https://github.com/makerlinck/antllm
A auto-classifier for files based on Langchain & Deepdanbooru-tagger-mini
https://github.com/makerlinck/antllm
automatic classification computer-vision cv eagle efficiency filesystem langchain langgraph llm utility
Last synced: 5 months ago
JSON representation
A auto-classifier for files based on Langchain & Deepdanbooru-tagger-mini
- Host: GitHub
- URL: https://github.com/makerlinck/antllm
- Owner: makerlinck
- Created: 2025-03-06T08:58:13.000Z (7 months ago)
- Default Branch: master
- Last Pushed: 2025-03-25T23:02:53.000Z (7 months ago)
- Last Synced: 2025-03-26T00:22:28.796Z (7 months ago)
- Topics: automatic, classification, computer-vision, cv, eagle, efficiency, filesystem, langchain, langgraph, llm, utility
- Language: Python
- Homepage:
- Size: 14.4 MB
- Stars: 6
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README-EN.md
Awesome Lists containing this project
README
# AntLLM 🐜 **Smart File Manager**
[中文版](README.md) | [English](README.md)
[//]: # ([](https://github.com/makerlinck/AntLLM))
[](https://opensource.org/licenses/MIT)
---
## Project Overview 📌 | Overview
A smart file management system based on **⚡ FastAPI** and **Deepdanbooru-tagger-mini**, featuring:
- 🖼️ Automated media file archiving and classification
- 🖼️ Smart tracking image source, download, and classification automation
- 🏷️ Intelligent semantic tag generation
- 🚀 Coming soon: Eagle media library integration
- 🚮 Enhanced LangChain integration for extended capabilities
### Technology Stack & Architecture
Frontend-backend separation architecture:
- **Backend**: Built with FastAPI framework, integrated with lightweight image recognition model Deepmini to provide high-performance tagging services.
- **Frontend**: Provides a visual interface via Eagle media library Node.js plugins, supporting tag management and categorized browsing.
```
Future Expansion Plan: Integrate LangChain to build intelligent automated workflows,
combining multi-modal task processing for full automation from file recognition to semantic analysis and archival.
```
## Core Features 🚀 | Features
#### Key Capabilities
| **Module** | **Function Description** | **Status** |
|:----------:|-------------------|------|
| 🧠 TAG Engine | Deepmini image recognition | ✅ |
| 🏷️ Media Tagging | Deepmini tag generation | ✅ |
| ⚙ Eagle Plugin | Plugin support | ✅ |
| ⚙ Office Plugin | Plugin support | 🚧 |
| 🧠 AI Engine | LangChain multi-modal task understanding | 🚧 |
| 🖼️ File Classification | Semantic-based smart archiving | 🚧|
| 🖼️ One-click Source Search | Customized search/download/classification workflow | 🚧|
| 🖼️ Visualization UI | More user-friendly experience | 🚧|
---
## Quick Deployment ⚡ | Quick Deployment
### Environment Setup | Environment Preparation
Note: This setup is currently for development/debugging only
``` bash
Install dependencies
$ pip install -r requirements.txt
$ wget https://github.com/KichangKim/DeepDanbooru/releases/download/v4-20200814-sgd-e30/deepdanbooru-v4-20200814-sgd-e30.zip
```
Unzip the file and place contents into `./src/models/Deepmini/resnet-models`
### Basic Usage | Basic Usage
``` bash
-----------------Start Service----------------
uvicorn main:app --host --port
-----------------Run Tests----------------
$ python ./src/test_run_tensorflow.py # Test TensorFlow
$ python ./src/test_run_deepmini.py # Test Deepmini
```
Here's a Node.js example for API calls:
``` javascript
async function processChunk(uris, objs) {
try {
const response = await fetch('http://127.0.0.1:8000/api/tagger', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ tag_language: DEFAULT_LANGUAGE,query_uris: uris }),
signal: abortController.signal
});
const { response: results } = await response.json();
results.forEach((item, index) => {
objs[index].tags = item.img_tags;
objs[index].save();
addLog(`已处理: ${objs[index].name}`);
});
} catch (error) {
if (error.name !== 'AbortError') throw error;
}
}
```
---
## Directory Structure 🌳 | Directory Structure
```text
AntLLM/
├── src/ - Core source code
│ ├── api/ - API modules
│ │ └── __init__.py
│ ├── core/ - Core functionality
│ │ ├── __init__.py
│ │ └── configure.py
│ ├── models/ - Model files
│ │ ├── deepmini/
│ │ │ ├── resnet-models/ - Tag model files
│ │ │ └── __init__.py
│ ├── schemas/ - Data schemas
│ │ ├── __init__.py
│ │ └── tagger.py
│ ├── service/ - Service modules
│ │ └── __init__.py
│ ├── utils/ - Utility functions
│ │ ├── constant/
│ │ │ ├── __init__.py
│ │ │ └── file_manager.py
│ │ ├── __init__.py
│ │ └── main.py - FastAPI entry point
│ └── settings.py - Application settings
├── data/ - Data files
│ ├── tagger_model/ - Tag model related files
│ │ ├── .settings_bak.yaml
│ │ └── settings.yaml
├── docs/ - Project documentation
├── requirements.txt - Python project dependencies
└── README.md - Project overview
```
---
## Contribute The World ! 🤝 | Contribute
Join us by submitting ideas via Issues or contributing code through PRs:
1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Push to your branch
5. Open a Pull Request
We welcome all contributors to help make AntLLM better!
---
## License 📄 | License
This project is licensed under the MIT License