lhumina_code/home
23
0
Fork 0
Code Issues 14 Pull requests Projects Releases Packages Wiki Activity Actions

Voice transcription always returns "Thank you" regardless of input #108

New issue
Open
opened 2026-04-12 13:55:21 +00:00 by mik-tf · 0 comments
mik-tf commented 2026-04-12 13:55:21 +00:00
Owner
Copy link

Problem

Hero Voice transcription via Groq Whisper always returns "Thank you." regardless of what the user actually says. Tested with multiple sentences ("what is geomind", longer phrases) — all return "Thank you."

Expected

Accurate transcription of spoken words, as it worked before the sessions 17-18 refactoring.

Context

  • Version: v0.9.0-dev on herozero.gent04.grid.tf
  • Provider: Groq Whisper API
  • Groq API key: confirmed working (spend limit was hit earlier, now raised)
  • Pipeline: audio reaches Groq, response comes back — but content is wrong
  • Worked before: transcription was accurate in v0.8.7-dev and earlier

Investigation

  1. Check if audio format/encoding changed during the hero_voice refactoring (Cargo.toml version pin updates, sync with development)
  2. Compare hero_voice_server audio handling between development and development_mik01 commits
  3. Check if the Groq Whisper model parameter changed (whisper-large-v3 vs other)
  4. Test with the previous working commit to confirm regression vs Groq API issue
  5. Check hero_aibroker voice/transcription routing — did the URL routing refactoring (sessions 17-18: //<socket_type>/) break the audio payload forwarding?

Repos to check

  • hero_voice (version pin changes, Cargo.toml sync)
  • hero_aibroker (transcription routing)
  • hero_zero/services/hero_voice.toml (env var changes)
  • hero_zero/services/hero_aibroker.toml (env var changes)
## Problem Hero Voice transcription via Groq Whisper always returns "Thank you." regardless of what the user actually says. Tested with multiple sentences ("what is geomind", longer phrases) — all return "Thank you." ## Expected Accurate transcription of spoken words, as it worked before the sessions 17-18 refactoring. ## Context - **Version:** v0.9.0-dev on herozero.gent04.grid.tf - **Provider:** Groq Whisper API - **Groq API key:** confirmed working (spend limit was hit earlier, now raised) - **Pipeline:** audio reaches Groq, response comes back — but content is wrong - **Worked before:** transcription was accurate in v0.8.7-dev and earlier ## Investigation 1. Check if audio format/encoding changed during the hero_voice refactoring (Cargo.toml version pin updates, sync with development) 2. Compare hero_voice_server audio handling between development and development_mik01 commits 3. Check if the Groq Whisper model parameter changed (whisper-large-v3 vs other) 4. Test with the previous working commit to confirm regression vs Groq API issue 5. Check hero_aibroker voice/transcription routing — did the URL routing refactoring (sessions 17-18: /<service>/<socket_type>/<path>) break the audio payload forwarding? ## Repos to check - hero_voice (version pin changes, Cargo.toml sync) - hero_aibroker (transcription routing) - hero_zero/services/hero_voice.toml (env var changes) - hero_zero/services/hero_aibroker.toml (env var changes)
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
development
main
No results found.
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
lhumina_code/home#108
No description provided.