Introducing the Nanakoe Nina Voice Avatar - powered by our state-of-the-art RVC v2 model, with 700 epochs of intensive training. At Weights, we've meticulously trained this AI Voice model, using two minutes of original speaking samples from Yuki Takada, striving to recreate the unique texture and tone of Nanakoe Nina's voice. Our innovative pitch extraction system - RMVPE, is employed meticulously through 9.7k steps to ensure unmatched audio clarity. We've leveraged a batch size of 4 and regular v2 pretraining over 40k to provide a seamless and realistic AI Voice experience. This model speaks volumes of the strides being made in AI Music and showcases the powerful applications of RVC models in transforming audio experiences. Our AI technology presents a brilliant opportunity to re-imagine musical possibilities by creating stunning AI Covers. Invite you to explore our free AI tools and start creating captivating text-to-speech conversions and AI Covers, for a truly immersive audio experience.

we brought this bitch back to life let's goooo trained on 2 minutes of her speaking cv. Yuki Takada pitch extraction: rmvpe steps: 9.7k batch size: 4 pretrain: regular v2 / 40k huggingface:

New Creation

Nanakoe Nina (VOICE AVATAR)

Audio