Twirre Meulenbelt
262376fb58
fix: break LLM response with fewer types of punctuation
...
ref: N25B-207
2025-11-05 15:01:01 +01:00
Twirre Meulenbelt
06e9e4fd15
chore: ruff format
2025-11-05 14:15:03 +01:00
Twirre Meulenbelt
1b58549c2a
test: fix expected test value after changing audio token allowance
...
ref: N25B-209
2025-11-05 12:41:48 +01:00
Twirre Meulenbelt
689a0d35da
Merge remote-tracking branch 'origin/dev' into demo
2025-11-05 12:38:08 +01:00
Twirre Meulenbelt
5c228df109
fix: allow Whisper to generate more tokens based on audio length
...
Before, it sometimes cut off the transcription too early.
ref: N25B-209
2025-11-05 10:41:11 +01:00
Twirre Meulenbelt
8e4d8f9d1e
Merge remote-tracking branch 'origin/dev' into demo
...
# Conflicts:
# src/control_backend/agents/bdi/behaviours/receive_llm_resp_behaviour.py
# src/control_backend/agents/llm/llm.py
# src/control_backend/agents/ri_command_agent.py
# src/control_backend/agents/transcription/speech_recognizer.py
2025-11-02 21:07:50 +01:00
Twirre Meulenbelt
4ffe3b2071
fix: make VAD unit tests work after changes
...
Namely, the Streamer has to be marked ready.
ref: N25B-216
2025-10-30 16:40:45 +01:00
Twirre Meulenbelt
86938f79c0
feat: end to end connected for demo
...
Includes the Transcription agent. Involved updating the RI agent to receive messages from other agents, sending speech commands to the RI agent, and some performance optimizations.
ref: N25B-216
2025-10-30 10:42:25 +01:00
