output and inferred shape have different ranks - when running the example in readme

[bamtheboozle]

following the basic example in readme, (swap man1 and man2 into mans1.jpeg) it does not generate the face swap and it throws:

running on a mac, CPU provider

2024-04-10 10:17:33.073537 [E:onnxruntime:, sequential_executor.cc:514 ExecuteKernel] Non-zero status code returned while running CoreML_11777492068329204276_6 node. Name:'CoreMLExecutionProvider_CoreML_11777492068329204276_6_6' Status Message: Exception: /Users/runner/work/1/s/onnxruntime/core/providers/coreml/model/model.mm:71 InlinedVector<int64_t> (anonymous namespace)::GetStaticOutputShape(gsl::span<const int64_t>, gsl::span<const int64_t>, const logging::Logger &) inferred_shape.size() == coreml_static_shape.size() was false. CoreML static output shape ({1,1,1,800,1}) and inferred shape ({3200,1}) have different ranks.

[Toshi711]

any solution?

[stefanotto]

I'm having the same issue.

[virtuemantra]

is there any solution for this error its happening on Mac

[randylu]

any solution?

[hirozexal]

@bamtheboozle are you able to find any solution to this? Not sure how to fix this at runtime

[antekone]

Same issue, any solutions?

[hopye]

I was using reactor in comfyui perfectly yesterday, today is giving me this same error :
[ONNXRuntimeError] : 1 : FAIL : Non-zero status code returned while running CoreML_5101909898278375113_6 node. Name:'CoreMLExecutionProvider_CoreML_5101909898278375113_6_6' Status Message: Exception: /Users/runner/work/1/s/onnxruntime/core/providers/coreml/model/model.mm:66 InlinedVector<int64_t> onnxruntime::coreml::(anonymous namespace)::GetStaticOutputShape(gsl::span, gsl::span, const logging::Logger &) inferred_shape.size() == coreml_static_shape.size() was false. CoreML static output shape ({1,1,1,800,1}) and inferred shape ({3200,1}) have different ranks.

[hopye]

im still having this issue using comfyui-reactor , it was working and then I know what happened ,lol

[ashleykleynhans]

@hopye this has nothing to do with comfyui-reactor, it is a stand-alone application.

[leiy88]

My Solution: I switched from using the CoreMLExecutionProvider (for Apple Silicon MPS) to the CPUExecutionProvider.In my specific case, I’m working with ComfyUI-Reactor, and I modified the reactor_swapper.py file to implement this change.

[CaledoniaProject]

Just go back to CPU provider, verified on Mac M3 computer.

def getFaceSwapModel(model_path: str):
    model = insightface.model_zoo.get_model(model_path)
    if hasattr(model, 'session'):
        providers = ['CPUExecutionProvider']
        sess_options = onnxruntime.SessionOptions()
        model.session = onnxruntime.InferenceSession(
            model_path,
            providers=providers,
            sess_options=sess_options
        )
    return model

def process(source_img: Union[Image.Image, List],
            target_img: Image.Image,
            source_indexes: str,
            target_indexes: str,
            model: str):
    # Force CPU provider for consistency
    providers = ['CPUExecutionProvider']

    # load face_analyser
    face_analyser = getFaceAnalyser(model, providers)