voices config in local file personas.json, e.g.
"tts_settings":{"language_code": "de-DE","name":"de-DE-Standard-E","ssml_gender": "texttospeech.SsmlVoiceGender.MALE","speaking_rate": 0.9,"pitch": -10},
additional (pre-recorded) audio files in /audio/personas/
-- directory logic
--- generated files in /audio/personas/ with a subdir e.g. quittenbaum for every persona
-- file name logic / categories:
--- greetings
--- "idle" > later introduced, difference to wait??
--- wait
--- bye
--- files generated with category name and added number or like specified in as filename in personas.json, e.g. wait1.mp3, wait2.mp3
-- how to create them with /utils/audiofile-generator.pycd ~/gptree/venv-gptree/bin/python3 utils/audiofile-generator.py
Fitting (environment) variable
adjust "known" environment variable
-- e.g. what kind of variables
-- what aggregation state, e.g. mean of the last 7 days, last month, ...
-- see utils/weather.py
Post installation there are some tasks to get GPTree running as you want it. Main tasks for customization are:
Prompt / Persona
Voice
personas.json
, e.g."tts_settings":{
"language_code": "de-DE",
"name":"de-DE-Standard-E",
"ssml_gender": "texttospeech.SsmlVoiceGender.MALE",
"speaking_rate": 0.9,
"pitch": -10
},
/audio/personas/
-- directory logic --- generated files in/audio/personas/
with a subdir e.g.quittenbaum
for every persona -- file name logic / categories: --- greetings --- "idle" > later introduced, difference to wait?? --- wait --- bye --- files generated with category name and added number or like specified in asfilename
inpersonas.json
, e.g.wait1.mp3
,wait2.mp3
-- how to create them with /utils/audiofile-generator.pycd ~/gptree/
venv-gptree/bin/python3 utils/audiofile-generator.py
Fitting (environment) variable
utils/weather.py