Am rummaging through voice samples to see if there's anything worth editing up for coqui #AI#TTS, to use in oobabooga. I have gigabytes of samples from the datasets used with the old TTS software from a few years ago, which I downloaded in a faulty state from #GitHub and fixed.
