Pretrained models - netease-youdao/EmotiVoice GitHub Wiki

Two types of pretrained models are needed for running EmotiVoice:

The first type consists of TTS models that offer over 2000 voices.

You can obtain them by downloading pretrained models. Alternatively, I highly recommend using the following command to acquire them, which I highly recommend.

git clone https://www.modelscope.cn/syq163/outputs.git

Make sure to arrange the directory structure as follows:

If you encounter any difficulties, please refer to issue #27 or #21 for assistance.

Another required pretrained model is 'simbert-base-chinese'.

You can obtain it using the following commands:

git lfs install
git lfs clone https://huggingface.co/WangZeJun/simbert-base-chinese WangZeJun/simbert-base-chinese

Alternatively, you can use the following command to get it:

git clone https://www.modelscope.cn/syq163/WangZeJun.git

If you encounter any difficulties, please refer to issue #4 for assistance.