Tweet

BigScience Research Workshop

3 Jan, 2 tweets, 2 min read

T0 generalizes to unseen natural language tasks in English, often outperforming GPT-3 while being 16x smaller. 🔥

We tested T0 on funky cooking recommendations and world knowledge! 🤠 What other tasks should we give T0?

Check out the new blog post: bigscience.huggingface.co/blog/t0

@huggingface

You can try T0 directly in your browser or download it from the @huggingface model repository. huggingface.co/bigscience/T0pp

• • •

Missing some Tweet in this thread? You can try to force a refresh

This Thread may be Removed Anytime!

Twitter may remove this content at anytime! Save it as PDF for later use!

More from @BigscienceW

BigScience Research Workshop

@BigscienceW

18 Oct 21

First modeling paper out of BigScience is here!

T0 shows zero-shot task generalization on English natural language prompts, outperforming GPT-3 on many tasks, while being 16x smaller!

Model: huggingface.co/bigscience/T0pp
Repo: github.com/bigscience-wor…
Paper: arxiv.org/abs/2110.08207

This was an international collaborative effort, with over 40 people across more than 25 organizations. The group included dedicated researchers and engineers from different universities, companies, and think tanks.

Our approach uses natural language prompts that allow us to share a format for a large variety of NLP tasks. We used a prompted format with the goal of allowing our model to generalize to unseen prompted tasks.

Read 10 tweets

Share this page!

BigScience Research Workshop

Try unrolling a thread yourself!

More from @BigscienceW

BigScience Research Workshop

Did Thread Reader help you today?

Like this author's thread?