Robert Bogan Kang

https://www.researchgate.net/profile/Zuheng-Kang [email protected]

@FirBo-Music-Studio Xiangyang hello rbk!

Results 2 comments of


                                            Robert Bogan Kang

wavlm extract features failed in loop

I tried a lot, and find the answer. Reference from [https://github.com/microsoft/UniSpeech/tree/main/downstreams/speaker_verification](url). We should add `with torch.no_grad():` when extracting features. For example: ```python with torch.no_grad(): feature = (audio) ``` In your...

RuntimeError: CUDA error: out of memory

I tried a lot, and find the answer. Reference from [https://github.com/microsoft/UniSpeech/tree/main/downstreams/speaker_verification](url). We should add `with torch.no_grad():` when extracting features. For example: ```python with torch.no_grad(): feature = (audio) ```