Jong Wook Kim

Hi, the training code is available at: https://github.com/jongwook/crepe but it is even less maintained than this one. One hack you could do is, just subsample every 3 (or 2) samples...

I don't have access to the dataset at the moment, but the dataset was not from the RWC dataset itself but re-synthesized vocal tracks as described in the [pYIN paper](https://www.eecs.qmul.ac.uk/~simond/pub/2014/MauchDixon-PYIN-ICASSP2014.pdf),...

It's one of the tricks used for the approach, which is not quite orthodox for classification tasks in ML - it also uses binary cross entropy with soft labels, whereas...

The labels I had contained Hz values (that doesn't necessarily align with semitone intervals), from which I calculated the soft labels using a Gaussian-shaped curve with a standard deviation of...

Correct, and thanks for reporting this! `FutureWarning` does not alter the behavior, so you can ignore them for now. I am not totally sure how `check_is_fitted` is getting called, it's...

Hi, thanks for the suggestion. While I could add the CLI argument in the future, you can manually select the GPU using an environment variable: ``` CUDA_VISIBLE_DEVICES=2 crepe audio.wav ```...

Does Intel CPU have `int8`-optimized operations? I thought they are specific to CUDA. Tensorflow already uses all CPU cores by default, and I'm not sure if there's practically any performance...

It's not precisely optimize but [you might be interested in some of these numbers reported here](https://github.com/marl/crepe/releases/tag/v0.0.6)

Yeah that's why I haven't put the table on README yet!

In that case we'd need a continuous emission probability model. For continuous emissions hmmlearn only supports Gaussian / Gaussian mixtures. We could of course use one of those instead of...