Improving Romanian LLM Pretraining Data using Diversity and Quality Filtering
Authors:
Preprint
Abstract
Large Language Models (LLMs) have recently exploded in popularity, often matching or outperforming human abilities on many tasks. One of the key factors in training LLMs is the availability and curation of high-quality data. Data quality is especially crucial for under-represented languages, where high-quality corpora are scarce. In this work we study the characteristics and coverage of Romanian pretraining corpora and we examine how they differ from English data. By training a lightweight multitask model on carefully LLM-annotated Romanian texts, we are able to analyze and perform multi-level filtering (e.g., educational value, topic, format) to generate high-quality pretraining datasets. Our experiments show noteworthy trends in the topics present in Romanian and English data, while also proving the effectiveness of filtering data through improved LLM pretraining performance across multiple benchmarks.Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Judging Quality Across Languages: A Multilingual Approach to Pretraining Data Filtering with Language Models2025
- Meta-rater: A Multi-dimensional Data Selection Method for Pre-training Language Models2025
- “Vorbești Românește?” A Recipe to Train Powerful Romanian LLMs with English Instructions2024
- OLMo: Accelerating the Science of Language Models2024
- A Pretrainer’s Guide to Training Data: Measuring the Effects of Data Age, Domain Coverage, Quality, & Toxicity2024
- The birth of Romanian BERT2020
- BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding2019
- HellaSwag: Can a Machine Really Finish Your Sentence?2019
- Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering2018