The original MoleculeNet paper (co-authored by @bharath), published in 2017, introduced a large-scale benchmark for molecular machine learning. It curates multiple public datasets, establishes metrics for evaluation, and offers high-quality open-source implementations of multiple previously proposed molecular featurization and learning algorithms (released as part of the DeepChem open-source library).
This project is the second iteration of MoleculeNet (a large-scale benchmark for molecular machine learning), which aims to improve the findings of existing papers with new datasets, benchmarks, and models that operate on molecular data.
This is the link for the proposal for MoleculeNet 2.0. Feel free to review it. I would be glad to have your feedback.