No better way to kick off the weekend than building with this incredible community!

Building with an open model enabled the team behind RadOnc-GPT to fine-tune the model on their records using local GPUs. This means that no patient data is ever shared outside of their secure network — a requirement for data security in healthcare settings.

There are few fields in medicine that require more precision or data than radiation oncology. RadOnc-GPT is a fine-tuned LLM, built with Meta Llama 2, that has the potential to significantly improve radiation therapy decision-making.

More details ➡️ go.fb.me/ma1vim

New research from FAIR: Better & Faster Large Language Models via Multi-token Prediction

Research paper ➡️ go.fb.me/wty7gj

We show that replacing next token prediction tasks with multiple token prediction can result in substantially better code generation performance…

The Meta Llama 3 Hackathon is this weekend in SF with @Cerebral_Valley! Get on the list ➡️ go.fb.me/em3p78

What to expect
• Two days of building alongside the best hackers in AI
• Hands on support from the Llama team
• Talks from some of the top names in the industry

AudioSeal is accepted at #ICML2024 ! 🚀 AI at Meta

AudioSeal is the state of art audio watermarking model designed for deepfakes mitigation. It is robust, lightning-fast & imperceptible⚡️

📄 paper: arxiv.org/abs/2401.17264

🔗 code (commercial friendly) github.com/facebookresear…

We continue to be amazed by how organizations of all sizes are building with Meta Llama models — from education to generative AI assistants for business and all the way to medical applications ⬇️
go.fb.me/9tspzb