Since becoming actively involved with the artificial intelligence ecosystem in early 2015, Facebook Inc. has made numerous contributions ranging from niche software modules to entire server blueprints. The social networking giant expanded its repertoire yet again this week by open-sourcing a toolkit called Torchnet that provides building blocks for deep learning projects.

As the name implies, it’s designed for use with Torch, a popular AI development framework that has been adopted by several of Facebook’s engineering teams. Torchnet’s main selling point is a set of five programming abstractions meant to common tasks involved in implementing deep learning functionality. One module provides logic for training models and testing their accuracy, while another helps assess the results. The remaining components are focused on more mundane activities like managing the large volumes of data that are required to carry out such experiments.

Torchnet can dramatically reduce the amount of new code that has to be produced for a deep learning project, which saves valuable time and resources. As a result, organizations developing software using Torch will be able to reduce delivery times while enabling their engineers to move onto other tasks quickler, speeding up operations all around. The functionality should help make Facebook much more competitive against the other tech giants that are working to expand their presence in the artificial intelligence world.

One of the company’s biggest rivals in this space is Alphabet Inc., which last year released an open-source Torch competitor called TensorFlow that is starting to make serious inroads. Meanwhile, Microsoft Corp. and Amazon Inc. are also offering free AI development tools in a bid to replicate the search giant’s success. The web-scale crowd can be expected to continue churning out new technologies as interest in deep learning increases among enterprises.

Image via Pixabay