Devstral: Fine-tuning Language Models for Coding Agent Applications

Rastogi, Abhinav, Yang, Adam, Jiang, Albert Q., Liu, Alexander H., Sablayrolles, Alexandre, Héliou, Amélie, Martin, Amélie, Agarwal, Anmol, Ehrenberg, Andy, Lo, Andy, Roux, Antoine, Darcet, Arthur, Mensch, Arthur, Bout, Baptiste, Rozière, Baptiste, De Monicault, Baudouin, Bamford, Chris, Wallenwein, Christian, Renaudin, Christophe, Lanfranchi, Clémence, Denoix, Clément, Barreau, Corentin, Mizelle, Darius Dabert Devon, Casas, Diego de las, Chane-Sane, Elliot, Fugier, Emilien, Hanna, Emma Bou, Berrada, Gabrielle, Delerce, Gauthier, Guinet, Gauthier, Novikov, Georgii, Neubig, Graham, Lample, Guillaume, Martin, Guillaume, Jaju, Himanshu, Ludziejewski, Jan, Rute, Jason, Delignon, Jean-Malo, Chabran, JeanHadrien, Studnia, Joachim, Barmentlo, Joep, Amar, Jonas, Roberts, Josselin Somerville, Denize, Julien, Saxena, Karan, Yadav, Karmesh, Khandelwal, Kartik, Chandu, Khyathi Raghavi, Jain, Kush, Lavaud, Lélio Renard, Blier, Léonard, Zhao, Lingxiao, Martin, Louis, Saulnier, Lucile, Gao, Luyu, Pellat, Marie, Guillaumin, Mathilde, Felardos, Mathis, Dinot, Matthieu, Darrin, Maxime, Augustin, Maximilian, Seznec, Mickaël, Gupta, Neha, Raghuraman, Nikhil, Duchenne, Olivier, Wang, Patricia, von Platen, Patrick, Saffer, Patryk, Jacob, Paul, Wambergue, Paul, Kurylowicz, Paula, Chagniot, Philomène, Stock, Pierre, Agrawal, Pravesh, Delacourt, Rémi, Soletskyi, Roman, Sauvestre, Romain, Vaze, Sagar, Gandhi, Sanchit, Subramanian, Sandeep, Dalal, Shashwat, Gandhi, Siddharth, Ghosh, Soham, Mishra, Srijan, Aithal, Sumukh, Antoniak, Szymon, Scao, Teven Le, Lavril, Thibaut, Schueller, Thibault, Foubert, Thomas, Robert, Thomas, Wang, Thomas, Lacroix, Timothée, Bewley, Tom, Nemychnikova, Valeriia, Paltz, Victor, Richard, Virgile, Li, Wen-Ding, Marshall, William, Wang, Xingyao, Zhang, Xuanyu, Wan, Yihan, Tang, Yunhao

arXiv.org Artificial Intelligence 

We introduce Devstral-Small, a lightweight open source model for code agents with the best performance among models below 100B size. In this technical report, we give an overview of how we design and develop a model and craft specializations in agentic software development. The resulting model, Devstral-Small is a small 24B model, fast and easy to serve. Despite its size, Devstral-Small still attains competitive performance compared to models more than an order of magnitude larger.

Duplicate Docs Excel Report

Title
None found

Similar Docs  Excel Report  more

TitleSimilaritySource
None found