Api not releasing memory after inference

RVC-Project / Retrieval-based-Voice-Conversion

in preparation...

MIT License

240 stars 37 forks source link

Api not releasing memory after inference #6

Open briangrider opened 7 months ago

briangrider commented 7 months ago

Hi there,

I believe I almost have this all figured out and it's working great. One issue I'm having is that after infering one time using the API, memory usage stays very high (12.7gb out of 16), even though no processing is happening. This is happening on a macbook air m1 16gb. Is there a way to force rvc to release that memory usage after each api call? Thanks in advance!

Tps-F commented 6 months ago

It is still under development and I have not addressed that yet. Fortunately, I have the same environment, so I'll fix it when I'm done implementing it for now!

mmlingyu commented 3 months ago

i am also met it.

m-pektas commented 1 month ago

I faced the same issue. I tried to profile memory using memory_profiler and a part of the output following;

Mem usage    Increment  Occurrences   Line Contents
-------------------------------------------------------
2150.9 MiB   1049.0 MiB           1        _, (sr, output) = vc.vc_single(audio_path, const.f0method0, index_path)