FireRedTTS

FireRedTTS copied to clipboard

comfyui节点管理器无法安装这个缺失节点，其他text节点都无法连接。。

1

 comfyui节点管理器无法安装这个缺失节点，其他text节点都无法连接。。

ss958

english

1

is it working with english?

ekswathi

1. Fixed Clone the repo shell's cloning command. 2. Fixed grammar mistakes in 'Tips'. 3. Made 'Tips' sounding more professional.

SiddharthGuptaPyDev

希望會有

ioritree

I tested the CER of the pretrained model, the cer i get is 6.09 in char level，it is not as good as claimed in ther paper, and has many “delete”...

Liujingxiu23

hello，a great job。 I have some questions after reading this paper。 1、what is acoustic tokens？ 2、must train a multi-stream LM？what is its structure？ and melCodec？

Pydataman

请求cURL： `curl --location 'http://172.21.50.136:8081/v1/audio/speech' \ --header 'Content-Type: application/json' \ --data '{ "input": "小红书，是中国大陆的网络购物和社交平台，成立于2013年6月。", "response_format": "mp3", "use_tn": true }'` output文件只有“成立于2013年6月” 缺前边的字这怎么解决，是模型的问题吗

flystar8

我照着Basic Usage里面的结果跑的。我看fireredtts/models/fireredtts.py文件下实际应该转换的是prompt_text + text，但我发现转换为语音后总是缺几个字。 [example.wav](https://github.com/user-attachments/files/23156319/example.wav)

caozhenxiang-kouji

Does FireRedTTS support Chinese dialect synthesis?

1

Superman-Valencia

AttributeError: 'NoneType' object has no attribute 'detach' If it's in Chinese, the inference is possible, but in English, the error mentioned above will be reported

1simpleuser