9 C
Wednesday, September 28, 2022

Meta’s new AI brings us one step closer to a universal language translator

Must read

Trump fan who attacked cop Fanone on Jan. 6 sentenced to more than 7 years in prison

WASHINGTON — A Donald Trump fan who brought his teenage son with him when he attacked then-D.C. police officer Mike Fanone and another officer...

Can crypto make the world a better place? • londonbusinessblog.com

Image Credits: londonbusinessblog.com The crypto world has never shied away from making big promises, but as the industry matures and the public expands, now is...

Dotcom Crash Lessons, Lower CAC, Product Driven Sales • londonbusinessblog.com

On a recent Twitter Space, M13 partner Anna Barber and I looked back at the dotcom crash in search of lessons operators can use...

Limit reached – Join the EU Startups CLUB

€147/quarter This option is ideal for companies and investors who want to keep up to date with Europe's most promising startups, have full access...
Shreya Christinahttps://londonbusinessblog.com
Shreya has been with londonbusinessblog.com for 3 years, writing copy for client websites, blog posts, EDMs and other mediums to engage readers and encourage action. By collaborating with clients, our SEO manager and the wider londonbusinessblog.com team, Shreya seeks to understand an audience before creating memorable, persuasive copy.

Meta has taken another step towards creating a universal language translator.

The company has an open source AI model that translates more than 200 languages, many of which are not supported by existing systems.

The research is part of a Meta initiative launched earlier this year.

Greetings, Humanoids

Sign up for our newsletter now for a weekly roundup of our favorite AI stories delivered to your inbox.

“We call this project No language left behindand the AI ​​modeling techniques we’ve used from NLLB help us create high-quality translations on Facebook and Instagram for languages ​​spoken by billions of people around the world,” Meta CEO Mark Zuckerberg said in a Facebook post.

NLLB focuses on languages ​​with fewer resources, such as Maori or Maltese. Most people in the world speak these languages, but they lack the training data that AI translations typically need.

Meta’s new model is designed to overcome this challenge.

To do this, the researchers first interviewed speakers of disadvantaged languages ​​to understand their needs. They then developed a new data mining technique that generates training sentences for: languages ​​with few resources.

They then trained their model on a mix of the mined data and human-translated data.

The result is NLLB-200 — a massive multilingual translation system for 202 languages.

The team assessed the model’s performance against the FLORES-101 dataset, which evaluates translations of low-resource languages.

“Despite doubling the number of languages, our final model performs 40% better than the prior art on Flores-101,” the study authors wrote:

SOTA comparison