Tensor shapes and dimensions in PyTorch - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a tensor of shape (3, 4).

PyTorch

import torch
x = torch.zeros([1])
print(x.shape)

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

A(4, 3)

B[3, 4, 5]

C(3, 4)

D[4, 3]

Attempts:

3 left

💡 Hint

Common Mistakes

Using square brackets instead of parentheses for shape.

Swapping the dimensions (4, 3) instead of (3, 4).

The shape (3, 4) creates a tensor with 3 rows and 4 columns.

2fill in blank

medium

Complete the code to get the number of dimensions of tensor x.

PyTorch

import torch
x = torch.randn(2, 5, 7)
dimensions = x.[1]()
print(dimensions)

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

Ashape

Bndim

Csize

Ddim

Attempts:

3 left

💡 Hint

Common Mistakes

Using shape which returns the size tuple, not the number of dimensions.

Using ndim which is not a PyTorch method.

The dim() method returns the number of dimensions of a tensor.

3fill in blank

hard

Fix the error in the code to reshape tensor x to shape (6, 10).

PyTorch

import torch
x = torch.randn(3, 20)
x = x.[1]((6, 10))
print(x.shape)

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

Aview

Bresize

Creshape

Dresize_

Attempts:

3 left

💡 Hint

Common Mistakes

Using reshape() which can sometimes cause errors if tensor is not contiguous.

Using resize() which changes the tensor size in-place and can cause unexpected behavior.

The view() method reshapes a tensor without copying data, but requires the tensor to be contiguous.

4fill in blank

hard

Fill both blanks to create a tensor of shape (2, 3) and then flatten it to 1D.

PyTorch

import torch
x = torch.randn([1])
x_flat = x.[2]()
print(x_flat.shape)

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

A(2, 3)

B(3, 2)

Cflatten

Dview

Attempts:

3 left

💡 Hint

Common Mistakes

Swapping the shape dimensions.

Using view() without specifying the new shape.

First, create a tensor with shape (2, 3). Then use flatten() to convert it to 1D.

5fill in blank

hard

Fill all three blanks to create a tensor, get its shape, and access the size of the second dimension.

PyTorch

import torch
x = torch.randn([1])
shape = x.[2]
second_dim = shape[[3]]
print(second_dim)

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

A(4, 5, 6)

Bshape

C1

Dsize

Attempts:

3 left

💡 Hint

Common Mistakes

Using size() method instead of shape attribute.

Accessing index 0 or 2 instead of 1 for the second dimension.

Create a tensor with shape (4, 5, 6). Use shape to get its size tuple. Access index 1 for the second dimension.