Trainer v0.12.0 is released

I recently released a new version of the Python package for training models.

As usual, the newest version is automatically used when training with Colab; folks making models locally on their own computers can install the update by typing

pip install --upgrade neural-amp-modeler

Full release notes are available on GitHub.

Just a couple things, but let's go through them quick.

What's new?

• WaveNets can be loaded from a .nam file. This is just a little convenience function, but if you've ever wanted to build using some .nam files and you don't have the .ckpt files from training, this makes things easier. The function is nam.models.init_from_nam.

• A bunch of bug fixes. Bugs in MPS and CUDA are now worked around (I'll try to contribute some upstream fixes in PyTorch, but I want NAM to work regardless of what you've got installed with it)

• A couple breaking changes. Just some keyword arguments and such here and there, but they're a bit deep. If this is you, then you're probably already looking at the release notes above (they're not too long).

Enjoy!