Hayeong

icon indicating a website link https://scholar.google.com/citations?user=Jw3X6UgAAAAJ&hl=en&oi=ao icon indicating an email [email protected]

icon location Seoul, Republic of Korea icon location Speech Synthesis (TTS, VC), Generative Models | AI Researcher

Results 2 comments of Hayeong

Not possible to build

Hello, I was able to build the code for that repo quite well with another version as well. Would you like to try using this? ``` torch==1.13.1+cu117 torchaudio==0.13.1+cu117 ```

The speech file generated by cosyvoice.inference_zero_shot does not match the order of the words. The generated audio file is basically wrong.

@cckamiya I encountered a similar issue. When I trained a model initialized with Qwen-2.5 and the training did not converge well, the word order became misaligned and some words were...