Repository containing the project for the course on Business and Project Management at the University of Pisa (A.Y. 2022/2023) realized by Fabiano Pilia, Emanuele Tinghi and Matteo Dal Zotto.

Finetuned FLAN-T5 to translate English to Hawaiian Pidgin

This project summarizes large text from any article to a smaller version without any loss in context. It uses the T5 Base transformer model.

Multilingual Text Summarizer is a web application that summarizes text, PDFs, and images in multiple languages using a T5 transformer model. The application is built with Streamlit, EasyOCR, and Hugging Face Transformers.

Creating a RAG model and Fine tuning for custom purposes

Fine-Tuning LLM for summarization in Portuguese (T5)

Fine-tuning the T5-base model using parts of the CNN/DailyMail dataset using PyTorch Lightning.

A NLP based summarizer which can summarize news article.

German-to-English fine-tuned T5 base model with dependency-parsing enhancement.

In this we explore the large language models like Pegasus ,ERNIE and T5 Large

A django application that allows you to summarize the content of a press article found by a user. It uses the T5 model for summarization, the LDA algorithm for topic modeling and Selenium to scrape the content of the linked article.

Institute Technical Summer Project -23/24

finetuning t5-base model for detoxifying texts.

A RAG (Retrieval augmented generation)-based FAQ Chat-Bot, designed to operate within an organization's internal domain. - Jul 2023 - Oct 2023

Thesis scope: Train and Develop a Table-to-Text Transformer-based model for contextual summarization of tabular data. To achieve this T5-small , T5-base, Bart-base and Llama2 7B chat were finetuned on ToTTo and QTSumm. Regarding ToTTo, the models outperformed the benchmark.

This repository contains the code and outputs for the CS505: Natual Language Processing course project. The objective of this work is to explore the performance of different machine learning models in generating commit messages from changes in code.

An agent-based, peer-to-peer, chat/social messaging application, intended for users that speak differing native languages. Built atop Fetch.ai's uAgents framework and t5-base integration.

This repository consists of the implementation of the paper "Automatic Ellipsis Reconstruction in Coordinated German Sentences based on Text-To-Text Transfer Transformers" which is accepted at 27th International Conference on Text, Speech and Dialogue.

Python codes generation from latex expressions. Using synthetic dataset and CodeT5-base model.

An agent-based reader view plugin for summarizing an article or web page, using Fetch.ai's uAgents framework and the t5-base integration.