PlayVoice/vits_chinese issues and pull requests

#100 - message = infosub[1] IndexError: list index out of range

Issue - State: closed - Opened by SRbone over 1 year ago - 7 comments

#100 - message = infosub[1] IndexError: list index out of range

Issue - State: closed - Opened by SRbone over 1 year ago - 7 comments

#99 - 模型大小

Issue - State: closed - Opened by huangmeimao over 1 year ago - 3 comments

#99 - 模型大小

Issue - State: closed - Opened by huangmeimao over 1 year ago - 3 comments

#98 - 请问bert模型prosody_model.pt每次更换自己的数据时，需要用自己的数据重新训练prosody_model.pt吗

Issue - State: closed - Opened by huangmeimao over 1 year ago - 6 comments

#98 - 请问bert模型prosody_model.pt每次更换自己的数据时，需要用自己的数据重新训练prosody_model.pt吗

Issue - State: closed - Opened by huangmeimao over 1 year ago - 6 comments

#97 - 请问具体怎么微调模型

Issue - State: open - Opened by huangmeimao over 1 year ago - 4 comments

#96 - 请问readme里提到的模型微调怎么实现？

Issue - State: closed - Opened by huangmeimao over 1 year ago

#95 - 请教速度问题

Issue - State: open - Opened by gandolfxu over 1 year ago - 4 comments

#95 - 请教速度问题

Issue - State: open - Opened by gandolfxu over 1 year ago - 4 comments

#94 - 想基于预训练模型继续训练，应该怎么做

Issue - State: open - Opened by jmaxzh over 1 year ago - 10 comments

#94 - 想基于预训练模型继续训练，应该怎么做

Issue - State: open - Opened by jmaxzh over 1 year ago - 10 comments

#93 - vocab.txt里没有的字报错

Issue - State: closed - Opened by yihuitang over 1 year ago - 2 comments

#93 - vocab.txt里没有的字报错

Issue - State: closed - Opened by yihuitang over 1 year ago - 2 comments

#92 - 关于 some Natural Speech Features Of Microsoft代码

Issue - State: open - Opened by chusevip8 over 1 year ago - 2 comments

#92 - 关于 some Natural Speech Features Of Microsoft代码

Issue - State: open - Opened by chusevip8 over 1 year ago - 2 comments

#91 - 基于VITS-fast-finetuning加入bert

Issue - State: closed - Opened by caifangvip over 1 year ago - 10 comments

#91 - 基于VITS-fast-finetuning加入bert

Issue - State: closed - Opened by caifangvip over 1 year ago - 10 comments

#90 - 怎么优化儿化音？

Issue - State: open - Opened by caifangvip over 1 year ago - 2 comments

#90 - 怎么优化儿化音？

Issue - State: open - Opened by caifangvip over 1 year ago - 2 comments

#89 - train.txt格式

Issue - State: closed - Opened by caifangvip over 1 year ago - 1 comment

#88 - 老师您好，还有一个问题，就是使用标贝数据训练时，如果我把训练数据集减小至7000条以下，就会报这个错，您能帮我分析一下吗？

Issue - State: closed - Opened by liumingda over 1 year ago - 1 comment

#87 - 老师您好，请问您迭代了多少iteration？

Issue - State: closed - Opened by liumingda over 1 year ago - 10 comments

#86 - BERT造成发音不准

Issue - State: open - Opened by wizardk over 1 year ago - 4 comments

#85 - 请问我要训练方言，需要如何训练？

Issue - State: open - Opened by zoremax over 1 year ago - 1 comment

#84 - Where is the Natural Speech Features Of Microsoft mentioned on the title “with some Natural Speech Features Of Microsoft”

Issue - State: closed - Opened by dawenxi-only over 1 year ago - 1 comment

#83 - 训练1400多轮了，合成不出任何声音是为什么呢

Issue - State: closed - Opened by jmaxzh over 1 year ago - 9 comments

#82 - 试了下效果还可以，就是默认模型不支持读数字和英文，感觉有点限制

Issue - State: closed - Opened by futureflsl over 1 year ago - 3 comments

#81 - 请教一个关于【迁移】项目的问题

Issue - State: open - Opened by aylitat over 1 year ago - 2 comments

#80 - error: could not create 'monotonic_align\core.cp39-win_amd64.pyd': No such file or directory

Issue - State: closed - Opened by aylitat over 1 year ago - 1 comment

#79 - 不支持数字吗？

Issue - State: closed - Opened by aylitat over 1 year ago - 1 comment

#78 - RuntimeError: The expanded size of the tensor (50) must match the existing size (0) at non-singleton dimension 1. Target sizes: [192, 50]. Tensor sizes: [192, 0]

Issue - State: open - Opened by m1258218761 over 1 year ago - 5 comments

#77 - 数字念不出来啊

Issue - State: closed - Opened by stevezhang88 over 1 year ago - 1 comment

#76 - baker data 需要训练多久？

Issue - State: open - Opened by 1040003585 over 1 year ago - 2 comments

#75 - 刚接触vits，请问训练自己的模型只需要把标贝的数据替换成自己的数据集就可以了吗

Issue - State: closed - Opened by vingin007 over 1 year ago - 8 comments

#74 - infer loss

Issue - State: closed - Opened by zdj97 over 1 year ago - 3 comments

#73 - 大佬，有个推理的问题请赐教！

Issue - State: open - Opened by YuruW over 1 year ago - 12 comments

#72 - Train 的vits_prepare.py 运行错误：IndexError: Replacement index 2 out of range for positional args tuple

Issue - State: open - Opened by 1040003585 over 1 year ago - 6 comments

#71 - 换了一套音素模型，但是推理的时候出错

Issue - State: open - Opened by zdj97 over 1 year ago - 6 comments

#70 - 建模单元

Issue - State: closed - Opened by zdj97 over 1 year ago - 4 comments

#69 - 一定要16k的语音吗

Issue - State: closed - Opened by zdj97 over 1 year ago

#68 - How can train on English?

Issue - State: open - Opened by nivibilla over 1 year ago - 6 comments

#67 - segment_size 对不同数据集有影响吗？

Issue - State: open - Opened by liroda over 1 year ago - 6 comments

#66 - 中英混合TTS

Issue - State: open - Opened by flame110 over 1 year ago - 4 comments

#65 - 多说话人模型发音混乱？

Issue - State: closed - Opened by hildazzz over 1 year ago - 16 comments

#64 - segment_size问题

Issue - State: closed - Opened by pgolds over 1 year ago - 6 comments

#63 - 是否有计划支持多发音人？

Issue - State: closed - Opened by alphonz over 1 year ago - 1 comment

#62 - 语速问题

Issue - State: open - Opened by TinaChen95 over 1 year ago - 3 comments
Labels: enhancement

#61 - 关于quantization

Issue - State: closed - Opened by pgolds over 1 year ago - 5 comments

#60 - 关于ZeroShot

Issue - State: closed - Opened by suzhenghang over 1 year ago - 3 comments

#59 - 请问音频采样率一定要16khz吗？

Issue - State: closed - Opened by zhao-cs-debug over 1 year ago - 4 comments

#58 - loss小的有点过分

Issue - State: open - Opened by RachelLee1010 over 1 year ago - 3 comments

#57 - Which paper is Prosody Bert based on?

Issue - State: closed - Opened by OswaldoBornemann over 1 year ago - 3 comments

#56 - 输入非常规字词生成的时候会出错

Issue - State: open - Opened by xhappy123 over 1 year ago - 8 comments

#55 - Other languges implementation?

Issue - State: open - Opened by NK990 over 1 year ago - 3 comments

#54 - 句尾吞音

Issue - State: closed - Opened by AwakingsWings over 1 year ago - 13 comments

#53 - 数据prepare包缺失

Issue - State: closed - Opened by RachelLee1010 over 1 year ago - 4 comments

#52 - 大神你好，我是tts小白，想求教些问题，希望不吝赐教，感谢

Issue - State: closed - Opened by YangangCao over 1 year ago - 6 comments

#51 - bert 模型的来源是什么呢

Issue - State: open - Opened by shuishu over 1 year ago - 1 comment

#50 - How to get Prosody pretrain model?

Issue - State: closed - Opened by rechawine over 1 year ago - 3 comments

#49 - RuntimeError: The expanded size of the tensor (50) must match the existing size (0) at non-singleton dimension 1. Target sizes: [192, 50]. Tensor sizes: [192, 0]

Issue - State: closed - Opened by showmelater over 1 year ago - 8 comments

#48 - A-Z字母和数字0-9发声不了

Issue - State: closed - Opened by showmelater over 1 year ago - 2 comments

#47 - RuntimeError: The size of tensor a (76) must match the size of tensor b (84) at non-singleton dimension 1

Issue - State: closed - Opened by jayne-ho over 1 year ago - 2 comments

#46 - 请教支持流式输出吗？

Issue - State: open - Opened by kendo6666 over 1 year ago - 23 comments
Labels: enhancement

#45 - 支持单人自定义数据集，格式兼容moegoe;支持windows训练和推理;更好的符号支持。

Pull Request - State: closed - Opened by AwakingsWings over 1 year ago

#44 - 这个项目能实现声音克隆功能吗？

Issue - State: open - Opened by liubin3702 over 1 year ago - 1 comment

#43 - hi, CLONE的稳定性和效果好像比VITS好，有计划实现下这个框架么

Issue - State: open - Opened by GuangChen2016 over 1 year ago

#42 - 关于音频中的静音段问题

Issue - State: closed - Opened by hermanseu over 1 year ago - 7 comments

#41 - 关于模型压缩

Issue - State: open - Opened by uloveqian2021 over 1 year ago - 29 comments
Labels: help wanted

#40 - 多说话人效果

Issue - State: open - Opened by suzhenghang over 1 year ago - 2 comments

#39 - 关于multispeaker改动训练问题

Issue - State: closed - Opened by pgolds over 1 year ago - 7 comments

#38 - 关于音素生成一些疑问

Issue - State: closed - Opened by suzhenghang over 1 year ago - 3 comments

#37 - 如何制作自己的数据集。

Issue - State: open - Opened by ttkrpink over 1 year ago - 3 comments

#36 - Baker Loss在18左右，声音没有预训练模型的干净

Issue - State: closed - Opened by startreker-shzy over 1 year ago - 11 comments

#35 - 能否增加：角色1、角色2。。。，请问如何修改呢？

Issue - State: open - Opened by cypinpai over 1 year ago - 12 comments

#34 - Some questions for the KL_loss and KL_loss_r and model behaviors

Issue - State: open - Opened by feng-yufei over 1 year ago - 13 comments
Labels: good first issue

#33 - Infer loss from NaturalSpeech

Issue - State: closed - Opened by wizardk over 1 year ago - 4 comments

#32 - 如何使用预训练模型开始训练？

Issue - State: open - Opened by VacantHusky over 1 year ago

#31 - Is needed to insert prosody label under Chinese data?

Issue - State: closed - Opened by asr-pub over 1 year ago - 3 comments

#30 - 支持中英混吗

Issue - State: open - Opened by hdmjdp over 1 year ago - 12 comments

#29 - Can I use this to train with 32K sampling audio data?

Issue - State: closed - Opened by ewayboy over 1 year ago - 5 comments

#28 - training for custom dataset

Issue - State: closed - Opened by huydang2106 almost 2 years ago - 4 comments

#27 - 求问微调/Finetuning？

Issue - State: open - Opened by godspirit00 almost 2 years ago - 20 comments

#26 - 请问BERT模型是如何训练的呢？

Issue - State: closed - Opened by TinaChen95 almost 2 years ago - 2 comments

#25 - 问题：推理错误，语调不对

Issue - State: closed - Opened by lucasjinreal almost 2 years ago - 4 comments

#24 - 大佬，您试一下这句话

Issue - State: closed - Opened by lucasjinreal almost 2 years ago - 1 comment

#23 - 要加韵律预测模型和文本前端，我还在梳理、会释放~星星好少，没动力更新~~~

Issue - State: closed - Opened by cat1222 almost 2 years ago

#22 - 大佬，用项目提供的数据集，训了250 epoch+时，感觉loss都不怎么下降了，这是训练到位了吗？

Issue - State: closed - Opened by cat1222 almost 2 years ago - 15 comments

#21 - stft.py line 199报错：runtimeerror: exp does not support automatic differentiation for outputs with complex dtype.

Issue - State: closed - Opened by cat1222 almost 2 years ago - 6 comments

#20 - NameError: name 'm_frame' is not defined

Issue - State: closed - Opened by DWGIII almost 2 years ago - 4 comments

#19 - question about the phone seqs

Issue - State: closed - Opened by FanhuaandLuomu over 2 years ago - 2 comments

#18 - 有出Java版的计划吗？

Issue - State: closed - Opened by watersoft123 over 2 years ago - 3 comments

#17 - 'monotonic_align.monotonic_align'

Issue - State: closed - Opened by zhengyongchoa over 2 years ago - 3 comments

#16 - 大佬我训练出来的音频语速太快，读的之间好像没有间隔。这个怎么解决？

Issue - State: closed - Opened by donkey1020 over 2 years ago - 1 comment

#15 - 我训练出来的人说话都连在一起很快。吐字不清楚

Issue - State: closed - Opened by donkey1020 over 2 years ago - 1 comment

#14 - 大佬是不是没办法训练一次然后通过调整音色音高之类的改变人物声音？？？？

Issue - State: closed - Opened by donkey1020 over 2 years ago - 3 comments

GitHub / PlayVoice/vits_chinese issues and pull requests