.. title

Python Client for MLflow

|status| |PyPI| |PyPI License| |PyPI Python Version| |ReadTheDocs| |Build| |Coverage| |pre-commit.ci|

.. |status| image:: https://www.repostatus.org/badges/latest/active.svg :target: https://www.repostatus.org/#active .. |PyPI| image:: https://badge.fury.io/py/mlflow-rest-client.svg :target: https://badge.fury.io/py/mlflow-rest-client .. |PyPI License| image:: https://img.shields.io/pypi/l/mlflow-rest-client.svg :target: https://github.com/MobileTeleSystems/mlflow-rest-client/blob/main/LICENSE.txt .. |PyPI Python Version| image:: https://img.shields.io/pypi/pyversions/mlflow-rest-client.svg :target: https://badge.fury.io/py/mlflow-rest-client .. |ReadTheDocs| image:: https://img.shields.io/readthedocs/mlflow-rest-client.svg :target: https://mlflow-rest-client.readthedocs.io .. |Build| image:: https://github.com/MobileTeleSystems/mlflow-rest-client/workflows/Tests/badge.svg :target: https://github.com/MobileTeleSystems/mlflow-rest-client/actions .. |Coverage| image:: https://codecov.io/gh/MobileTeleSystems/mlflow-rest-client/branch/main/graph/badge.svg :target: https://codecov.io/gh/MobileTeleSystems/mlflow-rest-client .. |pre-commit.ci| image:: https://results.pre-commit.ci/badge/github/MobileTeleSystems/mlflow-rest-client/main.svg :target: https://results.pre-commit.ci/latest/github/MobileTeleSystems/mlflow-rest-client/main

Python client for MLflow <https://mlflow.org>_ REST API.

Features:

• Minimal dependencies

• Unlike MLflow Tracking client <https://mlflow.org/docs/latest/python_api/mlflow.tracking.html>__ all REST API methods and params are exposed to user.

• MLflow URL is passed via constructor argument instead of env variable, so multiple client instances could be created in the same Python interpreter.

• Basic and Bearer auth are supported (via constructor args too).

• All class fields are validated with pydantic <https://pydantic-docs.helpmanual.io>_.

• All methods and classes are documented.

Limitations:

• There is no integration with ML frameworks and libraries. You should use official MLflow client <https://mlflow.org/docs/latest/python_api/mlflow.html>__ instead.

• There is no integration with S3 or other artifact storage type. You should access it directly with boto3 <https://boto3.amazonaws.com>_ or other client.

• Supported MLflow versions: from 1.17.0 to 1.23.0. It is possible to use client with older MLflow versions (e.g. 1.10.0), but this is not guaranteed.

• Only Python 3.7+ is supported. Python 3.6 and lower already reached end of life.

.. documentation

Documentation

See https://mlflow-rest-client.readthedocs.io/

.. contribution

Contribution guide

See <CONTRIBUTING.rst>__

Security

See <SECURITY.rst>__

.. install

Installation

Stable release

Stable version is released on every tag to ``master`` branch. Please use stable releases on production environment.
Version example: ``2.0.0``

.. code:: bash

    pip install mlflow-rest-client==2.0.0 # exact version

    pip install mlflow-rest-client # latest release

Development release

Development version is released on every commit to dev branch. You can use them to test some new features before official release. Version example: 2.0.0.dev5

.. code:: bash

pip install mlflow-rest-client==2.0.0.dev5 # exact dev version

pip install --pre mlflow-rest-client # latest dev version

.. develop

Development

Clone repo:

.. code:: bash

git clone [email protected]:MobileTeleSystems/mlflow-rest-client.git

cd mlflow-rest-client

Install dependencies for development:

.. code:: bash

pip install -r requirements-dev.txt

Install pre-commit hooks:

.. code:: bash

pre-commit install
pre-commit autoupdate
pre-commit install-hooks

Test pre-commit hooks run:

.. code:: bash

pre-commit run --all-files -v

.. usage

Usage

Make sure you have an MLflow Tracking Server <https://mlflow.org/docs/latest/tracking.html#running-a-tracking-server>_ running.

.. code:: python

from mlflow_rest_client import MLflowRESTClient

client = MLflowRESTClient("https://mlflow.domain", ignore_ssl_check=True)

experiment = client.get_or_create_experiment("experiment_name")
run = client.create_run(experiment.id)

See sample.py <https://github.com/MobileTeleSystems/mlflow-rest-client/blob/main/samples/sample.py>_ for more examples.