Quick Start Guide¶

Get your first OpenGPU task up and running in just a few minutes! This guide will walk you through creating, deploying, and running a simple task.

🎯 What We'll Build¶

We'll create a simple task that multiplies two numbers and deploy it on the OpenGPU network.

📋 Prerequisites¶

Before we start, make sure you have:

✅ OpenGPU SDK installed (Installation Guide)
✅ Python 3.8+ running
✅ Text editor or IDE ready

🏗️ Step 1: Create Your Task Handler¶

Create a new file called my_first_task.py:

import ogpu.service
from pydantic import BaseModel

# Define input and output data models
class MultiplyInput(BaseModel):
    a: int
    b: int

class MultiplyOutput(BaseModel):
    result: int
    message: str

# Create your task handler
@ogpu.service.expose()
def multiply_numbers(input_data: MultiplyInput) -> MultiplyOutput:
    """
    A simple task that multiplies two numbers.
    """
    result = input_data.a * input_data.b
    message = f"{input_data.a} × {input_data.b} = {result}"

    return MultiplyOutput(
        result=result,
        message=message
    )

if __name__ == "__main__":
    print("🚀 Starting OpenGPU service...")
    ogpu.service.start()

What's Happening Here?¶

Import: We import the OpenGPU service module and Pydantic for data models
Models: We define input and output data structures using Pydantic BaseModel
Decorator: The @ogpu.service.expose() decorator registers our function as a task
Function: Our task logic - simple multiplication with a descriptive message
Serve: ogpu.service.start() starts the service and makes it available on the network

🚀 Step 2: Start Your Service¶

Run your service to make it available on the OpenGPU network:

python my_first_task.py

Keep It Running

Leave this terminal window open. Your service needs to be running to receive and process tasks.

🌐 Quick Test with REST API Documentation¶

Once your service is running, you can immediately test it using the built-in API documentation:

Open your browser and navigate to: http://localhost:5555/docs
Explore the API - You'll see an interactive Swagger UI with your task endpoints
Test directly - Click on your multiply_numbers endpoint and use the "Try it out" button
Send a request - Use the example payload:
```
{
  "a": 15,
  "b": 7
}
```

Interactive Testing

The /docs endpoint provides a complete interactive interface where you can test your tasks without writing any client code!

🎯 What's Next?¶

Congratulations! You've successfully:

✅ Created your first OpenGPU task
✅ Published and retrieved task results

Real-World Examples¶

Try building these next:

Image Processing: Resize, filter, or analyze images
Data Analysis: Process CSV files or perform calculations
AI Models: Deploy machine learning inference tasks
Web Scraping: Extract and process web data

🆘 Need Help?¶

Documentation: You're reading it! Check other sections for advanced topics
GitHub Issues: Report bugs or ask questions
Community: Join discussions and get help from other developers

Happy building with OpenGPU! 🚀