Facebook has open sourced its advanced deep learning framework Caffe2. The framework offers great flexibility to build high-performance developments along with scalability and performance.
By leveraging artificial intelligence, Facebook’s Caffe2 can be used to recognise text, images, video and speech in smartphones and tablets. The framework is not an AI program but rather a tool that uses AI to program apps for smartphones. You can use Caffe2 even to write self-learning models that can be bundled into apps.
Original Caffe, the deep learning project started at the University of California, Berkeley, was developed to bring machine learning to data centres. But now, Facebook wants to enable developers with features image recognition, computer vision and natural language processing through Caffe2.
“Caffe2 is shipping with tutorials and examples that demonstrate learning at a massive scale which can leverage multiple GPUs in one machine or many machines with one or more GPUs,” Facebook said in a blog post.
Collaborative development model
Alongside Facebook’s efforts, Caffe2 is receiving contributions from Amazon, Intel, Microsoft, Nvidia and Qualcomm. The collaborative development enables a community experience that can serve complex models and deploy the next-generation of AI-enhanced apps and services for both cloud as well as mobile environments.
The Caffe2 code is available on GitHub to let you begin with your AI developments. Furthermore, you can read the Caffe2 documentation and tutorials from its official site to enhance your future projects.
[…] with technologies like the latest Object Detection API, Facebook is offering solutions including Caffe2 and PyTorch and Microsoft has its separate Cognitive […]