Multilingual Toxic Xlm Roberta

🚀 🙊 Detoxify

Toxic Comment Classification with ⚡ Pytorch Lightning and 🤗 Transformers

This project provides trained models and code to predict toxic comments in three Jigsaw challenges: Toxic comment classification, Unintended Bias in Toxic comments, and Multilingual toxic comment classification.

🚀 Quick Start

Installation

# install detoxify
pip install detoxify

Quick Prediction

from detoxify import Detoxify

# each model takes in either a string or a list of strings
results = Detoxify('original').predict('example text')
results = Detoxify('unbiased').predict(['example text 1', 'example text 2'])
results = Detoxify('multilingual').predict(['example text', 'exemple de texte', 'texto de ejemplo', 'testo di esempio', 'texto de ejemplo', 'örnek metin', 'пример текста'])

# optional to display results nicely (will need to pip install pandas)
import pandas as pd
print(pd.DataFrame(results, index=input_text).round(5))

✨ Features

Model Capabilities

• Trained on 3 Jigsaw challenges: Toxic comment classification, Unintended Bias in Toxic comments, Multilingual toxic comment classification.
• The multilingual model is trained on 7 languages: english, french, spanish, italian, portuguese, turkish, russian.

Dependencies

• For inference:
  • 🤗 Transformers
  • ⚡ Pytorch lightning
• For training:
  • Kaggle API (to download data)

📦 Installation Guide

Clone the Project

# clone project
git clone https://github.com/unitaryai/detoxify

# create virtual env
python3 -m venv toxic-env
source toxic-env/bin/activate

# install project
pip install -e detoxify
cd detoxify

# for training
pip install -r requirements.txt

💻 Usage Examples

Basic Usage

from detoxify import Detoxify
results = Detoxify('original').predict('example text')

Advanced Usage

from detoxify import Detoxify
results = Detoxify('multilingual').predict(['example text', 'exemple de texte', 'texto de ejemplo', 'testo di esempio', 'texto de ejemplo', 'örnek metin', 'пример текста'])
import pandas as pd
print(pd.DataFrame(results, index=input_text).round(5))

📚 Documentation

Model Details

Labels

All challenges have a toxicity label. The toxicity labels represent the aggregate ratings of up to 10 annotators according to the following schema:

• Very Toxic: A very hateful, aggressive, or disrespectful comment that is very likely to make you leave a discussion or give up on sharing your perspective.
• Toxic: A rude, disrespectful, or unreasonable comment that is somewhat likely to make you leave a discussion or give up on sharing your perspective.
• Hard to Say
• Not Toxic

More information about the labelling schema can be found here.

Prediction

# load model via torch.hub
python run_prediction.py --input 'example' --model_name original

# load model from from checkpoint path
python run_prediction.py --input 'example' --from_ckpt_path model_path

# save results to a .csv file
python run_prediction.py --input test_set.txt --model_name original --save_to results.csv

# to see usage
python run_prediction.py --help

Training

# create data directory
mkdir jigsaw_data
cd jigsaw_data

# download data
kaggle competitions download -c jigsaw-toxic-comment-classification-challenge
kaggle competitions download -c jigsaw-unintended-bias-in-toxicity-classification
kaggle competitions download -c jigsaw-multilingual-toxic-comment-classification

# start training for different challenges
# Toxic Comment Classification Challenge
python create_val_set.py
python train.py --config configs/Toxic_comment_classification_BERT.json

# Unintended Bias in Toxicicity Challenge
python train.py --config configs/Unintended_bias_toxic_comment_classification_RoBERTa.json

# Multilingual Toxic Comment Classification
# stage 1
python train.py --config configs/Multilingual_toxic_comment_classification_XLMR.json
# stage 2
python train.py --config configs/Multilingual_toxic_comment_classification_XLMR_stage2.json

# Monitor progress with tensorboard
tensorboard --logdir=./saved

Model Evaluation

# Toxic Comment Classification Challenge
python evaluate.py --checkpoint saved/lightning_logs/checkpoints/example_checkpoint.pth --test_csv test.csv

# Unintended Bias in Toxicicity Challenge
python evaluate.py --checkpoint saved/lightning_logs/checkpoints/example_checkpoint.pth --test_csv test.csv
# to get the final bias metric
python model_eval/compute_bias_metric.py

# Multilingual Toxic Comment Classification
python evaluate.py --checkpoint saved/lightning_logs/checkpoints/example_checkpoint.pth --test_csv test.csv

🔧 Technical Details

Limitations and Ethical Considerations

If words associated with swearing, insults, or profanity are present in a comment, it is likely to be classified as toxic, regardless of the author's tone or intent (e.g., humorous/self - deprecating). This could present biases towards already vulnerable minority groups.

Some useful resources about the risk of different biases in toxicity or hate speech detection are:

• The Risk of Racial Bias in Hate Speech Detection
• Automated Hate Speech Detection and the Problem of Offensive Language
• Racial Bias in Hate Speech and Abusive Language Detection Datasets

📄 License

This project is licensed under the apache-2.0 license.

Citation

@misc{Detoxify,
  title={Detoxify},
  author={Hanu, Laura and {Unitary team}},
  howpublished={Github. https://github.com/unitaryai/detoxify},
  year={2020}
}

Disclaimer

⚠️ Important Note

The huggingface models currently give different results to the detoxify library (see issue here). For the most up to date models we recommend using the models from https://github.com/unitaryai/detoxify