|
|
--- |
|
|
license: apache-2.0 |
|
|
language: |
|
|
- en |
|
|
- kha |
|
|
- grt |
|
|
- pbv |
|
|
tags: |
|
|
- preservation |
|
|
- research |
|
|
- nlp |
|
|
- ai |
|
|
- low-resource-languages |
|
|
- india |
|
|
- chatbots |
|
|
library_name: transformers |
|
|
title: Tynrai-AI |
|
|
sdk: static |
|
|
colorFrom: green |
|
|
colorTo: red |
|
|
--- |
|
|
# Tynrai |
|
|
|
|
|
**Tynrai (Tynrai-AI)** is an initiative dedicated to the **preservation of language** through technology. We focus on digitizing, documenting, and revitalizing the indigenous languages of Meghalaya, India. |
|
|
|
|
|
We build, curate, and release datasets and models including conversational agents that prioritize **real-world impact** for the Khasic and Garo languages. |
|
|
--- |
|
|
## Mission |
|
|
|
|
|
- **Preserve and digitize indigenous languages** |
|
|
- **Research in low-resource NLP** |
|
|
- **Build high-quality datasets and reproducible models** |
|
|
--- |
|
|
## Areas of Focus |
|
|
|
|
|
- **Neural Machine Translation (NMT)**: Specializing in Khasi, Garo, and Pnar |
|
|
- **Automatic Speech Recognition (ASR)**: Speech-to-text for indigenous languages |
|
|
- **Text-to-Speech (TTS)**: Natural speech generation for local languages |
|
|
- **Conversational AI**: Chat Bots and dialogue systems |
|
|
- **Language Preservation**: Documentation & corpus creation |
|
|
--- |
|
|
## What You’ll Find Here |
|
|
|
|
|
- **Chat Bots**: Interactive conversational agents for learning and assistance |
|
|
- **Datasets**: Parallel corpora, annotated text, speech resources, and QA Datasets |
|
|
- **Models**: Fine-tuned and experimental NLP models |
|
|
- **Spaces**: Demos and interactive experiments |
|
|
--- |
|
|
## Contact |
|
|
For collaboration or questions: |
|
|
- Hugging Face Discussions |
|
|
--- |
|
|
> Low-resource does not mean low-impact. |
