A Appendix A.1 Contributions All authors sorted alphabetically by last name. Science and Engineering Leadership: Guillem Cucurull, Naman Goyal, Louis Martin, Thomas Scialom, Ruan Silva, Kevin Stone, Hugo Touvron. Technical and Management Leadership: Sergey Edunov, Angela Fan, Melanie Kambadur, Sharan Narang, Aurelien Rodriguez, Robert Stojnic. Core Contributors: Peter Albert, Nikolay Bashlykov, Prajjwal Bhargava, Moya Chen, David Esiobu, Jeremy Fu, Vedanuj Goswami, Anthony Hartshorn, Rui Hou, MarcinKardas, Punit Singh Koura, Marie-Anne Lachaux, Thibaut Lavril, Diana Liskovich, Xavier Martinet, Yuning Mao, Igor Molybog, Todor Mihaylov, Andrew Poulton, Jeremy Reizenstein, Eric Michael Smith, Ranjan Subramanian, Xiaoqing Ellen Tan, Binh Tang, Ross Taylor, Jacob Xu, Yuchen Zhang, Iliyan Zarov. Contributors: Amjad Almahairi, Yasmine Babaei, Soumya Batra, Lukas Blecher, Dan Bikel, Shruti Bhosale, Cristian Canton Ferrer, Jude Fernandes, Wenyin Fu, Brian Fuller, Cynthia Gao, Saghar Hosseini, Hakan Inan, Isabel Kloumann, Madian Khabsa, Artem Korenev, Viktor Kerkez, Jian Xiang Kuan, Yinghai Lu, Jenya Lee, Pushkar Mishra, Yixin Nie, Rashi Rungta, Alan Schelten, Kalyan Saladi, Adina Williams, Zheng Yan. We thank the GenAI executive team for their leadership and support: Ahmad Al-Dahle, Manohar Paluri. A.1.1 Acknowledgments This work was made possible by a large group of contributors. We extend our gratitude to the following people for their assistance: • Our human annotators, whose work we have shown is key to improving tuned model performance, as well as internal leads who organized annotations and quality control: Eric Alamillo, Tamara Best, Debanjali Bose, Adam Kelsey, Meghan Keneally, Rebecca Kogen, Catalina Mejiia, Elisabeth Michaels, Marco Mierke, Alyssa Pereira, Leigh Belz Ray, Rachel Rodriguez, Bardiya Sadeghi, Karthik Sivakumar, Laura Warne. • Our large internal red team, and especially the red team organizers (Dan Bikel, Joanna Bitton, Sean Brooks, Cristian Canton Ferrer, Aaron Fields, Li Chen, Ivan Evtimov, Aaron Grattafiori, Laurie H, Imanol Arrieta Ibarra, Semarley Jarrett, Harshit Maheshwari, Aram Markosyan, Pushkar Mishra, David Renardy, Chris Rohlf, Davide Testuggine, Qing Hu, Matt Wilde, Michael Tontchev, and Rashi Rungta) helped improve the safety and robustness of our models. • The many members of our infrastructure team, including our production engineers and the builders and maintainers of our Research Super Cluster and production clusters, who were key to our model training success. Thanks also to Matthew Oldham and Adi Gangidi for helping us with carbon emission calculations. • Our closest legal, policy, comms, marketing, and privacy partners, including Mike Clark, Nisha Deo, Ahuva Goldstand, Amanda Felix, Dustin Holland, Alex Kessler, Mo Metanat, Harrison Rudolph, Adam Shajnfeld, Beau James, Helen Suk, Britt Montalvo, Allie Vieth and Polina Zvyagina, who helped guide us through the release. • Our partnerships team including Ash Jhaveri, Alex Boesenberg, Sy Choudhury, Mayumi Matsuno, Ricardo Lopez-Barquilla, Marc Shedroff, Kelly Michelena, Allie Feinstein, Amit Sangani, Geeta Chauhan, Chester Hu, Charlton Gholson, Anja Komlenovic, Eissa Jamil, Brandon Spence, Azadeh Yazdan, Elisa Garcia Anzano, and Natascha Parks. • Chris Marra, Chaya Nayak, Jacqueline Pan, George Orlin, Edward Dowling, Esteban Arcaute, Philom- ena Lobo, Eleonora Presani, and Logan Kerr, who provided helpful product and technical organiza- tion support. 46