cs.CLDec 3, 2025

Adapting Large Language Models to Low-Resource Tibetan: A Two-Stage Continual and Supervised Fine-Tuning Study

Authors: Lifeng Chen, Chenyang Zhong, Ryan Lai, Tianming Liu

Abstract

Adapting large language models (LLMs) to low-resource languages remains a major challenge due to data scarcity and cross-lingual drift. This work presents a two-stage adaptation of Qwen2.5-3B to Tibetan, a morphologically rich and underrepresented language. We employ Continual Pretraining (CPT) to establish Tibetan linguistic grounding, followed by Supervised Fine-Tuning (SFT) for task and translation specialization. Empirical evaluations demonstrate a consistent decrease in perplexity (from 2.98 →\rightarrow 1.54) and substantial improvements in Chinese→\rightarrowTibetan translation quality (BLEU: 0.046 →\rightarrow 0.261; chrF: 2.2 →\rightarrow 6.6). Layer-wise analysis across 435 layers in Qwen3-4B reveals that adaptation primarily concentrates on embedding and output heads, with mid--late MLP projections encoding domain-specific transformations. Our findings suggest that CPT constructs a Tibetan semantic manifold while SFT sharpens task alignment with minimal representational disruption. This study provides the first quantitative exploration of Tibetan adaptation dynamics for LLMs, and offers an open, reproducible framework for extending multilingual foundation models to low-resource settings.

Figures & tables

Explore similar work

CardsList
  1. Modular Monolingual Adaptation using Pretrained Language Models

    Jun 4, 2026Nalin Kumar, Ondřej DušekLow-Resource Language ProcessingTokenizer Adaptation

  2. Tibetan-TTS:Low-Resource Tibetan Speech Synthesis with Large Model Adaptation

    May 4, 2026Jiaxu He, Chao Wang, Jie Lian +4TTS SynthesisLow-Resource TTS Synthesis

  3. Can Linguistically Related Languages Guide LLM Translation in Low-Resource Settings?

    Mar 17, 2026Aishwarya Ramasethu, Niyathi Allu, Rohin Garg +2Machine TranslationLow-Resource MT