daggerml

DaggerML Python library for creating and managing DAGs (Directed Acyclic Graphs) for machine learning workflows.

Getting started

Installation

Install daggerml in whichever virtual environment you prefer.

pip install daggerml

For the CLI functionality, you have two options:

Option 1: Install CLI with daggerml (recommended for most users)

pip install daggerml[cli]

Option 2: Install CLI separately with pipx (keeps dependencies isolated)

pipx install daggerml-cli

Setting up a repo

Now we create a repo using the commandline.

dml config user ${EMAIL}
dml repo create ${REPO_NAME}
dml config repo ${REPO_NAME}

Now we can create dags or whatever we want using this repo.

from daggerml import Dml

with Dml().new("test", "this dag is a test") as dag:
  dag.result = 42

Now we can list repos, dags, etc.

dml dag list

Clean up

dml repo delete ${REPO_NAME}

Docs

For more info, check out the docs at daggerml.com.

Contributing

If you want to contribute, please check out the contributing guide.

License

daggerml is distributed under the terms of the MIT license.