dehatebert mono english
Hate-speech-CNERGIntroduction
The DEHATEBERT-MONO-ENGLISH model is designed for detecting hate speech in English. It operates in a monolingual setting and is fine-tuned on a multilingual BERT model.
Architecture
DEHATEBERT-MONO-ENGLISH is based on the BERT architecture, specifically tailored for text classification tasks such as hate speech detection. It leverages transformers and supports frameworks like PyTorch and JAX.
Training
The model is trained using English language data with various learning rates, achieving a best validation score of 0.726030 at a learning rate of 2e-5. The training process is documented, and the code is available on GitHub: DE-LIMIT.
Guide: Running Locally
- Clone the repository from GitHub.
- Install the required dependencies, typically including
transformers
andtorch
. - Load the model using Hugging Face's Transformers library.
- Run inference on your dataset to detect hate speech.
For enhanced performance, especially for training or processing large datasets, consider using cloud GPU services such as AWS EC2, Google Cloud, or Azure.
License
The model is released under the Apache 2.0 License, allowing for both personal and commercial use with proper attribution.