Additive quantization for truly tiny compressed diffusion models

• Main Author: Hasan, Adil
• Other Authors: Thomas Peyrin
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Machine learning
• Online Access: https://hdl.handle.net/10356/181210
• Institution: Nanyang Technological University

Tremendous investments have been made towards the commodification of diffusion models for generation of diverse media. Their mass-market adoption is however still hobbled by the intense hardware resource requirements of diffusion model inference. Model quantization strategies tailored specificall...