Never imagined MUM would end up in I/O when we first started it ;-) ;-)
Multitask Unified Model (MUM) — our latest AI milestone — has the potential to transform how Google helps you with complex information tasks. #GoogleIO
00:00
Afroz Mohiuddin
532 posts
Afroz Mohiuddin
@afrozenator
- Got published in Nature Communications :D nature.com/articles/s4146… With awesome collaborators: @alvin_rajkomar Eric Loreaux, Yuchen Liu, Jonas Kemp, Benny Li, Ming-Jun Chen, Yi Zhang & @Mysiak ...
- Extremely proud to have pioneered large scale distillation for Maverick and really delighted to be working alongside an extremely talented team. We truly hope the OSS community enjoys the fruits of our labour.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
- “To get what you want, you have to deserve what you want. The world is not yet a crazy enough place to reward a whole bunch of undeserving people.” — Charlie Munger RIPHighlight reel of Charlie Munger spitting one banger after another.
- Neat quotes from technical documentation. "Reusing the same [PRNG] state will cause sadness and monotony, depriving the end user of lifegiving chaos." jax.readthedocs.io/en/latest/note…
- "I have always found it strange that a stock that falls is seen as risky but a stock that fell (so in the past) becomes an opportunity." - @FromValue in an interview with @InvestmentTalkk
- Really proud to present this model to the world and really excited on what is coming ahead 🔥🦾🚀BREAKING: Meta's Llama 4 Maverick just hit #2 overall - becoming the 4th org to break 1400+ on Arena!🔥 Highlights: - #1 open model, surpassing DeepSeek - Tied #1 in Hard Prompts, Coding, Math, Creative Writing - Huge leap over Llama 3 405B: 1268 → 1417 - #5 under style control
- Congratulations to dear friends @YiTayML @PiotrPadlewski @DaniYogatama and everyone at Reka AI for an amazing multimodal model in such a short time! Eagerly looking forward for more awesomeness ahead!
We are excited to share Reka Flash ✨, a new state-of-the-art 21B multimodal model that rivals Gemini Pro and GPT 3.5 on key language & vision benchmarks 📈. We've trained this model from scratch and ground zero with a small (but amazingly capable team 🧙♂️) and relatively finite - “I never lose. I either win or learn.” — Nelson Mandela
- #Tensorflow in the #Numpy API - great work by @_agarwal_ashish, Wang Peng, Akshay Modi and team!! Works beautifully with #Trax ! Check it out :-)New in tf-nightly: the NumPy API. - GPU and TPU-accelerated NumPy code - Interoperable with the rest of the TF ecosystem Documentation: tensorflow.org/api_docs/pytho…
- Replying to @afrozenatorOfficial Blog - blog.google/products/searc…
- So many gems in here compiled by @polina_marinova The Profile Dossier: Hamdi Ulukaya, the Shepherd-Turned-Billionaire CEO by @ProfileRead theprofile.substack.com/p/the-profile-…
- Also reminds of Munger’s maxim of not taking a side on a debate unless you can put the opposite argument better than the best supporter of the said counter argument. (This also led me without an opinion on most topics 🙃)Daniel Dennett (28th March 1942 - 19th April 2024) RIP Dennett on how to criticise wisely
