RealEdit dataset

[BaherMh]

First of all, thank you for the excellent work.
I have a question about RealEdit dataset, is there a simple way to get all the audio files that are used in this dataset, or at least know exactly from which split of each dataset they are extracted, so to make it easier to evaluate on it.

[DHY02]

Hey! Did you figure it out eventually? I'm stuck on the same issue.

[BaherMh]

there doesn't seem to be easy way to do so.
you can get them from LibriTTS (dev-clean and dev-other) splits, gigaspeech dev split.
but as for spotify part of the dataset, it seems that the dataset is deleted unfortunately, if you know something else please let us know.
And for the evaluation code, I couldn't find the code they used.