Edit Readme

This commit is contained in:
Henry Hofmann 2025-09-18 13:40:22 +02:00
parent 455eac542c
commit aebde64c3b

174
README.md
View File

@ -1,13 +1,167 @@
# DOMYWO
A fully automated Video Generation Tool using Text Generation, Image Generation, Video Generation and a VLM to create Social Media Shorts from content with a colloquial sentence and a click. # 🌿DOMYWO
*A fully automated Video Generation Tool using Text Generation, Image Generation, Video Generation and a VLM to create Social Media Shorts from content with a colloquial sentence and a click*
## Features ---
- Text generation
- Image generation
- Video generation
- VLM integration
- Oneclick social media sharing
## Getting Started ## 🚀 Overview
...
DOMYWO is a onestop, endtoend pipeline that turns raw text into engaging short videos ready for TikTok, Reels, Shorts, etc.
With a single command you can:
| Step | What happens | Tech |
|------|--------------|------|
| 1⃣ | **Text***Story* | GPT4 / LLM |
| 2⃣ | *Story***Images** | Stable Diffusion / SDXL |
| 3⃣ | **Images** + *Story**Video* | FFmpeg + SRT / VTT |
| 4⃣ | Add a **colloquial sentence** as a caption | BLIP / LLaVA (VLM) |
| 5⃣ | **Export** as MP4 / WebM | FFmpeg |
> “Create a viral short in 60s, no editing skills required.”
---
## 🎯 Why DOMYWO?
| Problem | DOMYWOs Solution |
|---------|-------------------|
| Content creators spend hours writing scripts → **DOMYWO** autowrites story outlines. |
| Finding royaltyfree images is tedious → **DOMYWO** generates custom visuals on demand. |
| Video editing is timeconsuming → **DOMYWO** stitches everything automatically. |
| Captions that match tone & context are hard → **DOMYWO** injects colloquial captions via a VLM. |
| Need a quick shareable clip → **DOMYWO** outputs readytopost videos. |
---
## 🛠️ Tech Stack
| Layer | Technology | Reason |
|-------|------------|--------|
| **Text Generation** | OpenAI GPT4 (or local LLM) | Natural language scripting |
| **Image Generation** | Stable Diffusion 1.5 / 2.1 / SDXL | Highquality visuals |
| **Video Assembly** | FFmpeg + Python `moviepy` | Robust, crossplatform media processing |
| **VLM Captioning** | BLIP2 / LLaVA | Contextaware subtitle generation |
| **CLI** | Click / Typer | Userfriendly command line |
| **Containerization** | Docker | Reproducible environment |
| **CI** | GitHub Actions | Automated tests & release |
---
## 📦 Installation
> ⚡ **Prerequisites**
> • Python3.10+ (tested on 3.103.12)
> • `git`
### 1. Clone & Setup
```bash
git clone https://gitea.zephyre.one/Pakobbix/DOMYWO.git
cd DOMYWO
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -e .[dev] # Installs core + dev dependencies
```
---
## 🎬 Usage
### Basic OneCommand Demo
```bash
domywo generate \
--title "The Future of Remote Work" \
--prompt "Explain why remote work is here to stay" \
--output ./output/future-remote-work.mp4
```
#### Flags
| Flag | Description | Default |
|------|-------------|---------|
| `--title` | Title of the video | Autoderived |
| `--prompt` | Text prompt for generation | None (interactive) |
| `--style` | Image style (e.g., “minimalist”, “illustration”) | “photo” |
| `--length` | Video length in seconds (max 120) | 60 |
| `--output` | Output path | `./video.mp4` |
| `--offline` | Use local models only | `False` |
### Interactive Mode
If you omit `--prompt`, the CLI will ask you for a prompt:
```bash
$ domywo generate
Enter a prompt (or type 'exit'): The benefits of mindfulness for students
```
### Batch Mode
Create a `prompts.json`:
```json
[
{"title": "Healthy Snacks", "prompt": "List 5 quick healthy snacks"},
{"title": "Morning Routine", "prompt": "Explain a 10min morning routine"}
]
```
Run:
```bash
domywo batch --file prompts.json --output-dir ./videos/
```
---
## 🤝 Contributing
We ❤️ community contributions!
1. Fork the repo.
2. Create a branch: `git checkout -b feature/myawesomefeature`.
3. Run tests: `pytest`.
4. Commit and push.
5. Open a Pull Request → describe your changes clearly.
### Coding Guidelines
- Follow PEP8 (flake8).
- Write docstrings for public functions.
- Add tests for new features.
### Reporting Issues
- Use the GitHub Issues tracker.
- Include reproducible steps, logs, and any relevant screenshots.
---
## 📄 License
MIT © 2025 Pakobbix feel free to use, modify, and distribute.
---
## 📌 Roadmap (v1.0)
| Feature | Status |
|---------|--------|
| 🔜 CLI wrapper | 🚧 |
| 🔜 Text → Image → Video pipeline | 🚧 |
| 🔜 VLM captioning | 🚧 |
| 🔜 Support for 4K output | 🚧 |
| 🔜 UI dashboard | 🚧 |
| 🔜 Automatic subtitle sync | 🚧 |
---
## 📞 Get in Touch
- **Issue tracker**: <https://gitea.zephyre.one/Pakobbix/DOMYWO/issues>
- **Email**: pakobbix@example.com
---
Happy creating! 🚀