The Hub as Infrastructure. From Open PyTorch Models, to a Safe and Perfor...
summary
Hugging Face runs a distribution hub with 2.74M public models and 1.44M PyTorch-compatible checkpoints, where the top 200 models drive 50% of downloads and the top 1.4M models still get ≥200 downloads each. The hub records ~4.47M uploads with ~8–9% containing malicious files, so SafeTensors moves to the PyTorch Foundation after audits and will keep a stable format. Roadmap focuses on device-aware loading, new datatypes and kernel pages on the hub to speed checkpoint loading and hardware-specific execution.
AI-written summary · about
chapters (5)
- 0:00Hub as Infrastructure
- 1:27Hugging Face Hub evolution
- 3:37Model usage and security
- 5:22Safe Tensors in PyTorch
- 7:51Future of execution and kernels
related talks
14mDiving in Hugging Face Hub; Share Your Model Weights on the #1 AI Hub, Home of 700k+...
11mBuilding Frontier Intelligence in the Open
21mWhat's New for PyTorch Developer Infrastructure
47mGet to Know PyTorch - A Cutting Edge Open Source AI Framework for Deep Learning
18mThe State of the PyTorch Ecosystem in 2026: Global Trends and China’s Rising Role
43mPyTorch Foundation: 1 Year of Open Governance
01h 18mPutting Tensorflow Models into Production
19mGet to know PyTorch - a cutting edge open source AI framework for Machine Learning
