search

tsurumeso / vocal-remover

Vocal Remover using Deep Neural Networks
MIT License
1.47k stars 215 forks source link

Colab issue , Is there any way to set the segment size? #167

Open Difkyear opened 7 months ago

Difkyear commented 7 months ago

After running colab for 2 minutes, the result is that the running program is automatically terminated. loading model... done loading wave source... done stft of wave source... done ^C

My audio file is an hour old and seems to exceed the RAM limit. Is there any way to set the segment size?

Difkyear commented 7 months ago
image
Goldp-js commented 7 months ago

I have 16 Gb of local RAM and it only uses 8 gb (7.80) for inference

KatieBelli commented 6 months ago

After running colab for 2 minutes, the result is that the running program is automatically terminated. loading model... done loading wave source... done stft of wave source... done ^C

My audio file is an hour old and seems to exceed the RAM limit. Is there any way to set the segment size?

If you haven't found a solution yet, this one should work:

Install Ultimate Vocal Remover GUI (Github) on your PC. Import your pth model to the model folder of VR Architecture and choose the right parameter option. Now you can choose your model, its segment size and run it on several audio files. This could work for long audio, too. It's worth a try.