CLI Reference

KladML includes a command-line interface for common tasks.

Global Commands

`kladml version`

Show the installed version.

kladml version
# KladML version 0.1.0

`kladml --help`

Show all available commands.

kladml --help

Project Commands

`kladml init`

Initialize a new KladML project.

kladml init <project-name> [OPTIONS]

Arguments:

Argument	Description
`project-name`	Name of the project directory to create

Options:

Option	Default	Description
`--template`, `-t`	`default`	Project template to use

Example:

kladml init my-forecaster
cd my-forecaster

Run Commands

`kladml run native`

Run a training script using your local Python environment.

kladml run native <script> [OPTIONS]

Arguments:

Argument	Description
`script`	Path to the Python script to run

Options:

Option	Default	Description
`--experiment`, `-e`	`default`	Experiment name for tracking

Example:

kladml run native train.py --experiment baseline

`kladml run local`

Run a training script inside a Docker/Podman container.

kladml run local <script> [OPTIONS]

Arguments:

Argument	Description
`script`	Path to the Python script to run

Options:

Option	Default	Description
`--device`, `-d`	`auto`	Device to use: `auto`, `cpu`, `cuda`, `mps`
`--runtime`, `-r`	`auto`	Container runtime: `auto`, `docker`, `podman`
`--image`, `-i`	(auto)	Custom Docker image to use

Examples:

# Auto-detect runtime and device
kladml run local train.py

# Force CUDA and Docker
kladml run local train.py --device cuda --runtime docker

# Use custom image
kladml run local train.py --image my-registry/my-image:latest

Default Images:

Device	Image
`cpu`	`ghcr.io/kladml/worker:cpu`
`cuda`	`ghcr.io/kladml/worker:cuda12`
`mps`	`ghcr.io/kladml/worker:cpu` (fallback)

Environment Variables

KladML respects these environment variables:

Variable	Description
`KLADML_TRAINING_DEVICE`	Override default device (`cpu`, `cuda`, `mps`)
`KLADML_STORAGE_ARTIFACTS_DIR`	Directory for saving artifacts
`KLADML_EXPERIMENT`	Default experiment name