用轻量方法让10亿参数中文模型更稳定，不乱蹦外文字符。
Efficient Training of Robust Traditional Chinese LLaMA-1B on a Single Consumer GPU: Continual Pre-training, SFT, and DPO
- 三阶段优化：持续预训练+指令微调+偏好优化，全靠小适配器。
- 非中文字符输出减少51.3%，跨语言错误降低77.2%以上。
- 单张消费级显卡可运行，适合想做中文AI的开发者。
小型语言模型在成本敏感、低延迟场景中具有优势，但在繁体中文（TC）应用中受限于词元级不稳定问题——模型会随机输出非繁体中文字符或发生语言切换。本文通过构建纯繁体中文1B模型（PureTC-1B），采用三阶段稳定化流程对Meta发布的Llama-3.2-1B-Instruct模型进行优化。方法结合基于繁体语料的持续预训练（CPT）、指令数据上的监督微调（SFT）以及利用繁体中文遵循度偏好进行的直接偏好优化（DPO），实现无需全模型重训的单语种鲁棒性提升。在模拟真实使用场景的基准测试中，相比基线模型，PureTC-1B在非繁体输出词元上实现了51.3%的相对减少（微平均）。在命名实体翻译（NET）任务中，相较Llama-3B和Qwen-1.5B，其错误语言词元分别降低77.2%和57.2%。该流程仅依赖适配器、可复现且对硬件友好，为繁体中文及其他非英语语言提供实用的稳定性增强方案。
原文摘要 · Abstract (English)
Small Language Models (SLMs) enable cost-effective, on-device and latency-sensitive AI applications, yet their deployment in Traditional Chinese (TC) remains hindered by token-level instability - models unpredictably emit non-TC characters or code-switch into other languages. We address this practical reliability gap by creating PureTC-1B, a three-stage stabilization pipeline for Llama-3.2-1B-Instruct (an open-weight, instruction-tuned model released by Meta) using parameter-efficient LoRA adapters. Our method combines Continual Pre-Training (CPT) on TC-centric corpora, Supervised Fine-Tuning (SFT) with instruction data, and Direct Preference Optimization (DPO) using TC-adherence preferences to improve monolingual robustness without full-model retraining. On a benchmark designed to simulate real-world usage, PureTC-1B achieves a 51.3% relative reduction (micro-average) in non-TC output tokens versus the base model. On a Named Entity Translation (NET) task, PureTC-1B further reduces incorrect-language tokens by 77.2% relative to Llama-3B and 57.2% relative to Qwen-1.5B, indicating that robust TC adherence is attainable even at the 1B scale. The pipeline is reproducible, adapter-only, and hardware-friendly, offering practitioners a practical recipe to enhance language stability for TC and potentially other non-English languages.
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。