Simplifying and Optimizing the Use of Deep Learning Frameworks

 

As we all push forward with the development of artificial intelligence (AI) solutions, software developers and data scientists increasingly want to leverage deep learning frameworks. To back up a bit, deep learning is a type of machine learning that can enable more complex solutions based on evaluation of abstractions of data. Scaling through added layers and processing, deep learning can build in aggregate from user input and experiences, much the way people learn. Deep learning frameworks enable algorithms to continually improve their performance on complex tasks like speech and image recognition.

To get on this path to a new generation of AI solutions, developers and data scientists need to find ways to reduce the steep learning curve that comes with the deployment and configuration of deep learning frameworks. Then, find ways to accelerate the development, training, and deployment of models. This is where the new Intel® Deep Learning SDK comes into play.

Intel® Deep Learning SDK

The Intel Deep Learning SDK is a free package of tools that data scientists and software developers can use to experiment with innovating around deep learning solutions. The SDK encompasses a training tool and a deployment tool that can be used separately or together in a complete deep learning workflow.

Training Tool

When deployed as a training tool, the SDK enables data scientists to easily prepare training data, design models, and train models with automated experiments and advanced visualizations. It simplifies the installation and usage of popular deep learning frameworks optimized for Intel platforms. In fact, the SDK comes prepackaged with ready-to-run Intel® Distribution for Caffe*, one of the popular frameworks used for deep learning. The team behind the Intel Deep Learning SDK will be updating it with other frameworks optimized for Intel® architecture on an ongoing basis – TensorFlow* will be added soon.

Frameworks included in the Intel Deep Learning SDK are already optimized to take advantage of the underlying Intel chipset. This makes it easy to not only get started with a framework, but to benefit from performance, even if you are just at the “play around with it” phase.

You can download the SDK onto a Mac* laptop or Linux* server, and as part of the beta, you can install and run the SDK on Amazon Web Services with an account. After a quick setup process, choose the framework you want to use. Then, go for it—in your very own deep learning sandbox. It’s now that easy to start discovering the power of deep learning frameworks.

Deployment Tool

When using the deployment tool, the SDK helps application developers optimize trained deep learning models through model compression and weight quantization, which are tailored to end-point device characteristics. It delivers a unified API to integrate the inference with application logic.

As Chief Developer Advocate for AI and Analytics solutions at Intel, I’m excited about the Intel Deep Learning SDK and what it can do for software developers and data scientists. What we do is all about enabling innovation and making technology as accessible as possible to our community. AI is a huge focus right now and this is a great on-ramp to deep learning frameworks. Through its ability to simplify the development, training, deployment, and optimization of deep learning solutions, the SDK can greatly accelerate the path to innovative AI solutions.

I think it’s important to add that this wasn’t something Intel did in isolation. Like so many software initiatives, the SDK was created in close collaboration with a broad community of developers and data scientists who focus on AI solutions, including many talented people who are active in the Intel® Developer Zone. As we push forward with further enhancements of the SDK, we welcome feedback from the community.

Getting Started with the Intel® Deep Learning SDK

Ready to get started? Download the tool from the product website. For a tutorial on using the Intel Deep Learning SDK Training Tool, visit our Getting Started Website. It provides detailed step-by-step instructions on how to put the tool to work to accelerate deep learning training. You can also get a quick overview of how to get started with the tool in our Getting Started Video, which demonstrates model training using LeNet* topology and the MNIST dataset. For help, ask questions in our forum.

Look for the Intel Deep Learning SDK to be showcased at our many meetups and sponsored events this year. Check the Intel® NervanaTM AI Academy for the latest schedule and other great tools accompanying the Intel Deep Learning SDK.