How Good Are Low-bit Quantized LLaMA3 Models? An Empirical Study

Wei Huang, Xudong Ma,Haotong Qin, Xingyu Zheng,Chengtao Lv, Hong Chen,Jie Luo, Xiaojuan Qi,Xianglong Liu, Michele Magno

arxiv(2024)

引用 0|浏览50
暂无评分
摘要
Meta's LLaMA family has become one of the most powerful open-source Large Language Model (LLM) series. Notably, LLaMA3 models have recently been released and achieve impressive performance across various with super-large scale pre-training on over 15T tokens of data. Given the wide application of low-bit quantization for LLMs in resource-limited scenarios, we explore LLaMA3's capabilities when quantized to low bit-width. This exploration holds the potential to unveil new insights and challenges for low-bit quantization of LLaMA3 and other forthcoming LLMs, especially in addressing performance degradation problems that suffer in LLM compression. Specifically, we evaluate the 10 existing post-training quantization and LoRA-finetuning methods of LLaMA3 on 1-8 bits and diverse datasets to comprehensively reveal LLaMA3's low-bit quantization performance. Our experiment results indicate that LLaMA3 still suffers non-negligent degradation in these scenarios, especially in ultra-low bit-width. This highlights the significant performance gap under low bit-width that needs to be bridged in future developments. We expect that this empirical study will prove valuable in advancing future models, pushing the LLMs to lower bit-width with higher accuracy for being practical. Our project is released on https://github.com/Macaronlin/LLaMA3-Quantization and quantized LLaMA3 models are released in https://huggingface.co/LLMQ.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要