'NSInvalidArgumentException', reason: '+[FIRInstanceIDCheckinPreferences preferencesFromKeychainContents:]: 'unrecognized selector sent to class '

Question

When I try to use the MLModelInterpreter from Firebase to load a tflite-File in my Xcode project I'm getting the following error after the Launchsreen is fully visible:

*** Terminating app due to uncaught exception 'NSInvalidArgumentException', reason: '+[FIRInstanceIDCheckinPreferences preferencesFromKeychainContents:]: unrecognized selector sent to class 0x10235ca58

My code is based on the tutorial from Firebase, thus I’m surprised that it doesn’t work. Here is the code:

import Firebase


    var interpreter: ModelInterpreter?

    func convertTheModel(){


        guard let modelPath = Bundle.main.path( forResource: "model", ofType: "tflite", inDirectory: ""
            )
            else {
                print("not able to load model")
                return

        }

        let localModel = CustomLocalModel(modelPath: modelPath)
        interpreter =  ModelInterpreter.modelInterpreter(localModel: localModel)

        let ioOptions = ModelInputOutputOptions()
        do {
            try ioOptions.setInputFormat(index: 0, type: .float32, dimensions: [1, 22])
            try ioOptions.setOutputFormat(index: 0, type: .float32, dimensions: [1, 1])
        } catch let error as NSError {
            print("Failed to set input or output format with error: \(error.localizedDescription)")
        }



    let inputs = ModelInputs()
    var inputData = Data()

        do {
            for _ in 0 ..< 22 {

              // require fixed-point values or the original bytes.
              var RandomNumber = Float.random(in: 0 ..< 1)

              // Append normalized values to Data object.
              let elementSize = MemoryLayout.size(ofValue: RandomNumber)
              var bytes = [UInt8](repeating: 0, count: elementSize)
              memcpy(&bytes, &RandomNumber, elementSize)
              inputData.append(&bytes, count: elementSize)
          }
          try inputs.addInput(inputData)
        } catch let error {
          print("Failed to add input: \(error)")
        }


        interpreter!.run(inputs: inputs, options: ioOptions) { outputs, error in
            guard error == nil, let outputs = outputs else { return }

            print(outputs)
        }

    }

After some debugging I have a vague guess that the error is caused by this line of code: interpreter = ModelInterpreter.modelInterpreter(localModel: localModel)

Additional Information

Before the issue described above I had another issue: When building the project the compiler showed this error:

/Users/Path/Application_Name/Update/Pods/GoogleMobileVision/Detector/Frameworks/GoogleMobileVision.framework/GoogleMobileVision(VisionExtension.pbobjc_0848a2b53cac8a49ea32ab4e6cb931d4.o)
ld: 46 duplicate symbols for architecture arm64
clang: error: linker command failed with exit code

Here is a summary of what I did:

1. I cleaned the buildingfolder
2. Reinstall all pods: pod deintegrate and than pod install
3. set No Common Blocks to no
4. remove all -ObjC as in this Link described.

My Podfile is:

platform :ios, '13.0'

target 'My Project' do
  # Comment the next line if you don't want to use dynamic frameworks
  use_frameworks!

  # Pods for My Project
  pod 'Firebase/Database'
  pod 'Firebase/Auth'
  pod 'GoogleMobileVision/FaceDetector'
  pod 'Firebase/Firestore'
  pod 'Firebase/MLModelInterpreter'


end

I would appreciate if anyone can help

Edit

If anybody has an other solution to load an tflite-file (or an h5-file: check out my previous Question) into an Xcode project it would be also helpful

@ Paul Beusterien

This Is the location where I pasted the -ObjC:

Answer 1

-ObjC must be in the app's Other Linker Flags for Firebase to work correctly. See https://github.com/firebase/firebase-ios-sdk/issues/4776 for more details about this issue.