Typo in README - input tensor shape of OpenUnmix

[sevagh]

Hello, I believe the true input shape of OpenUnmix (the spectrogram model, not the on-the-fly waveform one) is this, taken from the code:

(nb_samples, nb_channels, nb_bins, nb_frames)

This corresponds to the (I, F, T) that I've seen in the oracle code (I = channels, F = frequency bins, T = time frames).

The README describes the shape in a different order:

models.OpenUnmix: The core open-unmix takes magnitude spectrograms directly (e.g. when pre-computed and loaded from disk). In that case, the input is of shape (nb_frames, nb_samples, nb_channels, nb_bins)

[sevagh]

Huh, in fact the code right after says:

# get current spectrogram shape
nb_frames, nb_samples, nb_channels, nb_bins = x.data.shape

So I think the documentation block of the code might need adjusting. README might be correct?

[faroit]

yes, nb_samples, nb_channels, nb_bins, nb_frames is correct. Feel free to open a PR to correct the readme – we changed this in the recent model.