Open-source software
VulnTrain
A training toolkit for creating and publishing machine-learning models and datasets for vulnerability intelligence.
Training models for vulnerability intelligence
VulnTrain provides a reproducible workflow for preparing vulnerability data and training machine-learning models that can support vulnerability analysis. It brings the steps required to build and publish these resources together in an open-source toolkit.
The software is developed by CIRCL (Computer Incident Response Center Luxembourg) as part of its contribution to AIPITCH and is maintained openly on GitHub.
Models and datasets
The resources generated with VulnTrain are published through CIRCL’s Hugging Face organisation:
- Explore the machine-learning models generated by VulnTrain.
- Browse the associated datasets used to support training and research.
Publishing the models and their supporting datasets makes the resulting work easier to discover, evaluate and reuse in other cybersecurity tools and research workflows.
Get involved
The project is open source. You can review the implementation, report issues and propose improvements in the VulnTrain GitHub repository. Consult the repository README for current setup, training and usage instructions.