Silence Suzuka (Uma Musume Pretty Derby)

Create

Silence Suzuka (Uma Musume Pretty Derby)

JapaneseAnimeRVC V2Fictional
maryishi user image
maryishi
1 year ago
👀

3.6k

👍

43

🪄

1.2k

Description

thanks to chinosk6 for their umamusume-voice-text-extractor to extract voices from Silence Suzuka! Credits to Uma Musume Pretty Derby for its drama-delivering scenes and interesting gameplays! Feel free to give credits to me if you want.

Comments

sxndypz user image
sxndypz
1 year ago

i made the model first

sxndypz user image
sxndypz
1 year ago

anyway how long is the dataset?

maryishi user image
maryishi
1 year ago

about 17 minutes of her voices of story, races, team, etc.

maryishi user image
maryishi
1 year ago

Where can I find your Silence Suzuka model?

sxndypz user image
sxndypz
1 year ago

me too

sxndypz user image
sxndypz
1 year ago

it's in weights.gg

maryishi user image
maryishi
1 year ago

Oof, my apologies of making another one.

sxndypz user image
sxndypz
1 year ago

it's aight

sxndypz user image
sxndypz
1 year ago

what batch size?

maryishi user image
maryishi
1 year ago

Can't see it, could you give the link?

maryishi user image
maryishi
1 year ago

I forgot lol, probably 7 iirc.

sxndypz user image
sxndypz
1 year ago

my model is a bit

sxndypz user image
sxndypz
1 year ago

softer

sxndypz user image
sxndypz
1 year ago

compared to yours

maryishi user image
maryishi
1 year ago

Holy moly, I didn't know it was a bit same information about the 17 minutes one

sxndypz user image
sxndypz
1 year ago

did you train your model on colab or local?

maryishi user image
maryishi
1 year ago

On colab, used Minatolsuki's RVC training.

sxndypz user image
sxndypz
1 year ago

how did you not get disconnected?

maryishi user image
maryishi
1 year ago

I paid.

sxndypz user image
sxndypz
1 year ago

i see

maryishi user image
maryishi
1 year ago

Now I regretted of making her vocal since I didn't know there was the original one 😢

sxndypz user image
sxndypz
1 year ago

i mean

sxndypz user image
sxndypz
1 year ago

yours is better

sxndypz user image
sxndypz
1 year ago

there's nothing to regret, really

maryishi user image
maryishi
1 year ago

Thanks, but my instincts tell me that lower epochs could be more accurate than higher epochs.

sxndypz user image
sxndypz
1 year ago

your model is better at singing at least

maryishi user image
maryishi
1 year ago

Ah, I didn't test it for speaking if it works lol.

mustar22 user image
mustar22
1 year ago

<a:KaedeLoudButFast:522364705768079360>

maryishi user image
maryishi
1 year ago

Apologies for a sudden message, but where did you train your ai model?

sxndypz user image
sxndypz
1 year ago

here

sxndypz user image
sxndypz
1 year ago

but the post got deleted due to uh... a random error in the channel itself

Add a comment

Samples

New
Classic
1. Singing
Male
English
2. Singing
Female
English
3. Singing (Dry)
Female
English
4. Singing (High)
Female
English
5. Singing 2
Male
English
6. Singing (Dry)
Male
English
7. Singing (Dry, High)
Male
English

Pitch

More to explore

Loading more

Selected Audio
Selected Audio