8 repos
Generative image synthesis and manipulation using diffusion-based deep learning models, primarily implemented in Python with PyTorch. This cluster covers techniques for training, fine-tuning, and controlling diffusion models—including custom training approaches, instruction-based editing (instruct-nerf2nerf), efficient inference optimization (pruna), and multi-model composition (mixture-of-diffusers). Projects here span from foundational model adaptation to practical applications in 3D generation and image personalization.