akanametov / musegan Goto Github PK

In my case when I try to process the train_x_lpd_5_phr.npz dataset The operation always ends up crashing. What is the work around this?
below is a screenshot showing my VM Ram filling up

Hi,

First off let me say thanks so much for this amazing package.

I'm interested in running the pytorch implementation using the data from the original Tensorflow MuseGAN project that links to this project (https://github.com/salu133445/musegan). The data I am trying to use can be found at:

https://docs.google.com/uc?export=download&id=14rrC5bSQkB9VYWrvt2IhsCjOKYrguk3S

The issue I am running into is that the input data is of a different shape than the model currently supports. How should I go about changing the code to accept this input data?

Thanks so much!

Thanks for your code, it is very nice. But I found a little bug in the bar_generator code.
In fact, the last Reshape() layer in the network definition has an incorrect size. As you mentionned, the output shape of the last conv is (batch_size, out_channels, 8*hid_features//hid_channels, n_pitches) and the shape of Reshape layer is output shape: (batch_size, out_channels, 1, n_steps_per_bar, n_pitches). Such configuration works very well only when n_steps_per_bar is set to 16 but One would run experiment with a different n_steps_per_bar. (I run it with 48 steps per bar as far as I am concerned in fact).

For a more general purpose, I propose to add a linear mapping before that Reshape layer as shown in the figure attached.

Yours sincerely.