Tensors: Tensors are the fundamental data structure used in deep learning frameworks like PyTorch. A tensor is a multi-dimensional array that can store and manipulate numerical data efficiently. Tensors provide the foundation for performing computations in deep learning models. They can represent various types of data, including input features, model parameters, and intermediate results during computations.

torch: In the context of NLP, "torch" refers to the PyTorch library, which is a popular open-source deep learning framework. PyTorch provides a wide range of functionalities for building and training deep learning models, including NLP models. It offers various modules, functions, and classes for tensor operations, neural network architecture, optimization algorithms, loss functions, and more.

Tensors are multidimensional arrays that can store and manipulate numerical data efficiently. They are the fundamental data structure used for computations in deep learning models. However, Django can be used in combination with these deep learning frameworks to create web applications that utilize tensors for various tasks, such as data preprocessing, model training, and inference.

Here are some common functionalities and operations associated with tensors in deep learning frameworks, along with examples and their corresponding output:

Creation: Tensors can be created from various sources such as Python lists, NumPy arrays, or directly using built-in functions.

import torch

Create a tensor from a Python list
tensor1 = torch.tensor([1, 2, 3, 4, 5])
print(tensor1)

 Output: tensor([1, 2, 3, 4, 5])

Create a tensor from a NumPy array

import numpy as np
array = np.array([1, 2, 3, 4, 5])
tensor2 = torch.tensor(array)
print(tensor2)

Output: tensor([1, 2, 3, 4, 5])

Create a tensor of zeros

tensor3 = torch.zeros(3, 2)
print(tensor3)

 Output: tensor([[0., 0.],
                 [0., 0.],
                [0., 0.]])

Create a tensor of random values

tensor4 = torch.randn(2, 2)
print(tensor4)

 Output: tensor([[-0.5863,  0.2836],
                 [-0.4969, -0.4520]])

Operations: Tensors support a wide range of mathematical operations such as addition, subtraction, multiplication, division, and more.

import torch

tensor1 = torch.tensor([1, 2, 3])
tensor2 = torch.tensor([4, 5, 6])

Element-wise addition

result1 = tensor1 + tensor2
print(result1)

Output: tensor([5, 7, 9])

Element-wise multiplication

result2 = tensor1 * tensor2
print(result2)

Output: tensor([ 4, 10, 18])

Matrix multiplication

tensor3 = torch.tensor([[1, 2], [3, 4]])
tensor4 = torch.tensor([[5, 6], [7, 8]])
result3 = torch.matmul(tensor3, tensor4)
print(result3)

 Output: tensor([[19, 22],
                [43, 50]])

Indexing and Slicing: Tensors allow indexing and slicing operations to access specific elements or subsets of elements.
python

import torch

tensor = torch.tensor([[1, 2, 3], [4, 5, 6], [7, 8, 9]])

Access a specific element

element = tensor[0, 1]
print(element)

Output: tensor(2)

Slice a row

row_slice = tensor[1, :]
print(row_slice)

Output: tensor([4, 5, 6])

Slice a column

column_slice = tensor[:, 2]
print(column_slice)

Output: tensor([3, 6, 9])

Slice a submatrix

submatrix = tensor[:2, :2]
print(submatrix)

 Output: tensor([[1, 2],
                 [4, 5]])

import torch

tensor = torch.tensor([[1, 2, 3], [4, 5, 6], [7, 8, 9]])


Get the shape of a tensor
shape = tensor.shape
print(shape)

Output: torch.Size([3, 3])

reshape a tensor

reshaped_tensor = tensor.view(9)
print(reshaped_tensor)
Output: tensor([1, 2, 3, 4, 5, 6, 7, 8, 9])

Transpose a tensor
transposed_tensor = tensor.transpose(0, 1)
print(transposed_tensor)

 Output: tensor([[1, 4, 7],
                 [2, 5, 8],
                 [3, 6, 9]])

Mathematical Functions: Tensors provide various mathematical functions for element-wise operations, reduction operations, and more.

import torch

tensor = torch.tensor([1.5, 2.7, 3.1, 4.9])

Element-wise square root

sqrt_tensor = torch.sqrt(tensor)
print(sqrt_tensor)

Output: tensor([1.2247, 1.6432, 1.7607, 2.2136])

Sum of all elements

sum_tensor = torch.sum(tensor)
print(sum_tensor)

Output: tensor(12.2)

Mean of all elements

mean_tensor = torch.mean(tensor)
print(mean_tensor)

Output: tensor(3.05)