Welcome to the Hate Speech Classifiers repository. Here, we focus on refining various models tailored for classifying hate speech in Portuguese. To ensure seamless execution and avoid versioning or dependency issues, we highly recommend running these notebooks on Google Colab, leveraging the power of an A100 machine.
The repository is organized into two main sections: training and testing. Within the training section, notebooks are dedicated to specifying the model used for refinement. Additionally, they indicate whether the task is classical binary classification or multiclass classification.
Feel free to explore the notebooks to delve deeper into our process of refining hate speech classifiers in Portuguese. Let's combat hate speech together!
The TuPy project is the result of the development of Felipe Oliveira's thesis and the work of several collaborators. This project is financed by the Federal University of Rio de Janeiro (UFRJ) and the Alberto Luiz Coimbra Institute for Postgraduate Studies and Research in Engineering (COPPE).