substrate-ai / substrate

Substrate AI

Substrate AI

Deploy your ML code in the cloud in under 5 minutes

Hosted Platform • Running Locally

---

Simplify your ML training on cloud GPUs with Substrate AI. Our platform allows you to set up and run your models quickly, bypassing the usual setup hassles.

🚀 Getting Started

Installation

Install the package:

pip install substrate-ai

Configure your folder to use substrate

substrate-ai init

Run your code on the cloud

substrate-ai run

🎬 Demo

View Demo

Initialize your folder to be able to run with Substrate

substrate-ai init

Create a file with some Python code

print("Welcome to substrate!")

modify your substrate.yaml if necessary

project_name: Example Project
hardware:
type: cpu
main_file_location: ./main.py

Then run your code

substrate-ai run

The result

substrateai.mp4

✨ Features

• Fully open-source
  • Self-host our platform
• Quick deployment of ML code in the cloud
• No vendor lock-in
• AWS-powered GPUs without extra cost

More features on the horizon

💡 Examples

Explore example projects in our example folder

📚 Documentation

Access our documentation

🗺️ Roadmap

• v0 launched
• Filesystem integration
• Remote image building
• Cloud job analytics
• Inference support
• Spot training
• Decreased compute cost (as we do not profit from computing, our focus is minimizing expenses)

🖥 Running Locally

Stay tuned for detailed instructions

🌟 Interested?

Please leave us a star and/or sign up for the hosted version