Documentation

Flatten

Input

x:
  • Required: yes
  • Type: float16, float32, int8, uint8, int16, uint16, int32, uint32, int64, uint64
  • Description: input tensor
  • Restrictions: none

Attribute

  • axis:
    • Required: no
    • Type: int
    • Description: the start axis to flatten (default = 1). May be negative to index from the end (for example, –1 for the last axis).
    • Restriction: Must be in the range [–rank(x), rank(x)).
  • end_axis:
    • Required: no
    • Type: int
    • Description: the end axis to flatten (default = 1). May be negative to index from the end (for example, –1 for the last axis).
    • Restriction: Must be in the range [0, rank(x)).

Output

y:
  • Required: yes
  • Type: float16, float32, int8, uint8, int16, uint16, int32, uint32, int64, uint64
  • Description: output tensor
  • Restrictions: none