Modzy Core is an edge AI server that provides model serving, data connections, and APIs for custom development. Modzy Core must be linked to a Modzy server to receive updates, but can operate offline on a wide range of edge devices.
There are 2 ways to to use Modzy Core:
The recommended way to use Modzy Core is to install it as a systemd daemon on Linux. This option provides the most resilient and scalable way to run ML models on fleets of edge devices. Installation instructions can be found here.
The Modzy Core binary can be downloaded as part of a Modzy Edge deployment. Full instructions for configuring a device for use with Modzy Core, including download instructions, can be found in the guide How to Deploy a Model to the Edge
$ modzy-core completion
|Generate the autocompletion script for the specified shell
$ modzy-core server
|Starts a HTTP and gRPC server that implements Modzy's inference API.
$ modzy-core direct
|Starts a HTTP and gRPC server that implements Modzy's inference API with minimal overhead to minimize latency
$ modzy-core register
|Register Modzy Core on the device with an installation of Modzy
$ modzy-core install
|Install Modzy Core as a systemd daemon on the device
$ modzy-core logs
|View status logs
$ modzy-core start
|Starts the Modzy Core service when it's been installed as a systemd daemon.
$ modzy-core stop
|Stops the Modzy Core service when it's been installed as a systemd daemon.
$ modzy-core restart
|Restarts the Modzy Core service when it's been installed as a systemd daemon.
$ modzy-core help
|Help about any command
These flags can be used with most modzy cli commands for things like in-line authentication, access to command-specific manuals, etc.
|Help for modzy-core
|Log format. Valid values are 'console' and 'json'. (default "console")
|Disable TLS verification. Default is false
./modzy-core server --tls-skip-verify
Updated 3 months ago