28 citations · 33 across the 8 of their papers we have counts for
9 papers
TildeOpen LLM: Leveraging Curriculum Learning to Achieve Equitable Language Representation
Toms Bergmanis, Martins Kronis, Ingus Jānis Pretkalniņš +5
Large language models often underperform in many European languages due to the dominance of English and a few high-resource languages in training data. This paper presents TildeOpe…
Localizing AI: Evaluating Open-Weight Language Models for Languages of Baltic States
Jurgita Kapočiūtė-Dzikienė, Toms Bergmanis, Mārcis Pinnis
Although large language models (LLMs) have transformed our expectations of modern language technologies, concerns over data privacy often restrict the use of commercially available…
From Zero to Production: Baltic-Ukrainian Machine Translation Systems to Aid Refugees
Toms Bergmanis, Mārcis Pinnis
In this paper, we examine the development and usage of six low-resource machine translation systems translating between the Ukrainian language and each of the official languages of…
Open Terminology Management and Sharing Toolkit for Federation of Terminology Databases
Andis Lagzdiņš, Uldis Siliņš, Mārcis Pinnis +3
Consolidated access to current and reliable terms from different subject fields and languages is necessary for content creators and translators. Terminology is also needed in AI ap…
Dynamic Terminology Integration for COVID-19 and other Emerging Domains
Toms Bergmanis, Mārcis Pinnis
The majority of language domains require prudent use of terminology to ensure clarity and adequacy of information conveyed. While the correct use of terminology for some languages…
Facilitating Terminology Translation with Target Lemma Annotations
Toms Bergmanis, Mārcis Pinnis
Most of the recent work on terminology integration in machine translation has assumed that terminology translations are given already inflected in forms that are suitable for the t…
Mitigating Gender Bias in Machine Translation with Target Gender Annotations
Artūrs Stafanovičs, Toms Bergmanis, Mārcis Pinnis
When translating "The secretary asked for details." to a language with grammatical gender, it might be necessary to determine the gender of the subject "secretary". If the sentence…
Robust Neural Machine Translation: Modeling Orthographic and Interpunctual Variation
Toms Bergmanis, Artūrs Stafanovičs, Mārcis Pinnis
Neural machine translation systems typically are trained on curated corpora and break when faced with non-standard orthography or punctuation. Resilience to spelling mistakes and t…
Training Data Augmentation for Context-Sensitive Neural Lemmatization Using Inflection Tables and Raw Text
Toms Bergmanis, Sharon Goldwater
Lemmatization aims to reduce the sparse data problem by relating the inflected forms of a word to its dictionary form. Using context can help, both for unseen and ambiguous words.…