Name: Kayn (League of Legends - EN) - RIN_E3 | AI RVC Model
Brand: Weights
Rating: 4.4 (16 reviews)

WebKit

 Introducing Kayn (League of Legends - EN) RVC V2 - RIN\_E3 | RMVPE, an advanced AI Voice Model powered by Retrieval-Based Voice Conversion (RVC) technology. With over 1640 epochs of training, this state-of-the-art model provides highly accurate text-to-speech capabilities, perfect for creating AI covers or generating realistic voiceovers for your projects.

Our League of Legends themed Kayn model has been specifically designed to deliver engaging, authentic experiences for fans of the popular gaming franchise. Through the use of RVC technology, we have ensured precise and natural sounding speech synthesis, making it easy to integrate into various applications such as video games, animations, and virtual assistants.

Furthermore, this model excels at producing high-quality AI music, allowing you to generate customized compositions based on its extensive vocabulary drawn from over two minutes of rips from real matches featuring regular Kayn dialogues. By harnessing the power of artificial intelligence and deep learning algorithms, Weights empowers creators like you to craft immersive audio experiences effortlessly.



My most trained model yet... (and yes, graphs were good, can show) ~2mins, ripped from game, regular kayn (not shadow or rhaast) Note: I don't think singing is the strongest suit for the model, but the testing audios I've got aren't cleaned up that well so... :shrug: If it doesn't sound good, then you likely need to lower transpose/pitch.

Kayn (League of Legends - EN) - RIN_E3

Kayn (League of Legends - EN) | RVC V2 - RIN_E3 | RMVPE | 1640 Epochs

talktalkrouter

yigit halil

Just saw that @Grimoire had recently also made a model lol ( I think theirs is probably better at singing than my model. Dataset is around 19.5 mins from this video: It's likely that you will need to adjust transpose/pitch.

Alan Wake (Alan Wake 2) - RIN_E3

Alan Wake (Alan Wake 2) | RVC V2 - RIN_E3 | RMVPE | 390 Epochs

Using RIN_E3, batch size 4 (prob should have done 8, this took way too long lol). Dataset audio (cleaned up after) comes from this video: Sounds much better than my previous model:

Smolder (League of Legends - EN) - RIN_E3

Smolder (League of Legends - EN) | RVC V2 - RIN_E3 | RMVPE | 425 Epochs

Trained using rmvpe. Dataset made from voice lines ripped from game. Using rmvpe is recommended. Einhar has a deeper voice, so if output doesn't sound like Einhar, set transpose to a lower value. Credit would be appreciated.

Einhar (Path of Exile)

Einhar (Path of Exile) RVC v2 - 250 Epochs

Trained using mangio-crepe. *32 seconds* of usable voice lines ripped from the game Using model with mangio-crepe/rmvpe + default settings works fine Credit would be appreciated

Kayn (League of Legends - EN) - RIN_E3

Description

Samples

Saiba Momoi (Blue Archive)

Adel imam (Egyptian Actor) عادل امام (ممثل مصري)

JP Furina / Focalors CV: Inori Minase (Genshin Impact)

Ariana Grande AI

Hu Tao JP (Genshin Impact)

JENNIE of BLACKPINK [Strong Ver.]

Saiba Momoi (Blue Archive) (VA: Tokui Sora)

Hatsune Miku

SpongeBob SquarePants (Talking And Singing)

Takanashi Hoshino (from Blue Archive)

Satoru Gojo (JJK) [VA Yuichi Nakamura]

ENHYPEN Heeseung

Sunaokami Shiroko (Blue Archive)

Goku (Mario Castañeda) (Latin American Spanish) (UPDATE)

Villager (Minecraft)

Mortis [Brawl stars]

Jungkook (BTS)

Melodie - (Brawl Stars) - (English & Korean Voices)

Tendou Arisu (Blue Archive)

Kanye West