[{"_1":2,"_170":-5,"_171":-5},"loaderData",{"_3":4,"_13":14,"_21":22,"_144":145},"root",{"_5":-5,"_6":-7,"_7":8,"_9":-7,"_10":11,"_12":-7},"currentUser","primaryEmail","siteVerification","6LeIxAcTAAAAAJcZVRqyHh71UMIEGNQ_MXjiZKhI","toast","enableDev",false,"incidentStatus","routes/_layout",{"_15":-5,"_16":17,"_18":17,"_19":11,"_20":11},"username","isAskGordonEnabled",true,"isUsingNewTemplate","showMinimalLanding","showHoneypotTrap","routes/_layout.r.$namespace.$reponame",{"_23":24,"_25":-5,"_26":11,"_27":11,"_28":29,"_30":31,"_32":-5,"_33":11,"_34":35,"_88":-5,"_89":90,"_135":-5,"_136":11,"_137":17,"_138":11,"_139":-5,"_140":11,"_141":-5,"_142":143},"canonicalUrl","https://hub.docker.com/r/agoddrie/srtgen/tags","mcpServer","canSeeDockerfile","canSeeAutobuilds","namespace","agoddrie","reponame","srtgen","cloudBuildSource","isAutobuildsAvailable","repo",{"_36":37,"_43":44,"_46":47,"_48":49,"_50":11,"_51":52,"_53":54,"_55":56,"_39":31,"_28":29,"_58":59,"_60":45,"_61":62,"_63":64,"_65":66,"_67":-5,"_68":62,"_69":70,"_71":11,"_72":29,"_73":74,"_79":11,"_80":81,"_85":29,"_86":87},"categories",[38],{"_39":40,"_41":42},"name","Machine learning & AI","slug","machine-learning-and-ai","content_types",[45],"image","date_registered","2025-11-21T23:17:22.254994Z","description","AI-powered subtitle generator and translator with NVIDIA GPU acceleration\n","is_private","last_updated","2025-11-27T08:16:11.19498Z","last_modified","2025-11-23T21:12:37.405578Z","media_types",[57],"application/vnd.docker.container.image.v1+json","pull_count",1699,"repository_type","star_count",0,"status",1,"status_description","active","source","collaborator_count","full_description","# SRTGEN - AI-Powered Subtitle Generator\n\n

\n\n![SRTGEN Logo](static/favicon.svg)\n\n**GPU-accelerated subtitle generation for video files using OpenAI Whisper AI and NLLB translation**\n\n[![Docker Hub](https://img.shields.io/docker/pulls/agoddrie/srtgen?style=flat-square)](https://hub.docker.com/r/agoddrie/srtgen)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg?style=flat-square)](LICENSE)\n[![Docker Image Size](https://img.shields.io/docker/image-size/agoddrie/srtgen/latest?style=flat-square)](https://hub.docker.com/r/agoddrie/srtgen)\n\n[Features](#-features) • [Quick Start](#-quick-start) • [Usage](#-usage) • [Configuration](#-configuration) • [Docker Hub](https://hub.docker.com/r/agoddrie/srtgen)\n\n\n\n

\n\n---\n\n## ✨ Features\n\n- 🎯 **AI-Powered Transcription** - OpenAI Whisper with GPU acceleration\n- 🌍 **Multi-Language Translation** - Facebook's NLLB-200-1.3B for 200+ languages\n- 🎬 **Triple SRT Output** - Original language, English, and target language\n- 📝 **Word-Level Timestamps** - Precise subtitle timing synchronization\n- 🌐 **Modern Web UI** - Browse files, monitor jobs, configure settings\n- 🚀 **Background Processing** - Queue multiple transcription jobs\n- 🔍 **Smart File Browser** - Search, filter, path memory, existing SRT detection\n- 🐳 **Docker Ready** - One-click deployment on Unraid\n- 💾 **Persistent Settings** - Language preferences saved in browser\n- ⚡ **GPU Optimized** - CUDA acceleration for 10x faster processing\n\n## 🚀 Quick Start\n\n### Docker Run (Fastest)\n\n```bash\ndocker run -d --name srtgen \\\n --gpus all \\\n -p 5000:5000 \\\n -v /path/to/media:/media \\\n -e WHISPER_MODEL=medium \\\n agoddrie/srtgen:latest\n```\n\n**Access:** `http://localhost:5000`\n\n### Docker Compose (Recommended)\n\nCreate `docker-compose.yml`:\n\n```yaml\nversion: '3.8'\nservices:\n srtgen:\n image: agoddrie/srtgen:latest\n container_name: srtgen\n ports:\n - \"5000:5000\"\n volumes:\n - /mnt/user/TV-Series:/media/tv:rw\n - /mnt/user/Movies:/media/movies:rw\n environment:\n - WHISPER_MODEL=medium\n deploy:\n resources:\n reservations:\n devices:\n - driver: nvidia\n count: all\n capabilities: [gpu]\n restart: unless-stopped\n```\n\nStart:\n```bash\ndocker-compose up -d\n```\n\n### First Run Setup\n\nInstall translation dependencies (one-time):\n```bash\ndocker exec -it srtgen pip install transformers sentencepiece protobuf sacremoses\n```\n\n### Unraid Template\n\n1. Go to **Docker** tab → **Add Container**\n2. Click **Template repositories** → Add: `https://github.com/agoddrie/srtgen`\n3. Search for **SRTGEN**\n4. Configure paths and click **Apply**\n\n## 📋 Requirements\n\n- **NVIDIA GPU** (recommended for speed)\n- Docker with NVIDIA Container Runtime\n- 8GB+ VRAM for medium Whisper model + NLLB-1.3B\n\n## 🎮 Usage\n\n### Web Interface (Recommended)\n\n1. **Open browser:** `http://localhost:5000`\n2. **Browse media folders** or use search\n3. **Select video file** (MKV, MP4, AVI)\n4. **Choose language** or enable auto-detect\n5. **Configure overwrite** settings\n6. **Click \"Generate Subtitles\"**\n7. **Monitor progress** in sidebar\n\n### Output Files\n\nFor `movie.mkv`, generates:\n```\nmovie.en.srt # Original language transcription\nmovie.en.srt # English translation (if not English)\nmovie.nl.srt # Target language (e.g., Dutch)\n```\n\n### Supported Formats\n\n- **Video:** `.mkv`, `.mp4`, `.avi`\n- **Languages:** 99+ via Whisper (EN, NL, FR, DE, ES, IT, PT, PL, RU, JA, ZH, KO, AR, TR, etc.)\n- **Translation:** 200+ via NLLB-200\n\n## ⚙️ Configuration\n\n### Whisper Model Selection\n\nSet via environment variable:\n\n| Model | VRAM | Speed | Accuracy | Recommended |\n|-------|------|-------|----------|-------------|\n| `tiny` | ~1GB | Fastest | Good | Testing |\n| `base` | ~1GB | Very Fast | Better | Quick jobs |\n| `small` | ~2GB | Fast | Good | Balanced |\n| `medium` | ~5GB | Moderate | Excellent | ⭐ **Production** |\n| `large` | ~10GB | Slow | Best | Maximum quality |\n\n**Example:**\n```yaml\nenvironment:\n - WHISPER_MODEL=medium\n```\n\n### Settings Modal\n\nAccess via ⚙️ icon in web UI:\n- **Default Language** - Preset translation target\n- Settings persist in browser localStorage\n\n### Volume Mapping\n\n```yaml\nvolumes:\n - /host/path:/media/name:rw # Read-write required for SRT creation\n```\n\n**Important:** Use `:rw` flag to allow subtitle file creation\n\n## 🛠️ Technology Stack\n\n| Component | Technology | Purpose |\n|-----------|-----------|---------|\n| **AI Transcription** | OpenAI Whisper | Speech-to-text with 99+ languages |\n| **AI Translation** | Facebook NLLB-200-1.3B | Context-aware translation (200+ languages) |\n| **Backend** | Python 3.11 + Flask | REST API and job management |\n| **GPU Acceleration** | PyTorch + CUDA | 10x faster processing |\n| **Audio Processing** | FFmpeg | Extract audio from video containers |\n| **Frontend** | Vanilla JavaScript | Lightweight, no dependencies |\n| **Storage** | localStorage | Settings persistence |\n\n### How It Works\n\n```mermaid\ngraph LR\n A[Video File] -->|FFmpeg| B[Audio 16kHz WAV]\n B -->|Whisper| C[Original SRT + Language]\n C -->|Whisper Translate| D[English SRT]\n C -->|NLLB-200| E[Target Language SRT]\n D --> F[Triple SRT Output]\n E --> F\n```\n\n1. **Audio Extraction:** FFmpeg → 16kHz mono WAV in `/tmp`\n2. **Transcription:** Whisper with word-level timestamps + auto language detection\n3. **English Translation:** Whisper's built-in translation task\n4. **Target Translation:** NLLB batch processing (5 segments) with context\n5. **Timestamp Adjustment:** Extends duration for longer translations\n6. **SRT Generation:** Filters non-speech, formats timestamps, saves files\n\n## 🐛 Troubleshooting\n\n

Translation module not found

\n\n```bash\ndocker exec -it srtgen pip install transformers sentencepiece protobuf sacremoses\n```\n*Note: Required on first run. Will be added to Dockerfile in future release.*\n

\n\n

GPU not detected

\n\nVerify NVIDIA runtime:\n```bash\ndocker run --rm --gpus all nvidia/cuda:11.8.0-base-ubuntu22.04 nvidia-smi\n```\n\nInstall nvidia-container-toolkit:\n```bash\ndistribution=$(. /etc/os-release;echo $ID$VERSION_ID)\ncurl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -\ncurl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list\nsudo apt-get update && sudo apt-get install -y nvidia-container-toolkit\nsudo systemctl restart docker\n```\n

\n\n

Template changes not visible

\n\n1. Hard refresh browser: `Ctrl+Shift+R`\n2. Restart container: `docker restart srtgen`\n3. Clear browser cache\n

\n\n

Permission denied writing SRT

\n\nEnsure volumes use `:rw` flag:\n```yaml\nvolumes:\n - /path/to/media:/media:rw # Not :ro\n```\n

\n\n

Subtitle timing off by ~1 second

\n\nWord-level timestamps enabled by default. If issues persist:\n- Try different Whisper model\n- Check video/audio sync in source file\n- Report issue with example file\n

\n\n## 📦 Building from Source\n\n```bash\ngit clone https://github.com/agoddrie/srtgen.git\ncd srtgen\ndocker build -t agoddrie/srtgen:latest .\ndocker-compose up -d\n```\n\n## 🤝 Contributing\n\nContributions welcome! Please:\n\n1. Fork the repository\n2. Create feature branch (`git checkout -b feature/amazing`)\n3. Commit changes (`git commit -m 'Add amazing feature'`)\n4. Push to branch (`git push origin feature/amazing`)\n5. Open Pull Request\n\n### Development Setup\n\n```bash\n# Clone repo\ngit clone https://github.com/agoddrie/srtgen.git\ncd srtgen\n\n# Build image\ndocker build -t srtgen:dev .\n\n# Run with hot-reload\ndocker run -it --rm \\\n --gpus all \\\n -p 5000:5000 \\\n -v $(pwd)/app.py:/app/app.py \\\n -v $(pwd)/mkv_transcribe.py:/app/mkv_transcribe.py \\\n -v /media:/media:rw \\\n srtgen:dev\n```\n\n## 📊 Benchmarks\n\nTest file: 22-minute episode (The IT Crowd S01E01)\n\n| Model | GPU | Time | Accuracy | VRAM |\n|-------|-----|------|----------|------|\n| tiny | RTX 5070 Ti | 45s | 85% | 1.2GB |\n| base | RTX 5070 Ti | 1m 15s | 92% | 1.5GB |\n| small | RTX 5070 Ti | 2m 30s | 96% | 2.8GB |\n| **medium** | **RTX 5070 Ti** | **4m 20s** | **98%** | **5.2GB** |\n| large | RTX 5070 Ti | 8m 45s | 99% | 10.5GB |\n\n*Translation adds ~2-3 minutes for NLLB-1.3B processing*\n\n## 📄 License\n\nMIT License - See [LICENSE](LICENSE) file\n\n**Third-party components:**\n- OpenAI Whisper - [MIT License](https://github.com/openai/whisper/blob/main/LICENSE)\n- Facebook NLLB - [CC BY-NC 4.0](https://github.com/facebookresearch/fairseq/blob/nllb/LICENSE)\n- FFmpeg - [GPL/LGPL](https://ffmpeg.org/legal.html)\n\n## 🙏 Acknowledgments\n\n- [OpenAI Whisper](https://github.com/openai/whisper) - Robust speech recognition\n- [Facebook NLLB](https://github.com/facebookresearch/fairseq/tree/nllb) - High-quality translation\n- [FFmpeg](https://ffmpeg.org/) - Multimedia framework\n- [Flask](https://flask.palletsprojects.com/) - Web framework\n- [PyTorch](https://pytorch.org/) - Deep learning platform\n\n## ⭐ Support\n\nIf you find SRTGEN useful, please:\n- ⭐ Star the repository\n- 🐛 Report bugs via [Issues](https://github.com/agoddrie/srtgen/issues)\n- 💡 Request features\n- 📖 Improve documentation\n- 🔄 Share with others\n\n---\n\n

\n\n**Made with ❤️ for the Unraid community**\n\n[Docker Hub](https://hub.docker.com/r/agoddrie/srtgen) • [GitHub](https://github.com/agoddrie/srtgen) • [Issues](https://github.com/agoddrie/srtgen/issues)\n\n

\n","has_starred","hub_user","immutable_tags_settings",{"_75":11,"_76":77},"enabled","rules",[78],".*","is_automated","permissions",{"_82":17,"_83":11,"_84":11},"read","write","admin","user","storage_size",22125568161,"orgData","recentTags",[91,129],{"_92":45,"_93":94,"_95":96,"_97":98,"_99":100,"_101":102,"_51":118,"_119":94,"_120":29,"_121":57,"_39":122,"_123":124,"_125":109,"_126":118,"_127":66,"_128":17},"content_type","creator",13906668,"digest","sha256:df8a96ed035be55ec509ddd988430a37249ebea42eee5bb636bdcd072fa38417","full_size",4569081524,"id",1016398815,"images",[103],{"_104":105,"_95":96,"_106":107,"_108":109,"_110":111,"_112":113,"_114":107,"_115":-5,"_116":98,"_63":66,"_117":-5},"architecture","amd64","features","","last_pulled","2026-09-27T07:32:55.921280688Z","last_pushed","2025-11-27T08:16:05.555438089Z","os","linux","os_features","os_version","size","variant","2025-11-27T08:16:05.987982Z","last_updater","last_updater_username","media_type","latest","repository",29227430,"tag_last_pulled","tag_last_pushed","tag_status","v2",{"_92":45,"_93":94,"_95":96,"_97":98,"_99":130,"_101":131,"_51":133,"_119":94,"_120":29,"_121":57,"_39":134,"_123":124,"_125":109,"_126":133,"_127":66,"_128":17},1019704084,[132],{"_104":105,"_95":96,"_106":107,"_108":109,"_110":111,"_112":113,"_114":107,"_115":-5,"_116":98,"_63":66,"_117":-5},"2025-11-27T08:16:10.144545Z","1.3.0","sbxKitMetadata","isPotentialSpam","isHubSbxKitsEnabled","showSbxLogoOverride","sbxTrademarkText","displaySecurityMetadata","tagAttestations","tagsHaveAttestations",[],"routes/_layout.r.$namespace.$reponame.tags",{"_34":146,"_154":155,"_142":168,"_169":11},{"_36":147,"_43":149,"_46":47,"_48":49,"_50":11,"_51":52,"_53":54,"_55":150,"_39":31,"_28":29,"_58":59,"_60":45,"_61":62,"_63":64,"_65":66,"_67":-5,"_68":62,"_69":70,"_71":11,"_72":29,"_73":151,"_79":11,"_80":153,"_85":29,"_86":87},[148],{"_39":40,"_41":42},[45],[57],{"_75":11,"_76":152},[78],{"_82":17,"_83":11,"_84":11},"tags",{"_156":157,"_158":-5,"_159":-5,"_160":161},"count",2,"next","previous","results",[162,165],{"_92":45,"_93":94,"_95":96,"_97":98,"_99":130,"_101":163,"_51":133,"_119":94,"_120":29,"_121":57,"_39":134,"_123":124,"_125":109,"_126":133,"_127":66,"_128":17},[164],{"_104":105,"_95":96,"_106":107,"_108":109,"_110":111,"_112":113,"_114":107,"_115":-5,"_116":98,"_63":66,"_117":-5},{"_92":45,"_93":94,"_95":96,"_97":98,"_99":100,"_101":166,"_51":118,"_119":94,"_120":29,"_121":57,"_39":122,"_123":124,"_125":109,"_126":118,"_127":66,"_128":17},[167],{"_104":105,"_95":96,"_106":107,"_108":109,"_110":111,"_112":113,"_114":107,"_115":-5,"_116":98,"_63":66,"_117":-5},[],"canDeleteTags","actionData","errors"]

agoddrie/srtgen

Tags for agoddrie/srtgen