wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
7,455 Commits 276 Branches 111 Tags
ba3100be0d02fa77e8900cbfbc461aeab198fb57
Commit Graph

7455 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Paul Kompfner
ba3100be0d Another change to support Gemini Live in Pipecat Flows: rather than strip function call and response messages out of context before sending to Gemini Live when seeding conversation history, which we were doing to sidestep a seeming Gemini Live limitation (see https://stackoverflow.com/a/79851394), convert them to regular text messages with special formatting 2026-02-04 11:44:58 -05:00
Paul Kompfner
9e65e77095 Another change to support Gemini Live in Pipecat Flows: now even in the case of programmatically-appended context messages, do a full reconnection—I've tried many things to get "live updates" to context working reliably and have been unable to (Gemini code comments also warn against doing "live updates" after the initial context seeding prior to starting audio input). 2026-02-03 21:22:24 -05:00
Paul Kompfner
3183f9c077 Revert "Another change to support Gemini Live in Pipecat Flows: avoid triggering a model response twice when there is a Flows node transition that returns a value from the transition function and loads new context messages in the new node" 
This reverts commit 90e6f0dca8.
 2026-02-03 17:02:31 -05:00
Paul Kompfner
90e6f0dca8 Another change to support Gemini Live in Pipecat Flows: avoid triggering a model response twice when there is a Flows node transition that returns a value from the transition function and loads new context messages in the new node 2026-02-03 16:56:34 -05:00
Paul Kompfner
44b917c546 Another change to support Gemini Live in Pipecat Flows: if the only change to the context is newly-appended messages, send them to the server. 
This requires us to distinguish between newly-appended "bookkeeping" messages that just reflect what Gemini Live already said, and messages that were programmatically inserted, such as from the the transition to the new Pipecat Flows node.

This change makes it so that using `LLMMessagesAppendFrame` will have the desired effect, of updating the Gemini Live conversation.
 2026-02-03 15:45:00 -05:00
Paul Kompfner
c1ac1a6326 Changes to support Gemini Live in Pipecat Flows: 
- Detect when a newly-received context warrants a reconnection to the Gemini Live API; we need to reconnect in order to re-seed new conversation history or swap out the current set of tools. This reconnection occurs when Pipecat Flows transitions between conversational nodes, as context has been edited and/or tools added/removed.
- Strip function call and response messages out of context before sending to Gemini Live when seeding conversation history, to sidestep a seeming Gemini Live limitation (see https://stackoverflow.com/a/79851394)
 2026-02-02 16:34:12 -05:00
Mark Backman
ca49acfaa6 Merge pull request #3619 from pipecat-ai/mb/resemble-readme 
Resemble cleanup
 2026-02-02 09:20:11 -05:00
Mark Backman
86147f15f3 Renumber the Resemble foundational example 2026-02-02 09:07:05 -05:00
Mark Backman
5cda72d138 Add Resemble TTS to README 2026-02-02 09:05:03 -05:00
Mark Backman
54e62a8177 Merge pull request #3134 from pipecat-ai/mb/resemble-tts-draft 
Add ResembleAITTSService
 2026-02-02 08:59:27 -05:00
Mark Backman
a592b7fdf0 Update per PR 1789, align with ErrorFrame norms 2026-02-02 08:55:29 -05:00
Mark Backman
ba2b7c05d6 Add ResembleAITTSService 2026-02-02 08:55:27 -05:00
Aleix Conchillo Flaqué
f453227ba3 Merge pull request #3612 from pipecat-ai/aleix/use-kokoro-onnx 
KokoroTTSService: use kokoro-onnx instead of kokoro
 2026-01-31 21:03:55 -08:00
Aleix Conchillo Flaqué
52cc64019a Merge pull request #3611 from pipecat-ai/aleix/aicoustics-example-update 
examples: update 07zd to use vad_analyzer in LLMUserAggregator
 2026-01-31 21:02:50 -08:00
Aleix Conchillo Flaqué
95689cc81c KokoroTTSService: use kokoro-onnx instead of kokoro 2026-01-31 17:20:27 -08:00
Aleix Conchillo Flaqué
675c7c43e3 examples: update 07zd to use vad_analyzer in LLMUserAggregator 2026-01-31 15:31:15 -08:00
Aleix Conchillo Flaqué
bfd19e867c Merge pull request #3610 from pipecat-ai/aleix/dont-add-rtvi-observer-if-already-there 
PipelineTask: don't add RTVIObserver if already there
 2026-01-31 14:57:52 -08:00
Aleix Conchillo Flaqué
acc9923c0a PipelineTask: don't add RTVIObserver if already there 2026-01-31 14:54:29 -08:00
Mark Backman
bdc9e7e2e4 Merge pull request #3608 from pipecat-ai/mb/quickstart-0.0.101 
Update quickstart for 0.0.101
 2026-01-31 10:39:17 -05:00
Mark Backman
a587e1b99a Update quickstart for 0.0.101 2026-01-31 09:52:24 -05:00
Aleix Conchillo Flaqué
7853e5ca93 Merge pull request #3606 from pipecat-ai/changelog-0.0.101 
Release 0.0.101 - Changelog Update
v0.0.101 		2026-01-30 22:58:22 -08:00
aconchillo
614b8e1a62 Update changelog for version 0.0.101 2026-01-30 22:54:31 -08:00
Aleix Conchillo Flaqué
ef51c2a5c6 changelog: fix 3582 changed file 2026-01-30 22:48:26 -08:00
Aleix Conchillo Flaqué
f42dc0d38e Merge pull request #3605 from pipecat-ai/aleix/gemini-live-schedule-transcription-timeout-handler 
GeminiLiveLLMService: let the transcription timeout handler be scheduled
 2026-01-30 22:44:05 -08:00
Aleix Conchillo Flaqué
d87f3543c7 GeminiLiveLLMService: let the transcription timeout handler be scheduled 2026-01-30 22:41:10 -08:00
Aleix Conchillo Flaqué
fee633cb92 scripts(evals): disable kokoro for now 2026-01-30 21:23:42 -08:00
Aleix Conchillo Flaqué
607af91153 Merge pull request #3604 from pipecat-ai/mb/fix-ivr-navigator-aggregation 
Fix IVRNavigator to push AggregatedTextFrame when switching to conver…
 2026-01-30 21:22:20 -08:00
Mark Backman
e779233918 Fix IVRNavigator to push AggregatedTextFrame when switching to conversation mode 2026-01-30 21:07:49 -05:00
Aleix Conchillo Flaqué
604d5d0b14 examples: update 07zi and 07zj to use vad_analyzer form LLMUserAggregator 2026-01-30 16:14:02 -08:00
Mark Backman
342ae7af41 Merge pull request #3601 from pipecat-ai/mb/add-22-release-evals 
Add 22 foundational to release evals
 2026-01-30 15:31:54 -05:00
Mark Backman
c92ec1552e Add 22 foundational to release evals 2026-01-30 15:12:52 -05:00
Aleix Conchillo Flaqué
93160f1455 scripts(evals): remove vad_analyzer from transport 2026-01-30 12:08:12 -08:00
Aleix Conchillo Flaqué
e3158e1131 Merge pull request #3600 from pipecat-ai/aleix/llm-server-timeout-task-never-waited 
LLMService: make sure function call timeout handler is started
 2026-01-30 12:01:18 -08:00
Mark Backman
63a23246d5 Add UserTurnCompletionLLMServiceMixin (#3518) 
* Added UserTurnCompletionLLMServiceMixin class

* Added 22-filter-incomplete-turns.py foundational example

* Removed old 22 natural conversation foundational examples

* Added test_user_turn_completion_mixin.py
 2026-01-30 14:57:15 -05:00
Aleix Conchillo Flaqué
569ea9849a Merge pull request #3599 from pipecat-ai/aleix/release-evals-disable-rtvi 
scripts(evals): disable RTVI
 2026-01-30 11:44:46 -08:00
Aleix Conchillo Flaqué
a98ca9b65b LLMService: make sure function call timeout handler is started 2026-01-30 11:38:26 -08:00
Aleix Conchillo Flaqué
c9310789dc scripts(evals): use new vad_analyzer from LLMUSerAggregator 2026-01-30 10:57:17 -08:00
Aleix Conchillo Flaqué
b93e12d701 scripts(evals): disable RTVI 2026-01-30 10:52:38 -08:00
Aleix Conchillo Flaqué
3f77da627d Merge pull request #3583 from pipecat-ai/aleix/move-vad-analyzer-to-llm-user-aggregator 
VAD analyzer is now passed to LLMUserAggregator
 2026-01-30 10:46:10 -08:00
Aleix Conchillo Flaqué
35d265770d LLMUserAggregator: don't process certain self-queued frames 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
9632efec8c VADProcessor: broadcast frames 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
27dbfa1eda NvidiaTTSService: return AsyncIterator instead of AsyncIterable 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
183c0aa4ef LLMUserAggregator: queue frames internally so strategies and controllers can process them 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
a69a037ffa changelog: add updates for #3583 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
c46e7f5da0 TurnAnalyzerUserTurnStopStrategy: only update vad params if frame contains vad 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
307aeaeda0 examples: update with LLMUserAggregatorParams vad_analyzer and VADProcessor 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
305ab44132 tests: add unittest.main() call 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
b486f35c70 audio: add new VADProcessor 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
c92080b0d2 LLMUserAggregator: add vad_analyzer and use VADController 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
ddfedaf478 audio(vad): add new VADController 2026-01-30 10:07:34 -08:00