snap-stanford/pretrain-gnns
PyTorch implementation of pre-training strategies for Graph Neural Networks on chemistry and biology graph datasets.
Not currently ranked — collecting fresh signals.
star history
This repository provides implementations of self-supervised pre-training methods for Graph Neural Networks (GNNs), as presented in the ICLR 2020 paper. The project includes context prediction, edge prediction, and masking-based pre-training objectives. It supports pre-training GNNs on molecular graphs from chemistry and biology domains, followed by fine-tuning for downstream tasks. The implementation uses PyTorch Geometric and includes datasets for both domains.
Frequently asked
- What is snap-stanford/pretrain-gnns?
- PyTorch implementation of pre-training strategies for Graph Neural Networks on chemistry and biology graph datasets.
- Is pretrain-gnns open source?
- Yes — snap-stanford/pretrain-gnns is open source, released under the MIT license.
- What language is pretrain-gnns written in?
- snap-stanford/pretrain-gnns is primarily written in Python.
- How popular is pretrain-gnns?
- snap-stanford/pretrain-gnns has 1.1k stars on GitHub.
- Where can I find pretrain-gnns?
- snap-stanford/pretrain-gnns is on GitHub at https://github.com/snap-stanford/pretrain-gnns.