Sentiment-Analysis-Using-RoBERTa

The purpose of this project was to analyze a dataset for sentiment analysis using a pre-trained RoBERTa model. The model architecture was adjusted for multi-class classification, and the optimizer and learning rate scheduler were configured for optimal training performance. A training and evaluation loop was designed to monitor the model’s performance throughout fine-tuning, including saving and loading the trained model. The end result is a sentiment analysis model that leverages RoBERTa’s large-scale language understanding to classify text accurately.