RVC-Project
How Can I Create A Own Pretrained Model For specific Language
How Can I Create A Own Pretrained Model For an specific Language . because train an each voice model for my language take too much of time . is this possible ?
You can certainly create your own pretrained model, which requires you to train a model with a large amount of data and then delete the emb.g_weight layer from it. However, pretrained models cannot be directly used for inference and still need to be fine-tuned with data of a specific voice. Given this, creating your own pretrained model may seem unnecessary. The task of RVC is voice conversion, allowing you to input and output in any language, so it is not necessary to create a pretrained model for a specific language.
Would pre-training be done in the same way that a voice model is trained or is it done differently with other software?
