Skip to main content

The Python client for MFlux.ai

Project description

mflux-ai

https://img.shields.io/pypi/v/mflux_ai.svg https://img.shields.io/travis/AIAScience/mflux-ai-python.svg?branch=master

This is the official mflux-ai python library for MFlux.ai

Features

  • Fetch connection strings and tell MLflow how to connect with MFlux.ai
  • Download and upload objects/datasets from/to the MFlux.ai cloud service

Quickstart

Installation

pip install mflux-ai

Basic usage

import mflux_ai

mflux_ai.init("INSERT_YOUR_PROJECT_TOKEN_HERE")

# MLflow now knows how to connect with your project server, hosted on MFlux.ai

Store and retrieve datasets

my_dataset = np.zeros(shape=(10000, 100), dtype=np.float32)
dataset_filename = "my-dataset.pkl"

mflux_ai.put_dataset(my_dataset, dataset_filename)

my_loaded_dataset = mflux_ai.get_dataset(dataset_filename)

assert_array_equal(my_dataset, my_loaded_dataset)

History

v0.7.0 (2020-01-14)

  • Implement mflux_ai.get_best_run(), which returns the best run in a model group defined in MFlux.ai.

v0.6.0 (2019-12-13)

  • Move functions from mflux_ai.mflux_ai to mflux_ai.core. The API stays backwards-compatible for now, but warnings are shown if the old API is used. This change was applied to get cleaner and more consistent import statements.

v0.5.3 (2019-09-23)

  • Improve the performance and the support for special characters in object names in mflux_ai.get_dataset by unpickling in memory instead of using a temporary file on disk.
  • Don’t expose non-public variables and imports on the top-level package

v0.5.2 (2019-09-20)

  • Improve the performance of mflux_ai.put_dataset by pickling in memory instead of using a temporary file on disk.

v0.5.1 (2019-09-12)

  • Add support for MLflow authentication
  • Improve the performance of mflux_ai.put_dataset
  • Correctly reset the MinIO client when init completes successfully
  • Specify the desired API version and let the user know if an upgrade is needed

v0.4.0 (2019-09-01)

  • Mark mflux_ai.set_env_vars() as deprecated. Use mflux_ai.init() instead.
  • Remove support for Python 3.4
  • Add support for secure MinIO connections

v0.3.0 (2019-08-16)

  • Add a function init that will eventually replace set_env_vars
  • Check if the provided project token is valid.

v0.2.1 (2019-08-16)

  • Set licence to Apache License 2.0
  • Transition from pre-alpha to alpha.

v0.2.0 (2019-08-14)

  • Add convenience functions for storing and retrieving datasets

v0.1.1 (2019-08-14)

  • First release on PyPI. Has support for setting environment variables for MLflow based on an MFlux.ai project token.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for mflux-ai, version 0.7.0
Filename, size File type Python version Upload date Hashes
Filename, size mflux_ai-0.7.0-py2.py3-none-any.whl (11.0 kB) File type Wheel Python version py2.py3 Upload date Hashes View hashes
Filename, size mflux-ai-0.7.0.tar.gz (16.7 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page