Summary of Preparing Lessons For Progressive Training on Language Models, by Yu Pan et al.
Preparing Lessons for Progressive Training on Language Models
by Yu Pan, Ye Yuan, Yichun Yin, Jiaxin Shi, Zenglin Xu, Ming Zhang, Lifeng Shang, Xin Jiang, Qun Liu
First submitted to arxiv on: 17 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- 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 Apollo method addresses the challenges of training deep Transformers in artificial intelligence by prepaving lessons for expanding operations during training. It involves low-value-prioritized sampling to train different depths and weight sharing to facilitate efficient expansion. This approach achieves state-of-the-art acceleration ratios, rivaling methods using pretrained models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Apollo is a novel method that trains deep Transformers efficiently while reducing time, financial, and environmental costs. By prepaving lessons for expanding operations during training, Apollo learns high-layer functionality during low-layer training. It uses low-value-prioritized sampling to train different depths and weight sharing to facilitate efficient expansion. |
