utsavgarg

I am working for the CloudCV organization on a project named Fabrik. The project aims at bringing down the threshold for getting started with Deep Learning by providing a drag and drop way of building neural networks, while also providing functionality to pro users for visualizing the said graphs and allowing to export graphs imported from one Deep Learning framework to another.

The first couple of weeks of GSoC 2017 went by rather quickly, and the work was mainly targeted at fixing and completing unfinished features. The coding period began with us successfully resolving issues with and merging a pull request for a UI overhaul that had been pending for some time. Then I worked on fixing the integration for Travis builds followed by addition of guidelines to contribute to the project. I then worked on resolving some issues, namely cyclic graphs and one related to dropout layer before moving on to adding new functionality to the project.

Until now the project supported 20 of the 56 available layers in the Caffe framework, I added support for all the other layers and now Fabrik supports all layers available in the Caffe framework. Then I worked on adding unit tests for each of these layers and finally some UI changes to allow easy access to these layers when using the drag and drop functionality.

This coming week I’ll be working on adding support for Python layer in the Caffe framework which is a user defined layer and can have custom parameters.