---
library_name: peft
license: apache-2.0
datasets:
- venkycs/llm4security
language:
- en
metrics:
- accuracy
tags:
- security
- infosec
- cybersec
- hack
---
## Training procedure

This repository contains security-specific domain data collected using Stanford Alpaca. We want to express our gratitude to the Stanford Alpaca team for their valuable contributions. This dataset enhances our understanding of security-related language and supports the development of safer AI applications.

Acknowledgments - Special thanks to Stanford Alpaca for providing the tools and resources to gather security domain data.

For more information about Stanford Alpaca, visit their website.

### Framework versions


- PEFT 0.4.0 (Thanks to HuggingFace team)
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_venkycs__llama-v2-7b-32kC-Security)

| Metric                | Value                     |
|-----------------------|---------------------------|
| Avg.                  | 42.99   |
| ARC (25-shot)         | 49.83          |
| HellaSwag (10-shot)   | 77.33    |
| MMLU (5-shot)         | 44.41         |
| TruthfulQA (0-shot)   | 47.96   |
| Winogrande (5-shot)   | 71.74   |
| GSM8K (5-shot)        | 3.87        |
| DROP (3-shot)         | 5.81         |