Iver Jordal

Not sure if we can see the exact line that caused the issue, but it should be possible to do a `git bisect` to narrow down the issue to a...

I did a git bisect, and found that this is the "first bad commit": https://github.com/asteroid-team/asteroid/commit/51e68353f7b2f47708686d4c08b61a2ddec4dafc

I've narrowed it down a bit, only exporting encoder + decoder (i.e. without the `TDConvNet`): ``` import torch from asteroid.filterbanks import make_enc_dec class MySimpleEncoderDecoder(torch.nn.Module): def __init__(self): super().__init__() self.encoder, self.decoder =...

So maybe I should open an issue at https://github.com/asteroid-team/asteroid-filterbanks instead? What do you think @mpariente?

Or maybe not. There's this issue already: https://github.com/asteroid-team/asteroid-filterbanks/issues/10

A reverb augmentation is a welcome addition (I'd be happy to review and accept a PR), although we already have a transform for convolving with room impulse response wavs.

Maybe this is relevant? https://github.com/sweetcocoa/ddsp-pytorch/blob/master/components/reverb.py#L14

We should make examples of what the API should look like in each scenario

The last one, "Audio (length same as the input)" is supported already by doing this: ``` augment = ExampleTransform(...) augment(samples, sample_rate) augment.freeze_parameters() augment(other_samples, sample_rate) augment.unfreeze_parameters() ``` We can refine the...

Interesting! I get a feeling that it could blow up in complexity if we want to cover all use cases in one class. Maybe it makes sense to make separate...