run them the first step is to configure the way to launch the orchestrator and the cli in a .env file.

You can copy .env.template for an example of what's expected.

Module tests

These tests only rely on the sdk, no connection to an orchestrator is done.

To execute the module tests, simply run

poetry run task test

Integration tests

These tests launch and use an orchestrator they are slower but more in depth.

Run in a docker image

This is the easiest way to run the tests.

./scripts/run_integration_test.sh

By default it will use the latest public version cogment/orchestrator released on DockerHub. To use specific versions define COGMENT_ORCHESTRATOR_IMAGE.

./scripts/run_integration_test.sh

Run locally

poetry run task test --launch-orchestrator

You'll need to define:

• Either COGMENT_ORCHESTRATOR_IMAGE (as above) or COGMENT_ORCHESTRATOR as the path to a locally built Cogment orchestrator executable.

Lint

Run the linter using

poetry run task lint

Check conflicting dependencies with "popular" Python packages

./scripts/check_dependencies_conflicts.sh

This script will check for conflicts required by the cogment-py-sdk and the popular Python packages in the AI/ML/Data ecosystem.

Build the source package

Build the source package (this step will only be succesfull if poetry install succeeded)

poetry build -f sdist

Release process

People having mainteners rights of the repository can follow these steps to release a version MAJOR.MINOR.PATCH. The versioning scheme follows Semantic Versioning.

1. Run ./scripts/create_release_branch.sh MAJOR.MINOR.PATCH to create the release branch and update the version of the package,
2. On the release branch, check and update the changelog if needed,
3. Update ./.cogment-api.yaml to use the public release of the API,
4. Update ./.gitlab-ci.yml, integration_test, and COGMENT_ORCHESTRATOR_IMAGE to match the latest public releases of the orchestrator and cli
5. Make sure everything's fine on CI,
6. Run ./scripts/tag_release.sh MAJOR.MINOR.PATCH to create the specific version section in the changelog, merge the release branch in main, create the release tag and update the develop branch with those.

The rest, publishing the package to PyPI and updating the mirror repositories, is handled directly by the CI.