Commit Graph

1057 Commits

Author SHA1 Message Date
Kwindla Hultman Kramer
9c22f5b81b async google llm 2024-12-04 15:52:52 -08:00
Kwindla Hultman Kramer
f33f08d667 partially working audio+transcription parallel pipelines 2024-12-04 08:51:35 -08:00
Aleix Conchillo Flaqué
52caaa4afb transports(daily): use a task to execute callbacks
This commit fixes an issue where we were not waiting for
`asyncio.run_coroutine_threadsafe` to complete which can cause a series of
undesired issues (e.g. not actually executing the coroutine).
2024-12-03 18:58:54 -08:00
Aleix Conchillo Flaqué
1cf93f1dcb FrameProcessor: ignore other frames during CancelFrame 2024-12-03 16:26:29 -08:00
Aleix Conchillo Flaqué
322dd0cea1 AudioBufferProcessor: use on_audio_data event handler to retrieve audio 2024-12-03 16:12:40 -08:00
Aleix Conchillo Flaqué
a6a4910931 transports(services): incoming transport messages should be urgent 2024-12-03 14:30:15 -08:00
Aleix Conchillo Flaqué
52cefaa9d6 frames: remove AppFrame 2024-12-03 14:30:15 -08:00
Aleix Conchillo Flaqué
42658ecd92 frames: use mixins for audio and image data 2024-12-03 14:30:15 -08:00
Aleix Conchillo Flaqué
a6606a4040 transports(base_output): remove unused code 2024-12-03 14:30:15 -08:00
Aleix Conchillo Flaqué
d6c944cdc1 processors(audio): fix AudioBufferProcessor interruptions 2024-12-03 14:30:15 -08:00
Aleix Conchillo Flaqué
a5c7b02a73 frames: input frames are now system frames
Input frames from a transport should be processed fast and there's no need for
them to be queued internally in each element.
2024-12-03 14:30:15 -08:00
Aleix Conchillo Flaqué
c2135cbe11 transports(websockets): use frame serializers during interruptions 2024-12-02 19:17:17 -08:00
Aleix Conchillo Flaqué
32495ddd0b Merge pull request #769 from pipecat-ai/aleix/daily-subscribe-video-source
transports(daily): subscribe to the desired video source
2024-12-02 19:16:14 -08:00
Aleix Conchillo Flaqué
5e854c4d03 transports(daily): subscribe to the desired video source 2024-12-02 12:13:23 -08:00
Aleix Conchillo Flaqué
71cf94e936 transports(daily): warn if transcription enabled but no token provided 2024-12-02 09:55:17 -08:00
James Hush
5e22ef251d fix: add logging and error handling for issue #721 (#755) 2024-11-29 13:06:45 +08:00
Mark Backman
3c19a7ae3d Use Google Gemini message format when adding message to the LLM context 2024-11-27 12:46:51 -05:00
Kwindla Hultman Kramer
8ddac0ccd8 Testing with gemini-exp-1114. Bug fix 2024-11-21 10:33:12 -08:00
Mark Backman
106042bbb2 Make language support more robust 2024-11-20 12:56:11 -05:00
Mark Backman
676ff14913 Merge pull request #735 from pipecat-ai/vp-internal-push-frame-fix
internal push frame fix
2024-11-20 06:34:40 -05:00
Vanessa Pyne
14893ade92 Update src/pipecat/processors/frame_processor.py
Co-authored-by: Mark Backman <mark@daily.co>
2024-11-19 22:37:58 -06:00
Mark Backman
e79289454a Merge pull request #734 from pipecat-ai/mb/fix-cartesia 2024-11-19 21:27:52 -05:00
vipyne
e00c75ce3f fix: raise exception in internal_push_frame 2024-11-18 16:01:04 -06:00
Mark Backman
cf62167f54 Revert: services(cartesia): generated TTSStoppedFrame after no more audio 2024-11-18 12:25:04 -05:00
Aleix Conchillo Flaqué
635aa6eb5b Merge pull request #729 from pipecat-ai/aleix/fastapi-websocket-dont-close
transports(fastapi): don't try to close socket
2024-11-18 16:01:41 +01:00
Mark Backman
1ff17cc2b6 Merge pull request #733 from pipecat-ai/aleix/add-missing-init-files
processors: add missing __init__.py
2024-11-18 09:44:56 -05:00
Christian Stuff
5d7b3f2b38 add handler for disconnect-bot message 2024-11-18 09:33:30 -05:00
Aleix Conchillo Flaqué
23e5b1ec4d processors: add missing __init__.py 2024-11-18 11:32:20 +01:00
Aleix Conchillo Flaqué
7f5a8928b8 transports(fastapi): don't try to close socket
The websocket is passed from outside (in the transport constructor) so we should
not be trying to close it. FastAPI does actually close it later. We didn't see
any issue because these functions were not implemented properly. The value to
check was `application_state` instead of `client_state`. But in any case,
Pipecat should not be responsible for closing things passed from outside.
2024-11-18 01:15:19 +01:00
Aleix Conchillo Flaqué
5445bb0363 rtvi: add on_bot_started event 2024-11-17 22:40:00 +01:00
Aleix Conchillo Flaqué
67f975a2c8 transports(daily): add more dial-in events 2024-11-16 01:22:50 +01:00
Mark Backman
966974bfc6 Change STTMuteProcessor to STTMuteFilter 2024-11-14 19:47:37 -05:00
Mark Backman
f807f233bd Suppress UserStartedSpeakingFrame and UserStoppedSpeakingFrame when muted 2024-11-14 17:11:51 -05:00
Mark Backman
33108f5798 Code review feedback 2024-11-14 17:05:08 -05:00
Mark Backman
5fe679039c Add STTMuteProcessor to un/mute the STT 2024-11-14 13:35:02 -05:00
Kwindla Hultman Kramer
534f710f5d Merge pull request #688 from pipecat-ai/khk/natural-conversation
More work on llm-as-judge phrase endpointing
2024-11-14 09:15:16 -08:00
Mark Backman
72412cc0c4 Code review feedback 2024-11-14 09:31:04 -05:00
Aleix Conchillo Flaqué
4f88fc0eb8 services(tts): initialize language to the proper language code 2024-11-13 17:19:23 +01:00
Aleix Conchillo Flaqué
37d1f4c4e1 services(tts): some language to service language cleanup 2024-11-13 17:19:23 +01:00
Aleix Conchillo Flaqué
ef9e86d997 services(playht): make sure we only skip wav header no matter the size 2024-11-13 17:19:23 +01:00
Aleix Conchillo Flaqué
2d2ef5a417 services(playht): voice engine is Play3.0-mini 2024-11-13 17:19:23 +01:00
Aleix Conchillo Flaqué
c1fff00586 services(playht): fix language codes 2024-11-13 17:19:23 +01:00
Mark Backman
0af2196f50 Merge pull request #708 from pipecat-ai/mb/add-rime-ai
Add RimeTTSService
2024-11-12 18:29:53 -05:00
Mark Backman
cd42320788 Update changelog 2024-11-12 18:28:04 -05:00
Kwindla Hultman Kramer
ee53535f41 gemini audio-in with no transcription 2024-11-11 21:04:50 -08:00
Kwindla Hultman Kramer
b56c789ae4 fixes for proposed judge pipeline 2024-11-11 21:04:50 -08:00
Mark Backman
958ec42e8d Add Rime.ai TTS service 2024-11-11 21:58:09 -05:00
Kwindla Hultman Kramer
1438e5654a remove old comment 2024-11-10 16:08:10 -08:00
Aleix Conchillo Flaqué
d125a38737 services(cartesia): generated TTSStoppedFrame after no more audio
The TTSStoppedFrame should be generated when the TTS services stoped generating
audio not when the bot stops speaking.
2024-11-10 09:55:45 +01:00
Mark Backman
446bb0aeaf Merge pull request #702 from pipecat-ai/mb/azure-websocket
Add an Azure TTS websocket service
2024-11-09 17:41:53 -05:00