Aleix Conchillo Flaqué
ef3d732607
DailyTransport: allow capturing multiple simultaneous audio/video sources
2025-05-01 18:58:44 -07:00
Aleix Conchillo Flaqué
6d63cff1bf
DailyTransport: custom audio tracks support
2025-05-01 18:58:44 -07:00
Mark Backman
a6a4d3d71f
Merge pull request #1706 from rimelabs/matt/rime/update_url
...
[Rime] - Update url for Websockets API
2025-04-30 19:14:04 -04:00
Mark Backman
c52de0f5de
Merge pull request #1696 from pipecat-ai/mb/fix-gemini-live-context
...
Fix: GeminiMultimodalLiveLLMService was appending tokens to the context
2025-04-30 19:12:06 -04:00
Mark Backman
a1e1255f16
Strip newlines from generated user transcript
2025-04-30 18:27:46 -04:00
Mark Backman
c4f758725e
Ignore TranscriptionFrames too
2025-04-30 18:22:43 -04:00
Aleix Conchillo Flaqué
957fa5546d
RTVIObserver: add RTVIObserverParams to configure what to send
2025-04-30 15:09:02 -07:00
Aleix Conchillo Flaqué
039cb8fcae
Merge pull request #1690 from pipecat-ai/aleix/rtvi-function-call-single-param
...
RTVIProcessor: use single FunctionCallParams
2025-04-30 15:04:05 -07:00
Mark Backman
8e05f2f1a1
Merge pull request #1702 from pipecat-ai/mb/stt-mute-transcription-frames
...
Add InterimTranscriptionFrame and TranscriptionFrame to STTMuteFilter…
2025-04-30 17:54:24 -04:00
Matt Kim
8467aa1ed3
[Rime] - Update url for Websockets API
...
Rime has migrated their Websockets api to the base url `user.rime.ai` along with all other tts endpoints.
See the [docs](https://docs.rime.ai/api-reference/endpoint/websockets )
`users-ws.rime.ai` is deprecated and will not reflect upgrades to the rime ws api.
2025-04-30 14:20:13 -07:00
Mark Backman
9c5878af3d
OpenAI Realtime and Gemini Live push LLMTextFrame again, overwrite the assitant context aggregator for LLMTextFrame
2025-04-30 17:18:20 -04:00
Mark Backman
7e09933070
OpenAI Realtime should push TTSTextFrame only
2025-04-30 16:28:17 -04:00
Mark Backman
82a9d7f992
Gemini Mulitmodal Live to push TTSTextFrame only
2025-04-30 16:28:17 -04:00
Mark Backman
2ba60fc41f
Update TranscriptProcessor to handle GeminiMultimodalLiveLLMService changes
2025-04-30 16:28:16 -04:00
Mark Backman
685f951ae2
Fix: GeminiMultimodalLiveLLMService was appending tokens to the context
2025-04-30 16:28:16 -04:00
Mark Backman
20a59e8c56
Add InterimTranscriptionFrame and TranscriptionFrame to STTMuteFilter frame processing
2025-04-30 10:50:56 -04:00
Mark Backman
d9a0a93667
Add context_window_compression support to GeminiMultimodalLiveLLMService
2025-04-30 09:55:34 -04:00
Mark Backman
10cdc47e05
Merge pull request #1689 from pipecat-ai/mb/handle-http-smart-turn-errors
...
Handle case where Fal Smart Turn returns a 500 error
2025-04-29 14:21:45 -04:00
Filipi Fuchter
080f70d91c
Allowing to define the username and credential for the ice servers.
2025-04-29 10:32:42 -03:00
Mark Backman
ebed1fc6ea
Restructure _send_raw_request to raise errors early, then process the successful response
2025-04-29 09:05:14 -04:00
Aleix Conchillo Flaqué
6821b1cdab
RTVIProcessor: use single FunctionCallParams
2025-04-29 05:56:23 -07:00
Mark Backman
144ae9b611
Handle case where Fal Smart Turn returns a 500 error
2025-04-29 08:53:02 -04:00
Aleix Conchillo Flaqué
8accd3e387
Merge pull request #1681 from pipecat-ai/aleix/tts-service-llm-full-response-end-fix
...
TTSService: do not push LLMFullResponseEndFrame if not needed
2025-04-28 12:00:55 -07:00
Aleix Conchillo Flaqué
e3c965f4d5
TTSService: do not push LLMFullResponseEndFrame if not needed
2025-04-28 11:03:22 -07:00
Dev-Khant
5354e5d891
formatting
2025-04-28 22:57:18 +05:30
Dev-Khant
aa23a7b1e6
update changelog
2025-04-28 20:03:20 +05:30
Dev-Khant
c0c41789ab
Integrate with Mem0 OSS
2025-04-28 17:15:53 +05:30
Aleix Conchillo Flaqué
9cad6dfce9
RTVIProcessor: simplify handle_function_call() and depreacted handle_function_call_start()
2025-04-25 13:34:05 -07:00
Aleix Conchillo Flaqué
944bc23135
LLMService: use a single FunctionCallParams parameter for function calls
2025-04-25 13:34:03 -07:00
Aleix Conchillo Flaqué
1d863ee7de
Merge pull request #1669 from pipecat-ai/aleix/short-utterances-fixes
...
short utterances fixes
2025-04-25 13:25:15 -07:00
Aleix Conchillo Flaqué
03002ad685
LLMUserContextAggregator: reduce aggregation_timeout to 0.5
2025-04-25 13:21:00 -07:00
Aleix Conchillo Flaqué
99a4154cbc
LLMUserContextAggregator: ignore short uterrances while bot speaking
2025-04-25 13:21:00 -07:00
Aleix Conchillo Flaqué
3ac50b9902
Merge pull request #1664 from CerebriumAI/kyle/fix-ultravox-performance
...
Improved: Ultravox performance
2025-04-25 12:59:25 -07:00
Michael Louis
9557705b53
Changed warmup to internal function
2025-04-25 14:40:19 -04:00
Mark Backman
dfa10af6ed
Simplify VAD events to be detected and emitted from BaseInputTransport
2025-04-25 10:28:44 -04:00
Mark Backman
b298376766
Add VADUserStartedSpeakingFrame and VADUserStoppedSpeakingFrame
2025-04-25 09:06:54 -04:00
Mark Backman
d1086914fe
Add TranslationFrame and use in GladiaSTTService; add 13c-gladia-translation.py
2025-04-25 08:25:39 -04:00
Dan Berg
be8788e4da
Push gladia translations as a TranscriptionFrame
2025-04-25 08:24:51 -04:00
Kyle Gani
5c4079b286
Cleaned up: layout
2025-04-25 13:18:03 +02:00
Kyle Gani
5489ac5a73
Merge branch 'main' into kyle/fix-ultravox-performance
2025-04-25 13:16:48 +02:00
Kyle Gani
49fbcc86ac
Improved: Ultravox performance
2025-04-25 13:12:08 +02:00
Aleix Conchillo Flaqué
9fd76923fd
STTService: passthrough audio frames by default
2025-04-24 17:14:18 -07:00
Aleix Conchillo Flaqué
4ee6c4b59e
BaseOutputTransport: reword camera with video
2025-04-24 17:14:18 -07:00
Aleix Conchillo Flaqué
420912dd4b
transports: deprecate TransportParams.camera_* in favor of video_*
2025-04-24 17:14:18 -07:00
Aleix Conchillo Flaqué
8bfcfe8b1d
transports: deprecate TransportParams.vad_enabled
2025-04-24 17:14:18 -07:00
Aleix Conchillo Flaqué
9ee56bff9e
transports: push audio by default. s/vad_audio_passthrough/audio_in_passthrough/
2025-04-24 17:14:16 -07:00
vipyne
b7b2a5b7a1
mcp service fix and add multiple mcp example
2025-04-24 17:13:40 -05:00
vipyne
3384598e07
mcp_service: pr notes
2025-04-24 17:13:40 -05:00
vipyne
c420dbe57f
MCP Service
...
WIP getting mcp.run to work
add mcp[cli] to toml and lint
mcp stdio example
mcp sse example
mcp_run POC
ruff formatting
2025-04-24 17:13:40 -05:00
Mark Backman
70033ae00b
fix: SmartTurnMetricsData was reporting 0 for inference and processing time
2025-04-23 17:17:46 -04:00
