programindz/kashmiri-streaming-speech-recognition

🎙️ Real-Time Kashmiri Streaming ASR (FastAPI + Sherpa-ONNX)

This project demonstrates a real-time speech-to-text (ASR) web application with:

• —🎛️ Hugging Face Deployment taken from Luigi
• —🧠 Sherpa-ONNX streaming Zipformer model
• —🚀 FastAPI backend with WebSocket support
• —☁️ Docker-compatible deployment (CPU-only) on Hugging Face Spaces

🤖 Training

• —Model: Zipformer Small
• —Dataset: IndicVoices
• —WER: 36%

🧪 Local Development

1. 1.Install dependencies

pip install -r requirements.txt

1. 1.Run the app locally

uvicorn app.main:app --reload --host 0.0.0.0 --port 8501

Open http://localhost:8501 in your browser.

https://k2-fsa.github.io/sherpa/ncnn/endpoint.html

📁 Project Structure

.
├── app
│   ├── main.py
│   ├── asr.py
│   └── model parts
        └── All Model parts here (encoder, decoder, joiner, tokens)
    ├── index.html
├── requirements.txt
├── Dockerfile
└── README.md

🔧 Credits

• —Sherpa-ONNX
• —OpenCC
• —FastAPI
• —Hugging Face Spaces
• —AI4Bharat
• —Icefall-K2

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference