Tensor #

Tensor is the basic building block of all the supported operations. Tensor represents a $n$-dimensional array of type T numbers. If $n=0$, Tensor is a single number, i.e. scalar. If $n=1$, Tensor is a vector. If $n=2$, Tensor is a matrix.

The data type for a tensor can be any number or boolean which is used for mask tensors.

// tensor represents a multi-dimensional array of elements of type T.
type tensor[T Number | bool] struct {
    data []T
    shape Shape
    strides []int
    offset int
}

type Shape []int

Adding auto-gradient to Tensor requires adding two more methods to the Tensor interface. Calculating gradient is only supported by non-boolean numbers.

type Tensor[T tensor.Number] interface {
    tensor.Tensor[T]
    Grad() tensor.Tensor[T]
    Backward()
}