# Model card for base_model (medium)

* Language: en_US (English, United States)
* Speakers: 904
* Quality: medium
* Samplerate: 22,050Hz

## Dataset

* URL: http://www.openslr.org/141/
* License: CC BY 4.0

## Training

Trained from scratch on LibriTTS-R train-clean-360 with MRD and SDP
Uses vowel clusters, so only compatible with Piper 1.5.0+