I would like to use voice to enter the names of people in hiragana, by default the highest confidence prediction seems to select the most plausible Kanji for that name, but it would be more useful if it gave me the hiragana (phonetic sound) of the name. Is it possible to get this output? i.e. the hiragana for the highest confidence prediction.
@holyone2-8265 Unfortunately, this setting is not available with the baseline model as the response is dependent on the baseline training data which is a mix of all the above. If you can use custom training data to create a custom model then the response can be achieved in the required transcription.
5 people are following this question.
