Junity

> You can also post this issue to https://github.com/jaywalnut310/vits Yeah, but I think there is no contributor maintaining that repo, so I want to try if some repos based on...

> Very curious too. At first glance I thought about optimizations given the KL characteristics, but after scrambling over some papers about KL optimization and approximation I didn't find anything...

> Very curious too. At first glance I thought about optimizations given the KL characteristics, but after scrambling over some papers about KL optimization and approximation I didn't find anything...

> Hi @JunityZhan, sorry for the ping. Did you learn something new regarding this issue from the other VITS based repos? Still no, I didn't find any related information and...

These codes you are editting the parentheses are written by me, I think it will cause problems. In my environment, if the paranthesis is removed, then "-pg" and "-pd" will...

Thank you for your patience, I actually means for this line, ` "-pg %s" % pretrained_G14 if pretrained_G14 != "" else "", ` HOWEVER, I tested it and its totally...

> @JunityZhan You should let users see this exit info in WebUI info textbox. Not all users know console UI.  Hi, It actually can show info in the textbox,...

> 一下这么多，瑟瑟发抖 > PPG去年就试过，跑垃圾数据跑不起来的 我觉得不只是PPG，主要是提供了一个扩展的空间吧。（嗯，毕竟是可选，可以让用户默认hubert就好）。 我自己测试了一下，在inference用错模型了（这里我忘记做了，可以后面再说） 用ppg训练，然后inference的时候实际上是用hubert转换的，然后效果也就差一点点，因为inference用错模型了，所以我还是没法断定ppg就是不行，我可能还要再做点实验。 但主要还是通过一个point wise的投影层来让模型可以符合更多的speech representation模型。也增加了扩展性吧。 如果这个功能没问题的话我可以继续维护一下更多的speech representation模型加到里面。 然后请问您有没有了解过隔壁的diffusion model提升音质，我也在考虑整合进去，您怎么看？

> 你在缝的功能大多是已经经过了实验证明没用的，增加选项除了让用户懵逼外没啥大用，目前对于RVC来说基本是最优配置 diff是唯一可能和baseline互有优劣的选项 符合更多模型：没用的，用A训，用B推理本来也是魔幻行为；多encoder一起上：音色泄露的强度只会挑泄露最多的那个来；单encoder：直接对上输入通道就行了，没必要多加卷积层 可以提交到RVC-Project或者RVC的其他分支，但是肯定不会一起合并到主分支 好，我先关掉这个PR了