search

jarredou / MVSEP-MDX23-Colab_v2

Colab adaptation of MVSep Model for MDX23 music separation contest
271 stars 43 forks source link

UnboundLocalError: local variable 'vocals3' referenced before assignment #18

Closed changqingonly closed 7 months ago

changqingonly commented 7 months ago

Traceback (most recent call last): File "/$path/mvsep_v2/inference.py", line 978, in predict_with_model(options) File "/$path/mvsep_v2/inference.py", line 862, in predict_with_model result, sample_rates = model.separate_music_file(audio.T, sr, i, len(options['input_audio'])) File "/$path/mvsep_v2/inference.py", line 699, in separate_music_file vocals_high = lr_filter(vocals3.T, 12000, 'highpass') UnboundLocalError: local variable 'vocals3' referenced before assignment

branch: v2.4

deton24 commented 7 months ago

Provide used settings and song length.

jarredou commented 7 months ago

--use_InstVoc must be always set in command line if you're using directly inference.py without the related colab notebook. It is required whatever other model combination is chosen. InstVoc is used in all cases for high frequency band of the multiband ensemble. I'll add a safeguard for that for CLI usage (which I sometimes forget to fully test, as I'm using the notebook on Colab)

changqingonly commented 7 months ago

--use_InstVoc must be always set in command line if you're using directly inference.py without the related colab notebook. It is required whatever other model combination is chosen. InstVoc is used in all cases for high frequency band of the multiband ensemble. I'll add a safeguard for that for CLI usage (which I sometimes forget to fully test, as I'm using the notebook on Colab)

it's worked