HackMentor is a cybersecurity LLMs (Large Language Models) focused on domain-specific data fine-tuning. This project consists of three main parts: data construction, model training, and model evaluation.
Data construction: Methods and tools for creating domain-specific datasets (instructions & conversations) for fine-tuning LLMs.
Model training: Techniques and processes for training LLMs on the… See the full description on the dataset page:
https://huggingface.co/datasets/ukcli/hackmentor-instruction.
