Documentation
DeAI Architecture
Learn how Axionax integrates Python-based AI models directly into the consensus layer.
Python DeAI Layer
Validators run a dedicated Python sidecar process that executes AI models. This allows for seamless integration with PyTorch, TensorFlow, and Hugging Face libraries.
from axionax.deai import Validator
class MyWorker(Validator): ...
class MyWorker(Validator): ...
Proof of Probabilistic Checking
Unlike traditional PoS, PoPC verifies AI outputs using statistical sampling. Validators are scored based on the accuracy and latency of their inference results.
- Confidence Score > 99.9%
- Slashing for malicious results
Getting Started
01
Install the SDK
$ pip install axionax-deai
02
Initialize Worker
$ axionax-cli init --worker my-ai-node
03
Deploy Model
$ axionax-cli deploy ./model.pt --network testnet