Manual Installation & Python Setup

If you wish to develop custom engines, contribute to OneASR, or run natively on Apple Silicon Mac without Docker, follow this setup guide.

---

1. Prerequisites

• Python: 3.10 or 3.11 (Python 3.12+ support is in progress)
• FFmpeg: Required for audio decoding and resampling (brew install ffmpeg on macOS, apt install ffmpeg on Ubuntu)
• Git & Git LFS: For repository and model weights cloning

---

2. Clone & Environment Setup

# 1. Clone repository
git clone https://github.com/codeAndxv/OneASR.git
cd OneASR

# 2. Create Python virtual environment
python3 -m venv .venv
source .venv/bin/activate

# 3. Upgrade pip and install core dependencies
pip install --upgrade pip
pip install -r requirements.txt

---

3. Download Models

Place the model weights into the models/ directory according to your enabled engines in config.yaml:

# Faster-Whisper Medium
git clone https://huggingface.co/Systran/faster-whisper-medium models/faster-whisper-medium

# X-ASR Streaming Model (Sherpa-ONNX Zipformer)
# Place encoder, decoder, joiner, tokens.txt inside models/chunk-160ms-model/

---

4. Run Server

Start the FastAPI server via Uvicorn:

uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload

Output:

INFO:     Started server process [88412]
INFO:     Waiting for application startup.
INFO:     Preloading ASR Engine: xasr
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)

---

5. Web Console Development

The embedded frontend is built with Vue 3 + Vite:

cd web
npm install
npm run dev   # Runs on http://localhost:5173 with proxy to :8000