Skip to main content

Unified SDK experience for AILOOS federated AI nodes

Project description

AILOOS SDK

AILOOS is a robust Python library for participating in decentralized federated learning networks. This SDK enables nodes to join training sessions, contribute processing power, and interact with the Ailoos ecosystem securely.

Features

  • Federated Learning: Participate in distributed model training sessions using FedAvg and other advanced algorithms.
  • Privacy First: Secure key management and encrypted P2P communication (ECDSA + ECDH).
  • Decentralized Marketplace: Interact with the data marketplace to list or purchase datasets.
  • Hardware Integration: Smart capabilities detection for optimal workload assignment.

Node Roles

AILOOS supports 7 specialized node roles:

  1. Scout: Lightweight discovery and inference.
  2. Forge: heavy-duty model training.
  3. Validator: ZKP verification and auditing.
  4. Aggregator: Secure model aggregation.
  5. Oracle: Trusted external data provider.
  6. Archival: Long-term storage history.
  7. Relay: Cross-chain bridging.

See Full Node Roles Documentation for details.

Network Scales

AILOOS organizes nodes into 5 intelligence scales:

  1. Sovereign (El Átomo): 1 Node.
  2. Binary (Binario): 2 Nodes. (Small personal networks)
  3. Phalanx (La Escuadra): 3-20 Nodes.
  4. Foundry (La Fundición): 100-5k Nodes.
  5. Nexus (El Coloso): 100k+ Nodes.

Installation

AILOOS features a modular installation system to ensure high performance even on lightweight nodes:

# 1. Base Terminal & P2P Networking (For Scouts, Relays, Validators)
pip install ailoos

# 2. Heavy Computational Node (For Forge, requires PyTorch & Deep Learning engines)
pip install "ailoos[forge]"

Note: If you intend to train or run local inference via the ailoos-terminal, you MUST install the [forge] extension.

Initialize a Node

import asyncio
from ailoos.sdk.node_sdk import NodeSDK

async def main():
    # Initialize the node
    node = NodeSDK(node_id="my-ai-node-01")
    
    # Connect and start participating
    await node.initialize()
    await node.start()
    
    # Keep running
    while True:
        await asyncio.sleep(1)

if __name__ == "__main__":
    asyncio.run(main())

Advanced Usage

For granular control over components:

from ailoos.sdk.advanced_client import create_ailoos_node

async def run_advanced():
    # Create node with specific capabilities
    node = await create_ailoos_node(
        node_id="advanced-node-01",
        enable_marketplace=True,
        enable_federated=True
    )
    
    # Search for datasets
    datasets = await node.search_datasets(query="satellite imagery")
    print(f"Found {len(datasets)} datasets")

    # Join a training session
    await node.join_training_session("session_v1_xyz")

Requirements

  • Python 3.8+
  • Network connectivity for P2P and Federation

License

Proprietary and Confidential. Copyright © 2026 Empoorio INC. All Rights Reserved. See LICENSE for full terms.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ailoos-4.5.4.tar.gz (5.7 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

ailoos-4.5.4-py3-none-any.whl (6.4 MB view details)

Uploaded Python 3

File details

Details for the file ailoos-4.5.4.tar.gz.

File metadata

  • Download URL: ailoos-4.5.4.tar.gz
  • Upload date:
  • Size: 5.7 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.9.25

File hashes

Hashes for ailoos-4.5.4.tar.gz
Algorithm Hash digest
SHA256 c5ec1a031c0aa400251427c618529535c7ede9eb700a26a1e193a2fb0738867a
MD5 32a6ae91082eeff36042c10015732000
BLAKE2b-256 ede80f3f77eaa641382a731026664ffdad43a4e2bf104e21b07ad316a0d9721a

See more details on using hashes here.

File details

Details for the file ailoos-4.5.4-py3-none-any.whl.

File metadata

  • Download URL: ailoos-4.5.4-py3-none-any.whl
  • Upload date:
  • Size: 6.4 MB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.9.25

File hashes

Hashes for ailoos-4.5.4-py3-none-any.whl
Algorithm Hash digest
SHA256 d1868ac90e7b0436a1b3aad9a48296c757d861f12f763972d3617358d7cacc18
MD5 2ac77e5ee0a9fe7bff3fccbed5e0a001
BLAKE2b-256 266595583acca1471b8607d63818c55a6d06a1f26e6b0b02422731bb4ba353f2

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page