FunAudioLLM/CosyVoice

Issues

• 在参考run.sh微调cosyvoice 2.0时，发现qwen_pretrain_path是空的

  #748 opened 5 days ago by hixiaoxiong
  1

• [Bug] Tags < and > are removed during inference when text_frontend=True

  #743 opened 5 days ago by youngercloud
  1

• cosyvoice 2.0感觉好牛，但是试了一下推理速度太慢了，不知道有没有后续优化的计划？

  #739 opened 6 days ago by Jimmy-L99
  3

• Cosyvoice 2.0 has better non streaming performance, but the inference speed is too slow. Can we use an inference engine?

  #747 opened 5 days ago by Demon2958
  0

• CosyVoice 2.0 在此prompt音频下乱说话

  #740 opened 5 days ago by tomjamescn
  2

• CosyVoice2 会出现掉字的情况

  #746 opened 5 days ago by gaspire
  2

• 请问150ms的流式延迟实验条件是什么

  #723 opened 7 days ago by wanghuihhh
  9

• How to set output duration when using oneshot cross langual cloning?

  #741 opened 6 days ago by lucasjinreal
  2

• [BUG] `rtf` Greater Than 1 on V100 GPU

  #730 opened 7 days ago by SolomonLeon
  1

• CosyVoice2-0.5B does not have spk2info.pt

  #729 opened 7 days ago by ZxnSnowy
  7

• speech_tokenizer_v2.onnx GPU推理很慢

  #742 opened 6 days ago by zchoi
  0

• 更新了之后运行示例代码报错 KeyError: 'additional_special_tokens'

  #724 opened 7 days ago by laishujie
  10

• 使用CosyVoice2-0.5B模型webui预训练音色不显示

  #738 opened 6 days ago by Jandown
  4

• ttsfrd-0.4.2-cp310-cp310-linux_x86_64.whl 这个对应的 windows11下的文件是那个 是否可以给个下载地址

  #737 opened 6 days ago by anstonjie
  2

• 使用gitee上的操作运行, 已经启动了, 但是推理的时候还是报错AttributeError: 'ConditionalDecoder' object has no attribute 'static_chunk_size'

  #732 opened 7 days ago by 13576245149
  8

• zero-shot推理

  #718 opened 7 days ago by wwfcnu
  6

• can MBP M3 Pro run this cosyvoice2? can not install requirements.txt.

  #731 opened 6 days ago by hildazzz
  2

• pip install -r requirements.txt failed on MacOS for CosyVoice 2.0

  #734 opened 6 days ago by tonyjohnvan
  0

• about cosyvoice2

  #727 opened 7 days ago by shanhaidexiamo
  3

• cosy 2.0 0.5b disappearance on hf

  #733 opened 6 days ago by darkacorn
  0

• zero-shot voice cloning 支持生成广西方言吗？

  #716 opened 7 days ago by APeiZou
  0

• instruct text 支持

  #726 opened 7 days ago by Hunterhuan
  0

• AttributeError: 'NoneType' object has no attribute 'create_execution_context'

  #720 opened 7 days ago by donstang
  1

• KeyError: "tag" when running examples/libritts/cosyvoice/run.sh

  #714 opened 10 days ago by GrahamHuang3
  1

• transformer is not installed, please install it if you want to use related modules

  #711 opened 9 days ago by bk111
  1

• 能否在语音合成的同时出srt字幕，还是得合成完后再whisper识别出来呢？

  #712 opened 9 days ago by bk111
  1

• 如何实现批量并发执行？

  #713 opened 10 days ago by chenzhongzheng
  1

• 在运行run.sh进行到推理阶段时，在inference.py中调用CosyVoiceModel类创建对象时少传入了一个bool类型的参数是怎么回事呢？

  #708 opened 12 days ago by Nuyoah111111
  4

• 微调时，评估的loss一直上升

  #709 opened 12 days ago by wxzyd123
  1

• 多个请求的时候gpu利用率100%卡死

  #706 opened 13 days ago by xuejiale929
  3

• 在运行run.sh进行到推理阶段时，在inference.py中调用CosyVoiceModel类创建对象时少传入了一个bool类型的参数是怎么回事呢？

  #707 opened 12 days ago by Nuyoah111111
  0

• webui generates voice messages all the time

  #705 opened 13 days ago by ZxnSnowy
  1

• Several lines that frequently cause endless loops

  #704 opened 14 days ago by cksqs
  1

• can i retrain the flow and llm based on a new pre-trained vocoder? or further fine tune?

  #702 opened 17 days ago by hildazzz
  3

• 需要流式输入和流式输出

  #700 opened 17 days ago by Hkaisense
  3

• 是否有什么办法可以提高模型不陷入死循环的稳定性

  #701 opened 17 days ago by TheHonestBob
  2

• 音频生成失败，这是什么问题

  #703 opened 16 days ago by HuangBrant
  1

• Speech Tokenizer是否会开源torch推理代码

  #699 opened 18 days ago by Maokui-He
  1

• 如何禁止ttsfrd 打印 tn xxx to xxx 的log

  #698 opened 18 days ago by vra
  1

• 使用Instruct模型生成的音频内容非指定的tts_text，而是instruct_text

  #697 opened 19 days ago by nianniandoushabao
  1

• 配置bug

  #696 opened 19 days ago by Guanjunyun
  7

• python3执行没反应，python执行报错

  #695 opened 19 days ago by iloveuaa
  6

• 如何开启LOG日志，启动无反应

  #690 opened 19 days ago by iloveuaa
  2

• Win11 按文档安装启动webUI无任何反应 直接跳出了

  #689 opened 19 days ago by iloveuaa
  4

• 声音总是变音色变性

  #688 opened 20 days ago by SPRCSY
  2

• 模型支持商用么

  #692 opened 20 days ago by the-loki
  1

• why not to opensource checkpoint file of hifigan discriminator?

  #691 opened 20 days ago by CrisYangBW
  1

• 分段文本连续调用CosyVoice-300M的tts推理，音频后段电音越来越严重。

  #686 opened 21 days ago by ZzMLvzZ-792998470
  4

• 有4080显卡，却只使用CPU是怎么回事？GPU使用率一直是0%。

  #687 opened 22 days ago by ShawnZeng
  1

• ERROR: ttsfrd-0.3.6-cp38-cp38-linux_x86_64.whl is not a supported wheel on this platform.

  #685 opened 24 days ago by LIUKAI0815
  1