Forward pass computation in PyTorch - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to perform a forward pass using a PyTorch model.

PyTorch

output = model([1])

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

Aloss

Bmodel

Coptimizer

Dinput_tensor

Attempts:

3 left

💡 Hint

Common Mistakes

Passing the model itself instead of the input data.

Passing the loss or optimizer instead of input data.

The forward pass requires passing the input tensor to the model to get the output predictions.

2fill in blank

medium

Complete the code to apply the ReLU activation function after the linear layer.

PyTorch

x = torch.nn.functional.[1](linear_layer(input_tensor))

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

Arelu

Btanh

Csoftmax

Dsigmoid

Attempts:

3 left

💡 Hint

Common Mistakes

Using sigmoid or softmax which are used in different contexts.

Forgetting to apply any activation function.

ReLU is a common activation function used after linear layers to add non-linearity.

3fill in blank

hard

Fix the error in the forward pass by completing the missing method call.

PyTorch

output = model.[1](input_tensor)

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

Aforward

Bbackward

Ctrain

Deval

Attempts:

3 left

💡 Hint

Common Mistakes

Using backward which is for gradients, not forward pass.

Using train or eval which are mode settings.

The forward method defines how the input data passes through the model layers.

4fill in blank

hard

Fill both blanks to compute the output and apply softmax activation.

PyTorch

logits = model([1])
probabilities = torch.nn.functional.[2](logits, dim=1)

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

Ainput_tensor

Brelu

Csoftmax

Dsigmoid

Attempts:

3 left

💡 Hint

Common Mistakes

Using ReLU or sigmoid instead of softmax for probabilities.

Passing something other than input tensor to the model.

First, pass the input tensor to the model to get logits. Then apply softmax to convert logits to probabilities.

5fill in blank

hard

Fill all three blanks to perform a forward pass, apply ReLU, and compute loss.

PyTorch

logits = model([1])
activated = torch.nn.functional.[2](logits)
loss = loss_fn(activated, [3])

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

Ainput_tensor

Brelu

Ctarget_tensor

Dsigmoid

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing up input and target tensors.

Using sigmoid instead of ReLU for activation here.

Pass input tensor to model, apply ReLU activation, then compute loss comparing activated output to target tensor.