-
## Problem Overview
Currently it takes about 5-6 seconds to generate an audio below 10 seconds with prompt audio about 10+ second on one 3090 ti. It takes about 12G VRAM and 100% GPU util. So seems n…
-
### Description
We have to train the current mms model on multi speakers.
### Completion Criteria
A mms tts model with multi-speakers option using speaker id.
-
-
I am by no means a front-end developer, but I would like to provide a simple interface application with which my Thorsten voices can be easily used.
### Requirements:
- [ ] A single code base and …
-
got this node from comfyui manager, download & requirement installation are good, but failed to load. tried to uninstall and re-install many times, still no luck.
the log says:
Cannot import E:…
-
启动webui界面时python webui.py --port xxx --model_dir xxx，这种方式只能加载固定的单个模型，是否有什么方式能够同时加载多个模型，或执行不同TTS任务时选择模型呢？
-
### Self Checks
- [X] I have searched for existing issues [search for existing issues](https://github.com/langgenius/dify/issues), including closed ones.
- [X] I confirm that I am using English to su…
-
我是用官方文档的命令启动webui
`python3 webui.py --port 50000 --model_dir pretrained_models/CosyVoice-300M`
启动后界面可以正常打开，如下
上面是我的配置，点击运行后提示错误，终端输出如下：
```
2024-10-28 11:15:36,546 INFO get zero_shot inferenc…
-
Hello, thank you so much Mr @Nitrino for the amazing sensational revolutionary amazing exceptional life saver the great Easysubs extension! 💯 🥇
and a special thank and immense gratitude to great M…
-
### Describe the bug
It is not possible to run the server.py with the fairseq
### To Reproduce
```sh
docker run --rm -it -p 5002:5002 --platform linux/amd64 --entrypoint /bin/bash ghcr.io/idiap/co…