SEDIMARK Offering Generator

A streamlined offering generation system for the SEDIMARK infrastructure. This toolbox runs on QWEN-7B model and utilizes PEFT fine-tuning to produce SEDIMARK-compliant JSON-LD offerings. The weights for the models are here and are expected to be in the models/ directory.

Quick Start

1. Installation

# Clone the repository
git clone https://github.com/sedimark/offering-generator.git

# Clone or copy this directory
cd offering-generator

# Install dependencies
pip install -r requirements.txt

2. Model Setup

Place your PEFT checkpoint in the models/ directory. The expected structure would look something like this:

offering-generator/
└── models/
    └── checkpoint-40/
        ├── adapter_config.json
        ├── adapter_model.safetensors
        └── tokenizer.json

Update config.py if your checkpoint path differs.

3. Run the API Server

python run_api.py

The API will be available at http://localhost:8082

4. Run the Gradio Interface

python run_gradio.py

The interface will be available at http://localhost:7860

Project Structure

sedimark-clean/
├── api.py                 # FastAPI server
├── gradio_app.py         # Gradio interface
├── model_loader.py       # Model loading utilities
├── offering_generator.py # Core offering generation logic
├── config.py            # Configuration
├── run_api.py           # API runner script
├── run_gradio.py        # Gradio runner script
├── requirements.txt     # Dependencies
├── README.md           # This file
├── models/             # Model checkpoints (create this)
└── outputs/            # Generated offerings (auto-created)

Configuration

Edit config.py to customize:

• Model paths and checkpoint locations
• Generation parameters (temperature, top_p, etc.)
• Server ports and settings
• Output directories

API Endpoints

Core Endpoints

• GET / - Service information
• POST /generate - Generate offering with detailed control
• GET /api/offerings/generate - Simple generation endpoint
• GET /health - Health check
• GET /api/offerings/list - List generated offerings

Example API Usage

# Simple generation
curl "http://localhost:8082/api/offerings/generate?prompt=Generate%20IoT%20sensor%20offering"

# Advanced generation
curl -X POST "http://localhost:8082/generate" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Generate a SEDIMARK offering for smart city sensors",
    "temperature": 0.1,
    "max_new_tokens": 4096,
    "use_schema": false
  }'

Usage Examples

Generate IoT Sensor Offering

from offering_generator import SEDIMARKOfferingGenerator
from model_loader import ModelLoader

# Initialize
loader = ModelLoader()
loader.load_model()
generator = SEDIMARKOfferingGenerator(loader)

# Generate
offering = generator.generate_offering(
    prompt="Generate offering for temperature sensors in Tokyo",
    use_context=False,
    temperature=0.1
)

print(json.dumps(offering, indent=2))

Batch Generation

# Generate multiple offerings
curl "http://localhost:8082/api/offerings/generate?count=3&prompt=Smart%20city%20data"

This clean implementation provides the same powerful SEDIMARK generation capabilities with dramatically reduced complexity, making it ideal for production deployment.

License

This software is licensed under the European Union Public License 1.2 (EUPL-1.2).

This updated README reflects the new project structure, command-line options, storage choices, and includes information about the utility classes and debugging features.

Acknowledgement

This software has been developed by the University of Surrey and SAIReco under the SEDIMARK (SEcure Decentralised Intelligent Data MARKetplace) project. SEDIMARK is funded by the European Union under the Horizon Europe framework programme [grant no. 101070074]. This project is also partly funded by UK Research and Innovation (UKRI) under the UK government’s Horizon Europe funding guarantee [grant no. 10043699].