Skip to main content

Strong Deep-Learning Baseline algorithms for NLP

Project description


MEAD is a library for reproducible deep learning research and fast model development for NLP. It provides easily extensible abstractions and implementations for data loading, model development, training, experiment tracking and export to production.

It also provides implementations of high-performance deep learning models for various NLP tasks, against which newly developed models can be compared. Deep learning experiments are hard to reproduce, MEAD provides functionalities to track them. The goal is to allow a researcher to focus on model development, delegating the repetitive tasks to the library.


Tutorials using Colab




Baseline can be installed as a Python package.

pip install mead-baseline

You will need to have tensorflow_addons already installed or have it get installed directly with:

pip install mead-baseline[tf2]

From the repository

If you have a clone of this repostory and want to install from it:

cd layers
pip install -e .
cd ../
pip install -e .

This first installs mead-layers AKA 8 mile, a tiny layers API containing PyTorch and TensorFlow primitives, locally and then mead-baseline


We use Github CI/CD to automatically release TensorFlow and PyTorch via this project:

Links to the latest dockerhub images can be found there

A Note About Versions

Deep Learning Frameworks are evolving quickly and changes are not always backwards compatible. We recommend recent versions of whichever framework is being used underneath. We currently test on TF versions 2.1.0 and 2.4.1. The PyTorch backend requires at least version 1.3.0, though we recommend using a more recent version.


If you use the library, please cite the following paper:

  author =    "Pressel, Daniel
               and Ray Choudhury, Sagnik
               and Lester, Brian
               and Zhao, Yanjie
               and Barta, Matt",
  title =     "Baseline: A Library for Rapid Modeling, Experimentation and
               Development of Deep Learning Algorithms targeting NLP",
  booktitle = "Proceedings of Workshop for NLP Open Source Software (NLP-OSS)",
  year =      "2018",
  publisher = "Association for Computational Linguistics",
  pages =     "34--40",
  location =  "Melbourne, Australia",
  url =       ""

MEAD was selected for a Spotlight Poster at the NeurIPS MLOSS workshop in 2018. OpenReview link


  • Research supported with Cloud TPUs from Google's TPU Research Cloud (TRC)

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

mead-baseline-2.4.2.tar.gz (299.9 kB view hashes)

Uploaded source

Built Distribution

mead_baseline-2.4.2-py3-none-any.whl (431.7 kB view hashes)

Uploaded py3

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