Functor GNN Demo
Run the commit-indexed Functor GNN / mQSAR workflow and inspect generated training, commit, inference, equivalence, and rollback logs.
This demo showcases a deterministic GNN workflow where model learning is captured as versioned, reversible updates rather than opaque parameter changes. Each model state can be reproduced exactly, compared across versions, or rolled back without loss. Be sure to wait for the process to complete and you should see a log report.
The result is a transparent and auditable approach to AI, supporting reproducibility, traceability, and controlled evolution—capabilities that are especially valuable in domains such as mQSAR and other scientific applications.