Differentiable, trainable versionBy @yusuke_kizunaA differentiable version of the throat model that can be trained to match a voice.The original on GitHubMore tools for CHIMERA on MetaChimera.