An open framework for building recursive self-improving AI systems
NanoASI is a user-friendly framework for building recursive self-improving AI systems. It combines advanced MCTS (Monte Carlo Tree Search), consciousness flow tracking, and parallel universe exploration with an emphasis on accessibility and ease of use.
- 🎯 Simple to Start: One line to install, one line to run
- 🧩 Modular Design: Use only what you need, extend what you want
- 🔄 Self-Improving: Continuous learning and optimization
- 🌐 Universal Compatibility: Works everywhere - local, cloud, or notebook
- 🎨 Beautiful Interface: Built-in web UI for visualization and control
pip install nano_asifrom nano_asi import ASI
# Initialize and run with defaults
asi = ASI()
result = await asi.run("Generate an innovative solution for climate change")
print(result.solution)Run nano-asi directly in Google Colab:
# Run setup script
!wget https://raw.githubusercontent.com/TimeLordRaps/nano_asi/main/colab_setup.py
!python colab_setup.py
# Now you can use nano-asi
from nano_asi import ASI
import asyncio
import nest_asyncio
nest_asyncio.apply()
async def main():
asi = ASI()
result = await asi.run("Your task here")
print(result.solution)
asyncio.run(main())from nano_asi import ASI, Config
from nano_asi.modules import ConsciousnessTracker, LoRAGenerator
# Customize components
config = Config(
consciousness_tracking=True,
parallel_universes=3,
mcts_exploration_weight=1.5
)
# Initialize with custom components
asi = ASI(
config=config,
consciousness_tracker=ConsciousnessTracker(),
lora_generator=LoRAGenerator()
)
# Run with all features enabled
result = await asi.run(
task="Generate an innovative solution for climate change",
dataset="environmental_papers",
stream=True # Enable real-time output
)
# Access comprehensive results
print(result.solution)
print(result.consciousness_flow)
print(result.universe_explorations)
print(result.metrics)from nano_asi.web import launch_ui
# Start the web interface
launch_ui()- 🤖 ASI Class: Main interface for basic usage
- 🌐 Web UI: Visual interface for interaction
- 📊 Basic Metrics: Simple performance tracking
- 🎯 MCTS Engine: Guided exploration and optimization
- 🧬 LoRA Diffusion: Model adaptation and fine-tuning
- 📈 Judgment System: Multi-level evaluation framework
- 🌌 Universe Explorer: Parallel optimization strategies
- 🧠 Consciousness Tracker: Deep pattern analysis
- 🔧 Custom Components: Build your own modules
- Quick Start Guide - Get up and running in minutes
- Tutorials - Step-by-step guides with notebooks
- API Reference - Detailed technical documentation with examples
- Advanced Topics - Deep dives into theory and implementation
- Best Practices - Guidelines and recommendations
- Troubleshooting - Common issues and solutions
# Install development dependencies
pip install nano_asi[dev]
# Run tests
pytest
# Build documentation
pip install nano_asi[docs]
cd docs && make htmlWe welcome contributions! See our Contributing Guide for details.
GPL-3.0 License - see LICENSE for details.
🐦 Twitter/X: @TimeLordRaps