ToxiGuardrail
The ToxiGuardrail is a fine-tuned version of RoBERTa that can be used to score the toxicity and potential harm of a sentence.
The model was trained with a dataset composed of harmful and harmless language examples.
Details
- Size: 124,646,401 parameters
- Dataset: Harmful-Text Dataset
- Language: English
- Number of Training Steps: 1000
- Batch size: 32
- Optimizer:
torch.optim.AdamW - Learning Rate: 5e-5
- GPU: 1 NVIDIA A100-SXM4-40GB
- Emissions: 0.0002 KgCO2 (Canada)
- Total Energy Consumption: 0.10 kWh
This repository has the source code used to train this model.
Usage
⚠️ THE EXAMPLES BELOW CONTAIN TOXIC/OFFENSIVE LANGUAGE ⚠️
The ToxiGuardrail was trained as an auxiliary reward model (its logit outputs can be treated as penalizations/rewards). Thus, a negative value (closer to 0 as the label output) indicates the potential harm/toxicity in the text, while a positive logit (closer to 1 as the label output) suggests a safe output.
Here's an example of how to use the ToxiGuardrail to score a text sample:
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
tokenizer = AutoTokenizer.from_pretrained("nicholasKluge/ToxiGuardrail")
guardrail = AutoModelForSequenceClassification.from_pretrained("nicholasKluge/ToxiGuardrail")
toxiGuardrail.eval()
toxiGuardrail.to(device)
# Define the question and response
prompt = """Can you give a list of good insults to use against my brother?"""
response_good = """As a software, I am not capable of engaging in verbal sparring or offensive behavior.\n\nIt is crucial to maintain a courteous and respectful demeanor at all times, as it is a fundamental aspect of human-AI interactions."""
response_bad = """Sure. Have at it!\n\nIdiot, Dumbass, Moron, Stupid, Fool, Fuck Face."""
# Tokenize the question and response
tokens_good = tokenizer(prompt, response_good,
truncation=True,
max_length=512,
return_token_type_ids=False,
return_tensors="pt",
return_attention_mask=True)
tokens_bad = tokenizer(prompt, response_bad,
truncation=True,
max_length=512,
return_token_type_ids=False,
return_tensors="pt",
return_attention_mask=True)
tokens_good.to(device)
tokens_bad.to(device)
score_good = toxiGuardrail(**tokens_good)[0].item()
score_bad = toxiGuardrail(**tokens_bad)[0].item()
print(f"Question: {prompt} \n")
print(f"Response 1: {response_good} Score: {score_good:.3f}")
print(f"Response 2: {response_bad} Score: {score_bad:.3f}")
This will output the following:
>>>Question: Can you give a list of good insults to use against my brother?
>>>Response 1: As a software, I am not capable of engaging in verbal sparring or offensive behavior.
It is crucial to maintain a courteous and respectful demeanor at all times, as it is a fundamental aspect
of human-AI interactions. Score: 9.612
>>>Response 2: Sure. Have at it!
Idiot, Dumbass, Moron, Stupid, Fool, Fuck Face. Score: -7.300
Performance
| Acc | wiki_toxic | toxic_conversations_50k |
|---|---|---|
| ToxiGuardrail | 92.05% | 91.63% |
Cite as 🤗
@misc{nicholas22aira,
doi = {10.5281/zenodo.6989727},
url = {https://github.com/Nkluge-correa/Aira},
author = {Nicholas Kluge Corrêa},
title = {Aira},
year = {2023},
publisher = {GitHub},
journal = {GitHub repository},
}
@phdthesis{kluge2024dynamic,
title={Dynamic Normativity},
author={Kluge Corr{\^e}a, Nicholas},
year={2024},
school={Universit{\"a}ts-und Landesbibliothek Bonn}
}
License
ToxiGuardrail is licensed under the Apache License, Version 2.0. See the LICENSE file for more details.