Let's compare CG-1 with Nvidia's Israel-1. Israel-1 has 2048 H100 GPUs. But each GPU shows up as a unique device. It's your job to break apart your model and farm them out to each GPU. On Cerebras, 1 to 64 CS-2s show as one accelerator. Just implement mini-batches and train!
MACHINE LEARNING
-
AI uses deep learning to decode animal language patterns
By
–
Researchers are figuring out how to use AI to talk to animals: Deep learning models are able to derive patterns from digital bioacoustics to map out animal language just like any other.
-
Edward Grefenstette: AI and NLP Research Contributions
By
–
As of my last update in September 2021, Edward Grefenstette is a renowned researcher in the field of artificial intelligence and natural language processing, particularly known for his contributions to the area of deep learning. There is no information suggesting that he is
-
Weak Supervision for Open-Source LLM Instruction Tuning Data Curation
By
–
We used weak supervision to programmatically curate instruction tuning data for open-source LLMs like Llama 2 and RedPajama, enabling more granular error analysis and higher quality—without an army of manual annotators. Links to data and models on the blog!
-
How AI Could Supercharge Climatetech Innovation
By
–
How AI could supercharge climatetech
#AI #AIio #BigData #ML #NLU #Futureofwork @gp_pulipaka @stratorob @PetiotEric @EvanKirstel @Fgraillot @HaroldSinnott @HeinzVHoenen @helene_wpli http://
ow.ly/95Yc30swkQ0 -
FlashAttention-2: 2x Faster Attention with Better Parallelism
By
–
FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning FlashAttention-2 is 2x faster than previous version. Release blog: https://
crfm.stanford.edu/2023/07/17/fla
sh2.html
… Paper: https://
tridao.me/publications/f
lash2/flash2.pdf
… -
Weekly Thread on Latest AI and ML Developments
By
–
In the light of our short context window and lots of cool things that are coming out instantly, I am starting a weekly running thread of things that I find useful spanning a wide variety of topics from LLMs, vision, robotics, multimodals LLMs, etc… I have been jotting down
-
ADAS Algorithm Research Enhances Vehicle Occupant Safety
By
–
🚗🔬 Making strides in enhancing occupant safety through algorithm research with vehicle ADAS systems! #SafetyFirst🦺 https://t.co/VhZ5CfO0Ci
— MathWorks (@MathWorks) 20 juillet 2023Making strides in enhancing occupant safety through algorithm research with vehicle ADAS systems! #SafetyFirst
-
Progress in AI: Sudden Breakthroughs After Slow Development
By
–
Reality moves in jumps. Progress comes slowly, and then all at once.
-
Hugging Face Hiring ML Engineer Community Builder Research Support
By
–
Hiring for a new team at @huggingface role is a a mix between ML engineer/community builder working with the wider research community to train models, collect datasets, and building demos. Supporting throughout all the research phases dm me if interested