Accelerating diffusion with HuggingFace

What to expect?

The world of diffusion-based models has taken the internet by storm and is constantly evolving, making it difficult to leverage the latest trendy model for your generative workflows. However, with the introduction of HuggingFace diffusers, engineers can approach diffusion-based machine learning modeling in an easier, faster-to-train fashion with a set of modular tools for inference and training. In this talk, Zach Mueller and Suraj Patil of HuggingFace, alongside Anish Shah of Weights & Biases, will teach you how to utilize HuggingFace accelerate and diffusers to accelerate your diffusion modeling, with built-in experiment tracking, model checkpointing, and rich media logging provided by Weights & Biases.