2 OLMo 2 Furious
2025
Read 2 OLMo 2 Furious (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-00656.txt.
Description
Team OLMo presents in its 2025 report OLMo 2, a family of fully open language models at three sizes. Unlike open-weight releases, everything is public: weights, training data, code, logs, and thousands of checkpoints. The report details architecture changes for training stability and efficiency. A special late-stage data mix, introduced during annealing, lifted benchmark scores widely.
An instruct version follows best practices from an earlier project and adds reinforcement learning with verifiable rewards. The base models match or beat rival open-weight models while using less compute. The authors stress full transparency as the point of the release.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 137 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-00656