Questions about EleutherAI

Short answers, pulled from the story.

Who founded EleutherAI and when did they start?

Connor Leahy, Leo Gao, and Sid Black founded EleutherAI on the 7th of July 2020. They initially named their project LibreAI before changing it to EleutherAI later that month.

What is The Pile dataset released by EleutherAI?

EleutherAI released The Pile on the 31st of December 2020 as an 886 GB dataset designed for training large language models. This collection contained copyrighted material such as books and subtitles from over 170,000 YouTube videos across more than 48,000 channels.

When did EleutherAI release its first major model series?

The team released their first major model series in March 2021 with sizes ranging from 125 million to 6 billion parameters. GPT-J-6B arrived on the 9th of June 2021 as the largest open-source GPT-3-like model available at that time.

How does EleutherAI generate text-to-image synthesis models?

Katherine Crowson and Ryan Murdock developed these models by combining CLIP technology with VQGAN to create the VQGAN-CLIP system without requiring special equipment. Their approach built upon earlier ideas from Google's DeepDream project to generate art based solely on text descriptions.

Who runs EleutherAI as a non-profit institute today?

The organization later incorporated as a non-profit institute run by Stella Biderman, Curtis Huebner, and Shivanshu Purohit. They manage commercial partnerships necessary to maintain large-scale machine learning operations.