This repo contains the notebooks resulting from my bachelor's thesis developed at UB (Universitat de Barcelona).
The aim of the project was to explore how transformers worked, to analize BERTa and to extract the Semantic Textual Similarities between the Gran Enciclopèdia Catalana articles.
Tha graphs produced by the notebooks have been exported and have been uploaded to the 'graphs' folder.
The resulting streamlit app has been deployed and can be found at https://share.streamlit.io/t1emp0/tfg/main/GEC_streamlit_app.py