To get started with PureML, you’ll need to install the PureML Python SDK. This SDK will allow you to work with the PureML platform, from creating your account to deploying your models.
To install PureML, simply run the following command:
pip3 install pureml
If you`re a large company looking for a proof-of-concept, or an engineer looking to use the open-source version in a funky non-production way, then you’re in the right place! Our Docker compose deployment let’s you spin up a fresh Pureml instance in minutes.
Want more reliability? The easiest way to get started with PureML is to use PureML Cloud.
Self-hosted open-source deployment is Apache-2.0 licensed and provided without a guarantee.
- A subsystem with docker-engine installed. Installation guidelines
Setting up the compose
In a new directory where you want to setup the containers, create a new file
Add the following content from our official docker-compose example filedocker-compose.yml
version: "3" services: backend: image: puremlhq/pureml_backend:local-base environment: - PURE_SITE_BASE_URL=http://localhost:3000 ports: - 8080:8080 volumes: - pureml-data:/pureml_backend/data frontend: image: puremlhq/pureml_frontend environment: - BACKEND_URL=http://backend:8080/api/ ports: - 3000:3000 links: - backend volumes: pureml-data:
Run the following command to start your containers
docker compose up
Make sure your docker engine is running for the docker command to work.
Additionally, to run the containers in background you can use the command
docker compose up -d or
docker compose up --detach
If all goes well, you should have a PureML local instance setup and running
You can even checkout the auto generated Open API swagger documentation at /api/swagger/index.html
Update docker image
Make sure you are using latest docker image to use latest features of PureML. You can upgrade the package by using below command:
docker compose up --build