HPR4072: Piper text to speech engine

Hacker Public Radio

English - March 12, 2024 00:00 - 3.58 MB - ★★★★ - 34 ratings
Technology News Tech News community radio tech interviews linux open hobby software freedom Homepage Download Apple Podcasts Google Podcasts Overcast Castro Pocket Casts RSS feed

Previous Episode: HPR4071: Migration to digiKam as Digital Asset Management (DAM)

Next Episode: HPR4073: Is the 1990 documentary "Cyberpunk" worth watching today?

Using Piper
Github

Piper

Piper
Github

Thorsten-Voice

Thorsten-Voice Github

README

Piper Github README

Videos

Create your AI digital voice clone locally with Piper TTS |
Tutorial

Create voice clone with
Piper

Thorsten-Voice

Thorsten Mueller

Voices
Voice repo

Piper voices

Voice sample testing

Clone voices

Clone this model repository

Make sure you have git-lfs
installed

Git
lfs (Large filesystem)

In my case on Fedora 39

sudo dnf install git-lfs | teelog
Aside

I use an alias to create a log file with a date stamp

alias teelog='tee -a ~/logfiles/dump_`date +"%Y-%m-%d_%H%M%S"`.log'
git lfs install
git clone https://huggingface.co/rhasspy/piper-voices

# if you want to clone without large files – just their pointers
# prepend your git clone with the following env var:

GIT_LFS_SKIP_SMUDGE=1
Sample usage
echo 'Welcome to the world of speech synthesis!' | ./piper --model en_US-lessac-medium.onnx --output_file welcome.wav
Usage from a text file
cat sample/hamlet_to_be.txt | piper --model ~/git/piper/etc/test_voice.onnx --output_file hamlet_to-be_piper_sample_voice.wav