Combined, training these lightweight Llama models required 916K GPU hours on NVIDIA H100-80GB. As part of Meta’s commitment starting in 2020, we’ve maintained net zero greenhouse gas emissions in our global operations and matched 100% of the electricity use from the training
HARDWARE
-
Llama 3.2 Lightweight Models Launch with Mobile Platform Support
By
–
The lightweight Llama 3.2 models shipping today include support for @Arm
, @MediaTek & @Qualcomm to enable the developer community to start building impactful mobile applications from day one. -
Llama 1B & 3B: Lightweight Models for Mobile Edge Deployments
By
–
Even With their lightweight size, Llama 1B & 3B have a range of capabilities and were built to run on mobile devices & lightweight edge deployments. They empower developers to build personalized & private, on-device agentic applications.
-
Meta’s 1B Llama Model Enables Offline Mobile Chatbots
By
–
meta's 1B llama model, meant for running on devices, may be a game changer as far as getting things like powerful chatbots on phones without needing internet access (which could also help with some privacy concerns). super curious to see how it gets used and how well it works.
-
Llama 3.2: Lightweight Models for Edge Devices and Vision
By
–
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.
• -
Sam Altman’s Global Tour for Data Center and Chip Investment
By
–
Tracking the Sam Altman road show to drum up global support for a massive data-center and chip-factory investment plan. @CadeMetz + @trippmickle nytimes.com/2024/09/25/busin…
-
MathWorks STMicroelectronics Framework Streamlines TinyML Development
By
–
Advancing Edge AI and TinyML Development! Read how MathWorks and STMicroelectronics developed a framework that uses @MATLAB to streamline TinyML application development on resource-constrained devices https://
spr.ly/6015oWR1c -
Humanoid Robot Race: Which Company Will Win?
By
–
Which humanoid robot is going to win? How long before seeing them on the train/bus? The "humanoid robot race" mirrors the competition we've seen with large language models (LLMs). This race is complex, with different companies pushing the boundaries of what humanoid robots can
-
RDUs: Beyond GPUs for Generative and Agentic AI
By
–
Thank you to our partner @AITECHio for sharing the power of our RDUs! As #AI continues to mature, the industry is starting to move beyond traditional #GPUs. RDUs are an excellent alternative to power the most demanding generative & agentic AI workloads.
-
SolarPro: Most Powerful LLM on Single GPU
By
–
@upstageai #SolarPro has been featured in the most popular tldr newsletter among developers and startups (600,000 viewers). The article is titled "Most powerful LLM on a single GPU (12 minute read)". Please give it a read! https://
a.tldrnewsletter.com/web-version?ep
=1&lc=3bf8d792-bd90-11ee-817b-b777f755fa8b&p=703435f8-799a-11ef-8b11-371f32244928&pt=campaign&t=1727097241&s=dcda0944a0dad5602d8b27308915cf426959da99e53da7f4f12b2e6ffb0fd41f
…