Anne Valerie Ter sat on the short black bench in her piano teacher’s basement, her mother’s warm ... family’s home in Northern Virginia to train at Baltimore’s prestigious Peabody ...

(1) We released the 50 diffusion steps model (instead of 1000 steps) which runs 20X faster with comparable results. (2) Calling CLIP just once and caching the result runs 2X faster for all models.

Update EVA ViT (closest match) to support Perception Encoder models (https://arxiv.org/abs/2504.13181) from Meta, loading Hub weights but I still need to push ...