In this chapter, we will learn about Caffe2 operators and how we can compose networks using these operators. To learn how to use operators, we will start off by building a simple computation graph from scratch. After that, we will solve a real computer vision problem called MNIST (by building a genuine neural network with trained parameters) and use it for inference.
This chapter covers the following topics:
- Introduction to Caffe2 operators
- The difference between operators and layers
- How to use operators to compose a network
- Introduction to the MNIST problem
- Composing a network for the MNIST problem
- Inference through a Caffe2 network