Summary of Llm-neo: Parameter Efficient Knowledge Distillation For Large Language Models, by Runming Yang et al.
LLM-NEO: Parameter Efficient Knowledge Distillation for Large Language Models
by Runming Yang, Taiqiang Wu, Jiahao Wang, Pengfei Hu, Yik-Chung Wu, Ngai Wong, Yujiu Yang
First submitted to arxiv on: 11 Nov 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 re-examines the relationship between knowledge distillation (KD) and Low-Rank Adaption (LoRA), demonstrating that they share a common paradigm. Building on this insight, the authors propose LLM-NEO, a parameter-efficient KD method integrating LoRA to enhance knowledge transfer efficiency. The approach is evaluated on compressing Llama 2 and Llama 3.2, outperforming various baselines. Experimental results highlight the robustness of LLM-NEO on LoRA variants. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how we can make Large Language Models (LLMs) smaller and more efficient without losing their abilities. The researchers realized that two popular methods, knowledge distillation and low-rank adaptation, are actually doing similar things. They combined these ideas to create a new method called LLM-NEO, which helps us learn from one language model and apply it to another. Tests show that this new method works better than others at compressing certain language models. |
Keywords
» Artificial intelligence » Knowledge distillation » Language model » Llama » Lora » Low rank adaptation » Parameter efficient
