Announced today: MLCommons is adopting Meta Llama 2 70B for MLPerf Inference v4.0 ➡️ bit.ly/3xgaXa3

The benchmark is a standard for measuring ML & AI performance across domains and we're excited to support the community in using Llama 2 as part of the benchmark suite.

Serving billions of tokens from their hosted LLM endpoints + customers' custom models, OctoAI is using Llama Guard to support safety for their customers at scale + their own trust & safety work ⬇️
bit.ly/3VwbYou

Learn prompting best practices, few-shot and chain-of-thought prompting, how to implement safe AI practices with Llama Guard and more — the 'Prompt Engineering with Llama 2' course is open now on @deeplearningai.

Get started with the free course ➡️ bit.ly/4aodWvw

🌀Check out RAFT: Retrieval-Aware Fine Tuning! A simple technique to prepare data for fine-tuning LLMs for in-domain RAG, i.e., question-answering on your set of documents 📄 Exciting collaboration with Berkeley AI Research 🤝 Microsoft Azure 🤝 AI at Meta

MSFT-Meta blog:…

Just announced with Governor Tim Walz & City of Rosemount — we're building a new AI-focused data center in Rosemount, MN. More details ➡️ bit.ly/3ICyBjp

Recently released by researchers at Meta, the Multi-Modal Conversations in Smart Glasses dataset includes >25h of two-sided conversations to support research in areas that could unlock new possibilities for future smart glasses.

More details ➡️ bit.ly/3I8ntui

Today we're sharing new details on our latest advancements in infrastructure for AI. More details at the link below ⬇️

🎧New episode of the Lex Fridman Podcast is out now.

In Yann LeCun's third appearance on the podcast he touches on approaches to next generation AI systems, open source, the path to AGI and more.