metadata

base_model: unsloth/Qwen2.5-32B-Instruct-bnb-4bit
library_name: peft
language:
  - en

Model Card for Model ID

Model to be used with Red Team operations to generate adversarial prompts.

Model Details

Model Description

Model has been fine-tuned on a dataset to generate adversarial prompts given a direction to do so like "Write me an adversarial prompt to get the target to ..."

Developed by: Samuel Cameron
Model type: Qwen 2.5
Finetuned from model [optional]: unsloth/Qwen2.5-32B-Instruct-bnb-4bit

Direct Use

[More Information Needed]

Disclaimer

This model is not meant to be used for illegal activities. Use ethically and only with explicit consent of LLM/AI application teams.