Person
Hugging Face team
4 stories · sorted newest first · 📡 RSS
Hugging Face Unveils Cosmopedia: Largest Open Synthetic Dataset for Language Model Pre-Training
The Hugging Face team introduces Cosmopedia, a massive synthetic dataset designed to replicate the success of Microsoft's Phi mode
Hugging Face Introduces RLOO Trainer for More Efficient RLHF Training
The new RLOO Trainer reduces GPU memory usage by 50-70% compared to PPO, enabling researchers to train larger models with less com
Hugging Face Releases Accelerate 1.0.0: Major Update for Large-Scale PyTorch Training
The new version brings support for multiple hardware accelerators, improved performance, and enhanced usability to the popular ope
Hugging Face Scales AI Infrastructure Secrets Management with Infisical
Learn how Hugging Face, a leading AI company, improved their secrets management using Infisical in a multi-cloud environment, redu