Mirei Asaoka / Yoshiko Saki [朝岡実嶺 / 崎芳子] (RVC v2 - 2 versions & 300 Epochs)

Sincere greetings to all.

This might be your unusual “Request a Model” format, but I think that there should have been the “Actors / Actresses” category to add in this website, in the near future.

Yes, this is my very first AI voice model to release under my own work (and being the '1st test subject" for using RVC v2 technology - of course by me; but later being revised in several days due to strict “quality checks and control”.) - and hearing her very young voice back in her mid-teens up until I think late 30-ish (late 1980s to mid-2000s).

For those who don’t know, Mirei Asaoka (Japanese: 朝岡実嶺; real name: Yoshiko Saki (崎芳子)) is a former Japanese (top-class) sexy actress from 1991 to 1993 (and probably extended until the early 2000s during her normal acting career). She appeared in various movies and TV shows like…

• Juzo Itami’s 4 movies:
•• Minbo no Onna (1992)
•• Daibyoin (1993)
•• A Quiet Life (1995)
•• Marutai no Onna (1997)

• High School Teacher [Kōkō Kyōshi] (1993)
• Miseinen / Minor (1995)
••• and many more to mention here...

Indeed, even though there were struggles in her career, she’s a “blessing in disguise” for most of her fans (including long-time).

Image references:

Sample demonstration(s):

(↑ [SPOILER!] Jump to 26:04 for the voice sample ↑)

(↑ What if… she had the singing ability back in 1990s, or during her “best prime” years?)

There are 2 versions you may try (and have fun I guess. I also put the “Instructions_Directions.txt” file to provide much easier tutorial-esque than having nothing - so you may read it as well):

1. Normal tone:


300 epochs
Trained for 3 minutes and 54 seconds

2. High-pitched:


300 epochs
Trained for less than a minute, 36 seconds

•• All trained via Google Colab (local training), not using Gradio UI, due to restricted use for free users (as an author myself that is a free user)

Finally, if there’s anything else to say (including some complaints and whatever), then please feel free to comment and I’ll do my very best to provide you as soon as possible. After this, I might follow suit (to upload more AI voice models in the future with risks) indeed.

Thanks and have a great day!