Super excited to work with Andrew Ng and Sharon Zhou for this short course which shows how to improve accuracy of LLM applications! And it uses open source Llama 3-8b! model 🚀 to build an application that converts text to SQL with a custom schema. #Llama #LLMs

Anxiety is caused by who you think you should be. Peace is realizing who you are.

📣 Introducing Llama 3.2: Lightweight models for edge devices, vision models and more! What’s new? • Llama 3.2 1B & 3B models deliver state-of-the-art capabilities for their class for several on-device use cases — with support for Arm, MediaTek & Qualcomm on day one. •

We just unveiled Orion, our full AR glasses prototype that we’ve been working on for nearly a decade. When we started on this journey, our teams predicted that we had a 10% chance (at best) of success. This was our project to see if our dream AR glasses—wide FOV display, less

📣 New course now available on DeepLearning.AI: Introducing Multimodal Llama 3.2! The course covers both Llama 3.1 & Llama 3.2 and includes detailed rundowns on multimodal prompting, custom tool calling, Llama Stack + more. Take the 1h course for free ⬇️ bit.ly/3ZU80Ik

New Llama 3.2 course is here. This will help you get started on using Llama models for image reasoning, function calling, llama-stack and orchestration, applying multimodal prompting and more... in ~1 hr. Excellent ROI :)

Donkeys who understand physics know the easiest way to climb a steep staircase is to cross-climb. 🫏

Humans can learn something from these donkeys :)

Today is the start of a new era of natively multimodal AI innovation. Today, we’re introducing the first Llama 4 models: Llama 4 Scout and Llama 4 Maverick — our most advanced models yet and the best in their class for multimodality. Llama 4 Scout • 17B-active-parameter model

Llama 4 takes 43 seconds to analyse 900k tokens!

10 MIND-BLOWING things about Llama 4 that CRUSH the competition! 🤯   1. Outranks ChatGPT 4.0 on LM Arena leaderboard   2. Beats Grok, DeepSeek & Gemini 2.0 in benchmarks   3. Features MASSIVE 10M token context window   4. Completely FREE to use via Open Router   5. Comes in 3

Llama 4 independent tests suggest a Maverick is very solid model, but not enough to beat DeepSeek v3 (non-reasoner version), though size-performance trade-offs make it hard to do exact comparisons.

We evaluated Llama 4 ourselves: On GPQA Diamond, Maverick and Scout scored 67% and 52%, similar to Meta’s reported 57% and 69.8%. On MATH Level 5, Maverick and Scout got 73% and 62%. Maverick is competitive with leading open or low-cost models, and both outperform Llama 3.

🚨 No data? No problem. The new Absolute Zero framework trains LLMs to reason — with zero external data. 🤖 Self-generates tasks 🧠 Solves + verifies via code 🎯 Learns through reward signals 📄 arxiv.org/abs/2505.03335 #AI #LLM #ReinforcementLearning #AutonomousAI

🔥 Want to break into AI product management? This roadmap is gold — everything from basic ML concepts to fine-tuning, RAG, agents, prototyping, and evaluation loops. 📌 Save this. Study this. #AI #ProductManagement #LLMs #PromptEngineering

Introducing "Building with Llama 4." This short course is created with Meta AI at Meta, and taught by Amit Sangani, Director of Partner Engineering for Meta’s AI team. Meta’s new Llama 4 has added three new models and introduced the Mixture-of-Experts (MoE) architecture to its

Power of PyTorch where the model keeps training even with failures every 15s.

Tired of manual prompt tweaking? Watch the latest Llama tutorial on how to optimize your existing GPT or other LLM prompts for Llama with `llama-prompt-ops`, the open-source Python library! In this video, Partner Engineer Justin Lee demonstrates installation, project setup,