wanna-ml 0.2.8

CLI tool for managing ML projects on Vertex AI

WANNA-ML

---

Complete MLOps framework for Vertex-AI

---

About WANNA-ML

WANNA-ML is a CLI tool that helps researchers, data scientists, and ML Engineers quickly adapt to Google Cloud Platform (GCP) and get started on the cloud in almost no time.

It makes it easy to start a Jupyter notebook, run training jobs and pipelines, build a Docker container, export logs to Tensorboards, and much more.

We build on top of Vertex-AI managed services and integrate with other GCP services like Cloud Build and Artifact Registry to provide you with a standardized structure for managing ML assets on GCP.

Help

See the documentation for more details.

Get started

Installation

Install using pip install -U wanna-ml.

For more information on the installation process and requirements, visit out installation page in documentation

Use Docker Image

For your convenience, we have prepared a Docker image with everything you need to get started.

docker pull michalmrazek9/wanna-ml

docker run -it michalmrazek9/wanna-ml

$ wanna version

Authentication

WANNA-ML relies on gcloud for user authentication.

1. Install the gcloud CLI - follow official guide
2. Authenticate with the gcloud init
3. Set you Google Application Credentials gcloud auth application-default login

Docker Build

You can use a local Docker daemon to build Docker images, but it is not required. You are free to choose between local building on GCP Cloud Build. If you prefer local Docker image building, install Docker Desktop.

GCP IAM Roles and Permissions

Different WANNA-ML calls require different GCP permissions to create given resources on GCP. Our documentation page lists recommended GCP IAM roles for each wanna command.

Examples

Jump to the samples to see a complete solution for various use cases.

Issues

Please report issues to GitHub.

Contributing

Your contributions are always welcome, see CONTRIBUTING.md for more information. If you like WANNA-ML, don't forget to give our project a star!

Licence

Distributed under the MIT License - see LICENSE.