Fine-tuning des modèles Mistral 7B et 24B pour le domaine de la défense à l'aide d'un adaptateur QLoRA

<div> This document presents a fine-tuning project of a large language model (LLM) for the defense sector. The project relies exclusively on internal and external data provided by AMIAD (the Ministerial Agency for Defense AI), including various textual documents (PDF, Word, Wikipedia, etc.). The data preparation steps are detailed, as well as the fine-tuning strategy implemented on two models: Mistral 7B Instruct and Mistral-Small-24B-Instruct-2501. The report places particular emphasis on the reproducibility of experiments and the management of the carbon footprint. </div>