Strengths2
- End-to-end system covering data processing, model training, ensembling, and serving with explainability.
- Implementation of custom loss functions and a custom trainer demonstrates initiative beyond basic fine-tuning.
ML Application · Jupyter Notebook
Developed a sophisticated multi-label classification system to detect and prevent abusive content on social media platforms using ensemble of 4 fine-tuned transformer models (DeBERTa, ELECTRA, Twitter RoBERTa, HateBERT).
This repository presents an end-to-end system for detecting abusive comments in text using an ensemble of four fine-tuned transformer models (DeBERTa, ELECTRA, Twitter RoBERTa, and HateBERT). It includes model training scripts, an ensemble method for combining predictions, and a Streamlit-based web interface for real-time toxicity prediction with explainability features.
Merit describes the code at this commit. Later pushes count once the owner rescans. How merit is scored
Each field's share of the project's AI work, as scoring counts it. It ranks in the field it is most about.
Read from the project's code at its scored commit.
The owner keeps this project's limitations private.