Open Foundation Models and Datasets

Enabling an ecosystem of open foundation models, including those with multilingual and multi-modal capabilities, and open datasets.

We are responsibly enhancing the ecosystem of open foundation models and datasets. We are embracing multilingual and multimodal models, as well as science models tackling broad societal issues like climate change and education.

To aid AI model builders and application developers, we’re collaborating to develop and promote open-source tools for model training, tuning, and inference. We are also launching programs to foster the open development of AI in safe and beneficial ways, and hosting events to explore AI use cases.

Without good datasets, model training and tuning would be impossible. We are promoting the development of open datasets with clear governance and provenance controls so they can be used without concerns for legal and other risks.