Back to Blog

Hello World

Image of Sam Wookey
Sam Wookey

We're thrilled to announce that the Masterful AutoML platform is now available on the Python Package Index for everyone to use. Installation is as easy as pip install masterful. You get the full power of our deep learning platform to train your models to peak performance without manual tuning. With Masterful, developers can focus on solving problems with ML instead of wasting time and brain cells running endless training experiments.

Along with easy installation, we're also making Masterful free for personal or academic use. We’re excited to contribute to the community of developers, researchers, and ML practitioners by making state-of-the-art AutoML techniques widely accessible. 

What’s in this release?

We've added lots of great new features - here are just a few highlights:

Improved Training with Unlabeled Data through Semi-Supervised Learning (SSL)

We have improved our core SSL implementation for more consistent performance with unlabeled data. We support single label classification, binary classification, multi-label classification, and semantic segmentation today, with detection and instance segmentation coming very soon. We look forward to further building up this capability through our philosophy of doing the hard work of adapting inspiring research to perform reliably on real-world enterprise data.

Frontend Bundle

Masterful now installs our graphical frontend by default through the pip installation package. This gives you complete transparency into how the Masterful platform trained your model, the actual accuracy improvements, and deeper training insights.

All New Console Logging

We have completely revamped the console output of masterful.autofit for improved clarity and conciseness. Previously, we erred on the side of verbosity, with complete console access to in-depth training logs. However we found this was more confusing than helpful, so we streamlined the console output to give you a better understanding of the progress of your training, and better estimates for when it will complete.

What’s Next?

At Masterful, we believe that large amounts of labeled training data is a bug of modern machine learning, not a feature. As such, one of our goals is to reduce the amount of labeled training data you need to train a model and deploy it into production, while maintaining or even exceeding your existing quality metrics. We will continue to offer the latest in self-, semi-, and unsupervised training techniques, comprehensive regularization, and full utilization of available hardware, in an extensible and easy to use API.

Join Our Community

Have questions, or just want to chat about the latest ML research? Join the Masterful community at https://www.masterfulai.com/community. The team is available to answer your questions, work through issues, discuss feature requests, and generally work with you to solve your use cases with Masterful.

Get Started

Head on over to https://www.masterfulai.com/try-free to request your free account. Check out our installation guide guide and quickstart to get up and running in a few minutes!

 

 


Related Posts

Google Vertex vs Masterful AI (Part 1)

Image of Jack Lynch
Jack Lynch

Percent top-1 error reductions achieved when moving from Google Vertex to Masterful AI. 

Read more

Google Vertex vs Masterful AI (Part 2): Using unlabeled data

Image of Jack Lynch
Jack Lynch

In a previous post we compared Masterful AI’s computer vision training platform to Google’s Vertex...

Read more