English
Related papers

Related papers: SozKZ: Training Efficient Small Language Models fo…

200 papers

The rapid advancement of large language models (LLMs) has significantly propelled progress in natural language processing (NLP). However, their effectiveness in specialized, low-resource domains-such as Arabic legal contexts-remains…

Computation and Language · Computer Science 2025-08-25 Adil Bahaj , Mounir Ghogho

Speech tokenizers are foundational to speech language models, yet existing approaches face two major challenges: (1) balancing trade-offs between encoding semantics for understanding and acoustics for reconstruction, and (2) achieving low…

Language models have been effectively applied to modeling natural signals, such as images, video, speech, and audio. A crucial component of these models is the codec tokenizer, which compresses high-dimensional natural signals into…

Audio and Speech Processing · Electrical Eng. & Systems 2025-02-26 Shengpeng Ji , Ziyue Jiang , Wen Wang , Yifu Chen , Minghui Fang , Jialong Zuo , Qian Yang , Xize Cheng , Zehan Wang , Ruiqi Li , Ziang Zhang , Xiaoda Yang , Rongjie Huang , Yidi Jiang , Qian Chen , Siqi Zheng , Zhou Zhao

Pretraining massively multilingual Large Language Models (LLMs) for many languages at once is challenging due to limited model capacity, scarce high-quality data, and compute constraints. Moreover, the lack of language coverage of the…

A bottleneck to developing Semantic Parsing (SP) models is the need for a large volume of human-labeled training data. Given the complexity and cost of human annotation for SP, labeled data is often scarce, particularly in multilingual…

Computation and Language · Computer Science 2022-10-17 Andy Rosenbaum , Saleh Soltan , Wael Hamza , Amir Saffari , Marco Damonte , Isabel Groves

Large Language Models (LLMs) are widely deployed in real-world applications, yet little is known about their training dynamics at the token level. Evaluation typically relies on aggregated training loss, measured at the batch level, which…

Computation and Language · Computer Science 2024-10-17 Andrea Pinto , Tomer Galanti , Randall Balestriero

In this study, we introduce CT-LLM, a 2B large language model (LLM) that illustrates a pivotal shift towards prioritizing the Chinese language in developing LLMs. Uniquely initiated from scratch, CT-LLM diverges from the conventional…

We introduce Bielik v3, a series of parameter-efficient generative text models (1.5B and 4.5B) optimized for Polish language processing. These models demonstrate that smaller, well-optimized architectures can achieve performance comparable…

Machine Learning · Computer Science 2025-05-12 Krzysztof Ociepa , Łukasz Flis , Remigiusz Kinas , Krzysztof Wróbel , Adrian Gwoździej

Whisper and other large-scale automatic speech recognition models have made significant progress in performance. However, their performance on many low-resource languages, such as Kazakh, is not satisfactory. It is worth researching how to…

Audio and Speech Processing · Electrical Eng. & Systems 2025-05-08 Jinpeng Li , Yu Pu , Qi Sun , Wei-Qiang Zhang

Large language models (LLMs) for code rely on subword tokenizers, such as byte-pair encoding (BPE), learned from mixed natural language text and programming language code but driven by statistics rather than grammar. As a result,…

Computation and Language · Computer Science 2025-10-17 Yinxi Li , Yuntian Deng , Pengyu Nie

In this work, we explain our approach employed in the BabyLM Challenge, which uses various methods of training language models (LMs) with significantly less data compared to traditional large language models (LLMs) and are inspired by how…

Computation and Language · Computer Science 2025-03-07 Mohammad Amin Ghanizadeh , Mohammad Javad Dousti

We present Kakugo, a novel and cost-effective pipeline designed to train general-purpose Small Language Models (SLMs) for low-resource languages using only the language name as input. By using a large teacher model to generate synthetic…

Computation and Language · Computer Science 2026-01-21 Peter Devine , Mardhiyah Sanni , Farid Adilazuarda , Julieta Gil Loizaga , Barry Haddow

Training of a tokenizer plays an important role in the performance of deep learning models. This research aims to understand the performance of tokenizers in five state-of-the-art (SOTA) large language models (LLMs) in the Assamese language…

Computation and Language · Computer Science 2025-04-08 Sagar Tamang , Dibya Jyoti Bora

In this work, we introduce LokiLM, a 1.4B parameter large language model trained on 500B tokens. Our model performs strongly in natural language reasoning tasks and achieves state-of-the-art performance among models with 1.5B parameters or…

Computation and Language · Computer Science 2024-07-11 Justin Kiefel , Shrey Shah

As foundation AI models continue to increase in size, an important question arises - is massive scale the only path forward? This survey of about 160 papers presents a family of Small Language Models (SLMs) in the 1 to 8 billion parameter…

We introduce Xmodel-1.5, a 1-billion-parameter multilingual large language model pretrained on 2 trillion tokens, designed for balanced performance and scalability. Unlike most large models that use the BPE tokenizer, Xmodel-1.5 employs a…

Computation and Language · Computer Science 2024-12-05 Wang Qun , Liu Yang , Lin Qingquan , Jiang Ling

Large Language Models (LLMs) are becoming crucial across various fields, emphasizing the urgency for high-quality models in underrepresented languages. This study explores the unique challenges faced by low-resource languages, such as data…

Computation and Language · Computer Science 2024-05-09 Emre Can Acikgoz , Mete Erdogan , Deniz Yuret

Large language models (LLMs) for Arabic are still dominated by Modern Standard Arabic (MSA), with limited support for Saudi dialects such as Najdi and Hijazi. This underrepresentation hinders their ability to capture authentic dialectal…

Computation and Language · Computer Science 2025-08-20 Hassan Barmandah

We introduce PLaMo-100B, a large-scale language model designed for Japanese proficiency. The model was trained from scratch using 2 trillion tokens, with architecture such as QK Normalization and Z-Loss to ensure training stability during…

This work introduces TajPersLexon, a curated Tajik--Persian parallel lexical resource of 40,112 word and short-phrase pairs for cross-script lexical retrieval, transliteration, and alignment in low-resource settings. We conduct a…

Computation and Language · Computer Science 2026-05-11 Mullosharaf K. Arabov