I looked into JarbasAI some time ago but had some problems getting it running.
I see you have now included JarbasAI setup scripts so I will give it another try.
I am not a hardcore techie, hopefully I will get it to work. What I am very interested in is your client server concept in combination with your LILACS service/skills.
I have a couple of Mark1’s. The dream would be to be able to use these as clients (1 per floor in the house for the family) hooking up with a central the server…
if you have any trouble don’t hesitate to ask for help!
keep in mind that LILACS is early work and far from perfect, it can be quite slow at times
server/client i have been using with success and should be more or less stable
i have the nasty habit of pushing everything to the repo as a backup, this means sometimes i break things as fast as i add new features, open issues for any trouble you come across, i will try to make it as painless as possible for install and use, my aim is to have this ready to deploy anywhere for personal usage
I will take the time eventually to have a dedicated branch for a ready to use minimal client, jarbas fork is bigger than standard mycroft and you will not need most of the things in the mark 1s for the setup you described
Hello,
I am a beginner trying to work with Mycroft. Using git clone I have cloned the dictation skill on my linux machine. But when I command mycroft to “start dictating” it is giving the intent error “Sorry I didn’t catch that.Please rephrase your request”. Please guide me how should I go about this. Do I have to change any configuration file or any other changes.
Hi there, I took some time this weekend to get things going. Managed to get it installed on 2 Raspberry Pi’s. Then initiated one as server and one as client. That worked perfectly got them handshaked! Next challenge was to get the client fallback to the server fallback skills. That was more difficult as it kept being catched by the aiml fallback (looking at the config file this should not even happen as within the client config settings the amount of skills to load are limited).
I removed all skills, except for the hello_world skill and server fallback skill. But that was also not working. (was not picked up by the server fallback skill). I also got some python errors. The ‘client’ log was indicating ‘no intent failure or execution detected for 20 second’ => ‘not asking server’.
What is your preferred way to get the logs over so you can have a look into this?
do not try to use the jarbas fork, the code should work but is outdated and i have no time for giving support, I am in the process of remaking that fork with higher standards
when i started i did not really know how to use github and my branches are a mess, i am slowly migrating and fixing things over to this new github account
Jarbas_Ai, Thank you for the quick response. I was very intrigued by some of the development you have done with your fork of Mycroft. Particularly the snowboy integration, and peer to peer stuff. I will stick with my original Mycroft configuration for now and work at becoming proficient in skills development as I await your new GitHub cleanup. Thanks for your support and efforts with this project. I really enjoyed your ‘Guidelines for skills development’.
Cheers!
So I see where the hotword_factory.py supports snowboy but I can’t seem to find on the forum how to enable it or where to add the hotword model?
Any advice or point me in the right direction.
Thanks again.
not sure how to do it in the remote settings (home.mycroft.ai)
you can edit the config file like
"hotwords": {
"hey mycroft": {
"module": "snowboy",
// should support more than 1 model, but its failing for me with more than 1
"models": {"hey mycroft": "path/to/model.pmdl"},
"sensitivity": 0.5
}
}
you also need to have snowboy installed or it wont work, it does not come bundled with mycroft