Describe the feature you'd like to request
It is easily possible to pass additional parameters to the text generation model via the configuration GUI. However, such an option does not exist for image generation, transcription or TTS models. Particularly useful in transcription would be the ability to pass "diarization": "true".
Describe the solution you'd like
Add an equivalent input box for the other tasks in the Administration page.
Describe alternatives you've considered
N/A