Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

adding latest okay nabu wake word #216

Open
wants to merge 2 commits into
base: develop
Choose a base branch
from

Conversation

that1guy
Copy link
Collaborator

  • Adding latests Okay Nabu wake word to hopefully increase success rate

- testing to see if better UX
@formatBCE
Copy link

Did you try lowering probability cut-off? I have it on 0.8 for ok Nabu.

@that1guy
Copy link
Collaborator Author

Did you try lowering probability cut-off? I have it on 0.8 for ok Nabu.

@formatBCE where do I do this?

Currently in my testing this wake word is giving me pages of warnings, which then cause the TTS response to stutter when playing back. Sometimes the response takes 30 seconds to play back because mWW is eating up the ESPs processing power:

image

@formatBCE
Copy link

@formatBCE where do I do this?

models:
    - model: https://github.com/kahrendt/microWakeWord/releases/download/okay_nabu_20241223.2/okay_nabu.json
      probability_cutoff: 0.8
      id: okay_nabu

Currently in my testing this wake word is giving me pages of warnings, which then cause the TTS response to stutter when playing back. Sometimes the response takes 30 seconds to play back because mWW is eating up the ESPs processing power:

Damn, that's something weird! I don't have that on Respeaker, will try to reproduce on SAT1. Is it because of newest wake word? 0_o

@that1guy
Copy link
Collaborator Author

Yes. This appears to only reproduce in this PR with latest wake word to my knowledge. Please do let me know if you can reproduce this. I'm seeing this on two Sat1 devices in my home.

- This is performing well on my end.
- On desk near the user and in-ceiling 15ft from user.
- Please test
@that1guy
Copy link
Collaborator Author

that1guy commented Jan 3, 2025

@formatBCE @gnumpi I'm running the even newer Okay Nabu wake work model with a probability cutoff of 0.8. Please test it's performing very well on my end.

@formatBCE
Copy link

I tried it. For me it works worse than the first one... Much less responsive.

@that1guy
Copy link
Collaborator Author

that1guy commented Jan 3, 2025

I tried it. For me it works worse than the first one... Much less responsive.

how are you concluding it is "much less responsive"?

@that1guy that1guy requested a review from gnumpi January 11, 2025 21:10
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants