Logo
Sign in
Sign up

Sapphire "Midori" Kawashima (Sound! Euphonium) [RVC v2; RMVPE; 360 Epochs]

Created: June 12, 2024
Sapphire "Midori" Kawashima (Sound! Euphonium) [RVC v2; RMVPE; 360 Epochs]

This voice model is Sapphire Kawashima (CV: Moe Toyota) from the anime adaptation of the novel "Sound! Euphonium". She is a contrabassist and prefers to be called Midori. She's a mood maker for the Kitauji High School Concert Band and helps to keep the club's atmosphere warm. It can be challenging to handle such a large instrument, but she believes it plays a big role in the music.

The character descriptions are AI-generated text, so they may not be an accurate representation of their personalities. If you're interested, I recommend heading over to the official site for more details.

The model was trained locally on my machine using two GPUs, the RMVPE pitch extraction algorithm and the Titan Medium pretrain against Japanese HuBERT with a batch size of 4. Thus, you should replace your local HuBERT with the Japanese one for the best results.

As it currently stands, the singing dataset uses excerpts from her character song and anime ending themes; the speaking dataset uses dialog from episodes of the first season, including specials. I plan to expand the dataset from all three seasons. I've also cooked up another model using the KLM pretrain depending on your use case and preferences. Please credit me as **TheTrustedComputer** if you use this model; I highly appreciate it. Have fun!

https://huggingface.co/TheTrustedComputer/RVC_Models/resolve/main/MidoriKawashima_TitanMed_JPHuBERT.zip
https://huggingface.co/TheTrustedComputer/RVC_Models/resolve/main/MidoriKawashima_KLM4_JPHuBERT.zip

Additional Details:

Tags: No tags available

Download Link: https://huggingface.co/TheTrustedComputer/RVC_Models/resolve/main/MidoriKawashima_TitanMed_JPHuBERT.zip Hosted on Hugging Face

Convert this model in EasyAIVoice

Sample deck

Pitch previews

Listen across generated pitch samples, then Run. Hold Ctrl or Cmd to select multiple pitches.

samples ready
female base 路 pitch 0 Generated female reference sample
Run
male
female
female-base generated sample male-base generated sample Enrichment pending: no scores or voice accords yet Ctrl/Cmd-click = select multiple pitches