v0.1
This model is trained via extra conditioning on its own input. The next version of the model (v0.2) will have a slightly different architecture that seem to be providing deeper restoration, especially of narrow band audio.
This model is trained via extra conditioning on its own input. The next version of the model (v0.2) will have a slightly different architecture that seem to be providing deeper restoration, especially of narrow band audio.