search

JuanFMontesinos / VoViT

VoViT: Low Latency Graph-based Audio-Visual VoiceSeparation Transformer
https://ipcv.github.io/VoViT/
34 stars 9 forks source link

flerken version #4

Closed ahmadikalkhorani closed 1 year ago

ahmadikalkhorani commented 1 year ago

running python inference_interview.py raises the following error, which is probably due to flerken version incompatibility. I checked other versions but none of them has the requirement kwargs:

.../site-packages/flerken/models/UNet.py", line 287, in add_decoder TransitionBlock(self.dim[i], self.dim[i - 1], useBN=self.useBN, TypeError: init() got an unexpected keyword argument 'mode'

JuanFMontesinos commented 1 year ago

Hi, sorry for that. Flerken is a bit messy. I've moved the UNet used to the core model, so that it no longer depends on flerken. If you pull it should be working (I hope)

Best Juan