Skip to Content
DocumentationFunctionalView Opsdetach_op

detach_op

View the code on GitHub 

Structs

Struct: Detach

Fields

Methods

compute_shape(mut curr: ArrayShape, args: List[ArrayShape])
Computes the shape of an array after detaching.
Args

  • curr: ArrayShape The ArrayShape to store the result of the computation.

  • args: List[ArrayShape] The ArrayShape to detach.

__call__(mut curr: Array, args: List[Array])
Performs the forward pass for the detach operation. It sets the base of the argument to be the base of the current array and computes the shape of the current array via its dedicated ArraySahpe fwd fucntion.
Args

  • curr: Array The current array to store the result (modified in-place).

  • args: List[Array] The array to detach.

Note: The information of the shape computation is stored in the ArrayShape object of the curr array.

jvp(primals: List[Array], tangents: List[Array]) -> Array
more details
Args

  • primals: List[Array]

  • tangents: List[Array]

Returns
  • Array
vjp(primals: List[Array], grad: Array, out: Array) -> List[Array]
more details
Args

  • primals: List[Array]

  • grad: Array

  • out: Array

Returns
  • List[Array]
fwd(arg0: Array) -> Array
Detaches the input array from the computation graph.
Args
  • arg0: Array The input array.
Returns
  • Array - The detached array.

Note: This function is non-differentiable.

Functions

detach

detach(arg0: Array) -> Array
Detaches the input array from the computation graph.
Args
  • arg0: Array The input array.
Returns
  • Array - The detached array.

Note: This function is non-differentiable.

Last updated on
as_strided_opexpand_op