SubZeroAI's picture
Update README.md
48e6aef verified
---
license: cc-by-4.0
datasets:
- google/fleurs
- KTH/waxholm
- KBLab/rixvox
language:
- sv
pipeline_tag: text-to-speech
tags:
- piper
- piper_tts
- onnxruntime
---
# piper-swedish-tts-multispeaker
This model is trained on a subset of about 43000 (out of 866 000) audio files from a dataset that combines [NST](https://www.nb.no/sprakbanken/en/resource-catalogue/oai-nb-no-sbr-17/), google/fleurs, KTH/waxholm and KBLab/rixvox datasets. Currently only trained for 500 epochs so there is more work to be done, and in progress.