graphrag Docker & Self-Hosted Deployment

GitHub

A modular graph-based Retrieval-Augmented Generation (RAG) system

35,462 stars Python Python 1 file · ~40,698 tokens #gpt#gpt-4#gpt4#graphrag#llm#llms#rag
1. Pull Image (DOCKERHUB) docker pull microsoft/graphrag:latest
2. Docker Run (1-Click Launch) docker run -d --name graphrag -p 8501:8501 --restart always microsoft/graphrag:latest
3. Podman Run podman run -d --name graphrag -p 8501:8501 --restart always microsoft/graphrag:latest
4. Skopeo Inspect (Zero-Pull Manifest) skopeo inspect docker://docker.io/microsoft/graphrag:latest
5. ORAS Pull (OCI Artifact) oras pull microsoft/graphrag:latest
unified-search-app/Dockerfile 

# Copyright (c) Microsoft Corporation. All rights reserved.
# Dockerfile
# https://eng.ms/docs/more/containers-secure-supply-chain/approved-images
FROM mcr.microsoft.com/oryx/python:3.11

RUN curl -fsSL https://packages.microsoft.com/keys/microsoft.asc | gpg --dearmor -o /usr/share/keyrings/microsoft-prod.gpg
RUN apt-get update -y

# Install dependencies
WORKDIR ./
COPY . .
RUN curl -LsSf https://astral.sh/uv/install.sh | sh
ENV PATH="${PATH}:/root/.local/bin"
RUN uv sync --no-install-project

# Run application
EXPOSE 8501
ENTRYPOINT ["uv","run","poe","start_prod"]