Share feedback
Answers are generated based on the documentation.

Containerize a generative AI application

Table of contents

Prerequisites

Note

GenAI applications can often benefit from GPU acceleration. Currently Docker Desktop supports GPU acceleration only on Windows with the WSL2 backend. Linux users can also access GPU acceleration using a native installation of the Docker Engine.

  • You have installed the latest version of Docker Desktop or, if you are a Linux user and are planning to use GPU acceleration, Docker Engine. Docker adds new features regularly and some parts of this guide may work only with the latest version of Docker Desktop.
  • You have a git client. The examples in this section use a command-line based git client, but you can use any client.

Overview

This section walks you through containerizing a generative AI (GenAI) application using Docker Desktop.

Note

You can see more samples of containerized GenAI applications in the GenAI Stack demo applications.

Get the sample application

The sample application used in this guide is a modified version of the PDF Reader application from the GenAI Stack demo applications. The application is a full stack Python application that lets you ask questions about a PDF file.

The application uses LangChain for orchestration, Streamlit for the UI, Ollama to run the LLM, and Neo4j to store vectors.

Clone the sample application. Open a terminal, change directory to a directory that you want to work in, and run the following command to clone the repository:

$ git clone https://github.com/craig-osterhout/docker-genai-sample

You should now have the following files in your docker-genai-sample directory.

├── docker-genai-sample/
│ ├── .gitignore
│ ├── app.py
│ ├── chains.py
│ ├── env.example
│ ├── requirements.txt
│ ├── util.py
│ ├── LICENSE
│ └── README.md

Create Docker assets

Now that you have an application, you can create the necessary Docker assets to containerize it.

Tip

Gordon, Docker's AI assistant, can generate Docker assets for your project. Ask Gordon to create a Dockerfile, Compose file, and .dockerignore tailored to your application.

Create the following files in your docker-genai-sample directory.

Dockerfile
# syntax=docker/dockerfile:1

# Comments are provided throughout this file to help you get started.
# If you need more help, visit the Dockerfile reference guide at
# https://docs.docker.com/go/dockerfile-reference/

ARG PYTHON_VERSION=3.11.4
FROM python:${PYTHON_VERSION}-slim as base

# Prevents Python from writing pyc files.
ENV PYTHONDONTWRITEBYTECODE=1

# Keeps Python from buffering stdout and stderr to avoid situations where
# the application crashes without emitting any logs due to buffering.
ENV PYTHONUNBUFFERED=1

WORKDIR /app

# Create a non-privileged user that the app will run under.
# See https://docs.docker.com/go/dockerfile-user-best-practices/
ARG UID=10001
RUN adduser \
    --disabled-password \
    --gecos "" \
    --home "/nonexistent" \
    --shell "/sbin/nologin" \
    --no-create-home \
    --uid "${UID}" \
    appuser

# Download dependencies as a separate step to take advantage of Docker's caching.
# Leverage a cache mount to /root/.cache/pip to speed up subsequent builds.
# Leverage a bind mount to requirements.txt to avoid having to copy them into
# into this layer.
RUN --mount=type=cache,target=/root/.cache/pip \
    --mount=type=bind,source=requirements.txt,target=requirements.txt \
    python -m pip install -r requirements.txt

# Switch to the non-privileged user to run the application.
USER appuser

# Copy the source code into the container.
COPY . .

# Expose the port that the application listens on.
EXPOSE 8000

# Run the application.
CMD ["streamlit", "run", "app.py", "--server.address=0.0.0.0", "--server.port=8000"]
compose.yaml
# Comments are provided throughout this file to help you get started.
# If you need more help, visit the Docker Compose reference guide at
# https://docs.docker.com/go/compose-spec-reference/

# Here the instructions define your application as a service called "server".
# This service is built from the Dockerfile in the current directory.
# You can add other services your application may depend on here, such as a
# database or a cache. For examples, see the Awesome Compose repository:
# https://github.com/docker/awesome-compose
services:
  server:
    build:
      context: .
    ports:
      - 8000:8000

# The commented out section below is an example of how to define a PostgreSQL
# database that your application can use. `depends_on` tells Docker Compose to
# start the database before your application. The `db-data` volume persists the
# database data between container restarts. The `db-password` secret is used
# to set the database password. You must create `db/password.txt` and add
# a password of your choosing to it before running `docker compose up`.
#     depends_on:
#       db:
#         condition: service_healthy
#   db:
#     image: postgres
#     restart: always
#     user: postgres
#     secrets:
#       - db-password
#     volumes:
#       - db-data:/var/lib/postgresql/data
#     environment:
#       - POSTGRES_DB=example
#       - POSTGRES_PASSWORD_FILE=/run/secrets/db-password
#     expose:
#       - 5432
#     healthcheck:
#       test: [ "CMD", "pg_isready" ]
#       interval: 10s
#       timeout: 5s
#       retries: 5
# volumes:
#   db-data:
# secrets:
#   db-password:
#     file: db/password.txt
.dockerignore
# Include any files or directories that you don't want to be copied to your
# container here (e.g., local build artifacts, temporary files, etc.).
#
# For more help, visit the .dockerignore file reference guide at
# https://docs.docker.com/go/build-context-dockerignore/

**/.DS_Store
**/__pycache__
**/.venv
**/.classpath
**/.dockerignore
**/.env
**/.git
**/.gitignore
**/.project
**/.settings
**/.toolstarget
**/.vs
**/.vscode
**/*.*proj.user
**/*.dbmdl
**/*.jfm
**/bin
**/charts
**/docker-compose*
**/compose.y*ml
**/Dockerfile*
**/node_modules
**/npm-debug.log
**/obj
**/secrets.dev.yaml
**/values.dev.yaml
LICENSE
README.md

You should now have the following contents in your docker-genai-sample directory.

├── docker-genai-sample/
│ ├── .dockerignore
│ ├── .gitignore
│ ├── app.py
│ ├── chains.py
│ ├── compose.yaml
│ ├── env.example
│ ├── requirements.txt
│ ├── util.py
│ ├── Dockerfile
│ ├── LICENSE
│ └── README.md

To learn more about these files, see the following:

Run the application

Inside the docker-genai-sample directory, run the following command in a terminal.

$ docker compose up --build

Docker builds and runs your application. Depending on your network connection, it may take several minutes to download all the dependencies. You'll see a message like the following in the terminal when the application is running.

server-1  |   You can now view your Streamlit app in your browser.
server-1  |
server-1  |   URL: http://0.0.0.0:8000
server-1  |

Open a browser and view the application at http://localhost:8000. You should see a simple Streamlit application. The application may take a few minutes to download the embedding model. While the download is in progress, Running appears in the top-right corner.

The application requires a Neo4j database service and an LLM service to function. If you have access to services that you ran outside of Docker, specify the connection information and try it out. If you don't have the services running, continue with this guide to learn how you can run some or all of these services with Docker.

In the terminal, press ctrl+c to stop the application.

Summary

In this section, you learned how you can containerize and run your GenAI application using Docker.

Next steps

In the next section, you'll learn how you can run your application, database, and LLM service all locally using Docker.

Use containers for generative AI development »