AIAJ: Fine-Tuned Mistral-7B for French Legal Tasks

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.3 for French legal tasks, including legal Q&A, template generation, and personalized document creation. It was fine-tuned on a dataset of 541 examples using LoRA (r=4) on Kaggle.

Training Details

Dataset: 541 examples (reduced to 150 for training)
Epochs: 2
Training Loss: Decreased to 2.0857
Validation Loss: Decreased to 1.83909
Completion Date: 2025-05-20 15:02:47 CEST

For more details, contact the project team.