克隆后的音频播放时，杂音很重

**合成后的杂音很重怎么处理**

**Env & To Reproduce[复现与环境]**
encoder: pretrained_bak_5805000
ppg_extractor：24epoch.pt
ppg2mel：ppg2mel.yaml ppg2melbest_loss_step_322000.pth
vocoder_hifigan_24k：hifigan_24k.pt config.json

用到的模型如上所示，我运行了run.py得到合成后的音频，播放时发现杂音非常大，合成的音色和给定的音色还算接近，如何处理掉杂音问题？
另外我从README的2.3中下载了社区提供的预训练合成器，但不知道这个模型要用在何处？只有运行toolbox时指定了这个模型，run.py不需要用到吗？

烦请解答，谢谢！

Start with README section 2.3, run.py, and the toolbox entry point, then reproduce the reported noise with the listed encoder, PPG extractor, ppg2mel, and HiFi-GAN checkpoints and config. Compare how each entry point loads the models and document the compatible model usage and the conditions under which the synthesized audio is clear.

Written by the indexing model from the issue text.