Use Common Voice dataset
The dataset contains many sentences and pronunciations by native Japanese speakers. Although there are no translations, it could still be very useful.
The dataset contains many sentences and pronunciations by native Japanese speakers. Although there are no translations, it could still be very useful.