Hugging Face's logo
Activity Feed

Inference Provider

VERIFIED
3,354,896 monthly requests

AI & ML interests

None defined yet.

Papers

Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference

View all Papers

Organization Card
Community About org cards

Cerebras is the inventor of the Wafer-Scale Engine – the revolutionary processor at the heart of our Cerebras CS-X systems.

Our co-designed hardware/software stack enables training large language models upward of 1 trillion parameters using only data parallelism. Our massive on-chip memory enables us to fit all model weights in SRAM and run the World’s Fastest AI Inference.

Browse popular models enabled by Cerebras on Hugging Face!

Explore models trained on Cerebras!

Join the Cerebras Discord to discuss our work and research!

Collections 3

Cerebras REAP
Sparse MoE models compressed using REAP (Router-weighted Expert Activation Pruning) method
Cerebras REAP
Sparse MoE models compressed using REAP (Router-weighted Expert Activation Pruning) method
View 3 collections