Welcome to the repository for our trilingual context-aware question and answer generation model!
This repository hosts the code and resources for a state-of-the-art trilingual Q&A generation model. Our model is designed to generate context-aware questions and answers in three different languages, making it a versatile tool for various natural language processing tasks.
- Multilingual Support: Our model supports three languages: English, Hindi, and Yoruba. It is extendable to as many languages as desired.
- Context Awareness: It is capable of detecing the input language and generating questions and plausible answers based on context paragraph, making it suitable for tasks like reading comprehension, chatbots, and more.
- State-of-the-Art Performance: Our model has been trained with 11.5% portion of the datasets due to hardware constraint. We achieved impressive results in our experiments which you will find at the results section of our paper "Trilingual Context-aware Question and Answer Generation".
To use our model, follow these steps:
- Clone this repository:
git clone https://github.com/juliusco/triQGA.git - Install the required dependencies.
- Download the pre-trained model weights (available upon request).
- Run the evaluation script to generate questions and answers.
For more detailed instructions and examples, please refer to the documentation provided in the repository.
If you use our model or code in your research, please consider citing our paper. We will upload our trained model and the evaluation script we used for our paper at the time of acceptance by the editorial board.
For inquiries or requests regarding our model, feel free to contact us:
- Email: [email protected]
- GitHub Issues: Link to GitHub Issues
We appreciate your interest in our work and look forward to collaborating with the research community.