Getting Started

This guide walks you through installation, configuration, and your first interaction with ASPIRE.

Requirements

• Python 3.10+
• PyTorch 2.0+
• CUDA GPU with 16GB+ VRAM recommended (training is GPU-intensive)
• Anthropic API key for the Claude teacher, or an OpenAI API key for GPT-4 teachers

Windows is fully supported, including RTX 5080 / Blackwell GPUs.

Installation

Clone the repository and install in editable mode:

git clone https://github.com/mcp-tool-shop-org/aspire-si.git
cd aspire-si
pip install -e .

Set your API key

ASPIRE reads API keys from environment variables. Set the one that matches the teacher you plan to use.

# Windows
set ANTHROPIC_API_KEY=your-key-here

# Linux / macOS
export ANTHROPIC_API_KEY=your-key-here

You can also use OPENAI_API_KEY if you prefer GPT-4 as your teacher model.

Verify your environment

The doctor command checks Python version, CUDA availability, API key presence, and dependency health:

aspire doctor

If anything is missing the output will tell you exactly what to fix.

Try it out

List available teachers

aspire teachers

This prints every built-in teacher persona with a short description of its philosophy and what kind of thinking it produces.

Generate an adversarial dialogue

aspire dialogue "Explain why recursion works" --teacher socratic --turns 3

The student generates a response. The Socratic teacher challenges it. They go back and forth for three turns, with each round pushing the student toward deeper, clearer reasoning.

Initialize a training config

aspire init --output my-config.yaml

This creates a configuration file with sensible defaults. Edit it to set your model, teacher, dataset, and training hyperparameters before running aspire train.

Next steps

• Read How It Works to understand the four-stage pipeline.
• Explore Teachers to learn about each persona and composite strategies.
• See the full CLI Reference for every command and flag.