Label Studio is an open-source data labeling platform that supports annotation across text, images, audio, video, and time series, with collaborative workflows and ML-model integrations. This guide deploys Label Studio using Docker Compose with Traefik handling automatic HTTPS and persistent volumes for projects and labels. By the end, you'll have Label Studio serving an annotation workspace securely at your domain.
Set Up the Directory Structure
1. Create the project directory:
$ mkdir ~/labelstudio
$ cd ~/labelstudio
2. Create the environment file:
$ nano .env
DOMAIN=labelstudio.example.com
LETSENCRYPT_EMAIL=admin@example.com
Deploy with Docker Compose
1. Create the Compose manifest:
$ nano docker-compose.yaml
services:
traefik:
image: traefik:v3.6
container_name: traefik
command:
- "--providers.docker=true"
- "--providers.docker.exposedbydefault=false"
- "--entrypoints.web.address=:80"
- "--entrypoints.websecure.address=:443"
- "--entrypoints.web.http.redirections.entrypoint.to=websecure"
- "--entrypoints.web.http.redirections.entrypoint.scheme=https"
- "--certificatesresolvers.letsencrypt.acme.httpchallenge=true"
- "--certificatesresolvers.letsencrypt.acme.httpchallenge.entrypoint=web"
- "--certificatesresolvers.letsencrypt.acme.email=${LETSENCRYPT_EMAIL}"
- "--certificatesresolvers.letsencrypt.acme.storage=/letsencrypt/acme.json"
ports:
- "80:80"
- "443:443"
volumes:
- "./letsencrypt:/letsencrypt"
- "/var/run/docker.sock:/var/run/docker.sock:ro"
restart: unless-stopped
labelstudio:
image: heartexlabs/label-studio:1.23.0
container_name: labelstudio
expose:
- "8080"
environment:
- DJANGO_ALLOWED_HOSTS=${DOMAIN}
- CSRF_TRUSTED_ORIGINS=https://${DOMAIN}
- USE_X_FORWARDED_HOST=true
- SECURE_PROXY_SSL_HEADER=HTTP_X_FORWARDED_PROTO,https
volumes:
- ./data:/label-studio/data
labels:
- "traefik.enable=true"
- "traefik.http.routers.labelstudio.rule=Host(`${DOMAIN}`)"
- "traefik.http.routers.labelstudio.entrypoints=websecure"
- "traefik.http.routers.labelstudio.tls.certresolver=letsencrypt"
- "traefik.http.services.labelstudio.loadbalancer.server.port=8080"
restart: unless-stopped
2. Create the data directory with group write access:
$ mkdir data
$ sudo chown :0 data
3. Start the services:
$ docker compose up -d
4. Verify the services and view logs:
$ docker compose ps
$ docker compose logs
Create the Admin and a Sample Project
1. Open https://labelstudio.example.com and click Sign Up to create the first administrator.
2. From the dashboard, click Create Project and set:
-
Project Name:
Sentiment Analysis - Labeling Template: Text Classification
3. Click Save, then Import and paste the sample data:
[
{"text": "This product is amazing."},
{"text": "Worst experience ever."}
]
4. Annotate each task, click Submit, and confirm the Data Manager shows them as Completed.
Next Steps
Label Studio is running and served securely over HTTPS. From here you can:
- Configure ML backends (PyTorch, scikit-learn) for active learning loops
- Invite collaborators with role-based permissions per project
- Export annotated datasets in COCO, YOLO, JSON, or CSV for downstream training
For the full guide with additional tips, visit the original article on Vultr Docs.