It's worth noting that product is listed in the "Discontinued Products" section of the linked wiki.
Both of the ReSpeaker products in the non-discontinued section (ReSpeaker Lite, ReSpeaker 2-Mics Pi HAT) have only 2 mics, so it appears that things are converging in that direction.
I don't just want the hardware, I want the software too. I want something that will do STT on my speech, send the text to an API endpoint I control, and be able to either speak the text I give it, or live stream an audio response to the speakers.
That's the part I can't do on my own, and then I'll take care of the LLMs myself.
All of these components are available separately or as add-ons for Home Assistant.
I currently do STT with heywillow[0] and an S3-Box-3 which uses an LLM running on a server I have to do incredibly fast, incredibly accurate STT. It uses Coqui XTTS for TTS, with very high quality LLM based voice; you can also clone a voice by supplying it with a few seconds of audio (I tested cloning my own with frightening results).
Playback to a decent speaker can be done in a bunch of ways; I wrote a shim that captures the TTS request to Coqui and forwards it to a Pi based speaker I built, running MPD which then requests the audio from the STT server (Coqui) and plays it back on my higher quality speaker than the crappy ones built in to the voice-input devices.
If you just want to use what's available HA, there's all of the Wyoming stuff, openWakeword (not necessary if you're using this new Voice PE because it does on-device wakeword), Piper for TTS, or MaryTTS (or others) and Whisper (faster-whisper) for STT, or hook in something else you want to use. You can additionally use the Ollama integration to hook it into an Ollama model running on higher end hardware for proper LLM based reasoning.
I do the same, Willow has been unmaintained for close to a year, and calling it "incredibly fast" and "incredibly accurate" tells me that we have very different experiences.
It's a shame it's been getting no updates, I noticed that, but their secret sauce is all open stuff anyway so just replace them with the upstream components; their box-3 firmware and the application server is really the bit they built (as well as the "correction" service).
If it wasn't fast or accurate for you, what were you running it on? I'm using the large model on a Tesla GPU in a Ryzen 9 server, using the XTTS-2 (Coqui) branch.
The thing about ML based STT/TTS and the reasoning/processing is that you get better performance the more hardware you throw at it; I'm using nearly £4k worth of hardware to do it; is it worth it? No, is it reasonable? Also no, but I already had the hardware and it's doing other things.
I'll switch over to Assist and run Ollama instead now there's some better hardware with on-device wake-word from Nabu.
> ONLY A QUARTER OF PEOPLE WHO RECOVER HAVE EVER SOUGHT ASSISTANCE IN DOING SO (INCLUDING VIA 12-STEP PROGRAMS). THIS ACTUALLY MAKES ADDICTIONS THE PSYCHIATRIC DISORDER WITH THE HIGHEST ODDS OF RECOVERY.
You're mixing up capabilities with extra work with capabilities with no work.
If OS X also had tailor made kernel just for lower real time latency it could do even better.
But out of box, it's already better.
That said, the point is moot. 99% of pro audio/MIDI apps/VSTs don't run on Linux, so only experimental nerdy musicians and FOSS zealots would use it for real studio work. You do get the occasional profile on music tech websites e.g. "Techno musician X uses Linux", and somebody always jumps in a forum to say they use some Linux DAW (and usually add a list of ugly hacks, workarounds, and things that don't work), but it's clearly a total outlier situation...
Regular people aren't compiling kernels, they're installing something like Ubuntu Studio (https://ubuntustudio.org/) that comes with everything you need preconfigured out of the box.
The point outlined is that we don't know enough about how we identify objects to discard a simple adversarial attack; probably not a filter-based but maybe something else.
I am definitely using Weave Net for my trivial single-node cluster because it was the only CNI option that included a one-liner to install.
I have never had any reason to try a different one, the pods can talk to each other, "does what it says on the tin"... and on top of that but unrelated, their webinars tend to be star-studded and very informative.
Very cool. Eventually I'd like to turn Colossus into something more like what you have here, i.e. a "real" backend that does something meaningful. The good news is that the Bazel + gRPC plumbing was the hard part. Adding new, meaningful services will be pretty trivial going forward.
