🎬 Transcriptarr

AI-powered subtitle transcription service with REST API and Web UI

Transcriptarr is an AI-powered subtitle transcription service based on Subgen, featuring a modern FastAPI backend with 45+ REST endpoints, a Vue 3 web interface, and a distributed worker pool architecture.

---

✨ Features

🎯 Core Features

• Whisper Transcription - Support for faster-whisper and stable-ts
• Translation - Two-stage translation: Whisper to English, then Google Translate to target language
• CPU/GPU Workers - Scalable worker pool with CUDA support
• Persistent Queue - Priority-based queue manager with SQLite/PostgreSQL
• Library Scanner - Automatic scanning with configurable rules
• REST API - 45+ endpoints with FastAPI
• Web UI - Complete Vue 3 dashboard with 6 views
• Setup Wizard - Interactive first-run configuration
• Real-time Monitoring - File watcher, scheduled scans, and system resources

🔧 Technical Features

• Multiprocessing: Workers isolated in separate processes
• Priority Queuing: Queue with priorities and deduplication
• Graceful Degradation: Works without optional dependencies (Whisper, GPU)
• Thread-Safe: Row locking and context managers
• Auto-retry: Automatic retry of failed jobs
• Health Monitoring: Detailed statistics and health checks
• Database-backed Settings: All configuration stored in database, editable via Web UI

---

🚀 Quick Start

1. Install dependencies

# Basic dependencies
pip install -r requirements.txt

# Transcription dependencies (optional - required for actual transcription)
pip install stable-ts faster-whisper
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
pip install av>=10.0.0

2. First run (Setup Wizard)

# The setup wizard runs automatically on first start
python backend/cli.py server

# Or run setup wizard manually
python backend/cli.py setup

The setup wizard will guide you through:

• Standalone mode: Configure library paths, scan rules, and workers
• Bazarr mode: Configure as Bazarr subtitle provider (in development)

3. Start the server

# Development (with auto-reload)
python backend/cli.py server --reload

# Production
python backend/cli.py server --host 0.0.0.0 --port 8000 --workers 4

4. Access the application

URL	Description
http://localhost:8000	Web UI (Dashboard)
http://localhost:8000/docs	Swagger API Documentation
http://localhost:8000/redoc	ReDoc API Documentation
http://localhost:8000/health	Health Check Endpoint

---

📋 CLI Commands

# Server
python backend/cli.py server [options]
  --host HOST           Host (default: 0.0.0.0)
  --port PORT           Port (default: 8000)
  --reload              Auto-reload for development
  --workers N           Number of uvicorn workers (default: 1)
  --log-level LEVEL     Log level (default: info)

# Setup wizard
python backend/cli.py setup     # Run setup wizard

# Database
python backend/cli.py db init   # Initialize database
python backend/cli.py db reset  # Reset (WARNING: deletes all data!)

# Standalone worker
python backend/cli.py worker --type cpu
python backend/cli.py worker --type gpu --device-id 0

# Manual scan
python backend/cli.py scan /path/to/media [--no-recursive]

---

🏗️ Architecture

┌─────────────────────────────────────────────────────────┐
│                   FastAPI Server                         │
│  ┌─────────────────────────────────────────────────┐   │
│  │              REST API (45+ endpoints)            │   │
│  │  /api/workers  | /api/jobs     | /api/settings  │   │
│  │  /api/scanner  | /api/system   | /api/setup     │   │
│  └─────────────────────────────────────────────────┘   │
└──────────────────┬──────────────────────────────────────┘
                   │
    ┌──────────────┼──────────────┬──────────────────┐
    │              │              │                  │
    ▼              ▼              ▼                  ▼
┌────────┐   ┌──────────┐   ┌─────────┐      ┌──────────┐
│ Worker │   │  Queue   │   │ Scanner │      │ Database │
│  Pool  │◄──┤ Manager  │◄──┤ Engine  │      │ SQLite/  │
│ CPU/GPU│   │ Priority │   │ Rules + │      │ Postgres │
└────────┘   │  Queue   │   │ Watcher │      └──────────┘
             └──────────┘   └─────────┘

Data Flow

1. LibraryScanner detects files (manual/scheduled/watcher)
2. FileAnalyzer analyzes with ffprobe (audio tracks, subtitles)
3. Rules Engine evaluates against configurable ScanRules
4. QueueManager adds job to persistent queue (with deduplication)
5. Worker processes with WhisperTranscriber
6. Output: Generates .eng.srt (transcription) or .{lang}.srt (translation)

---

🖥️ Web UI

The Web UI includes 6 complete views:

View	Description
Dashboard	System overview, resource monitoring (CPU/RAM/GPU), recent jobs
Queue	Job management with filters, pagination, retry/cancel actions
Scanner	Scanner control, scheduler configuration, manual scan trigger
Rules	Scan rules CRUD with create/edit modal
Workers	Worker pool management, add/remove workers dynamically
Settings	Database-backed settings organized by category

---

🎛️ Configuration

Database-backed Settings

All configuration is stored in the database and manageable via:

• Setup Wizard (first run)
• Settings page in Web UI
• Settings API (/api/settings)

Settings Categories

Category	Settings
General	Operation mode, library paths, log level
Workers	CPU/GPU worker counts, auto-start, health check interval
Transcription	Whisper model, compute type, skip existing files
Scanner	Enable/disable, schedule interval, file watcher
Bazarr	Provider mode (in development)

Environment Variables

Only DATABASE_URL is required in .env:

# SQLite (default)
DATABASE_URL=sqlite:///./transcriptarr.db

# PostgreSQL (production)
DATABASE_URL=postgresql://user:pass@localhost/transcriptarr

---

📚 Documentation

Document	Description
docs/API.md	Complete REST API documentation (45+ endpoints)
docs/ARCHITECTURE.md	Backend architecture and components
docs/FRONTEND.md	Frontend structure and components
docs/CONFIGURATION.md	Configuration system and settings

---

🐳 Docker

# CPU only
docker build -t transcriptorio:cpu -f Dockerfile.cpu .

# GPU (NVIDIA CUDA)
docker build -t transcriptorio:gpu -f Dockerfile .

# Run
docker run -d \
  -p 8000:8000 \
  -v /path/to/media:/media \
  -v /path/to/data:/app/data \
  --gpus all \
  transcriptorio:gpu

---

📊 Project Status

Component	Status	Progress
Core Backend	✅ Complete	100%
REST API (45+ endpoints)	✅ Complete	100%
Worker System	✅ Complete	100%
Library Scanner	✅ Complete	100%
Web UI (6 views)	✅ Complete	100%
Settings System	✅ Complete	100%
Setup Wizard	✅ Complete	100%
Bazarr Provider	⏳ In Development	30%
Testing Suite	⏳ Pending	0%
Docker	⏳ Pending	0%

---

🤝 Contributing

Contributions are welcome!

📝 Credits

Based on Subgen by McCloudS.

Architecture redesigned with:

• FastAPI for REST APIs
• SQLAlchemy for persistence
• Multiprocessing for workers
• Whisper (stable-ts / faster-whisper) for transcription
• Vue 3 + Pinia for frontend

---

📄 License

MIT License - See LICENSE for details.