Posts

Post not yet marked as solved
4 Replies
This workaround at least prevents the load time issue, but of course using the bionic engine would be preferable, especially since we use the GPU for other things.MLModelConfiguration *config = [[MLModelConfiguration alloc] init]; config.computeUnits = MLComputeUnitsCPUAndGPU; MLModel *model = [MLModel modelWithContentsOfURL:modelUrl configuration:config error:&error];
Post not yet marked as solved
3 Replies
We only get this on the new devices with the Apple Neural Engine. I assume it could be some unsupported operation, but it runs fine in CoreML on other devices. It spits this error out on ANE compatible devices and then works fine (abeit the same speed as the other devices). I don't get any other info so it's basically impossible to determine which layer is causing a problem.