Tensor.linear()

Returns the linear transformation of the input tensor.

Usage

from tinygrad.tensor import Tensor
 
a1 = Tensor([[1, 2], [4, 5], [7, 8]])
weights = Tensor([[1, 2], [3, 4]])
bias = Tensor([1, 2])
 
tensor = a1.linear(weights, bias)
 
print(tensor.numpy())

Return value

[[ 8. 12.]
 [20. 30.]
 [32. 48.]]

leakyrelu log

;