I don’t see how you can go back to your old conversation mode without going back to an older version of HA. That is what I was getting at. From what I understand the old conversation has been deprecated and is now powered by almond.
Not great yet. I ask it to turn on the dining room light and it says I don’t understand.
Got down to “turn on dining room” and it says Ok I’ll turn the AC on cool, is that right?
Ummmmmm no.
So just waiting for it to mature a bit.
well I just restored my earlier snapshots and only then the Conversation element was free of Almond. This is unfortunate as I wanted to give Almond a try but not on the expense of sacrificing the existing conversation intents (which BTW work reliable and entire family have gotten used to the custom words done for each device / room in last year or so). Maybe I will setup another RPI with HA and experiment Almond and ADA,
@synesthesiam, I finally got to try Rhasspy, and I must say, it is amazing!
In two hours I have been able to set it up and it seems to be able to do everything I want to do.
I do have one problem though, when in the Speech tab clicking on Hold to Record, my intents are almost always recognized. However, when I use the Porcupine wake word, it is far less accurate. Does it somehow use a different STT engine then? Or is there some other difference?
1 Like
Hi @basnijholt, glad to hear you’ve had an (overall) positive experience with Rhasspy 
It goes through the same STT engine, so this almost certainly indicates a problem with the microphone configuration. That’s good news, since it should work great once you get it fixed! If you could post to the Rhasspy community site or open a Github issue with the details of your set up, I’d be happy to help.
Most commonly, switching between PyAudio and arecord for microphone input helps. In other cases, you’ll need to adjust your microphone volume or edit your asound.conf file.
Would it be possible to set up a synchronised set Ada and Almond servers? I’m thinking about having a “main” server installed next to my HomeAssistant on my home server, while maintaining a “slave” server hosted next to my IPv4 to IPv6 proxy (CGNAT, but have IPv6, so for external access I have an Azure VM running that forwards IPv4 connections to v6).
1 Like
Did Ada die in January? I haven’t been able to find any new information about it.
I was a Snips user that tried Mycroft which fell flat when it couldn’t recognize my kids or wife’s voices.
If you have voice samples (kids & wife), you can train your own wake word in Mycroft.
I really need an Alexa alternative that is more private and actually works. I don’t even do anything that tricky. Once again my Alexa quit working properly, and I keep getting 3 entries for everything whenever I discover after deleting everything from their web interface.
I had not heard of Ada until this thread. I had previously looked at Mycroft but it was not far enough along for me to want to use. When I think about it I just want to add timers, ask local weather, add items to some kind of remote shopping list, and turn Home Assistant items on and off. Is Mycroft or Ada a better choice for me to investigate?
-
Love this thread.
-
I have Almond setup and working, and it was really easy, so thank you to whoever made the HomeAssistant package so simple to get working.
-
Want to add Ada to it, but don’t know what to set for the tts and stt settings defined in the documentation for Ada. it says it must be setup in HomeAssistant in order for Ada to use it, but just shows
tts: cloud
stt: cloud
I mean is that really the actual setting? I have been looking for configuration options for stt and tts all over. The Home Assistant has google_translate set as tts currently in configuration.ynl. Am I supposed to change that? Nothing set fo rstt.
Thanks for any help / direction on this. Running on an R-pi 3.
1 Like
So in the effort to replace Alexa or Google Home with Almond and Ada, what are are options if we have a VM deployment of home assistant? Having another HA box to run these doesn’t seem to be the way and if you do deploy on a standalone HA instance, are you carrying the device around the house?
I have HA running as a VM on my homelab server and want to build a voice assistant that can run on a pi device (actually thinking of a Pi in a Big Mouth Billy Bass for … and giggles.)
i am also in same boat not getting how to use it
have you found any solution ?
know this is an older thread but Im also want to ditch Alexa
has anyone made any progress with this
If someone pass around here, I am interested too to remove alexa.
I setup almond, it was pretty easy but I need help with Ada starting with the hardware you are using.
Hi! Any progress on that? I have exactly the same questions.
Me too. Id tike to be able to run barebone (pref. no HA or docker on the pi, maybe someting like espHome?) on a satellite rPi, but having a hard time configuring Ada. It’s hard google for this term, and there’s been little development since release.
Ofcourse the devs been busy, but for me a google home replacement guide would be a(nother!) killer feature for HA 
Once I know how to configure Ada, wrapping in a dockerfile should be relatively easy.
2 Likes
So, this is the Ada dockerfiile from Hassio:
addons/Dockerfile at master · home-assistant/addons (github.com)
this compiles ADA, then we need to run it. I didn’t succeed on the first try, maybe ill pick it up later
I was looking at Ada/Almond as I eliminated Alexa. But it appears the development stopped on Almond and support is nearly not supported. A lot of hype early on but that’s about it.
Well, after installing some aptgeteand pip3 packages (from the requirements) I got the ada porcupine hotword to trigger successfully, so that’s a win.
Didn’t have time yet to test some more. Basically I know almond works great, and are just needs to fancy STT for me