3B参数小模型突破性能极限，媲美更大模型。
Nanbeige4-3B Technical Report: Exploring the Frontier of Small Language Models
- 分阶段训练调度+指令微调，提升小模型表现
- 推理增强与双偏好蒸馏，复杂任务能力显著提升
- 适合追求高效推理与高性价比的开发者使用
我们提出Nanbeige4-3B系列小型高性能语言模型。在23万亿高质量语料上预训练，并在超过3000万条多样化指令上进行微调，突破了小模型的缩放定律边界。预训练阶段设计细粒度暖启动-稳定-衰减（FG-WSD）调度器，逐步优化数据混合比例以提升性能；后训练阶段采用思辨生成优化与思维链重构联合机制，显著提升复杂任务表现。在SFT后，利用旗舰推理模型通过双偏好蒸馏（DPD）方法对Nanbeige4-3B进行知识蒸馏，进一步提升性能。最后，采用多阶段强化学习，结合可验证奖励与偏好建模，增强推理与人类对齐能力。大量评估显示，Nanbeige4-3B不仅显著优于同规模模型，更在多项基准上媲美更大模型。模型权重已开源：https://huggingface.co/Nanbeige。
原文摘要 · Abstract (English)
We present Nanbeige4-3B, a family of small-scale but high-performing language models. Pretrained on 23T high-quality tokens and finetuned on over 30 million diverse instructions, we extend the boundary of the scaling law for small language models. In pre-training, we design a Fine-Grained Warmup-Stable-Decay (FG-WSD) training scheduler, which progressively refines data mixtures across stages to boost model performance. In post-training, to improve the quality of the SFT data, we design a joint mechanism that integrates deliberative generation refinement and chain-of-thought reconstruction, yielding substantial gains on complex tasks. Following SFT, we employ our flagship reasoning model to distill Nanbeige4-3B through our proposed Dual Preference Distillation (DPD) method, which leads to further performance gains. Finally, a multi-stage reinforcement learning phase was applied, leveraging verifiable rewards and preference modeling to strengthen abilities on both reasoning and human alignment. Extensive evaluations show that Nanbeige4-3B not only significantly outperforms models of comparable parameter scale but also rivals much larger models across a wide range of benchmarks. The model checkpoints are available at https://huggingface.co/Nanbeige.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。