main branch may have a streaming issue

[HelloSzymon]

Hello,

I have i think weird problem. Code taken from ContentView of the SDK is not transcribing the voice. The contentView code was tested on iphone simulator/ iphone real device and apple vision pro simulator. On all same results so some random text:

I also tried with basic implementation which is provided in documentation and print(transcription) returns me nil. For now i do not know what and where i might find problem. Package depndencies is set to main branch. Audio file is set correctly.

import SwiftUI
import WhisperKit

struct ContentView: View {
    var body: some View {
        VStack {
            Image(systemName: "globe")
                .imageScale(.large)
                .foregroundStyle(.tint)
            Text("Hello, world!")
        }
        .padding()
        .onAppear{
            Task {
               let pipe = try? await WhisperKit()
                let path = Bundle.main.url(forResource: "MP3sample", withExtension: "mp3")!.absoluteString

               let transcription = try? await pipe!.transcribe(audioPath: path)?.text
                print(transcription)
            }

        }
    }

[atiorh]

@HelloSzymon

• Please depend on a release tag instead of main if this is for your product.
• We will create a new release this week with the latest main.
• For development, please report the issue with your OS version and real device name.
• I recommend testing "Transcribe > From File" (as opposed to streaming) first so we can narrow down your issue to the mic input source or the inference engine.
• Finally, the title of the issue is not descriptive.