AppleCoreMLInferenceMode

Inference modes for Apple CoreML on supported devices.

enum AppleCoreMLInferenceMode {
  CPU = 0,
  GPU = 1,
  ANE = 2,
}

Values

Enum	Value	Description
CPU	0	Uses only CPU for inference computations. Most compatible but potentially slower option
GPU	1	Prioritizes GPU for inference computations. Better performance for parallel processing tasks
ANE	2	Automatic selection with Apple Neural Engine (ANE) prioritized. Best performance on supported devices with neural engine hardware