# Efficient Continual Pre-training of LLMs for Low-resource Languages

arXiv: https://arxiv.org/abs/2412.10244 · HTML: https://mlsanity.com/paper/2412.10244/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 35 references that are in the map.
- First posted on arXiv 13 Dec 2024.

## Builds on (most cited first, 24 of 24)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Neural Machine Translation of Rare Words with Subword Units](https://mlsanity.com/paper/1508.07909/) (2015)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity](https://mlsanity.com/paper/2302.04023/) (2023)
- [Subword Regularization: Improving Neural Network Translation Models with Multiple Subword Candidates](https://mlsanity.com/paper/1804.10959/) (2018)
- [How Good Are GPT Models at Machine Translation? A Comprehensive Evaluation](https://mlsanity.com/paper/2302.09210/) (2023)
- [How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models](https://mlsanity.com/paper/2012.15613/) (2020)
- [Language Model Tokenizers Introduce Unfairness Between Languages](https://mlsanity.com/paper/2305.15425/) (2023)
- [Do All Languages Cost the Same? Tokenization in the Era of Commercial Language Models](https://mlsanity.com/paper/2305.13707/) (2023)
- [Is ChatGPT A Good Translator? Yes With GPT-4 As The Engine](https://mlsanity.com/paper/2301.08745/) (2023)
- [When Being Unseen from mBERT is just the Beginning: Handling New Languages With Multilingual Language Models](https://mlsanity.com/paper/2010.12858/) (2020)
- [Continual Pre-Training for Cross-Lingual LLM Adaptation: Enhancing Japanese Language Capabilities](https://mlsanity.com/paper/2404.17790/) (2024)
- [WECHSEL: Effective initialization of subword embeddings for cross-lingual transfer of monolingual language models](https://mlsanity.com/paper/2112.06598/) (2021)
- [Impact of Tokenization on Language Models: An Analysis for Turkish](https://mlsanity.com/paper/2204.08832/) (2022)
- [IndicGenBench: A Multilingual Benchmark to Evaluate Generation Capabilities of LLMs on Indic Languages](https://mlsanity.com/paper/2404.16816/) (2024)
- [Fast Vocabulary Transfer for Language Model Compression](https://mlsanity.com/paper/2402.09977/) (2024)
- [Tamil-Llama: A New Tamil Language Model Based on Llama 2](https://mlsanity.com/paper/2311.05845/) (2023)
- [Efficient Language Model Training through Cross-Lingual and Progressive Transfer Learning](https://mlsanity.com/paper/2301.09626/) (2023)
- [How Can We Effectively Expand the Vocabulary of LLMs with 0.01GB of Target Language Text?](https://mlsanity.com/paper/2406.11477/) (2024)
- [Exploring Design Choices for Building Language-Specific LLMs](https://mlsanity.com/paper/2406.14670/) (2024)
- [How do different tokenizers perform on downstream tasks in scriptio continua languages?: A case study in Japanese](https://mlsanity.com/paper/2306.09572/) (2023)
- [An Empirical Study on Cross-lingual Vocabulary Adaptation for Efficient Language Model Inference](https://mlsanity.com/paper/2402.10712/) (2024)
- [Cost-Performance Optimization for Processing Low-Resource Language Tasks Using Commercial LLMs](https://mlsanity.com/paper/2403.05434/) (2024)

## Built on by (most cited first, 8 of 8)

- [One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers](https://mlsanity.com/paper/2506.10766/) (2025)
- [Mitigating Catastrophic Forgetting in Target Language Adaptation of LLMs via Source-Shielded Updates](https://mlsanity.com/paper/2512.04844/) (2025)
- [Bridging the Knowledge Void: Inference-time Acquisition of Unfamiliar Programming Languages for Coding Tasks](https://mlsanity.com/paper/2602.06976/) (2026)
- [Small LLMs: Pruning vs. Training from Scratch](https://mlsanity.com/paper/2606.14150/) (2026)
- [AI-Driven Generation of Old English: A Framework for Low-Resource Languages](https://mlsanity.com/paper/2507.20111/) (2025)
- [Raising Bars, Not Parameters: LilMoo Compact Language Model for Hindi](https://mlsanity.com/paper/2603.03508/) (2026)
- [Reading the News: Adapting Large Language Models to Swedish Journalism Through Continued Pre-Training](https://mlsanity.com/paper/2608.30609/) (2026)
- [DeepEdu-v1: Efficient and Scalable Agentic LLMs for Vietnamese Education](https://mlsanity.com/paper/2609.31568/) (2026)

## Abstract (arXiv)

Open-source Large Language models (OsLLMs) propel the democratization of natural language research by giving the flexibility to augment or update model parameters for performance improvement. Nevertheless, like proprietary LLMs, Os-LLMs offer poorer performance on low-resource languages (LRLs) than high-resource languages (HRLs), owing to smaller amounts of training data and underrepresented vocabulary. On the other hand, continual pre-training (CPT) with large amounts of language-specific data is a costly proposition in terms of data acquisition and computational resources. Our goal is to drastically reduce CPT cost. To that end, we first develop a new algorithm to select a subset of texts from a larger corpus. We show the effectiveness of our technique using very little CPT data. In search of further improvement, we design a new algorithm to select tokens to include in the LLM vocabulary. We experiment with the recent Llama-3 model and nine Indian languages with diverse scripts and extent of resource availability. For evaluation, we use IndicGenBench, a generation task benchmark dataset for Indic languages. We experiment with various CPT corpora and augmented vocabulary size and offer insights across language families.