SearcharxivSearch

arXiv · 2507.04895

O_FT@EvalLLM2025 : \'etude comparative de choix de donn\'ees et de strat\'egies d'apprentissage pour l'adaptation de mod\`eles de langue \`a un domaine

Abstract

This paper presents the work carried out by the O_FT team, joint with Orange and Ouest-France, on adapting language models to the defense domain as part of the EvalLLM2025 challenge. This work focused on adapting the \texttt{Mistral-7B-Instruct-v0.3} model using classical techniques of continued pre-training and instruction-tuning. The core of our efforts is based on collecting, generating, and selecting data for these two stages as well as for model evaluation. Experiments show that our adapted models have better domain-specific knowledge and improved domain-specific task processing skills, along with comparable (or even superior) performance on general knowledge and skills. Considering the carbon footprint of our adaptations, this work demonstrates the feasibility of domain adaptation for relatively small models. -- Ce document pr\'esente les travaux r\'ealis\'es par l'\'equipe O_FT conjointe \`a Orange et Ouest-France sur l'adaptation de mod\`eles de langue au domaine de la d\'efense dans le cadre du challenge EvalLLM2025. Ces travaux se sont concentr\'es sur l'adaptation du mod\`ele \texttt{Mistral-7B-Instruct-v0.3} avec des techniques classiques de poursuite du pr\'e-entra\^inement et d'affinage sur instructions. L'essentiel de nos travaux a port\'e sur la constitution, g\'en\'eration et s\'election de donn\'ees pour ces deux \'etapes ainsi que pour l'\'evaluation des mod\`eles. Les exp\'eriences montrent que nos mod\`eles adapt\'es ont de meilleures de connaissances de fond et une meilleure capacit\'e de traitement de t\^aches sur le domaine de la d\'efense, ainsi que des performances comparables (voire sup\'erieures) sur des connaissances ou capacit\'es g\'en\'eralistes. Mis au regard des empreintes carbones de nos adaptations, ces travaux d\'emontrent ainsi la viabilit\'e de l'adaptation \`a un domaine de mod\`eles relativement petits.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Ismaël Rousseau, Claire Perroux, Pierre Adam, Thomas Girault, Lionel Delphin-Poulat, Morgan Veyret, Gwénolé Lecorvé, Géraldine Damnati. 2025-07-07. O_FT@EvalLLM2025 : \'etude comparative de choix de donn\'ees et de strat\'egies d'apprentissage pour l'adaptation de mod\`eles de langue \`a un domaine. https://arxiv.org/abs/2507.04895

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Auto-RecSys: Harnessing Autonomous Research Agents for Industry-Scale Recommender System

Auto-research agents have shown the potential to automate hypothesis generation, experiment execution, and iterative refinement. However, scaling this paradigm to industry-scale recommendation models introduces two challenges: (1) long feedback loops, where model training can take days, making serial iteration prohibitively slow and requiring parallel exploration across multiple research directions; and (2) system complexity, where large configurations, fragile infrastructure dependencies, and multi-day GPU jobs require robust and recoverable execution. We present Auto-RecSys, an autonomous research system for long-horizon experimentation on industry-scale recommendation models. Auto-RecSys addresses these challenges through three harness designs: (1) distributed asynchronous execution for running multiple experiments in parallel across servers, (2) centralized cross-server memory for persistent and recoverable execution across sessions and failures, and (3) cognitive-procedural separation, where natural-language skill files guide LLM reasoning while deterministic scripts enforce operational correctness. Auto-RecSys further employs a dual-loop self-evolving architecture: an Execution Evolution Loop in which model-specific playbooks accumulate operational knowledge by recording failed attempts and crystallizing successful pipelines, and an Idea Evolution Loop in which experimental outcomes inform subsequent ideation. Evaluated on recommendation models, Auto-RecSys significantly reduces the human time required per experiment cycle and improves execution reliability as its playbooks mature.

cs.CL

Structurally Speaking: Motif-Oriented Graph Captioning through Bidirectional Graph-Text Translation

Graph captions should help readers understand graph structure, rather than simply translate adjacency matrices into long textual edge lists. A useful graph caption abstracts connectivity into recognizable motifs, such as hubs, paths, cycles, cliques, and bridges, because these motifs provide compact structural units that are easier to read, compare, and recover. In this paper, we study motif-oriented graph captioning as a bidirectional graph-text translation task, where captions must both preserve enough topology for graph recovery and express the graph through concise motif-level descriptions. We show that direct prompting of GPT-5.1 often produces graph-recoverable captions by enumerating node-to-node connections, but these captions are verbose and can contain inconsistent motif interpretations. To address this gap, we introduce Structurally Speaking, a lightweight structured prompting protocol that guides translation between explicit connectivity and motif-level abstraction. Experiments on a synthetic motif-based dataset show that structured prompting produces shorter and more motif-consistent captions while maintaining comparable graph recovery. These results suggest that explicit topology-to-motif reasoning guidance can make LLM-generated graph captions more interpretable without model fine-tuning.

cs.CL

Using Semantic Uncertainty to Estimate Transition Relevance in Turn-taking

Turn-taking is a fundamental mechanism that governs when interlocutors speak and listen. Although Spoken Dialogue Systems (SDS) exploit a range of linguistic, acoustic, and non-verbal cues, they produce ill-timed responses in unscripted interaction. A central challenge is anticipating Transition Relevance Places (TRPs), or opportunities, not obligations, for a listener to take the floor. Human listeners do not wait for turn endings; as an utterance unfolds, they use expectations about its developing meaning to anticipate TRPs and decide whether to take the floor. We examine whether these evolving expectations can be modeled through semantic uncertainty -- an LLM-derived measure of how strongly a turn so far constrains what may plausibly come next. To do so, we sample possible continuations of an ongoing turn and use changes in semantic dispersion to identify TRPs within turns. We evaluate this account on a dataset with TRP labels derived from real-time listener responses, rather than retrospective annotation. Our approach substantially outperforms prompt-based and fine-tuned text-only baselines, providing empirical support for the view that evolving semantic constraints inform perceived turn-taking opportunities in unscripted interaction.

cs.CL