Documentation

L2Normalize

Input

x:
  • Required: yes
  • Type: float16
  • Description: input tensor to be normalized
  • Restriction: ND

Attribute

axis:
  • Required: no
  • Type: list of ints
  • Description: the list of axes to be normalized.
  • Restriction: input dimension size [–dim, dim – 1]
epsilon:
  • Required: no
  • Type: float
  • Description: non-zero floating-point constant
  • Restriction: It is usually set to 1e-4.

Output

y:
  • Required: yes
  • Type: float16
  • Description: output tensor after normalization
  • Restriction: ND