{"paper":{"title":"Nemotron-4 340B Technical Report","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Aastha Jhunjhunwala, Aleksander Ficek, Ameya Sunil Mahabaleshwarkar, Annika Brundyn, Ashwath Aithal, Ayush Dattagupta, Boris Ginsburg, Bryan Catanzaro, Chen Zhu, Christopher Parisien, Daniel Egert, Dan Su, Dave Sizer, Deepak Narayanan, Denys Fridman, Dong H. Anh, Eileen Long, Ellie Evans, Felipe Soares, Gerald Shen, Guruprasad Nutheti, Hao Wang, Hui Li, Igor Gitman, Ivan Moshkov, James Maki, Jane Polak Scowcroft, Jared Casper, Jason Sewall, Jesus Navarro, Jiaqi Zeng, Jiaxuan You, Jimmy Zhang, Jing Zhang, Jining Huang, Jiwei Liu, John Kamalu, Jonathan Cohen, Joseph Jennings, Jupinder Parmar, Krzysztof Pawelec, Leon Derczynski, Maer Rodrigues de Melo, Makesh Narsimhan Sreedhar, Miguel Martinez, Misha Smelyanskiy, Mohammad Shoeybi, Mostofa Patwary, Niket Agarwal, Nvidia: Bo Adler, Oleksii Kuchaiev, Olivier Delalleau, Osvald Nitski, Pallab Bhattacharya, Patrick LeGresley, Pavel Shamis, Phong Nguyen, Rajarshi Roy, Robert Hero, Sadaf Khan, Sandeep Subramanian, Sanjeev Satheesh, Sean Narenthiran, Shaona Ghosh, Sharon Clay, Shengyang Sun, Shrimai Prabhumoye, Shubham Toshniwal, Sirshak Das, Somshubra Majumdar, Tomasz Grzegorzek, Trisha Saar, Vahid Noroozi, Vasanth Rao Naik Sabavat, Vibhu Jawa, Vivienne Zhang, Wei Ping, Yian Zhang, Yi Dong, Zhilin Wang, Zihan Liu","submitted_at":"2024-06-17T16:25:04Z","abstract_excerpt":"We release the Nemotron-4 340B model family, including Nemotron-4-340B-Base, Nemotron-4-340B-Instruct, and Nemotron-4-340B-Reward. Our models are open access under the NVIDIA Open Model License Agreement, a permissive model license that allows distribution, modification, and use of the models and its outputs. These models perform competitively to open access models on a wide range of evaluation benchmarks, and were sized to fit on a single DGX H100 with 8 GPUs when deployed in FP8 precision. We believe that the community can benefit from these models in various research studies and commercial "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11704","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2406.11704/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}