File size: 5,733 Bytes
fc8c40e |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 |
# SAP Finance Dashboard with RPT-1-OSS Model
A comprehensive financial dashboard application built with Gradio that integrates the SAP-RPT-1-OSS model for predictive analysis on SAP finance datasets. Features include synthetic data generation, interactive visualizations, OData connectivity, AI-powered insights, and a playground for training custom models.
## Features
- **Multiple Synthetic SAP Finance Datasets**: General Ledger accounts, Financial Statements (P&L, Balance Sheet), and Sales Orders
- **Data Upload**: Upload custom CSV, Parquet, or JSON datasets for analysis
- **Interactive Visualizations**: Financial charts and graphs using Plotly
- **SAP-RPT-1-OSS Model Integration**: AI-powered predictions and analysis
- **Live OData Connection**: Connect to SAP systems to fetch real-time sales order data
- **Playground Tab**: Upload datasets, configure model parameters, train, and download predictions
- **Modern UI**: Built with Gradio, a Python-based web framework
## Installation
### Prerequisites
- Python 3.11 or higher
- Hugging Face account (for model access)
- SAP OData credentials (optional, for live data connection)
### Setup Steps
1. **Clone the repository**:
```bash
git clone <repository-url>
cd SAP-RPT-1-OSS-App
```
2. **Create a virtual environment**:
```bash
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
```
3. **Install dependencies**:
```bash
pip install -r requirements.txt
```
4. **Install SAP-RPT-OSS package**:
```bash
pip install git+https://github.com/SAP-samples/sap-rpt-1-oss
```
5. **Set up environment variables**:
- Copy `.env.example` to `.env`
- Fill in your SAP OData credentials
- Add your Hugging Face token for model access
6. **Authenticate with Hugging Face**:
```bash
pip install huggingface_hub
huggingface-cli login
```
Or set the `HUGGINGFACE_TOKEN` in your `.env` file.
7. **Generate synthetic data** (optional, if not already generated):
```bash
python -c "from utils.data_generator import generate_all_datasets; generate_all_datasets()"
```
## Usage
### Running the Application
Start the Gradio application:
```bash
python app_gradio.py
```
The application will be available at `http://localhost:7862` (default Gradio port).
### Application Tabs
1. **Dashboard**: Overview with key financial metrics and visualizations
2. **Data Explorer**: Browse and filter datasets (GL, Financial Statements, Sales Orders)
3. **Upload**: Upload custom CSV datasets for analysis
4. **Predictions**: Use SAP-RPT-1-OSS model for predictions and analysis with pre-configured scenarios
5. **OData**: Connect to SAP OData services and fetch live data
6. **Playground**: Upload datasets, configure model parameters (task type, target column, test split, context size, bagging, GPU), train models, and download predictions
## SAP OData Connection Setup
1. Set the following environment variables in your `.env` file:
- `SAP_USERNAME`: Your SAP username
- `SAP_PASSWORD`: Your SAP password
- `SAP_SERVER`: SAP server URL (default: `https://sapes5.sapdevcenter.com/`)
- `SAP_CLIENT`: SAP client number (default: `002`)
2. The OData connector uses the base URL:
`https://sapes5.sapdevcenter.com/sap/opu/odata/IWBEP/GWSAMPLE_BASIC`
3. Available endpoints:
- Sales Orders: `SalesOrderSet`
- Products: `ProductSet`
- Line Items: `SalesOrderLineItemSet`
- Business Partners: `BusinessPartnerSet`
## Model Configuration
The SAP-RPT-1-OSS model supports both classification and regression tasks. For best performance:
- **Recommended**: GPU with at least 80 GB memory, context size 8192, bagging factor 8
- **Lightweight**: CPU with context size 2048, bagging factor 1
The application automatically detects available resources and adjusts settings accordingly.
## Deployment
### GitHub Pages
Since Gradio requires a Python backend, GitHub Pages deployment requires alternative approaches:
1. **GitHub Actions**: Set up a workflow to run the Gradio app as a service
2. **Alternative Hosting**: Deploy to platforms that support Python web apps:
- Hugging Face Spaces (recommended for Gradio apps)
- Streamlit Cloud
- Heroku
- Railway
- Render
### Deploy to Hugging Face Spaces
The easiest way to deploy this Gradio app is using Hugging Face Spaces:
1. Create a new Space on [Hugging Face Spaces](https://huggingface.co/spaces)
2. Select "Gradio" as the SDK
3. Push this repository to your Space
4. Add your `HUGGINGFACE_TOKEN` as a secret in Space settings
5. The app will automatically deploy
See `.github/workflows/deploy.yml` for deployment configuration examples.
## Project Structure
```
SAP-RPT-1-OSS-App/
├── app_gradio.py # Main Gradio application
├── models/
│ └── rpt_model.py # SAP-RPT-1-OSS model wrapper
├── data/
│ ├── synthetic_gl_accounts.csv
│ ├── synthetic_financial_statements.csv
│ └── synthetic_sales_orders.csv
├── utils/
│ ├── data_generator.py # Generate synthetic SAP finance data
│ ├── visualizations.py # Chart generation functions
│ ├── odata_connector.py # OData connection utilities
│ └── playground.py # Playground utilities for model training
├── requirements.txt
├── README.md
└── .env.example
```
## License
This project is licensed under the Apache Software License, version 2.0.
## Support
For issues or questions, please create an issue in this repository.
## Acknowledgments
- SAP-RPT-1-OSS model: [Hugging Face](https://huggingface.co/SAP/sap-rpt-1-oss)
- Gradio framework: [Gradio Documentation](https://www.gradio.app/docs/)
|