Projects


MaLA - Massive Language Adaptation of Large Language Models

January 1, 2023

MaLA focuses on the adaptation of large language models (LLMs) to better understand and generate human language across diverse contexts and applications in the massively multilingual scenario. It aims to enhance the capabilities of existing LLMs b...


High-Performance Language Technologies

January 1, 2023

Embark on a journey of linguistic innovation with the High Performance Language Technologies (HPLT) Project. At the forefront of language and translation modeling, HPLT is dedicated to building the largest collection of free and reproducible model...


Pretrained Language Models for Mental Health

October 13, 2021

Revolutionizing mental health with pretrained language models, unraveling the potential of MentalBERT, MentalRoBERTa, MentalXLNet, and MentalLongformer, and navigating the landscape of large language models.


Deep Learning for Medical Code Assignment from Clinical Notes

March 1, 2020

Medical code assignment aims to predict medical codes from clinical notes. It is an essential task in medical information system management. This project conduct a series of studies on deep learning-based automatic coding methods.


Natural Language Processing for Healthcare

March 1, 2020

This project is my doctoral dissertation. It explores advanced techniques in deep learning for natural language processing, specifically focusing on improving clinical text encoding for intelligent medical information systems, multitask learning i...


Suicidal Ideation Detection in Online Social Content

February 10, 2020

In the digital age, detecting signs of suicidal intention from online posts is a crucial endeavor. Leveraging text feature processing and supervised learning methods, this project involves extracting meaningful insights from user-generated content...


Federated Deep Learning

March 30, 2018

Federated learning, a novel distributed learning paradigm, separates data collection and model training using multi-party computation and model aggregation. With the rapid advancement of deep neural networks, this approach has evolved into federat...