Ctrl K
ComunidadGuías…

Ayuda y problemas de computah

Pregunta a la comunidad o invita a KiBot con contexto del repositorio.

Publicar un problema

No publiques contraseñas, tokens ni claves privadas. HostingByte aplica una redacción básica de secretos evidentes antes de guardar el texto.

Problemas en HostingByte

Todavía no hay problemas publicados.

Issues abiertas upstream

#92 Common spoken confirmations and refusals reprompt to the floor instead of deciding · 1 comentarios#100 Barge-in: run a wake/VAD detector during playback and guard self-trigger · 3 comentarios#123 Verify PowerConf capture during reply playback for barge-in · 0 comentarios#14 Cut speech-to-text latency on the Pi · 3 comentarios#16 Add barge-in so the user can interrupt playback · 4 comentarios#25 Convert the test scripts into a pytest suite · 2 comentarios#118 Run test_mic_controls.py in the fast CI job · 0 comentarios#113 Make wake-word deployment validation independent and class-aware · 0 comentarios#17 Tune wake-word false-accept and false-reject behavior · 2 comentarios#37 Improve STT accuracy: vocabulary biasing and fine-tuning before heavier models · 1 comentarios#36 Evaluate open USB mic array vs PowerConf-over-USB for the Pi · 2 comentarios#35 Energy endpointing over-captures under continuous background noise · 1 comentarios#33 Stop defaulting the committed wake word to bundled hey_jarvis · 1 comentarios#48 syl drops the reply block on some voice turns (reply never written to FileOutbound) · 4 comentarios#46 Explore deeper interface options for the Anker PowerConf S3 mic · 1 comentarios#42 Confirm-before-act readback handshake (voice conversation mode) · 1 comentarios#38 Evaluate higher-quality TTS voices (Piper variants and open-source models) · 1 comentarios#59 #19 producer-side rollout requirements: stamp event_id and forward the correlation id · 2 comentarios#63 live_driver ignores the output_device config key (only the file pipeline honors it) · 0 comentarios#72 wake-eval: 1.6s refractory floor undercounts the re-arm gap for continuous-speech backgrounds · 0 comentarios