DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention

Open in new window