Skip to main content

IBM Services Framework for ML Applications Python 3 framework for building robust, production-ready machine learning applications. Official ML accelerator within the larger RAD-ML methodology.

Project description

MLApp · pip version Build Status License

MLApp is a Python library for building scalable data science solutions that meet modern software engineering standards.

MLApp was built and hardened in an enterprise context, to solve scalability issues for mid-size to Fortune 50 companies. It is applicable to a variety of data science use cases including machine learning, deep learning, NLP and optimization.

  • Embedded MLOps: Standardizes the way models and their metadatas are registered, stored and deployed.
  • Project scaffolding: Generates an opinionated project file structure that enforces modern engineering standards and improves readability and documentation across solutions.
  • Boilerplates: Includes a library of pre-built model templates that can be easily customized to accelerate development of common use cases.
  • Utilities: Includes an extendable set of utilities that increase developer productivity - including functions for selecting features and optimizing hyperparameters.
  • Connectors: Allows developers to easily integrate their projects with common data and analytics services.
  • Deployment integration: Applications built using MLApp can easily be deployed on common open and proprietary platforms, including Kubernetes and Azure Machine Learning.

Getting started

Install MLApp via pip:

pip install mlapp

Navigate to an empty project folder and generate the project scaffold:

mlapp init

Install a working example using boilerplates:

mlapp boilerplates install basic_regression

Update the run.py file in your project directory to point to the Basic Regression asset that you just installed:

configs = [
    {
        'config_path': "assets/basic_regression/configs/basic_regression_train_config.py",
        'asset_name': "basic_regression",
        'config_name': "basic_regression_config"
    }
]

Execute the run.py script:

python run.py

Congrats! You've trained your first model in MLApp. Take a look at the output directory to see the results.

Next steps

A great place to start is the crash course.

You should also check out the full project documentation.

Contributing

We welcome contributions from the community. Please refer to CONTRIBUTING for more information.

Project details


Download files

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

Source Distribution

mlapp-0.1.4.tar.gz (195.5 kB view hashes)

Uploaded Source

Built Distribution

mlapp-0.1.4-py3-none-any.whl (262.5 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page