ML2P CLI Reference

ml2p

Minimal Lovable Machine Learning Pipeline.

A friendlier interface to AWS SageMaker.

ml2p [OPTIONS] COMMAND [ARGS]...

Options

--cfg <cfg>

Project configuration file. Default: ./ml2p.yml.

--version

Show the version and exit.

dataset

Create and manage datasets.

ml2p dataset [OPTIONS] COMMAND [ARGS]...

create

Create a dataset.

ml2p dataset create [OPTIONS] DATASET

Arguments

DATASET

Required argument

delete

Delete a dataset.

ml2p dataset delete [OPTIONS] DATASET

Arguments

DATASET

Required argument

dn

Download a file SRC from the dataset and save it in DST.

If DST is omitted, the source file is downloaded as its own name.

ml2p dataset dn [OPTIONS] DATASET SRC [DST]

Arguments

DATASET

Required argument

SRC

Required argument

DST

Optional argument

generate

Launch a processing job that generates a dataset.

ml2p dataset generate [OPTIONS] DATASET

Options

-m, --model-type <model_type>

The name of the type of model.

Arguments

DATASET

Required argument

list

List datasets for this project.

ml2p dataset list [OPTIONS]

ls

List the contents of a dataset.

ml2p dataset ls [OPTIONS] DATASET

Arguments

DATASET

Required argument

rm

Delete a file from a dataset.

ml2p dataset rm [OPTIONS] DATASET FILENAME

Arguments

DATASET

Required argument

FILENAME

Required argument

up

Upload a file SRC to a dataset as DST.

If DST is omitted, the source file is uploaded under its own name.

ml2p dataset up [OPTIONS] DATASET SRC [DST]

Arguments

DATASET

Required argument

SRC

Required argument

DST

Optional argument

endpoint

Create and inspect endpoints.

ml2p endpoint [OPTIONS] COMMAND [ARGS]...

create

Create an endpoint for a model.

ml2p endpoint create [OPTIONS] ENDPOINT_NAME

Options

-m, --model-name <model_name>

The name of the model to base the endpoint on. Defaults to the endpoint name without the live/analysis/test suffix.

Arguments

ENDPOINT_NAME

Required argument

delete

Delete an endpoint.

ml2p endpoint delete [OPTIONS] ENDPOINT_NAME

Arguments

ENDPOINT_NAME

Required argument

describe

Describe an endpoint.

ml2p endpoint describe [OPTIONS] ENDPOINT_NAME

Arguments

ENDPOINT_NAME

Required argument

invoke

Invoke an endpoint (i.e. make a prediction).

ml2p endpoint invoke [OPTIONS] ENDPOINT_NAME JSON_DATA

Arguments

ENDPOINT_NAME

Required argument

JSON_DATA

Required argument

list

List endpoints for this project.

ml2p endpoint list [OPTIONS]

wait

Wait for an endpoint to be ready or dead.

ml2p endpoint wait [OPTIONS] ENDPOINT_NAME

Arguments

ENDPOINT_NAME

Required argument

init

Initialize the project S3 bucket.

ml2p init [OPTIONS]

model

Create and inspect models.

ml2p model [OPTIONS] COMMAND [ARGS]...

create

Create a model.

ml2p model create [OPTIONS] MODEL_NAME

Options

-t, --training-job <training_job>

The name of the training job to base the model on. Defaults to the model name without the patch version number.

-m, --model-type <model_type>

The name of the type of model.

Arguments

MODEL_NAME

Required argument

delete

Delete a model.

ml2p model delete [OPTIONS] MODEL_NAME

Arguments

MODEL_NAME

Required argument

describe

Describe a model.

ml2p model describe [OPTIONS] MODEL_NAME

Arguments

MODEL_NAME

Required argument

list

List models for this project.

ml2p model list [OPTIONS]

notebook

Create and manage notebooks.

ml2p notebook [OPTIONS] COMMAND [ARGS]...

create

Create a notebook instance.

ml2p notebook create [OPTIONS] NOTEBOOK_NAME

Arguments

NOTEBOOK_NAME

Required argument

delete

Delete a notebook instance.

ml2p notebook delete [OPTIONS] NOTEBOOK_NAME

Arguments

NOTEBOOK_NAME

Required argument

describe

Describe a notebook instance.

ml2p notebook describe [OPTIONS] NOTEBOOK_NAME

Arguments

NOTEBOOK_NAME

Required argument

list

ml2p notebook list [OPTIONS]

presigned-url

Create a URL to connect to the Jupyter server from a notebook instance.

ml2p notebook presigned-url [OPTIONS] NOTEBOOK_NAME

Arguments

NOTEBOOK_NAME

Required argument

start

Start a notebook instance.

ml2p notebook start [OPTIONS] NOTEBOOK_NAME

Arguments

NOTEBOOK_NAME

Required argument

stop

Stop a notebook instance.

ml2p notebook stop [OPTIONS] NOTEBOOK_NAME

Arguments

NOTEBOOK_NAME

Required argument

repo

Describe and list code repositories.

ml2p repo [OPTIONS] COMMAND [ARGS]...

describe

Describe a code repository SageMaker resource.

ml2p repo describe [OPTIONS] REPO_NAME

Arguments

REPO_NAME

Required argument

list

List code repositories.

ml2p repo list [OPTIONS]

training-job

Create and inspect training jobs.

ml2p training-job [OPTIONS] COMMAND [ARGS]...

create

Create a training job.

ml2p training-job create [OPTIONS] TRAINING_JOB DATASET

Options

-m, --model-type <model_type>

The name of the type of model.

Arguments

TRAINING_JOB

Required argument

DATASET

Required argument

describe

Describe a training job.

ml2p training-job describe [OPTIONS] TRAINING_JOB

Arguments

TRAINING_JOB

Required argument

list

List training jobs for this project.

ml2p training-job list [OPTIONS]

wait

Wait for a training job to complete or stop.

ml2p training-job wait [OPTIONS] TRAINING_JOB

Arguments

TRAINING_JOB

Required argument