Read the documentation.
Positive:
- the software for training and inference of this model is open source
- the trained model is freely available, and thus can be run on-premises
Negative:
- the training data is not freely available, limiting the ability of external parties to check and correct for bias or optimise the model’s performance and CO2 usage.
Learn more about the Nextcloud Ethical AI Rating in our blog.
AGPL-3.0