Tag: Low-Resource

A research note on how to study the effect of domain data size in LoRA-based machine translation adaptation.

A research-facing explanation of why LoRA is a good fit for low-resource domain machine translation: controlled adaptation, lower experimental cost, and reduced overfitting risk.

如何研究领域数据规模对 LoRA 机器翻译适配的影响：数据多少才够，什么时候收益递减，什么时候是数据质量问题。

从研究角度解释为什么 LoRA 适合低资源领域机器翻译：受控适配、更低实验成本，以及较低过拟合风险。

Domain-specific machine translation is not only a modeling problem. It is a data, terminology, evaluation, and risk problem.