Summary of Diffusion Model Patching Via Mixture-of-prompts, by Seokil Ham et al.
Diffusion Model Patching via Mixture-of-Prompts
by Seokil Ham, Sangmin Woo, Jin-Young Kim, Hyojun Go, Byeongjun Park, Changick Kim
First submitted to arxiv on: 28 May 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents Diffusion Model Patching (DMP), a novel method to improve pre-trained diffusion models that have already reached convergence. DMP inserts a small set of learnable prompts into the model’s input space, keeping the original model frozen. The dynamic gating mechanism selects and combines a subset of prompts at each timestep, drawing on their distinct expertise to “patch” the model’s functionality. This approach enhances the FID score of converged DiT-L/2 by 10.38% on FFHQ with only a 1.43% parameter increase and additional training iterations. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper shows how to make already-trained AI models better. It’s like adding a new skill or expertise to an existing model, but without using too many extra resources. The new method is called Diffusion Model Patching (DMP). DMP adds a few special prompts to the model that help it improve its performance on specific tasks. This means the model can get even better results with minimal extra effort. |
Keywords
* Artificial intelligence * Diffusion model
