Facebook is open-sourcing PyText for faster NLP development

Facebook AI Research is open-sourcing PyText, a modeling framework that blurs the boundaries between experimentation and large-scale deployment (built on PyTorch, its open source deep learning framework).

FB wants to make to make it easier to build and deploy NLP systems for developers, according to a blog post by Ahmed Aly Hegazy and Christopher Dewan today. That includes a simplified workflow for faster experimentation; prebuilt model architectures and utilities for text processing and vocabulary management. 

AI researchers and engineers can use PyText to more quickly and easily experiment with and deploy systems to perform document classification, sequence tagging, semantic parsing, multitask modeling, and other tasks.

Facebook today uses PyText for more than a billion daily predictions and is publishing its work, open-sourcing the PyText framework, and sharing pretrained models and tutorials for training and deploying PyText models at scale.

Below are some of the other open-sourced Facebook AI tools, including PyTorch, with their respective websites and github.

To read more

Register below for access to the Warren Enskat client portal

Written by

Daniel Enskat

Daniel has written over a dozen books on the global asset management industry and has lectured at universities around the world alongside speakers such as Secretary of State John Kerry, Dr. Mark Mobius, ex-Fed Chairman Alan Greenspan, Professor KC Chan and former Prime Minister Gerhard Schroeder.

He is widely sought after for presentations, discussions and his perspective on the global asset management industry, and in the last two decades has advised hundreds of investment management CEOs on strategy and global expansion.

Scroll to Top