wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
4,558 Commits 276 Branches 111 Tags
5c3f4180b9bee2cc30f7ec91411f23f7303e5ae2
Commit Graph

2589 Commits

Author SHA1 Message Date
Filipi Fuchter
5c3f4180b9 Refactored VAD analyzer to process multiple audio frames in a single iteration if needed. 2025-07-11 10:59:32 -03:00
Filipi Fuchter
1a146c2a64 Not serializing a JSON in case we have no audio. 2025-07-11 10:15:09 -03:00
Filipi Fuchter
eaeb9e6efa Not creating InputAudioRawFrame in case we don't have bytes. Fixed for Pilvo, Exotel and Telnyx. 2025-07-11 09:51:38 -03:00
Filipi Fuchter
1200aa4fb8 Not creating InputAudioRawFrame in case we don't have bytes. 2025-07-10 21:56:34 -03:00
Filipi Fuchter
b2ead325c4 Fixed an issue in ParallelPipeline that caused errors when attempting to drain the queues. 2025-07-10 21:50:35 -03:00
kompfner
b610ee26ba Merge pull request #2181 from pipecat-ai/pk/fix-aws-nova-sonic-pipeline-freeze 
Fix a pipeline freeze when using AWS Nova Sonic. The freeze occurs if…
 2025-07-10 16:30:55 -04:00
Paul Kompfner
2b867f1613 Fix a pipeline freeze when using AWS Nova Sonic. The freeze occurs if the user starts speaking before we've finished sending the "trigger " audio (AWS Nova Sonic can only start speaking in response to a user utterance, so we have a simulated user utterance to "trigger" the bot speaking without the user having actually spoken first). 2025-07-10 15:57:05 -04:00
Mark Backman
7b8fe565c7 Merge pull request #2182 from pipecat-ai/mb/run-example-usage 
run.py: Add example usage to the module docstring
 2025-07-10 12:48:29 -07:00
Mark Backman
a246862910 run.py: Add example usage to the module docstring 2025-07-10 11:41:49 -07:00
carolin-tavus
f0d8499f7e feat: avoid checking microphone enabled 2025-07-10 09:40:27 +00:00
Mark Backman
d98b6b418d Release prep for 0.0.75 (#2169) 
* Update CHANGELOG for 0.0.75 release

* Add RTVI changes to CHANGELOG

* Update CHANGELOG, add deprecation directives to rtvi.py

---------

Co-authored-by: mattie ruth backman <mattieruth@gmail.com>
 2025-07-08 15:39:44 -04:00
Mark Backman
b55b9c257b RivaSTTService: remove reset_watchdog, which is handled in the WatchdogQueue already 2025-07-08 11:19:47 -07:00
Mark Backman
a9d824753b Fix: NeuphonicTTSService to use latest websocket API 2025-07-08 11:08:08 -07:00
Filipi da Silva Fuchter
3c6a208101 Merge pull request #2148 from pipecat-ai/filipi/aws_bedrock 
Refactoring AWSBedrockLLMService to work async
 2025-07-08 12:14:28 -03:00
Mark Backman
931f34fccd OLLamaLLMService: Pass kwargs 2025-07-08 07:17:18 -07:00
Mark Backman
f2509adec1 Merge pull request #2162 from pipecat-ai/mb/tts-service-aggregate-sentences 
TTS services: Add aggregate_sentences arg
 2025-07-08 07:14:58 -07:00
Filipi Fuchter
74da197304 Refactored AWSBedrockLLMService and AWSPollyTTSService to work asynchronously using aioboto3 instead of the boto3 library. 2025-07-08 07:28:23 -03:00
mattie ruth backman
2959029151 PR Review fixes 2025-07-07 15:56:08 -07:00
mattie ruth backman
e590441b7b Add support for about info in ready messages and add deprecation comments to deprecated types 2025-07-07 15:56:08 -07:00
mattie ruth backman
43049c865c Add support for new RTVI client message protocol: handling and responding 2025-07-07 15:56:08 -07:00
Mark Backman
f53f45a6cd TTS services: Add aggregate_sentences arg 2025-07-07 15:38:31 -07:00
shahrukhx01
a84e7e30da WhisperSTTService: Add additional whisper model variants 2025-07-07 21:43:48 +02:00
Mark Backman
1375211610 UserIdleProcessor: Account for function calls in progress 2025-07-04 14:05:05 -07:00
Mark Backman
f9e8748a96 TwilioFrameSerializer: Handle user hanging up before the serializer 2025-07-04 09:42:16 -07:00
Filipi Fuchter
b573f9dab2 Removing duplicated code inside Gemini. 2025-07-04 10:57:53 -03:00
Aleix Conchillo Flaqué
02b63c28a5 FrameProcessor: remove unnecessary push task 
When we call `FrameProcessor.push_frame()` we end up calling
`FrameProcessor.queue_frame()` on the next or previous processor which already
uses the input queue and guarantees frame ordering. So, there's no need to have
a two queues next to each other.
 2025-07-03 13:57:32 -07:00
Mark Backman
2deca816ae Merge pull request #2137 from pipecat-ai/mb/fish-audio-normalize 
FishAudioTTSService: arg cleanup, add new InputParam and arg
 2025-07-03 13:29:14 -07:00
Mark Backman
b8d2fceced Merge pull request #2138 from pipecat-ai/mb/fix-google-llm-import-order 
GoogleLLMService: Linting fixes
 2025-07-03 13:26:32 -07:00
Sam Sykes
7596d71460 Speechmatics STT + multi-speaker conversations (#2036) 
* initial config

* skeleton

* Added a README (to be added to).

* Payloads coming from the ASR.

* doc update

* handle the partials and finals

* enable diarization in the example

* support sending messages to pipecat pipeline

* requirements fix in README

* updated example (with amusement)

* updated example to match master

* updated docs

* support for diarization tags

* logic fix for wrapper

* Use an internal SpeechFrame for speaker_id (not user_id).

* only include speaker tags on finalised transcript (as this may skew end of utterance detection)

* updated docs

* correction to docs and updated example

* updated requirement

* Fix for using default EU server.

* Updates from PR comments.

* Refactor based on comments in the original PR.

Primary focus on documentation, naming conventions and how `user_id` is used.

* Check for SMX installed when importing.

* Variable name change

* Comment correction.

* Support for Esporanto and Uyghur

* Impoved language support

* function name change

* Locale fix

* intercept

* interim changes

* pass the pipeline task to the module for adding events to the top of the pipeline

* logging for the pipeline

* Reduce timeout for content aggregator.

* staged update

* testing with Azure

* Updated context (Azure was dropping punctuation) and using better ElevenLabs model.

* Updated to RT 0.3.0 and use OpenAI (not Azure).

* Missing OpenAI import; parameter name change for output locale validation.

* Revert to `0.2.0` of RT SDK.

* fix for assignment of `output_locale_code`.

* update Speechmatics library to 0.3.1

* new transcription example

* updated asyncio task handling

* Updated doc strings

* enable OpenTelemetry logging

* removed import from stt for __init__

* updated examples and default values

* updated examples

* prevent lock up when closing the STT connection
 2025-07-03 17:25:13 -03:00
Mark Backman
096067b097 GoogleLLMService: Linting fixes 2025-07-03 13:23:13 -07:00
Mark Backman
ec09505f6b FishAudioTTSService: Add normalize as InputParam, model_id as arg 2025-07-03 13:14:15 -07:00
Mark Backman
251ea756c8 FishTTSService: deprecate model, add reference_id 2025-07-03 12:56:24 -07:00
otaqwawi
6045a8ad8c Add option to change the base URL for Google Generative AI. (#2113) 
* Add option to change the base URL for Google Generative AI.
This would be useful to support private instance or gateway of the API

* fix: add proper type hints for http_options in Google LLM service
 2025-07-03 11:12:35 -07:00
Mark Backman
c230d94ff0 Merge pull request #2125 from pipecat-ai/mb/deprecate-handle-function-call-start 
Add docs deprecation for handle_function_call_start
 2025-07-03 12:27:17 -04:00
Aleix Conchillo Flaqué
e7b02773f5 Merge pull request #2131 from pipecat-ai/aleix/dtmf-aggregator-dangling-tasks 
DtmfAggregator: cancel interruption task to avoid a dangling task
 2025-07-03 08:34:50 -07:00
Aleix Conchillo Flaqué
af8b4901d4 DtmfAggregator: cancel interruption task to avoid a dangling task 2025-07-03 08:18:48 -07:00
Aleix Conchillo Flaqué
64c8230960 PipelineTask: cancel idle queue before cancelling task 2025-07-03 08:18:21 -07:00
carolin-tavus
cb81f3d50e format 2025-07-03 14:38:20 +00:00
carolin-tavus
30a3b24287 Add persona validation (check that microphone is enabled) 2025-07-03 14:04:04 +00:00
Victor
72d503d3a3 Azure TTS fixed by clearing the audio queue before synthesizing the next text 2025-07-03 10:48:26 -03:00
Mark Backman
4ae045d704 Add docs deprecation for handle_function_call_start 2025-07-02 19:53:48 -07:00
Mark Backman
8c71939425 Merge pull request #2122 from pipecat-ai/mb/deprecation-docstrings 
Add deprecation directives, add indexing, only autodoc members
 2025-07-02 21:31:02 -04:00
Mark Backman
abee0f853c Add deprecation directives, add indexing, only autodoc members 2025-07-02 15:44:02 -07:00
Aleix Conchillo Flaqué
c5d54d06bb add run_llm to LLMMessagesAppendFrame and LLMMessagesUpdateFrame 2025-07-02 13:53:13 -07:00
Filipi Fuchter
76388a10b5 Deprecating the create_default_resampler and adding the changelog. 2025-07-02 16:20:58 -03:00
Filipi Fuchter
38bcc033a2 Improving the docs about when to use: SOXRAudioResampler x SOXRStreamAudioResampler 2025-07-02 16:20:48 -03:00
Filipi Fuchter
5af563cd91 Configured the services to use create_stream_resampler instead of create_default_resampler 2025-07-02 16:20:34 -03:00
Filipi Fuchter
c19f9bc43a Creating a new stream resampler which avoids clicks. 2025-07-02 16:19:47 -03:00
Aleix Conchillo Flaqué
f9086ee3a2 Merge pull request #2110 from pipecat-ai/aleix/daily-add-virtual-speaker-support 
DailyTransport: allow receiving audio in a single track
 2025-07-02 09:50:02 -07:00
Vanessa Pyne
d80e228c6f Merge branch 'main' into mcp-http-gemini-support 2025-07-02 11:47:18 -05:00