kalliope_tts_piper

Piper-TTS TTSModule for Kalliope. Refer to rhasspy/piper for details

##Installation:

kalliope install --git-url https://github.com/1001111github/kalliope_tts_piper.git

You will need to download at least one voice model with it's config from here or here. Place the model and json config in the desired directory. Rename the json config to match "full_model_file_name".json

Parameters

parameter	required	type	choices	comment
language	Yes	string	"your_model_choice"	filename, without extension
model	Yes	string	"path/your_model_choice.[onnx,tflite]	Full path to model file
name	No	string	Or any wake word	Relates voices to names used as trigger wake words
voiceN	No	string	Any label that contains 'voice'	Identifies a voice (language/model) related to a wake word
languageN	No	string	"your_model_choiceN"	filename, without extension for voice N
modelN	No	string	"path/your_model_choiceN.[onnx,tflite]	Full path to model file for voice N
nameN	No	string	Any wake word	Wake word associated with voice N
cache	Yes	BOOL	'True', 'False'

Example settings

#

default_text_to_speech: "piper"
text_to_speech
  - piper:
      language: "en_US-hfc_male-medium"
      model: "resources/tts/en_US-hfc_male-medium.onnx"
      name: 'default'
      voice1: 
         language: "en_US-kristin-medium"
         model: "resources/tts/en_US-kristin-medium.onnx"
         name: "alexa"
      cache: True