Wikuña - Preserving and Revitalizing the Quechua Language with AI

Introduction

Wikuña is an innovative project dedicated to the preservation and revitalization of the Quechua language, an indigenous language with a history spanning over 5000 years. Spoken by millions in the Andean region of South America, Quechua faces significant challenges due to socioeconomic and cultural factors. Wikuña leverages advanced AI technologies to create a digital ambassador for the Quechua language and culture.

Project Overview

Wikuña utilizes state-of-the-art Large Language Models (LLMs) to develop an AI agent capable of interacting in Quechua. This agent can teach the language, translate words and phrases, and generate text, making it a comprehensive tool for learners and enthusiasts. Additionally, Wikuña aims to promote the rich cultural heritage of the Andes by providing users with insights into traditions, folklore, and history. Based on LLaMA 3 8B

Features

Language Learning: Provides explanations on Quechua grammar and vocabulary.
Translation: Capable of translating words and phrases to and from Quechua.
Text Generation: Generates coherent and contextually appropriate text in Quechua.
Cultural Insights: Offers information about Andean traditions, folklore, and history.

Data Collection

The datasets used to train Wikuña include:

Grammar books and dictionaries.
Traditional songs and texts.
Personal notes and contributions from Quechua language experts.

These resources ensure that the AI model is well-rounded and accurate in its language capabilities.

Impact and Relevance

Wikuña has already made significant strides in promoting the Quechua language:

Educational Tool: Partnering with organizations to integrate Wikuña into language learning programs.
Cultural Preservation: Helping to maintain the rich heritage of the Andes by making it accessible to a broader audience.
Community Engagement: Showcased at science fairs, receiving positive feedback from students, teachers, and native speakers.

Ethical Considerations

Wikuña is developed with a strong emphasis on cultural sensitivity and ethical AI practices:

Cultural Sensitivity: Respectful representation of Andean culture, with consent from cultural experts and community leaders.
Data Fidelity: Ensuring the accuracy and unbiased nature of the information provided by the AI.

Future Plans

We are continuously working to improve Wikuña by:

Expanding the dataset with more diverse and comprehensive sources.
Enhancing the AI's capabilities with ongoing feedback and fine-tuning.
Developing partnerships with educational and cultural organizations to further integrate Wikuña into language preservation efforts.

Contributing

We welcome contributions from the community. If you have resources, feedback, or expertise in Quechua language and culture, please reach out to us.

Acknowledgements

Wikuña is inspired by the vision of Professor Yann LeCun and supported by numerous individuals and organizations dedicated to cultural preservation and technological innovation.

Wikuña - Bridging Ancient Wisdom with Modern Innovation