Re: [orca-list] Interesting text to speech announcement

[Kyle <kyle free2 ml>]

Quoting Peter Vágner:
> Easiest is perhaps installing it on a debian or debian derivative into 
> python virtualenv from PIP


That unfortunately didn't work for me. I got the same onnxruntime error 
that I got in Fedora. I did get the docker version working here, but 
that only exposes the web interface. I couldn't figure out how to run 
the client, so the speech-dispatcher configuration probably won't work. 
The easiest local install was on Ubuntu 22.04, probably the latest 
Debian will also work, using the .deb package at

https://github.com/MycroftAI/mimic3/releases/download/release%2Fv0.2.3/mycroft-mimic3-tts_0.2.3_amd64.deb

https://github.com/MycroftAI/mimic3/releases/download/release%2Fv0.2.3/mycroft-mimic3-tts_0.2.3_arm64.deb

or

https://github.com/MycroftAI/mimic3/releases/download/release%2Fv0.2.3/mycroft-mimic3-tts_0.2.3_armhf.deb

depending on the machine where you want it to run. The client, server 
and voice downloader all work in those packages. Oddly enough, I got the 
same onnxruntime error about not being able to find the correct version 
when I tried installing from source into the virtualenv in Ubuntu 22.04, 
but the packages I linked above work almost perfectly, with the 
exception of the epitran languages not having all the necessary files, 
which I'm hoping to find in the source tarball.

~Kyle