Getting Started with Trace2

Welcome! This guide will help you install, configure, and start using Trace2.

What You'll Learn

In this section, you'll learn how to:

✅ Install Trace2 and its dependencies
✅ Configure LLM backends (OpenAI, Anthropic, etc.)
✅ Build your first optimizable function
✅ Run your first optimization

Installation

Install Trace2 via pip and set up your environment

Quick Start

Build your first optimizable AI agent in 5 minutes

Configuration

Configure LLM backends and API keys

Installation (30 seconds)

pip install trace-opt

That's it! Trace2 is now installed.

Set Up Your API Key (1 minute)

import os
os.environ['OPENAI_API_KEY'] = 'your-api-key-here'

Your First Example (2 minutes)

from opto.trace import node, bundle
from opto.optimizers import OptoPrime

@bundle(trainable=True)
def greet(name):
    """Greet the user."""
    return f"Hey {name}"

optimizer = OptoPrime(greet.parameters())

# Optimize the greeting to be more professional
output = greet("Alice")
feedback = "Make it more professional and formal"

optimizer.zero_feedback()
optimizer.backward(output, feedback)
optimizer.step()

# Try again
print(greet("Alice"))  # Should be more formal now!