Nan

Thanks for your great work! I kewn model code based on wenet, could you tell what are used pipline ?

No punctuation for Chinese

4

When I recognized one minute of Chinese audio, I found that there was no punctuation

I can't use language='zh' when I use large-v3

3

warning: The current model is English-only but the language parameter is set to 'zh'; using 'en' instead.

NotImplementedError

1

when I fintune fellow https://github.com/huggingface/distil-whisper/tree/main/training. NotImplementedError: The model type whisper is not yet supported to be used with BetterTransformer.

Accelerate

1

Accelerate is a tool for multi-machine，but why you use it in single gpu？

I think this is a great work ! But there is exits limit as introduction. I want to know is there any better work based on SpeechGPT recently?

About ASR

4

Thanks for your excellent work！ I want to ask how the Discrete tokenizer's perform on the ASR？Can you tell me your understand？ Thanks！

感谢LLama-Omni的工作！本着学习的目的，我们尝试复现了这篇论文的训练代码，希望能对做端到端语音对话的朋友有一点点帮助，也希望大家能提意见，一起交流，共同进步。https://github.com/wntg/LLaMA-Omni

Streaming infer

3

I want to use wavtokenizer to speech AI. Is avtokenizer apply streaming infer?