docker-model-runner

Kiwix ZIM-to-vector RAG system for local, offline LLM knowledge retrieval

A template for your kickstart into GenAI!

A privacy-first, automated code review tool powered by local LLMs. Analyse Git diffs using Docker or Ollama models — no cloud and no data sharing required, just intelligent code reviews on your machine.

A streamlined chat application that leverages Docker Model Runner to serve Large Language Models (LLMs) through a modern Streamlit interface. This project demonstrates containerized LLM deployment with a user-friendly web interface.

Run an LLM like a container local (Docker Model Runner) to production (vLLM on EKS) with one env var change.

A local RAG-based Kubernetes documentation assistant using Python, Streamlit, SmolLM2, and Docker Model Runner. Retrieves relevant information from local Markdown and JSONL files to generate context-grounded answers without web search or hosted LLM APIs.

Demo of Docker Model Runner in both development and production environments.

Personal Knowledge Base (PKB) Links tool to organize, store and search effectively

Permissions-aware RAG reference implementation over Canadian federal public documents. Docker Model Runner, pgvector, Compose, Mattermost and MCP front doors, inference benchmarked locally and on Baseten.

About A template for your kickstart into GenAI!

This project demonstrates how to configure Spring AI to interact with Ollama and Docker Model Runner

A grounded support triage agent with hybrid retrieval, evidence citations, escalation policies, and multi-provider LLM fallback.

Local AI Chat Agent using Microsoft Agent Framework integrated with Docker Model Runner (DMR) locally

Reference implementation: LangGraph multi-agent pipeline for code generation with local LLMs

Configurable Streamlit workflows for querying local AI models through Docker Model Runner.