The official Docker image for Loud ML can be downloaded from our DockerHub repo. It can be used to install Loud ML on any operating system that supports Docker containers.
To install Loud ML Open Source version, run the command below:
docker pull loudml/loudml:latest-devel
To run the image, and in particular the
loudmld process giving
access to the HTTP API you can run:
docker run -ti -p 8077:8077 -v $VOLUME/etc/loudml/config.yml:/etc/loudml/config.yml:ro -v $VOLUME/var/lib/loudml:/var/lib/loudml:rw loudml/loudml
loudmld stores all training information to
This directory is local to the container and you can define $VOLUME in the
above shell command to keep these files on your host.
Docker users can use more advanced settings thanks to
An example for your reference is available in the loudml Github repository, directory docker-compose/. You can fork the example and create your own as required.
The Docker image places config files, logs, and the data directory in the appropriate locations for a Debian-based system:
Loud ML home directory or
Configuration files including
Environment variables including heap size, file descriptors.
The location of the data files of each model defined on the node.