LightOn Newsletter, September 2021

Dear reader,

You probably have noticed our new website.

Let me provide you with some context as to why we are prominently featuring Muse, our API for foundation models.

In 2012, Deep Neural Networks rocked the Computer Vision world. Eight years later, the release of GPT-3 is similarly changing the course of Natural Language Processing and Artificial Intelligence as witnessed by the recent rare ArXiv report of more than 100 Stanford researchers on the opportunities of these foundation models.

Up until a year ago, LightOn’s main focus had been to get hardware technology to open new avenues of computing. Our photonic technology started in the lab and now works in Data Centers while performing massively parallel computations. We are continuing its scaling so that LightOn OPUs can eventually train and even perform inference for these massive models.

Another side to LightOn’s story comes out of our hardware-algorithm in-house co-design efforts: Better hardware requires being deeply knowledgeable about software and algorithms. For the past year, our AI team has focused on getting to be the best at building these extremely complex foundation models. We released PAGnol, a french GPT-3/foundation model, and have continued our efforts by releasing a Spanish and an Italian one. LightOn is also part of the team involved in the Big Science project that aims at building a 200 billion parameter model. As the Akronomicon leaderboard shows, we enjoy being in the company of giants 😉.

Our next step is to make these models available and we are doing so through Muse, the API for the foundation models available at LightOn. Our vision is to provide these foundation models to the 80%+ of the world population that does not interact daily in English. Nuances in an Urdu joke given in Urdu will probably not translate well in English and that is why, we believe, every model ought to be language-specific. While our future offering will feature an English model, it will eventually cover natively 40 major languages. Initial commercial availability with few languages will start in early 2022. Muse’s name was chosen because the API is expected to be a conduit for inspiration to its users in the same way Zeus’ daughters have been for literature, science, and the arts for the past 2600 years.

The Muse API is currently running in private beta with several customers. If you or your business are interested in accessing it, please contact us.

Forward we go!

Igor Carron, CEO, LightOn

Muse Press release:

Create, Process, Understand and Learn with LightOn Muse 🎭

At LightOn we are proud to unveil Muse, an extreme-scale language API to enable the integration of human-like intelligence in any product or application, in any language.

Muse is already available in private beta and we are working closely with our beta users and customers to fine-tune our offering in order to serve their business needs in the best way possible.

Our existing customers span across multiple industries including SaaS platforms, SEO, or EdTech, taking advantage of the broad capabilities of the LightOn Muse API.

Some of the Muse highlights include:

• business-ready intelligence primitives accessible in just a few lines of code, such as best in class summarization, steerable generation, text classification, and search,
• powerful and unique foundation models trained on civilization-scale high-quality curated data,
• for the first time, the ability to serve the latest AI model, natively trained in many languages, starting with French, Spanish, and Italian — with more coming soon!

You can access Muse in private beta today to empower your products with business-ready intelligence primitives, accessible in just a few lines of code.

Learn more about the Muse API

Next LightOn AI Meetup on Thursday, September 16

For our 15th AI Meetup, we will host Federico Bianchi, a Postdoctoral Researcher at Università Bocconi. Federico will talk about Contrastive Language-Image Pre-training for the Italian Language (CLIP). Join us on Thursday, September 16, 2021, from 4:30 PM to 5:30 PM CEST.

Subscribe to our free AI Meetup here

The next episode of the BigScience workshop

The next big live event BigScience Episode 2 is happening on September 20th and you can register here. The event will be co-located with INLG 2021, the 14th International Conference on Natural Language Generation. Our very own Julien Launay will talk about Scaling Modelling Decisions to the >100B scale.

Register for the BigScience workshop

Igor Carron @ TechBBQ

Our CEO and Co-founder, Igor Carron will be at #TechBBQ21 on 16–17 September 2021 in Øksnehallen, Copenhagen to talk about The next generation of sustainable machine learning.

Laurent Daudet @ ECOC

On Monday 13 September, our CTO and co-founder Laurent Daudet will talk at the Neuromorphic Computing workshop at ECOC 2021. Don’t miss the chance to watch his presentation about Scaling up AI with Photonics Computing.