Tutorial 2 questions

[Githeo]

Hello,

I'm trying to check the Tutorial 2 - MultiLayer Split Neural Network

In this tutorial, the model is split among Alice, Bob, and Claire, but it looks like Bob has neither features nor labels. So why Bob is there? I could have more appreciated an example where the features are split among the 3. Did you already provide such example?

Also I don't understand why, when you send the images data to the first model

images = images.send(models[0].location)

the data goes to the entry part of models[0]:

models = [
    nn.Sequential(
                nn.Linear(input_size, hidden_sizes[0]),
                nn.ReLU(),
    ),
    nn.Sequential(
                nn.Linear(hidden_sizes[0], hidden_sizes[1]),
                nn.ReLU(),
    ),
    nn.Sequential(
                nn.Linear(hidden_sizes[1], output_size),
                nn.LogSoftmax(dim=1)
    )
]

Instead, when labels are sent (labels = labels.send(models[-1].location)), these go to the ending part of models[-1]. Infact, the shape of labels is equal to output_size, not hidden_sizes[1].

Good job, Cheers,