Octavian-ai/synthetic-graph-data

This project is a 'zombie' repository, nearly 3,000 days old with zero recent activity and very low adoption (22 stars). While Octavian-ai was an early pioneer in graph-based deep learning, the functionality provided here—generating synthetic graphs—has been completely commoditized and absorbed into modern standard libraries. Specifically, PyTorch Geometric (PyG) and Deep Graph Library (DGL) now offer significantly more robust, GPU-accelerated, and diverse synthetic generators (Erdős–Rényi, Watts–Strogatz, etc.) as built-in utilities. The defensibility is near zero as the code represents a standard implementation of known graph algorithms with no proprietary data or community moat. Frontier labs have no interest in standalone synthetic graph generators of this scale; they focus on large-scale relational simulators or generative models for molecular/biological graphs. For a technical user, this repository serves only as a historical artifact of early Graph ML experimentation.