What is it?
Tensors are the basis of every Deep Learning framework, like PyTorch and TensorFlow, being used to store almost everything, like input data, weights, biases, etc. Tensors in other fields, like Mathematics, are objects with properties, similar to the object paradigm used in OOP.
In Deep Learning, however, tensors can be interpreted as n-dimensional arrays with the assumption that they can run on a GPU. This is important because, although you can also created multi-dimensional arrays with NumPy, it can’t be processed on the GPU.
