a scuffed voice model, but improved.
trained on 3 mins and 20 seconds of her speaking from the fourth and fifth episodes (xminus mel-roformer + rx de-reverb)
cv. Yurika Kubo
pitch extraction: rmvpe
steps: 11.6k/11.7k
batch size: 4
pretrain: klm 4.2 / 40k
huggingface: https://huggingface.co/sxndypz/rvc-v2-models/resolve/main/nekoyamada_nekov2.zip