PaddlePaddle / PaddleSpeech

Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA/Streaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.
https://paddlespeech.readthedocs.io
Apache License 2.0
11.21k stars 1.86k forks source link

[TTS]tts server config set “lang: 'mix'” #3466

Open damon-93 opened 1 year ago

damon-93 commented 1 year ago

这个项目非常棒,但是我遇到了一些问题,想请教一下 在例子中我能顺利启动TTS server 的'zh'模型 现在我想启动TTS server的mix模式,替换了对应的频谱模型以及voc模型,确定了这两个模型存在并自动下载了(fastspeech2_mix-mix,hifigan_male-mix),但是运行报错: ' Failed to warm up on tts engine.'TTSServerExecutor' object has no attribute 'frontend'' 我想启动TTS的‘mix’模式,我该如何配置?

5b4f2c4a8d06cabfccf4d075ab53c52

damon-93 commented 1 year ago

image image image

lixikun commented 1 year ago

anyone help?