CoreML inference crash at some model

When building MLModel, it is set to use NPU. It seems that GPU is used during inference, but it crashes during Compile.

The stack is as follows:

+1

CoreML inference crash at some model
 
 
Q