Spaces:
Sleeping
Sleeping
File size: 3,243 Bytes
102ae18 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 |
# ConGr Visualizer Web Interface
A web-based interface for exploring and visualizing ConGrs (Consensus Graphs) from research datasets.
## Features
### Browse Existing Graphs
- **Dataset Selection**: Choose from available datasets (BIO, FP, HIST, REFS, MATH, AIME)
- **Entity Selection**: Browse entities within each dataset
- **Model Information**: See which language models were used for each graph
- **Graph Visualization**: Interactive network visualization using vis.js
- **Metadata Display**: View graph statistics and consensus text
### Create New Graphs
- **Text Input**: Enter multiple text sequences to create new ConGrs
- **Real-time Visualization**: See the graph structure as it's created
- **Save Functionality**: Save created graphs to pickle files
## Available Datasets
- **BIO**: Biography datasets with various public figures
- **FP**: False Presupposition datasets
- **HIST**: Historical events datasets
- **REFS**: Reference datasets
- **MATH**: Mathematical problem datasets
- **AIME**: American Invitational Mathematics Examination datasets
## Models
The graphs are generated using various language models:
- olmo7b
- qwen72b
- llama70b
- llama8b
## Installation
1. Install dependencies:
```bash
pip install -r requirements.txt
```
2. Start the server:
```bash
python server.py
```
3. Open your browser and navigate to:
```
http://localhost:8080
```
## Usage
### Browsing Existing Graphs
1. **Select Dataset**: Choose a dataset from the dropdown menu
2. **Select Entity**: Choose an entity from the available options
3. **View Graph**: The graph will be automatically loaded and displayed
4. **View Information**: Graph metadata and consensus text will be shown
### Creating New Graphs
1. **Enter Text**: Input multiple text sequences (one per line)
2. **Create Graph**: Click "Create Graph" to generate a new ConGr
3. **Save Graph**: Optionally save the graph to a pickle file
## API Endpoints
- `GET /api/datasets` - Get available datasets
- `GET /api/entities?dataset=<dataset>` - Get entities for a dataset
- `POST /api/load_existing_graph` - Load an existing graph
- `POST /api/create_graph` - Create a new graph from text sequences
- `POST /api/save_graph` - Save a graph to file
## Testing
Run the test script to verify the server is working correctly:
```bash
python test_server.py
```
## File Structure
```
web_interface/
├── server.py # Flask server
├── index.html # Web interface
├── requirements.txt # Python dependencies
├── test_server.py # Test script
└── README.md # This file
```
## Graph Information
When viewing a graph, you can see:
- **Dataset**: The source dataset
- **Entity**: The specific entity or topic
- **Model**: The language model used
- **Sequences**: Number of input sequences
- **Nodes**: Number of nodes in the graph
- **Edges**: Number of edges in the graph
- **Consensus**: The consensus text generated from the graph
## Visualization Features
- **Hierarchical Layout**: Graphs are displayed in a hierarchical structure
- **Color Coding**: Consensus nodes are highlighted in green
- **Interactive**: Zoom, pan, and hover for more information
- **Responsive**: Works on desktop and mobile devices
|