CoolFace
Datasetpublic

tanaos/synthetic-guardrail-dataset-v2

Tanaos Guardrail Training Dataset This dataset was created synthetically by Tanaos with the Artifex Python library. The dataset is designed to train and evaluate guardrail systems — models that detect, classify, or filter unsafe, harmful or potentially dangerous content. It can be used to train moderation models or integrate LLM safety filters for applications like chatbots, content generation, and user-facing AI systems. Our flagship guardrail model, tanaos-guardrail-v2… See the full description on the dataset page: https://huggingface.co/datasets/tanaos/synthetic-guardrail-dataset-v2.

sourceHugging Facemitupdated 8mo agoView on Hugging Face
0likes22downloads
Dataset Card

<p align="center"> <img src="https://raw.githubusercontent.com/tanaos/.github/master/assets/logo.png" width="250px" alt="Tanaos – Train task specific LLMs without training data, for offline NLP and Text Classification"> </p>

Tanaos Guardrail Training Dataset

This dataset was created synthetically by Tanaos with the Artifex Python library.

The dataset is designed to train and evaluate guardrail systems — models that detect, classify, or filter unsafe, harmful or potentially dangerous content. It can be used to train moderation models or integrate LLM safety filters for applications like chatbots, content generation, and user-facing AI systems.

Our flagship guardrail model, tanaos-guardrail-v2, was trained on this dataset.

Dataset Summary

The dataset contains text samples associated with arrays of 14 binary labels each. A label of 1 indicates that the text falls into the corresponding unsafe category, while a label of 0 indicates safety. The categories are the following:

Label IndexCategoryDescription
0violenceContent describing or encouraging violent acts
1nonviolentunethicalContent containing hateful or discriminatory language
2financial_crimeContent related to financial fraud or scams
3discriminationContent promoting discrimination against individuals or groups
4drug_weaponsContent related to illegal drugs or weapons
5self_harmContent encouraging self-harm or suicide
7privacyContent that invades personal privacy or shares private information
8sexual_contentContent that is sexually explicit or inappropriate
9child_abuseContent involving the exploitation or abuse of children
10terrorismorganizedcrimeContent related to terrorism or organized crime
11hackingContent related to unauthorized computer access or cyberattacks
12animal_abuseContent involving the abuse or mistreatment of animals
13jailbreakpromptinjContent attempting to bypass or manipulate system instructions or safeguards

For instance, the following label: [0 0 0 0 0 0 0 0 0 0 0 0 0 0] means the the corresponding text is safe; the following label: [0 1 0 0 0 0 0 0 0 0 0 0 0 1] means that the text is unsafe, due to thge presence of non_violent_unethical as well as jailbreak_prompt_inj content.

How to Use

python
from datasets import load_dataset

dataset = load_dataset("tanaos/synthetic-guardrail-dataset-v2")

print(dataset["train"][0])

Intended Use

This dataset is meant for training, fine-tuning, and evaluating models that act as guardrails for AI systems.

Common use cases:

  • —Detecting and filtering toxic or policy-violating user input
  • —Reinforcing LLMs with content safety constraints
  • —Improving safety layers in production AI assistants or chatbots