What should be the prompt to transcribe in another language than Chinese or English? #45

The file with example prompts only give examples in Chinese and English. I played a bit with those prompts and have difficulty on enforcing the model to transcribe in French (when the language in the audio is French).
Is there a way to specify the (source and target) language in the prompt? For example in French, what should I use for speaker-only transcription and timestamped diarization?