Summary of Low-rank Attention Side-tuning For Parameter-efficient Fine-tuning, by Ningyuan Tang et al.
Low-rank Attention Side-Tuning for Parameter-Efficient Fine-Tuning
by Ningyuan Tang, Minghao Fu, Ke Zhu, Jianxin Wu
First submitted to arxiv on: 6 Feb 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 The proposed Low-rank Attention Side-Tuning (LAST) method addresses the limitations of parameter-efficient fine-tuning (PEFT) techniques by disentangling trainable parameters from a large pretrained model. LAST freezes both model parameters and outputs, allowing a side-network to focus on learning task-specific knowledge. This approach is highly parallelizable across multiple optimization objectives, making it efficient for downstream task adaptation. Experiments demonstrate that LAST outperforms previous state-of-the-art methods on visual adaptation tasks while reducing GPU memory consumption by 70% and training time by 40%. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary LAST helps fine-tune large models with few trainable parameters. It freezes both model parts and adds a side-network. This lets the side-network focus on learning task-specific info without changing the big, already-trained model. LAST is good at adapting to new tasks and uses less memory and time than other methods. |
Keywords
» Artificial intelligence » Attention » Fine tuning » Optimization » Parameter efficient
