MachineLabs has released the entire code base of their machine learning platform as open source under the MIT license.
Following this announcement, all work at the organization will happen completely publicly, giving everyone the chance to join the effort and fork and fiddle with the code.
MachineLabs is an open source online platform for Machine Learning. It is accessible, sharable and explorable via the web. It has ready to use environments for main ML frameworks including Tensorflow, Theano, PyTorch, and Caffe. Also, it has an online code Editor and access to blazingly fast GPU hardware. Moreover, it can expose generated assets via API, enabling users to request trained models for browser-based front-end apps. Their core mission, MachineLabs says, “has always been to empower the Machine Learning community, we believe it is critical for us to be as open and transparent as possible.”
The open source announcement is MachineLabs’ first move to make their ML platform 100% open and transparent and entirely owned and governed by the community. In addition, the community that creates, runs and funds the ML platform will essentially also own and steer the revenue generating MachineLabs service. In order for this to happen, they are looking for platforms such as Aragon and Habour to implement a decentralized governance model.
MachineLabs researchers have also started working on a Machine Learning Online Course with a radical stance on simplicity to further help people getting started within the Machine Learning field.
They have created a Developer Guide to make it as easy as possible for everyone to contribute to the platform. They also ask contributors to go through the Contributing Guidelines, to make every contribution as smooth as possible.
To know more, about the open source announcement, go through the official MachineLabs blog.