Summary of Nora: Nested Low-rank Adaptation For Efficient Fine-tuning Large Models, by Cheng Lin et al.
NoRA: Nested Low-Rank Adaptation for Efficient Fine-Tuning Large Models
by Cheng Lin, Lujun Li, Dezhi Li, Jie Zou, Wei Xue, Yike Guo
First submitted to arxiv on: 18 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 introduces Nested Low-Rank Adaptation (NoRA), a novel approach to fine-tuning pre-trained models that leverages the capabilities of Low-Rank Adaptation (LoRA) techniques while reducing tunable parameters. NoRA adopts a dual-layer nested structure with Singular Value Decomposition (SVD) to effectively utilize original matrix knowledge and control model optimization. This approach allows for precise task adaptation with a compact parameter space, outperforming LoRA and its variants in tasks such as commonsense reasoning, fine-tuning vision-language models, and subject-driven generation. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary NoRA is a new way to make pre-trained models better at specific jobs while using fewer computer resources. It does this by combining two types of matrix calculations: Singular Value Decomposition (SVD) and Low-Rank Adaptation (LoRA). This helps the model learn more quickly and accurately without needing too many parameters. |
Keywords
» Artificial intelligence » Fine tuning » Lora » Low rank adaptation » Optimization
