🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Had a fun conversation with Pascal Hartig on designing sustainable data centers. Check out the podcast to learn more about Carbon Explorer!

Jevon's paradox can be annoying--you think you made the world better by reducing cost&energy for datacenter / #HPC / deep learning workloads just to discover that this increases demand even more! via bilge (ACM SIGHPC award runner up) at MLsys seminar. youtube.com/watch?v=LPrFL2…

#HotCarbon22 agenda exceeds expectations! Looking forward to presenting out work on 'Carbon Dependencies in Datacenter Design and Management'. Check out the program 👇

Hi all, I am co-organizing ACM SIGEnergy Workshop on Society, Climate, and Sustainability (SIGEnegy WeCan) with Prashant Shenoy and Minghua Chen. Find the program at lnkd.in/ewNXxXsU. A must attend for people interested in Climate and Sustainability. #sustainability

We are going to feature amazing key note speakers! On June 30th, Carole Jean Wu from Facebook AI will talk about the carbon footprint of AI and ways to reduce it! Check out the full announcement and remember that registration is only $10 😉 energy.acm.org/conferences/ee… ACM SIGEnergy

#DataPerf arxiv.org/pdf/2207.10062… Benchmarks for Data-Centric AI Development Vijay Janapa Reddi MLCommons Colby Banbury Xiaozhe Yao Will Gaviria Rojas Douwe Kiela david jurado David Kanter Rafaelmosquera Lora Aroyo bilge Tzu-Sheng Kuo 郭子生 Tristan Thrush Margaret Warren Andrew Ng (and many more)

"Understanding Scaling Laws for Recommendation Models" For two years, the AI world has had this glorious period of believing that big tech companies just need more compute to make their models better, not more user data. That period is ending. Here's what happened: [1/14]

I'm looking for PhD research interns for summer 2023 to work in ML efficiency and sustainability areas. Reach out to me with your CV if interested!

Just when I was thinking about using Twitter more actively to write about research stuff, this place turned into a circus town. I guess I'll just continue not using it much as before. 🤐😅

Excited to share our ASPLOS'23 paper on carbon-aware datacenters. We study how renewable energy from diverse sources, energy storage, and workload scheduling can balance trade-offs between embodied and operational carbon. Congrats to bilge and team! seas.upenn.edu/~leebcc/docume…

Big earthquake in Southeast Turkey, populated area and at night—people will be caught asleep at home. Preliminary reports are M 7.8. Early photos already showed pancaked bulletin. #DEPREMOLDU is the hashtag. (Or #deprem). Almost certainly needs global rescue team mobilization.

The 1999 Izmit earthquake killed ~18,000. Magnitude 7.4. Now, *two* earthquakes in Turkey, ten hours apart: M 7.8 and 7.7. Richter is a log scale. Each one is ~2.5 times bigger and ~four times stronger. Fault break seems to be hundreds of kilometers. All populated areas.😢

The future of #ML is data-centric! That’s why we built #DataPerf, the leaderboard for data. It is the 1st platform and community for data-centric competitions. Together we will break through data limitations and unlock better ML for the world mlcommons.org/en/news/datape…

Today, Meta researchers together with MLCommons working group, are launching DataPerf, the first platform for building data & data-centric AI algorithm leaderboards. We're excited for how DataPerf will help to push the data-centric AI field forward ⬇️

10 years of FAIR. 10 years of advancing the state of the art in AI through open research. We're celebrating the 10th anniversary of Meta's Fundamental AI Research team and continuing that legacy by sharing our work on three exciting new research projects today. Details below 🧵

Breaking my X fast to post this. Need ~2x faster LLM inference? Check this out! 👇

Meta presents Is Flash Attention Stable? Finds that Flash Attention sees roughly an order of magnitude more numeric deviation as compared to Baseline Attention at BF16 when measured during an isolated forward pass arxiv.org/abs/2405.02803

😺