Nanye Ma
YOU?
Author Swipe
View article: Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps
Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps Open
Generative models have made significant impacts across various domains, largely due to their ability to scale during training by increasing data, computational resources, and model size, a phenomenon characterized by the scaling laws. Rece…
View article: SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers
SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers Open
We present Scalable Interpolant Transformers (SiT), a family of generative models built on the backbone of Diffusion Transformers (DiT). The interpolant framework, which allows for connecting two distributions in a more flexible way than s…