Prokopis Prokopidis

Leon Voukoutis, Dimitris Roussis, Georgios Paraskevopoulos et al.

We describe the development and capabilities of Meltemi 7B, the first open Large Language Model for the Greek language. Meltemi 7B has 7 billion parameters and is trained on a 40 billion token Greek corpus. For the development of Meltemi 7B, we adapt Mistral, by continuous pretraining on the Greek Corpus. Meltemi 7B contains up-to-date information up to September 2023. Furthermore, we have translated and curated a Greek instruction corpus, which has been used for the instruction-tuning of a chat model, named Meltemi 7B Instruct. Special care has been given to the alignment and the removal of toxic content for the Meltemi 7B Instruct. The developed models are evaluated on a broad set of collected evaluation corpora, and examples of prompts and responses are presented. Both Meltemi 7B and Meltemi 7B Instruct are available at https://huggingface.co/ilsp under the Apache 2.0 license.

Francisco Casacuberta, Alexandru Ceausu, Khalid Choukri et al.

This work presents the results of the machine translation (MT) task from the Covid-19 MLIA @ Eval initiative, a community effort to improve the generation of MT systems focused on the current Covid-19 crisis. Nine teams took part in this event, which was divided in two rounds and involved seven different language pairs. Two different scenarios were considered: one in which only the provided data was allowed, and a second one in which the use of external resources was allowed. Overall, best approaches were based on multilingual models and transfer learning, with an emphasis on the importance of applying a cleaning process to the training data.

Spyridon Mavromatis, Sokratis Sofianopoulos, Prokopis Prokopidis et al.

Machine Translation (MT) for Ancient Greek (AG) to Modern Greek (MG) is a low-resource task, constrained by the lack of large-scale, high-quality parallel data. We address this gap by introducing the AG-MG Parallel Corpus, a new resource containing 132,481 sentence-aligned pairs derived from literary, historical, and biblical texts. We present a novel corpus creation pipeline that combines web-scraped, excerpt-level data with a multi-stage sentence-level alignment, and refinement process. Our method uses VecAlign with LaBSE embeddings, which we first fine-tune on a manually-aligned AG-MG subset, followed by an LLM-based error/misalignment correction phase using Gemini 2.5 Flash to ensure high alignment quality. Furthermore, we provide the first comprehensive benchmark of modern MT models on this task, evaluating three fine-tuning strategies across NMT models (NLLB, M2M100) and a Greek LLM (Llama-Krikri-8B). Our experiments show that fine-tuning yields significant improvements over base models, increasing performance by up to +10.3 BLEU points. Specifically, full-parameter fine-tuning of Llama-Krikri-8B achieves the highest overall performance with a BLEU score of 13.16, while the QLoRA-adapted M2M100-1.2B model demonstrates the largest relative gains and highly competitive results. Our dataset and models represent a significant contribution to Greek NLP.

Dimitris Roussis, Leon Voukoutis, Georgios Paraskevopoulos et al.

We introduce Llama-Krikri-8B, a cutting-edge Large Language Model tailored for the Greek language, built on Meta's Llama 3.1-8B. Llama-Krikri-8B has been extensively trained on high-quality Greek data to ensure superior adaptation to linguistic nuances. With 8 billion parameters, it offers advanced capabilities while maintaining efficient computational performance. Llama-Krikri-8B supports both Modern Greek and English, and is also equipped to handle polytonic text and Ancient Greek. The chat version of Llama-Krikri-8B features a multi-stage post-training pipeline, utilizing both human and synthetic instruction and preference data, by applying techniques such as MAGPIE. In addition, for evaluation, we propose three novel public benchmarks for Greek. Our evaluation on existing as well as the proposed benchmarks shows notable improvements over comparable Greek and multilingual LLMs in both natural language understanding and generation as well as code generation.

Tyler A. Chang, Catherine Arnett, Abdelrahman Eldesokey et al. · uw

To date, there exist almost no culturally-specific evaluation benchmarks for large language models (LLMs) that cover a large number of languages and cultures. In this paper, we present Global PIQA, a participatory commonsense reasoning benchmark for over 100 languages, constructed by hand by 335 researchers from 65 countries around the world. The 116 language varieties in Global PIQA cover five continents, 14 language families, and 23 writing systems. In the non-parallel split of Global PIQA, over 50% of examples reference local foods, customs, traditions, or other culturally-specific elements. We find that state-of-the-art LLMs perform well on Global PIQA in aggregate, but they exhibit weaker performance in lower-resource languages (up to a 37% accuracy gap, despite random chance at 50%). Open models generally perform worse than proprietary models. Global PIQA highlights that in many languages and cultures, everyday knowledge remains an area for improvement, alongside more widely-discussed capabilities such as complex reasoning and expert knowledge. Beyond its uses for LLM evaluation, we hope that Global PIQA provides a glimpse into the wide diversity of cultures in which human language is embedded.

Núria Bel, Vassilis Papavasiliou, Prokopis Prokopidis et al.

The objective of the PANACEA ICT-2007.2.2 EU project is to build a platform that automates the stages involved in the acquisition, production, updating and maintenance of the large language resources required by, among others, MT systems. The development of a Corpus Acquisition Component (CAC) for extracting monolingual and bilingual data from the web is one of the most innovative building blocks of PANACEA. The CAC, which is the first stage in the PANACEA pipeline for building Language Resources, adopts an efficient and distributed methodology to crawl for web documents with rich textual content in specific languages and predefined domains. The CAC includes modules that can acquire parallel data from sites with in-domain content available in more than one language. In order to extrinsically evaluate the CAC methodology, we have conducted several experiments that used crawled parallel corpora for the identification and extraction of parallel sentences using sentence alignment. The corpora were then successfully used for domain adaptation of Machine Translation Systems.