Introducing Model Lab: a new tool we built to make sense of training LLMs! The app lets you easily: – compare different models (LLaMA, Pythia etc)
– estimate training FLOPS, memory, inference cost
– visualize loss curves
Give it a try here https://
cerebras.net/model-lab/
@cerebras
-
Cerebras Model Lab: Compare LLMs, Estimate Training Costs
By
–
-
Cerebras Wafer-Scale Cluster Simplifies LLM Training
By
–
Large language models are extremely challenging to train. Our latest blog discusses how the Cerebras Wafer-Scale Cluster eliminates the need for huge compute budgets and complex distributed compute techniques. Read how we train models with a few clicks: https://
hubs.li/Q01R433-0 -
Cerebras Wafer-Scale Architecture Featured in IEEE Micro
By
–
IEEE released the latest Micro magazine, delving into the Hot Chips 34 conference. One of the standout features of this edition is an article authored by Cerebras' Chief Hardware Architect, Sean Lie. Read about our unique wafer-scale architecture here: https://
hubs.li/Q01QWDSQ0 -
U-Net Models Now Train on 50 Megapixel Images
By
–
With our latest innovations, we now can train U-Net computer vision models on 50 megapixel images! Read more about how we accomplish this here: https://t.co/6jL94xTrdC pic.twitter.com/sN6OK35L8N
— Cerebras (@cerebras) 19 mai 2023With our latest innovations, we now can train U-Net computer vision models on 50 megapixel images! Read more about how we accomplish this here: https://
hubs.li/Q01Qy4rf0 -
Cerebras’ James Wang Discusses AI Hardware Innovation on ARK Podcast
By
–
ICYMI – Cerebras' James Wang joined the ARK Investment Management LLC podcast to discuss everything going on with #AI and how Cerebras is pushing the frontier of AI hardware. Listen here: https://
hubs.li/Q01QxyT90 -
Cerebras-GPT Developer Community Meeting and Model Showcase
By
–
ICYMI, we recently held a Cerebras-GPT based developer community meeting
— Cerebras (@cerebras) 18 mai 2023
Watch the discussion here: https://t.co/82I7DNF4Bj
Check out our models here: https://t.co/7xPDNdXXRX pic.twitter.com/NW6clahM9oICYMI, we recently held a Cerebras-GPT based developer community meeting Watch the discussion here: https://
hubs.li/Q01QqfBV0 Check out our models here: https://
hubs.li/Q01Qqrwz0 -
Cerebras-GPT Developer Community Meeting Best Practices
By
–
The Cerebras-GPT Developer Community Meeting starts at 9am PT today! Senior ML Engineer, Daria Soboleva, will share best practices using and deploying Cerebras-GPT. This is a great opportunity to learn about building applications with Cerebras-GPT. Join: https://
hubs.li/Q01Q27f50 -
Cerebras Connects Researchers at BioIT23 World Conference
By
–
Headed to the #BioIT23 World Conference in Boston next week? We'd love to connect with you. Why? Because we're allowing researchers to think big, removing roadblocks to research. Book a meeting with us to learn more: https://
hubs.li/Q01PqCMp0, or email bioit@cerebras.net. -
Simplified AI Training Code: Minibatches on Single GPU
By
–
It really is that simple. And the code is even simpler – just minibatches, as simple as running on one GPU!
-
Cerebras Demonstrates Sparse Neural Network Hardware at ICLR2023
By
–
Cerebras sponsored the #ICLR2023 workshop on Sparsity in Neural Networks and presented "Sparse Pre-training and Dense Fine-tuning." Our hardware natively supports structured/dynamic sparsity, and we welcome collaboration on sparsity research!