Skip to main content
LoomOSVersion Python 3.11+ License

​
Platform overview

LoomOS provides enterprise-grade infrastructure for AI model deployment, training, verification, and continuous improvement at scale. Below you’ll find focused documentation for each major area of the platform — start with the Platform Overview or jump directly to the SDK, API reference, or deployment guides.

Platform Overview

Architecture, quickstart, and high-level features

Core Modules

LoomDB, Scheduler, Security, Marketplace

Reinforcement Learning

RL Gym, PPO, WEAVE, and distributed training patterns

Nexus System

Master/worker architecture, failover, LoomCtl

Blocks & Adapters

Adapters for OpenAI, Hugging Face, Vision, and Diffusion models

SDK & CLI

Python SDK, job manifests, and CLI commands

Examples & Tutorials

Quick demos and full WEAVE examples

Advanced Features

Tapestry, Prism, Weaver, and container runtimes

API Reference

REST endpoints and Python client docs

Deployment Guide

Docker Compose, Kubernetes, and configuration
If you want any of these pages expanded into additional subpages (for example splitting ‘Core Modules’ into LoomDB, Scheduler, Security), tell me which one and I’ll create them.

​
How to use these docs

  • Start with “Platform Overview” for architecture and quickstart
  • Read “Core Modules” to understand LoomDB, Scheduler, and Security
  • Use “SDK & CLI” for programmatic integration and job manifests
  • Visit “API Reference” for endpoint and client details

​
Verification checklist

  • Follow the quickstart to run the demo
  • Confirm LoomCtl responds: GET /api/v1/cluster/status
  • Run an example training job and verify events appear in LoomDB
Verify success by completing the Quick Start and seeing example outputs in examples/.