Hugo Flores García

Hi Loreto, Apologies for the late response. I'm currently out of town, and wouldn't be able to look into the issue until tomorrow (Dec 2). Apologies if this causes any...

Giving it a quick look, it looks like the data loading code cannot find the .json files associated with preprocessed data entries. See this line to look at where the...

Instead of trying to debug with the *quite* verbose ray output, try running a single training script first: ``` python music_trees/train.py --model_name hprotonet --height 4 --d_root 128 --loss_alpha 1 --name...

Also, I noticed that the audio preprocessing is done within kaldi, and not with torch preprocessing. Would that mean that I wouldn't be able to run gradients through the mel...

We use a continuous transformer in our new paper: https://arxiv.org/pdf/2307.04686.pdf for music generation and find that it works well! we use the continuous representation of the VQ-VAE latents as the...

hmm, I feel like this could either be version problem with madmom or sndfile, since onset detection happens through madmom. what's the full call stack? is this libsndfile error happening...

Hmm, I don't have a windows machine to debug on atm, but it looks like it's failing to write the input audio file to a temp directory for onset processing:...

this could be it: https://stackoverflow.com/questions/23212435/permission-denied-to-write-to-my-temporary-file looks like we're trying to open the file twice: once when `NamedTemporaryFile()` is created, and another in `sig.write`. This solution from stackoverflow could work, you...

The mask is not the audio with muted parts (though we can represent the mask as that). A better way to think of the mask is an array with 1s...

hmm, might be a problem with audiotools or the encoding of your audio files since you mention it happening with many audio files. what format are your audio files encoded...