Why generative models create data in PyTorch - Test Your Understanding

Practice - 5 Tasks

1fill in blank

easy

Complete the code to define a simple generative model class in PyTorch.

PyTorch

import torch.nn as nn

class SimpleGenerator(nn.Module):
    def __init__(self):
        super(SimpleGenerator, self).__init__()
        self.model = nn.Sequential(
            nn.Linear(100, 256),
            nn.ReLU(),
            nn.Linear(256, 784),
            nn.Tanh()
        )

    def forward(self, [1]):
        return self.model(x)

Drag options to blanks, or click blank then click option'

Anoise

Binput

Cdata

Dx

Attempts:

3 left

💡 Hint

Common Mistakes

Using a variable name not defined in the method parameters.

Forgetting to include the input parameter in the forward method.

The forward method takes an input tensor, commonly named 'x', which is passed through the model to generate output data.

2fill in blank

medium

Complete the code to generate fake data from random noise using the generator model.

PyTorch

import torch

noise = torch.randn(16, 100)
fake_data = generator.[1](noise)

Drag options to blanks, or click blank then click option'

Agenerate

Bpredict

C__call__

Dforward

Attempts:

3 left

💡 Hint

Common Mistakes

Trying to call a non-existent 'generate' or 'predict' method.

Using 'forward' method directly instead of calling the model.

In PyTorch, calling the model instance directly (using __call__) runs the forward method, generating fake data from noise.

3fill in blank

hard

Fix the error in the training loop where the generator tries to create data.

PyTorch

for epoch in range(5):
    noise = torch.randn(32, 100)
    fake_data = generator.[1](noise)
    # further training steps...

Drag options to blanks, or click blank then click option'

Aforward

B__call__

Cgenerate

Drun

Attempts:

3 left

💡 Hint

Common Mistakes

Calling a method that does not exist on the model.

Using 'forward' method directly instead of calling the model.

Calling the generator instance directly (using __call__) is the correct way to generate data in PyTorch, not using a non-existent 'generate' or 'run' method.

4fill in blank

hard

Fill both blanks to create a dictionary of generated images with their indices.

PyTorch

generated_images = {i: fake_data[1] for i in range(10) if i [2] 5}

Drag options to blanks, or click blank then click option'

A[i]

B(i)

C<

D>

Attempts:

3 left

💡 Hint

Common Mistakes

Using parentheses instead of square brackets for indexing.

Using the wrong comparison operator in the condition.

To index tensors, square brackets are used. The condition filters indices less than 5.

5fill in blank

hard

Fill all three blanks to create a dictionary of generated images with their pixel sums greater than 1000.

PyTorch

filtered_images = {i: fake_data[1] for i in range(20) if fake_data[2].sum() [3] 1000}

Drag options to blanks, or click blank then click option'

A[i]

C>

D<

Attempts:

3 left

💡 Hint

Common Mistakes

Using parentheses instead of square brackets for indexing.

Using '<' instead of '>' in the condition.

Use square brackets to index the tensor at i, then sum the pixels and check if greater than 1000.