-
Step 1: Check our huggingface repository and download the model https://huggingface.co/unibuc-cs/CyberGuardian
-
Step 2: Use Ollama or llama.cpp to do local inference of the model, indepedently of the architecture you run on (e.g., you can use MacOS, CPU only, GPUs, etc.).
-
Step 3: Install the UI/requirements.txt packages to run the project.
-
Step 4: While the UI interface is currently built with Streamlit library, use streamlit run UI/Main_Page.py to start.
-
Notes:
- Check our video demos, create a profile, and test your skills.
- To run our demo in the presentations, define a OS env variable: DEMO_USE_CASE=hospital
- Our dataset is uploaded to https://huggingface.co/datasets/unibuc-cs/CyberGuardianDataset
- Note that the repository contains scripts to inject your own data, update the dataset and even include your data in RAG.
- Check the documentation in the Readme.
We created a set of Pycharm confirmation to sustain development. If you load the project you will see various setups to speed to things. For example:
- Use UITest_NO_LOGIN to test interaction with the model in UI, skipping all the account creation or login.
- Use CyberGuardianLLM_Training to start fine-tuning.
- Dataset handling can be done using DatasetUtils config
- To interact with the model and ask questions in a reproduceable way, use main_qa.
@conference{CyberGuardian,
author={Ciprian Paduraru. and Catalina Patilea. and Alin Stefanescu.},
title={CyberGuardian: An Interactive Assistant for Cybersecurity Specialists Using Large Language Models},
booktitle={Proceedings of the 19th International Conference on Software Technologies - ICSOFT},
year={2024},
pages={442-449},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012811700003753},
isbn={978-989-758-706-1},
issn={2184-2833},
}
Acknowledgements: This research was supported by European Union’s Horizon Europe research and innovation programme under grant agreement no. 101070455, project DYNABIC, where we use the code for the Chat4Operator component.