In my last year of studies at IDC I participated in the course NLP. In the final project me and a master's degree student designed and developed two deep learning models to predict emotions from sentences. One model was based on LSTM and another on Google’s BERT. Then, we made a comprehensive comparison between the results and concluded interesting insights.
The project was developed in a Python development environment called Google Colab. We used a variety of libraries such as PyTorch, Pandas, Matplotlib and Hugging Face' Transformers.
I invite you to read the final report that summarizes all the results.