chatbot_dhnl_v3

chatbot-dhnl-v3

Introduction 🎉

<!-- <p style="text-align: justify;"> Large Language Models (LLMs) are increasingly demonstrating their importance in addressing complex natural language processing tasks. However, they are still limited in generating text related to personalized datasets. Building a support system for answering questions in Vietnamese for students at Nong Lam University, Ho Chi Minh City, based on academic regulations, is a critical and practical task. This study focuses on researching methods for preprocessing Vietnamese data and fine-tuning Large Language Models (LLMs) to align with the specific language characteristics and content of the university's academic regulations. Additionally, the research team has constructed a dataset of the university's academic regulations and developed a Vietnamese text generation service to answer questions related to this dataset, which has been integrated into a chat website utilizing this service. </p> -->

Large Language Models (LLMs) are increasingly demonstrating their importance in addressing complex natural language processing tasks. However, they are still limited in generating text related to personalized datasets. Building a support system for answering questions in Vietnamese for students at Nong Lam University, Ho Chi Minh City, based on academic regulations, is a critical and practical task. This study focuses on researching methods for preprocessing Vietnamese data and fine-tuning Large Language Models (LLMs) to align with the specific language characteristics and content of the university's academic regulations. Additionally, the research team has constructed a dataset of the university's academic regulations and developed a Vietnamese text generation service to answer questions related to this dataset, which has been integrated into a chat website utilizing this service.

• Developed by:

  • Nguyễn Đăng Phước
  • Vũ Ngọc Thanh Trúc

• Model type: Multimodal Transformer with over 7B parameters

• Languages (NLP): Primarily Vietnamese with multilingual capabilities

• Fine-tuned from: nluai/dataset_dhnl_qna_v2

Examples 🧩

<div align="left"> <img src="assets/demo_chat1.png" width="2000"/> </div> <div align="left"> <img src="assets/demo_chat2.png" width="2000"/> </div>