base_model: unsloth/qwen2.5-coder-7b-instruct-bnb-4bit
tags:
- text-generation-inference
- transformers
- unsloth
- qwen2
- trl
- code-review
- static-analysis
license: apache-2.0
language:
- en
datasets:
- CyberNative/Code_Vulnerability_Security_DPO
Uploaded model
- Developed by: doss1232
- License: apache-2.0
- Finetuned from model : unsloth/qwen2.5-coder-7b-instruct-bnb-4bit
This qwen2 model was trained 2x faster with Unsloth and Huggingface's TRL library.
Model information
Offensive-Qwen2.5-Coder-7b is a fine-tuned Qwen-2.5-Coder-7b model and a part of Offensive-Engine, a collection of models fine-tuned for cybersecurity. It is train using the methods in Automating Code Review Activities by Large-Scale Pre-training paper. Which means, it is suitable for code review and static analysis tasks, while still maintain coding capability. Additionally, the model is further fine-tuned to find vulnerability in code, making it great for pentester and security professionals to use.
Further improvment
This model will be use for Auto-Pentest Framework.
Offensive-GenAI Framwork
Auto-Pentest is Agent system, built with Langchain to automate pentesting and web testing. It is designed to work in corporation's source code, where security and confidentiality are preserverd.