Command line interface (CLI)

This document provides a complete reference for the command line interface (CLI). It describes all available commands, options, arguments, and expected behavior. The CLI interface is the same across all inference snaps.

Overview

The CLI is the main entry point to interact with the inference snap, manage engines, configure settings, and run interactive sessions. The syntax used is:

<inference-snap> [COMMAND] [OPTIONS] [ARGS]

Global options

To show usage information about any of the commands, set the --help flag.

For example:

user@host:~$

qwen-vl --help

Usage:
  qwen-vl [command]

Basic Commands:
  status       Show the status
  chat         Start the chat CLI

Configuration Commands:
  get          Print configuration option
  set          Set configuration option

Management Commands:
  list-engines List available engines
  show-engine  Print information about an engine
  use-engine   Select an engine

Additional Commands:
  show-machine Print info about the host machine
  help         Help about any command

Flags:
  -h, --help   help for qwen-vl

Use "qwen-vl [command] --help" for more information about a command.

Basic commands

These provide the core functionality for everyday use such as checking tool status or starting an interactive chat session.

status

Displays information about the server and runtime environment.

<inference-snap> status

Details shown:

• Server status and active endpoints

• Currently selected engine

• Availability of better engines, e.g., after attaching new hardware or installing drivers.

• Missing drivers for available hardware

• Resource usage by engines, including cached data

chat

Launch the command-line chat application.

<inference-snap> chat

Configuration commands

Change configuration options including reading, and setting persistent parameters.

set

Set configuration values.

<inference-snap> set <key>=<value>

The Set command overrides configuration values. Unknown keys are rejected.

get

The get command retrieves configuration values.

<inference-snap> get [key]

If no key is given, all configurations are listed in tabular form.

Management commands

Manage available engines by listing, inspecting, and selecting them for use.

list-engines

List engines with details including name, description, vendor, and compatibility.

<inference-snap> list-engines

show-engine

Show engine details.

<inference-snap> show-engine <engine> [flags]

YAML is the default output format, but JSON is also available. You can use the --format <output-type> flag to specify a format.

Output includes:

• Supported devices

• Default configurations

use-engine

To switch to a specific engine:

<inference-snap> use-engine <engine>

To automatically select the best engine:

<inference-snap> use-engine --auto

Switching engines will trigger the download of engine resources (as snap components). A confirmation prompt will appear before downloading.

For regular installations, the --auto flag is applied by default. This command can also be used to revert to the original engine after manual changes, or to reselect the best engine after driver updates, or new hardware installation.

Additional commands

show-machine

Show information about the host machine.

<inference-snap> show-machine [flags]

YAML is the default output format, but JSON is also available. You can use the --format <output-type> flag to specify the output format as yaml or json.