How to Get Dutch Output for Dutch Audio Using Whisper Model via Hugging Face Inference Endpoint?

Inference Endpoints on the Hub
1

Hello Hugging Face community,

I’m working with the Whisper model for a speech-to-text task, specifically handling Dutch audio files. I’m utilizing the dedicated inference endpoint provided by Hugging Face. However, I’m facing a challenge in ensuring that the transcription output aligns with the Dutch language of the audio input.

Below is the code snippet I’m currently using:

import requests

API_URL = "-----"
headers = {
    "Accept": "application/json",
    "Authorization": "Bearer ----",
    "Content-Type": "audio/wav"
}

def query(filename):
    with open(filename, "rb") as f:
        data = f.read()
    response = requests.post(API_URL, headers=headers, data=data)
    return response.json()

output = query("rec.wav")
print(output)

This setup successfully sends the audio file to the Hugging Face inference endpoint and receives a response. However, the language of the output transcription is not in Dutch but in Chinese.
How can I add a language code for dutch?