wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
8,695 Commits 276 Branches 111 Tags
63254fe337986ae9060fa33db2007eaeec7fb1aa
Commit Graph

19 Commits

Author SHA1 Message Date
Paul Kompfner
d779a5b4ea Use "developer" role for programmatic conversation-kickoff messages 
These messages are developer instructions to the assistant (e.g. "Please
introduce yourself to the user"), not simulated user input. The
"developer" role is semantically correct for this purpose.
 2026-03-24 16:02:42 -04:00
Mark Backman
978a1a2083 Update the system_instruction wording in the foundational examples to not mention WebRTC call 2026-03-13 12:22:10 -04:00
Mark Backman
cd28c82de3 Update examples to use the class Settings alias 2026-03-07 09:15:24 -05:00
Mark Backman
c3794956ef Add deprecation version, fix foundational example double system message 2026-03-06 15:16:58 -05:00
Aleix Conchillo Flaqué
593b75bc8b Update foundational examples to use "user" role 
Use system_instruction on LLM service constructors instead of adding
system messages to LLMContext. Messages added to context now use
"user" role.
 2026-03-06 09:53:33 -08:00
Paul Kompfner
78deaa735d Move system_instruction into LLMSettings 
Add `system_instruction` field to `LLMSettings` so it is runtime-updatable via settings.
For Google (GoogleLLMService, GoogleVertexLLMService), deprecate the init-time arg since it was already shipped. For Anthropic, AWS Bedrock, and OpenAI, remove the init-time arg entirely since it was never shipped.

Still need to handle realtime services (OpenAI Realtime, Grok Realtime, Gemini Live).
 2026-03-06 09:57:08 -05:00
Aleix Conchillo Flaqué
0004a116d8 examples(foundational): use system_instruction in all examples 2026-03-04 17:37:32 -08:00
Aleix Conchillo Flaqué
83a8379401 examples: remove the now default turn analyzer user turn stop strategy 2026-02-09 14:33:38 -08:00
Mark Backman
a5fc2b1650 Set VADParams stop_secs to 0.2 by default 2026-02-06 23:49:08 -05:00
Aleix Conchillo Flaqué
307aeaeda0 examples: update with LLMUserAggregatorParams vad_analyzer and VADProcessor 2026-01-30 10:07:34 -08:00
Aleix Conchillo Flaqué
861588e4a3 examples: update all examples to use the new LLMContextAggregatorPair tuple 2026-01-13 14:01:53 -08:00
Aleix Conchillo Flaqué
2626154a64 update examples and tests copyright and use a proper dash in 2024-2026 2026-01-07 19:32:22 -08:00
Aleix Conchillo Flaqué
eb5a797b12 turns: rename bot turn start to user turn stop strategies 2025-12-30 14:33:58 -08:00
Aleix Conchillo Flaqué
5496aa722f turns: simplify imports and don't require full strategy module path 2025-12-28 16:20:15 -08:00
Aleix Conchillo Flaqué
8b861d9143 LLMUserAggregator: move turn_start_strategies from PipelineTask 2025-12-28 08:16:34 -08:00
Aleix Conchillo Flaqué
d22e1f18bb examples: update with new user and bot turn start strategies 2025-12-19 14:47:02 -08:00
Paul Kompfner
8851d18f92 Tweak the LLM prompt again to try to fix the issue of LLMs sometimes omitting punctuation in their output. 2025-11-13 10:02:33 -05:00
Paul Kompfner
1802f949ef Fix an issue with some examples where punctuation was missing from the LLM output, by tweaking the LLM prompt. 2025-11-12 17:12:03 -05:00
Paul Kompfner
5222ff99de Apply includes_inter_frame_spaces = True in all LLM and TTS services that need it. 
Note that for `LLMTextFrame`s, the right behavior is pretty much always `includes_inter_frame_spaces = True`. I decided *not* to go ahead and make that the default for `LLMTextFrame`s, though, simply to not introduce a subtle behavior change for creative/unexpected use-cases that were relying on text in hand-crafted `LLMTextFrame`s being handled a certain way. Ditto for `TTSTextFrame`s.

Also, fix an issue in `NeuphonicTTSService` where it wasn't pushing `TTSTextFrame`s.

Also, fix the broken `SarvamHttpTTSService` example.

Also, add a couple of missing examples.
 2025-11-12 15:10:11 -05:00