[object Object]

[object Object][object Object]undefined[object Object]

• Description: Counts the occurrence of each non-negative integer in an array. [object Object] is the minimum size of the output tensor. When [object Object] is a null pointer, each occurrence of element [object Object] increments the corresponding [object Object] value by 1. When weights is provided, the corresponding [object Object] value is incremented by [object Object]. Consequently, the size of [object Object] is the greater of (the maximum value in [object Object] + 1) and [object Object].

• Formulas:

  If [object Object] is the value of [object Object] at position [object Object] and [object Object] is specified, then:

  $$out[self_i] = out[self_i] + weights_i$$

  Otherwise:

  $$out[self_i] = out[self_i] + 1$$

[object Object]

Each operator has calls. First, [object Object] is called to obtain the workspace size required for computation and the executor covering the operator computation process. Then, [object Object] is called to perform computation.

• [object Object]
• [object Object]

[object Object]

• Parameters:

  • [object Object] (aclTensor*, computation input): aclTensor on the device. The data type can be INT8, INT16, INT32, INT64, or UINT8. The value must be a non-negative integer. The can be 1-dimensional ND. It supports .
  • [object Object] (aclTensor*, computation input): aclTensor on the device, which is the weight of each [object Object] value. It can be a null pointer. The data type can be FLOAT, FLOAT16, FLOAT64, INT8, INT16, INT32, INT64, UINT8, or BOOL. The can be 1-dimensional ND. The shape must be the same as that of [object Object]. It supports .
  • [object Object] (int64_t, computation input): an integer on the host, which specifies the minimum length of the output tensor. This parameter ensures the minimum length of the [object Object]. If the maximum value of [object Object] is less than [object Object], the length of [object Object] is [object Object]. Otherwise, the length of [object Object] is the maximum value of [object Object] + 1.
  • [object Object] (aclTensor*, computation output): aclTensor on the device. The data type can be INT32, INT64, FLOAT, or DOUBLE. The can be 1-dimensional ND. It supports . The length of [object Object] is the maximum value of [object Object] + 1 or [object Object], whichever is greater.
  • [object Object] (uint64_t *, output): size of the workspace to be allocated on the device.
  • [object Object] (aclOpExecutor **, output): operator executor, covering the operator computation process.

• Returns:

  [object Object]: status code. For details, see .

[object Object]

• Parameters:

  • [object Object] (void *, input): address of the workspace to be allocated on the device.
  • [object Object] (uint64_t, input): size of the workspace to be allocated on the device, which is obtained by calling the first-phase API [object Object].
  • [object Object] (aclOpExecutor*, input): operator executor, covering the operator computation process.
  • [object Object] (aclrtStream, input): stream for executing the task.

• Returns:

  [object Object]: status code. For details, see .

[object Object]

• Deterministic computation:
  • [object Object] defaults to a deterministic implementation.

[object Object]

The following example is for reference only. For details, see .