# Multi-speaker GlowTTS models for Talrómur 2 This release includes a partially trained multi-speaker model using the GlowTTS architecture in the Coqui TTS library, available [here](https://github.com/cadia-lvl/coqui-ai-TTS/tree/is). The model is trained on all of the speakers in the [Talrómur 2](http://hdl.handle.net/20.500.12537/167) corpus. The model included here is the best model available during the training at the time of publishing. At run time it is possible to choose any of the voices to produce a similar sounding synthesized voice. ## How to run To run the model follow the README instructions in [Coqui TTS](https://github.com/cadia-lvl/coqui-ai-TTS/tree/is). ## Files The files included are: - **the model**: best_model.pth - **model configuration**: config.json - **log file for the training**: trainer_0_log.txt - **the recipe used for training**: train_glowtts.py ## License This software is licensed under the CC by 4.0 public license. ## Authors Reykjavík University Þorsteinn Daði Gunnarsson (thorsteinng@ru.is) ## Acknowledgements This project was funded by the Language Technology Programme for Icelandic 2019-2023. The programme, which is managed and coordinated by [Almannarómur](https://almannaromur.is/), is funded by the Icelandic Ministry of Education, Science and Culture.