ref: remove old common voice datasets

Update README.md

fix: fix learning rate

fix: fix learning rates and loss

Update README.md

Update README.md

feat: add samples

ref: update readme

ref: update readme

feat: update eval

ref: use my hosting for models

feat: add hubconf

feat: add librilight

feat: better support for conditioning, ability to create voices from alignments, added three built-in voices

wip: woring on integrating with gpt network