Hey everyone!

October has come to a close and so has the 2024 Hacktoberfest! This was our first year participating and we had some awesome contributions (big and small) to KitOps. On behalf of the entire KitOps team, we want to say thank you to everyone who participated.

As always, if you would like to contribute content to the newsletter, let us know in the KitOps Discord #community-kit-content channel.

Happy reading, Jesse,
KitOps Project Lead

P.S. Our team will be joining the Dagger.io community call on Thursday, November 7th to discuss how you can create an MLOps pipeline by combining Dagger and KitOps. The call will be streamed on the Dagger.io YouTube account.

Release Notes

Last week we released v0.3.4.–This version includes this year's #Hacktoberfest contributions, updates to the KitOps documentation, and Readme updates.

In addition to this release, we published three Dagger modules to the Daggerverse:
kit: Use common kit commands such as pack, unpack, push, tag
huggingface: Download repositories from Huggingface
gguf: Convert to GGUF model serialization

Click here to view source

And finally, this month we made it easier for Mac and Linux users to install KitOps via Homebrew–Check out the docs

From around the web

From TestGuild Devops Toolchain Podcast
Simplifying the AI/ML to Production Pipeline with Görkem Ercan

Adventures in Machine Learning Podcast
The Role of Open Source in Modern Development Practices - ML 170

Beers and Bytes Podcast
Revolutionizing MLOps: Gorkem Ercan on Jozu's Game-Changing Solutions for AI Integration

The Jozu Blog
Top 5 Production-Ready Open Source AI Libraries for Engineering Teams

The Jozu Blog
Free Online Tutorials to Help You Develop Machine Learning Applications

The Jozu Blog
Building an MLOps pipeline with Dagger.io and KitOps

The Jozu Blog
10 MLOps Tools That Comply With the EU AI Act

From the community–Rohan Sharma
Top 5 open-source MLOps tool to boost your production ✨

Get involved

Interested in getting involved with KitOps? We encourage everyone to checkout our Getting started documentation. If you find a bug, have a feature idea, or have general feedback simply open a ticket on GitHub.

While you're there, check out our good first issues. Contributing code is the best way to support an open source project.

To start a new project, we suggest using our Quickstart ModelKits, hosted on Jozu Hub:
Meta Llama3.1
Google Gemma
Microsoft phi3
Fine-tuning
Rag pipeline
Object detection

Jozu Hub was purpose built to host ModelKits, and is free to use for individual developers.

Earn some KitOps swag

Love developer swag?

Every month we send swag to community members who actively create content or contribute to KitOps. We're always looking for video tutorials, how-to guides, and getting started content.

Simply share your content in the KitOps Discord #community-kit-content channel, and we will reshare you content and DM you for shipping information.