Summary

Olmo-core 3 is Ai2's open training infrastructure for large language models and mixture-of-experts systems. A redesigned MoE training stack distributes model development across large compute systems while targeting efficient scaling toward trillion-parameter models.

Key Features

Large-language-model training.
Mixture-of-experts infrastructure.
Distributed training.
Trillion-parameter scaling orientation.
Computational-efficiency improvements.
Open training framework.
Research customization.
Public code release.

About Olmo-core 3


Research teams can inspect and adapt the training infrastructure behind new Olmo models rather than relying only on finished checkpoints. Core capabilities include Large-language-model training, Mixture-of-experts infrastructure, Distributed training, and Trillion-parameter scaling orientation.


Public code or model weights are linked from the project, with setup instructions for running or adapting the system. Additional characteristics include Computational-efficiency improvements, Open training framework, Research customization, and Public code release.

Get more likes & reach the top of search results by adding this button on your site!

Embed button preview - Light theme
Embed button preview - Dark theme
TurboType Banner

Subscribe to the AI Search Newsletter

Get top updates in AI to your inbox every weekend. It's free!