programindz/kashmiri-streaming-speech-recognition

1---2title: Kashmiri Streaming ASR Zipformer3emoji: 💻4colorFrom: purple5colorTo: gray6sdk: docker7pinned: false8short_description: 'Online Streaming ASR System for Kashmiri Language'9license: apache-2.010---11 12# 🎙️ Real-Time Kashmiri Streaming ASR  (FastAPI + Sherpa-ONNX)13 14This project demonstrates a real-time speech-to-text (ASR) web application with:15* 🎛️ Hugging Face Deployment taken from [Luigi](https://huggingface.co/spaces/Luigi/Streaming-Zipformer)16* 🧠 [Sherpa-ONNX](https://github.com/k2-fsa/sherpa-onnx) streaming Zipformer model17* 🚀 FastAPI backend with WebSocket support18* ☁️ Docker-compatible deployment (CPU-only) on Hugging Face Spaces19 20 21## 🤖 Training22* Model: [Zipformer Small](https://github.com/k2-fsa/icefall)23* Dataset: [IndicVoices](https://huggingface.co/datasets/ai4bharat/IndicVoices)24* WER: 36%25  26## 🧪 Local Development27 281. **Install dependencies**29 30```bash31pip install -r requirements.txt32```33 342. **Run the app locally**35 36```bash37uvicorn app.main:app --reload --host 0.0.0.0 --port 850138```39 40Open [http://localhost:8501](http://localhost:8501) in your browser.41 42[https://k2-fsa.github.io/sherpa/ncnn/endpoint.html](https://k2-fsa.github.io/sherpa/ncnn/endpoint.html)43 44## 📁 Project Structure45 46```47.48├── app49│   ├── main.py50│   ├── asr.py51│   └── model parts52        └── All Model parts here (encoder, decoder, joiner, tokens)53    ├── index.html54├── requirements.txt55├── Dockerfile56└── README.md57```58 59## 🔧 Credits60 61* [Sherpa-ONNX](https://github.com/k2-fsa/sherpa-onnx)62* [OpenCC](https://github.com/BYVoid/OpenCC)63* [FastAPI](https://fastapi.tiangolo.com/)64* [Hugging Face Spaces](https://huggingface.co/docs/hub/spaces)65* [AI4Bharat](https://ai4bharat.iitm.ac.in/)66* [Icefall-K2](https://github.com/k2-fsa/icefall)67 68 69Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference