This repository hosts the code and documentation for the AI_Newbies team participating in the Upstage AI Hackathon, where we have developed a fine-tuned AI model designed to promote Jeju Island tourism. The model is deployed through a Streamlit app, offering interactive conversations that mimic the personalities of tourism ambassadors and provide insights into local attractions.
University of Malaya, Malaysia
- Jonathan Siew Zunxian
- Ooi Rui Zhe
- Wong Yoong Yee
The fine-tuned adapters, which tailor the large language model to specific tasks, can be found on GitHub - AI_Newbies-Upstage-AI-Hackathon.The repository includes our approach to training LLM adapters, providing detailed notebooks and resources that outline the methods used to customize the model's responses for promoting Jeju Island tourism. You can explore and download the fine-tuned adapters directly from the repository.
Follow these steps to get the project running.
Ensure you have Python installed on your machine. You can download Python from python.org.
- Clone the repository
git clone https://github.com/RextonRZ/AI_Newbies-Streamlit.git
- Navigate to the project directory
cd AI_Newbies-Upstage-AI-Hackathon
- Virtual environment Set up
python -m venv venv
Activate the virtual environment after creating:
#For Command Prompt:
venv\Scripts\activate
#For PowerShell:
.\venv\Scripts\Activate
Install Required Packages
- Install Streamlit Install the Streamlit package by running the following command:
pip install streamlit
Sometimes, you might be prompted to update pip during the installation process. If prompted, follow the instructions provided to update pip by running the command suggested in the prompt.
- Install Predibase To use the Predibase library, install it with the following command:
pip install predibase
- Start the Streamlit app
streamlit run jejubot.py
The AI model and fine-tuned adapters presented in this repository are intended for stimulation and educational purposes only. While we have made efforts to ensure the accuracy and relevance of the content, the use of these models in production environments or for critical applications should be approached with caution.