MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910468821483520 |
|---|---|
| author | Hu, Shengding Tu, Yuge Han, Xu He, Chaoqun Cui, Ganqu Long, Xiang Zheng, Zhi Fang, Yewei Huang, Yuxiang Zhao, Weilin Zhang, Xinrong Thai, Zheng Leng Zhang, Kaihuo Wang, Chongyi Yao, Yuan Zhao, Chenyang Zhou, Jie Cai, Jie Zhai, Zhongwu Ding, Ning Jia, Chao Zeng, Guoyang Li, Dahai Liu, Zhiyuan Sun, Maosong |
| author_facet | Hu, Shengding Tu, Yuge Han, Xu He, Chaoqun Cui, Ganqu Long, Xiang Zheng, Zhi Fang, Yewei Huang, Yuxiang Zhao, Weilin Zhang, Xinrong Thai, Zheng Leng Zhang, Kaihuo Wang, Chongyi Yao, Yuan Zhao, Chenyang Zhou, Jie Cai, Jie Zhai, Zhongwu Ding, Ning Jia, Chao Zeng, Guoyang Li, Dahai Liu, Zhiyuan Sun, Maosong |
| contents | The burgeoning interest in developing Large Language Models (LLMs) with up to trillion parameters has been met with concerns regarding resource efficiency and practical expense, particularly given the immense cost of experimentation. This scenario underscores the importance of exploring the potential of Small Language Models (SLMs) as a resource-efficient alternative. In this context, we introduce MiniCPM, specifically the 1.2B and 2.4B non-embedding parameter variants, not only excel in their respective categories but also demonstrate capabilities on par with 7B-13B LLMs. While focusing on SLMs, our approach exhibits scalability in both model and data dimensions for future LLM research. Regarding model scaling, we employ extensive model wind tunnel experiments for stable and optimal scaling. For data scaling, we introduce a Warmup-Stable-Decay (WSD) learning rate scheduler (LRS), conducive to continuous training and domain adaptation. We present an in-depth analysis of the intriguing training dynamics that occurred in the WSD LRS. With WSD LRS, we are now able to efficiently study data-model scaling law without extensive retraining experiments on both axes of model and data, from which we derive the much higher compute optimal data-model ratio than Chinchilla Optimal. Additionally, we introduce MiniCPM family, including MiniCPM-DPO, MiniCPM-MoE and MiniCPM-128K, whose excellent performance further cementing MiniCPM's foundation in diverse SLM applications. MiniCPM models are available publicly at https://github.com/OpenBMB/MiniCPM . |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_06395 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies Hu, Shengding Tu, Yuge Han, Xu He, Chaoqun Cui, Ganqu Long, Xiang Zheng, Zhi Fang, Yewei Huang, Yuxiang Zhao, Weilin Zhang, Xinrong Thai, Zheng Leng Zhang, Kaihuo Wang, Chongyi Yao, Yuan Zhao, Chenyang Zhou, Jie Cai, Jie Zhai, Zhongwu Ding, Ning Jia, Chao Zeng, Guoyang Li, Dahai Liu, Zhiyuan Sun, Maosong Computation and Language Machine Learning The burgeoning interest in developing Large Language Models (LLMs) with up to trillion parameters has been met with concerns regarding resource efficiency and practical expense, particularly given the immense cost of experimentation. This scenario underscores the importance of exploring the potential of Small Language Models (SLMs) as a resource-efficient alternative. In this context, we introduce MiniCPM, specifically the 1.2B and 2.4B non-embedding parameter variants, not only excel in their respective categories but also demonstrate capabilities on par with 7B-13B LLMs. While focusing on SLMs, our approach exhibits scalability in both model and data dimensions for future LLM research. Regarding model scaling, we employ extensive model wind tunnel experiments for stable and optimal scaling. For data scaling, we introduce a Warmup-Stable-Decay (WSD) learning rate scheduler (LRS), conducive to continuous training and domain adaptation. We present an in-depth analysis of the intriguing training dynamics that occurred in the WSD LRS. With WSD LRS, we are now able to efficiently study data-model scaling law without extensive retraining experiments on both axes of model and data, from which we derive the much higher compute optimal data-model ratio than Chinchilla Optimal. Additionally, we introduce MiniCPM family, including MiniCPM-DPO, MiniCPM-MoE and MiniCPM-128K, whose excellent performance further cementing MiniCPM's foundation in diverse SLM applications. MiniCPM models are available publicly at https://github.com/OpenBMB/MiniCPM . |
| title | MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2404.06395 |