ML Flow's open source nature allows for flexibility and adaptability in the MLOps space.
ML Flow offers significant flexibility and extensibility, but users may struggle with effectively structuring their work and managing multiple experiments, models, and projects.
ML Flow's community plays a crucial role in its success and evolution, with contributions from the community shaping the platform's development and direction.
Deep dives
Importance of Open Source in ML Flow
ML Flow's open source nature allows for flexibility and adaptability in the machine learning operations (MLOps) space. It provides a platform that captures the essence of an end-to-end ML life cycle while catering to a diverse set of user needs. The ability to integrate with various tools and software makes ML Flow a widely adopted and successful project. It strikes a balance between addressing user requirements, open source principles, and enterprise needs. ML Flow's focus on ease of use, with individual components that stand on their own, lowers the adoption threshold and allows users to gradually explore and expand their usage.
The Future of ML Flow
ML Flow has become a pivotal tool in the MLOps ecosystem and continues to grow alongside the evolving needs of users. The team behind ML Flow is actively working on improving platform features and compatibility with industry-standard tools. They are exploring the integration of ML Flow with feature stores and seeking ways to make feature management more accessible. Additionally, plans are being discussed to introduce better project management, enhanced logging capabilities, and improved user experiences. The goal is to provide a complete end-to-end ML platform while also catering to specific use cases and making ML Flow more robust and versatile.
Challenges and Considerations in ML Flow
ML Flow offers significant flexibility and extensibility, but this can sometimes present challenges for users. Deciding on the best workflows within the platform can be tricky, as the open and unopinionated nature of ML Flow allows for varying approaches. Users may struggle with effectively structuring their work and managing multiple experiments, models, and projects. However, the ML Flow team is actively working on addressing these challenges and considering more opinionated concepts like pipelines. The goal is to make it easier for users to adopt best practices and overcome common hurdles when operationalizing ML Flow within their stack.
Community Engagement and Growth
ML Flow's community has played a crucial role in its success and evolution. Community contributions, including third-party implementations and extensions, have enriched the ML Flow ecosystem. Open discussions, feedback, and collaborations with the community have shaped the direction of ML Flow's development. The team appreciates the collective efforts to build upon and improve the platform, and contributions from the community often lead to enhancements and innovative integrations. The goal is to foster a healthy community and continue providing value to users by addressing real-world challenges and use cases.
Future Roadmap and Involvement
ML Flow is actively working on a public roadmap to transparently communicate their plans for the project's future development. They actively encourage community participation and contribution. Incentives for involvement include recognition in the release notes, ML Flow merchandise like t-shirts and mugs, and even books in some cases. The team also expresses an interest in expanding the set of maintainers for ML Flow to ensure ongoing growth and sustainability of the project. Ultimately, ML Flow's roadmap and direction are shaped by user feedback, real-world needs, and the goal of continuous improvement in MLOps.
MLOps Coffee Sessions #103 with Corey Zumar, MLOps Podcast on Making MLflow co-hosted by Mihail Eric.
// Abstract
Because MLOps is a broad ecosystem of rapidly evolving tools and techniques, it creates several requirements and challenges for platform developers:
- To serve the needs of many practitioners and organizations, it's important for MLOps platforms to support a variety of tools in the ecosystem. This necessitates extra scrutiny when designing APIs, as well as rigorous testing strategies to ensure compatibility.
- Extensibility to new tools and frameworks is a must, but it's important not to sacrifice maintainability. MLflow Plugins (https://www.mlflow.org/docs/latest/plugins.html) is a great example of striking this balance.
- Open source is a great space for MLOps platforms to flourish. MLflow's growth has been heavily aided by: 1. meaningful feedback from a community of ML practitioners with a wide range of use cases and workflows & 2. collaboration with industry experts from a variety of organizations to co-develop APIs that are becoming standards in the MLOps space.
// Bio
Corey Zumar is a software engineer at Databricks, where he’s spent the last four years working on machine learning infrastructure and APIs for the machine learning lifecycle, including model management and production deployment. Corey is an active developer of MLflow. He holds a master’s degree in computer science from UC Berkeley.
--------------- ✌️Connect With Us ✌️ -------------
Join our slack community: https://go.mlops.community/slack
Follow us on Twitter: @mlopscommunity
Sign up for the next meetup: https://go.mlops.community/register
Catch all episodes, blogs, newsletters, and more: https://mlops.community/
Connect with Demetrios on LinkedIn: https://www.linkedin.com/in/dpbrinkm/
Connect with Mihail on LinkedIn: https://www.linkedin.com/in/mihaileric/
Connect with Corey on LinkedIn: https://www.linkedin.com/in/corey-zumar/
Timestamps:
[00:00] Origin story of MLFlow
[02:12] Spark as a big player
[03:12] Key insights
[04:42] Core abstractions and principles on MLFlow's success
[07:08] Product development with open-source
[09:29] Fine line between competing principles
[11:53] Shameless way to pursue collaboration
[12:24] Right go-to-market open-source
[16:27] Vanity metrics
[18:57] First gate of MLOps drug
[22:11] Project fundamentals
[24:29] Through the pillars
[26:14] Best in breed or one tool to rule them all
[29:16] MLOps space mature with the MLOps tool
[30:49] Ultimate vision for MLFlow
[33:56] Alignment of end-users and business values
[38:11] Adding a project abstraction separate from the current ML project
[42:03] Implementing bigger bets in certain directions
[44:54] Log in features to experiment page
[45:46] Challenge when operationalizing MLFlow in their stack
[48:34] What would you work on if it weren't MLFlow?
[49:52] Something to put on top of MLFlow
[51:42] Proxy metric
[52:39] Feature Stores and MLFlow
[54:33] Lightning round [57:36] Wrap up
Get the Snipd podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share & Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
