The implementation training and evaluation of a Structure Seer model designed for
reconstruction of adjacency of a molecular graph from the labelling of its nodes.
The detailed characterisation and disclosure of the model architecture is provided in:
Structure Seer - a machine learning model for chemical structure elucidation
from a node labelling of a molecular graph, Digital discovery, 2023