It aims to promote low-resource AI models and link to open-access, non-profit archival repositories and AI models which are mostly excluded from other listings. The filtering options and features of AI models captured in the repository are focussed on features of AI models that support innovation in music making and challenge existing dominant deep-learning models. This repository supports open and editable lists of AI models, their training datasets, possibility of fine-tuning and control, licensing, and input and output types. The collection can be filtered by these properties, helping identify models which could be trained from scratch or fine-tuned on small datasets.
