Population-based training is very effective at tuning hyper-parameters or at improving exploration in RL. However, costs and high runtimes often drive users away from these methods. We show in our @ICMLConf paper how to efficiently train populations with #JAX on a single GPU 🚀!

We’re proud to announce the release of our new open-source library, CATX, which provides an efficient implementation of the CATS algorithm for contextual bandits in JAX. >> You can access CATX here: bit.ly/3OlWo7G >> Read our blog here: bit.ly/3PRcNCl 1/3

We’re pleased to announce a stronger collaboration w/Google Cloud as part of their Cloud Partner Advantage program. We are now a certified Service and Build partner, and will be offering AI productized innovation tools to Google Cloud customers. Info: bit.ly/3SXe90G

1/ We’re happy to announce our latest contribution to the open-source ecosystem and RL community! We are releasing Jumanji, a suite of Industry-Driven Reinforcement Learning environments written in #JAX 📚 Doc: tinyurl.com/doc-jumanji ⭐ Github: tinyurl.com/code-jumanji

LLMs can also learn the language of Covid-19. Here, BioNTech SE and InstaDeep built a model to predict high-risk Covid-19 variants far before the WHO designated them as variants. The model ingests viral spike protein sequences and predicts their immune escape and fitness.

As the world’s population passes 8B, sustainable agriculture is more important than ever. In our mission to build AI that benefits everyone, and in partnership with Google Cloud, we trained the next generation of AI models for plant genomics 🌱🧬 bit.ly/3VFfjys (1/7)

🧬 🤖 TECHBIO BOOM! BioNTech SE to buy InstaDeep for £342M in cash + up to £200M more. In Feb ‘21, I wrote “I see exciting ways that ML can be integrated into rational and predictive AI-first mRNA vaccine design” Thx for delivering Karim Beguir 😉 investors.biontech.de/news-releases/…

Our mission to build AI that benefits everyone just took a giant leap forward! BioNTech SE is to acquire InstaDeep in a £362M deal🚀to accelerate AI-driven discovery, design & development of next-gen immunotherapies & vaccines at global scale!🌍1/N bit.ly/3VW5GLn

1/ We are happy to announce the open-source release of the inference code and weights of our four genomics #LLM, the nucleotide transformers, ranging from 500M to 2.5B parameters, trained in collaboration with NVIDIA and @TU_Munchen 🧬 bit.ly/3YRSUz1

Super excited to reveal “Cape to Carthage” our new 📽️ documentary now an official selection at @AIFilmFestOrg!🔥 For the 1st time we tell the story of an all African, female-led AI research team rising against the odds, and their incredible journey. Exclusive preview today at

Tech 24 - Ces deux chercheuses africaines en intelligence artificielle impressionnent le monde entier ➡️ go.france24.com/xS5

Generative AI is expert counsel for the masses.

It’s official! 🎉Today InstaDeep becomes a proud member of the BioNTech family. Want to apply your AI skills to make a difference? Join us, we’re hiring!

🎉 New release from InstaDeep Research! Meet Flashbax ⚡ a lightweight library of accelerated replay buffers for Reinforcement Learning, written in #JAX. Dive into the code and check out the examples! github.com/instadeepai/fl…

now seems as good a time as ever to remind people that the biggest breakthroughs at OpenAI came from a previously unknown researcher with a bachelors degree from olin college of engineering

🚀 Had fun creating GrahamGPT, inspired by Y Combinator 's Paul Graham using his essays. Aim: Understanding OpenAI's GPTs, noting their current limits. Next: Testing 'actions' with real-world applications. 👀 Try it here: chat.openai.com/g/g-PRUrSsHZS-… #AI #TechInnovation

🚀 Just launched PromptGPT! It's a handy GPT for prompt engineering, making it easier to create effective prompts for language models. Perfect for AI enthusiasts at any level. 🤖💡 chat.openai.com/g/g-tQfhIDCsk-… #AI #GPTs #OpenAI #ChatGPT

We're proud to be empowering the next generation of AI talent🎦Now you can watch the full documentary #CapeToCarthage at decisiveagents.com/capetocarthage

We can finally say it publicly: poolside got their models running on Amazon Web Services’s Trainium/Inferentia thanks to ZML. No code change. Promise kept.

Today, we’re thrilled to announce a transformative partnership with AWS that establishes poolside as a first-party AWS offering. This means poolside's product and models are now available to be contracted directly from AWS — a significant milestone in our trajectory.