Is mostly based on supervised/semi supervised learning, in order to find a point of low entropy (stability) in output. In other words it is purposefully designed and meant for research around open source models.
Unrelated but will be introduced to this project is, datasets in terms of quality. This will pertain to a specific case where i only work on having it do one specific task, not multi-modal inference, just pure datasets,probability & statistics.
This is not meant as a service or product, as it constitutes research and understanding of what is and what isnt useful / quality data. furthermore feel free to fork!