Summary of Neuzip: Memory-efficient Training and Inference with Dynamic Compression Of Neural Networks, by Yongchang Hao et al.
NeuZip: Memory-Efficient Training and Inference with Dynamic Compression of Neural Networks
by Yongchang Hao, Yanshuai Cao, Lili Mou
First submitted to arxiv on: 28 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
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 NeuZip, a novel weight compression scheme that reduces neural network model sizes without compromising performance. By leveraging the entropy of floating-point numbers, NeuZip enables efficient training and inference on devices with limited memory capacity. The authors demonstrate significant memory savings for large models like Llama-3 8B, reducing the required memory from 31GB to under 16GB during training, and by more than half in inference, while maintaining near-lossless performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary NeuZip is a new way to make neural networks smaller without losing their ability to learn. Normally, using too many parameters makes networks better, but this can be a problem when you don’t have enough memory on your device. NeuZip fixes this by making the weights in the network more compact, allowing it to use less memory while still doing its job well. |
Keywords
» Artificial intelligence » Inference » Llama » Neural network
