xianyuxm

> The specific reason may be that **llama.cpp/convert-hf-to-gguf.py** encountered issues during the rapid iteration process. I experienced the same problem when exporting and quantizing qwen2 in the [latest version of...

I encountered the same problem while running in the Windows 10 environment with Python 3.11. However, when using the WSL environment, I was able to install the dependencies successfully.

> 阅读相关源码发现流式的TTS服务返回的是pcm音频的base64，需要将base64 -> pcm byte array -> wav file，相关源码见paddlespeech/server/utils/audio_process.py 的save_audio函数 多谢大佬的分享，确实是这个原因。关键代码如下： ` for wav in self.infer( text=sentence, lang=self.config.lang, am=self.config.am, spk_id=spk_id, ): # wav type: float32, convert to pcm (base64) wav...