Hands-On Deep Learning with Go

Your first neural network

Let's start with the basic naming our package and importing the packages we will need. This process is carried out in the following steps:

1. For this example, we will be using a tensor library provided by the same developers as Gorgonia. We will use it for the backing tensors that are attached to their respective nodes in the computation graph:

package mainimport (    "fmt"    "io/ioutil"    "log"    . "gorgonia.org/gorgonia"    "gorgonia.org/tensor")

2. Create a variable that will catch errors with the following code:

var err error

We can now define the main struct for embedding the neural network's graph, weights, and prediction (output). In a deeper network, we would have w0, w1, w2, w3, and so on, until wn. There are additional ...