# Cabrita: closing the gap for foreign languages

arXiv: https://arxiv.org/abs/2308.11878 · HTML: https://mlsanity.com/paper/2308.11878/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- Builds on 19 of its 29 references that are in the map.
- First posted on arXiv 23 Aug 2023.

## Builds on (most cited first, 19 of 19)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions](https://mlsanity.com/paper/1905.10044/) (2019)
- [SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing](https://mlsanity.com/paper/1808.06226/) (2018)
- [mT5: A massively multilingual pre-trained text-to-text transformer](https://mlsanity.com/paper/2010.11934/) (2020)
- [GLU Variants Improve Transformer](https://mlsanity.com/paper/2002.05202/) (2020)
- [Root Mean Square Layer Normalization](https://mlsanity.com/paper/1910.07467/) (2019)
- [Scaling Language Models: Methods, Analysis & Insights from Training Gopher](https://mlsanity.com/paper/2112.11446/) (2021)
- [The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only](https://mlsanity.com/paper/2306.01116/) (2023)
- [The Flan Collection: Designing Data and Methods for Effective Instruction Tuning](https://mlsanity.com/paper/2301.13688/) (2023)
- [How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models](https://mlsanity.com/paper/2012.15613/) (2020)
- [Efficient and Effective Text Encoding for Chinese LLaMA and Alpaca](https://mlsanity.com/paper/2304.08177/) (2023)
- [PTT5: Pretraining and validating the T5 model on Brazilian Portuguese data](https://mlsanity.com/paper/2008.09144/) (2020)
- [Building a Sentiment Corpus of Tweets in Brazilian Portuguese](https://mlsanity.com/paper/1712.08917/) (2017)
- [Sabi'a: Portuguese Large Language Models](https://mlsanity.com/paper/2304.07880/) (2023)
- [BERTa'u: Ita'u BERT for digital customer service](https://mlsanity.com/paper/2101.12015/) (2021)

## Built on by (most cited first, 25 of 25)

- [Multilingual Large Language Model: A Survey of Resources, Taxonomy and Frontiers](https://mlsanity.com/paper/2404.04925/) (2024)
- [How Can We Effectively Expand the Vocabulary of LLMs with 0.01GB of Target Language Text?](https://mlsanity.com/paper/2406.11477/) (2024)
- [Advancing Generative AI for Portuguese with Open Decoder Gerv'asio PT*](https://mlsanity.com/paper/2402.18766/) (2024)
- [An Empirical Study on Cross-lingual Vocabulary Adaptation for Efficient Language Model Inference](https://mlsanity.com/paper/2402.10712/) (2024)
- [TeenyTinyLlama: open-source tiny language models trained in Brazilian Portuguese](https://mlsanity.com/paper/2401.16640/) (2024)
- [Tucano: Advancing Neural Text Generation for Portuguese](https://mlsanity.com/paper/2411.07854/) (2024)
- [Introducing Bode: A Fine-Tuned Large Language Model for Portuguese Prompt-Based Task](https://mlsanity.com/paper/2401.02909/) (2024)
- [FuxiTranyu: A Multilingual Large Language Model Trained with Balanced Data](https://mlsanity.com/paper/2408.06273/) (2024)
- [ptt5-v2: A Closer Look at Continued Pretraining of T5 Models for the Portuguese Language](https://mlsanity.com/paper/2406.10806/) (2024)
- [LlamaTurk: Adapting Open-Source Generative Large Language Models for Low-Resource Language](https://mlsanity.com/paper/2405.07745/) (2024)
- [A Survey of Large Language Models for European Languages](https://mlsanity.com/paper/2408.15040/) (2024)
- [Adapting Chat Language Models Using Only Target Unlabeled Language Data](https://mlsanity.com/paper/2412.11704/) (2024)
- [It's All About In-Context Learning! Teaching Extremely Low-Resource Languages to LLMs](https://mlsanity.com/paper/2508.19089/) (2025)
- [Juru: Legal Brazilian Large Language Model from Reputable Sources](https://mlsanity.com/paper/2403.18140/) (2024)
- [PoETa v2: Toward More Robust Evaluation of Large Language Models in Portuguese](https://mlsanity.com/paper/2511.17808/) (2025)
- [Why We Build Local Large Language Models: An Observational Analysis from 35 Japanese and Multilingual LLMs](https://mlsanity.com/paper/2412.14471/) (2024)
- [Curi'o-Edu 7B: Examining Data Selection Impacts in LLM Continued Pretraining](https://mlsanity.com/paper/2512.12770/) (2025)
- [Fine-tuning with HED-IT: The impact of human post-editing for dialogical language models](https://mlsanity.com/paper/2406.07288/) (2024)
- [SLIM-RAFT: A Novel Fine-Tuning Approach to Improve Cross-Linguistic Performance for Mercosur Common Nomenclature](https://mlsanity.com/paper/2408.03936/) (2024)
- [BRoverbs -- Measuring how much LLMs understand Portuguese proverbs](https://mlsanity.com/paper/2509.08960/) (2025)
- [Semantic Aware Linear Transfer by Recycling Pre-trained Language Models for Cross-lingual Transfer](https://mlsanity.com/paper/2505.10945/) (2025)
- [Vocabulary Customization for Efficient Domain-Specific LLM Deployment](https://mlsanity.com/paper/2509.26124/) (2025)
- [LGSE: Lexically Grounded Subword Embedding Initialization for Low-Resource Language Adaptation](https://mlsanity.com/paper/2603.22629/) (2026)
- [Synthetic Rewriting as a Quality Multiplier: Evidence from Portuguese Continued Pretraining](https://mlsanity.com/paper/2603.24826/) (2026)
- [Language Models for Portuguese: A Systematic Mapping Study](https://mlsanity.com/paper/2608.18138/) (2026)

## Abstract (arXiv)

The strategy of training the model from scratch in a specific language or domain serves two essential purposes: i) enhancing performance in the particular linguistic or domain context, and ii) ensuring effective tokenization. The main limitation inherent to this approach lies in the associated cost, which can reach six to seven-digit dollar values, depending on the model size and the number of parameters involved. The main solution to overcome the cost challenge is to rely on available pre-trained models, which, despite recent advancements such as the LLaMA and LLaMA-2 models, still demonstrate inefficiency for certain specific domain problems or prove ineffective in scenarios involving conversational memory resources, given the large number of tokens required to represent text. To overcome this issue, we present a methodology named Cabrita, which, as our research demonstrates, successfully addresses the performance and efficient tokenization problem, all at an affordable cost. We believe that this methodology can be applied to any transformer-like architecture model. To validate the study, we conducted continuous pre-training exclusively using Portuguese text on a 3-billion-parameter model known as OpenLLaMA, resulting in a model named openCabrita 3B. The openCabrita 3B also features a new tokenizer that results in a significant reduction in the number of tokens required to represent the text. In our assessment, for few-shot learning tasks, we achieved similar results with this 3B model compared to a traditional continuous pre-training approach as well as to 7B models English pre-trained models.