Tensor.layernorm()

Returns the layer normalization of the input tensor.

Usage

from tinygrad.tensor import Tensor
 
a1 = Tensor([[1, 2], [3, 4]])
a1 = a1.layernorm()
 
 
print(a1.numpy())

[[-0.99998  0.99998]
 [-0.99998  0.99998]]

;