Pytorch-this gives error of shape expand(torch.DoubleTensor{[100, 16, 16]}, size=[100]): the number of sizes provided (1) must be greater or equal to the number of dimensions in the tensor (3)

1

##############################################################################

TODO: Implement this function using only a single loop over x_train.

You may not use torch.norm (or its instance method variant), nor any

functions from torch.nn or torch.nn.functional.

##############################################################################

Replace “pass” statement with your code

for i in range(num_train):

sum_=torch.sum((x_test-x_train[i,:]) **2,dim=1)

cists[i] = torch.sqrt(sum_)
2

Please provide more information on what your task is and show us your code so far.

3

def compute_distances_no_loops(x_train, x_test):

“”"

Computes the squared Euclidean distance between each element of the training

set and each element of the test set. Images should be flattened and treated

as vectors.

This implementation should not use any Python loops. For memory-efficiency,

it also should not create any large intermediate tensors; in particular you

should not create any intermediate tensors with O(num_train*num_test)

elements.

Inputs:

  • x_train: Torch tensor of shape (num_train, C, H, W)

  • x_test: Torch tensor of shape (num_test, C, H, W)

Returns:

  • dists: Torch tensor of shape (num_train, num_test) where dists[i, j] is the

    squared Euclidean distance between the ith training point and the jth test

    point.

“”"

Initialize dists to be a tensor of shape (num_train, num_test) with the

same datatype and device as x_train

num_train = x_train.shape[0]

num_test = x_test.shape[0]

dists = x_train.new_zeros(num_train, num_test)

##############################################################################

TODO: Implement this function without using any explicit loops and without

creating any intermediate tensors with O(num_train * num_test) elements.

You may not use torch.norm (or its instance method variant), nor any

functions from torch.nn or torch.nn.functional.

HINT: Try to formulate the Euclidean distance using two broadcast sums

and a matrix multiply.

##############################################################################

Replace “pass” statement with your code

Output: sqrt((x-y)^2)

(x-y)^2 = x^2 + y^2 - 2xy

test_sum = torch.sum((x_test).pow(2), dim=1)

train_sum = torch.sum((x_train).pow(2), dim=1)

inner_product = (x_test)*(x_train)

dists = torch.sqrt( test_sum + train_sum-2*inner_product) # broadcast

##############################################################################

END OF YOUR CODE

##############################################################################

return dists
So I am supposed to calculate dists vector without using any loops,basically by the help of broadcasting and matrix multiplication .I think there need to be done reshaping but i am not getting how.I have done this using two loops as well as one loop

4

Basicaly x_train and x_test are two tensors here ,x_test has the testing data set where as the training data set is against which we are checking the similarity of our data set