Commit Graph

5568 Commits

Author SHA1 Message Date
kompfner
fca4ecc73c Merge pull request #2675 from pipecat-ai/pk/service-switcher-logic-simplification
Simplify a bit of logic in `ServiceSwitcher`
2025-09-18 09:17:22 -04:00
Paul Kompfner
cfa333508b Simplify a bit of logic in ServiceSwitcher 2025-09-17 21:03:38 -04:00
Mark Backman
9e7260393a Merge pull request #2671 from pipecat-ai/mb/fix-asyncai-ttstextframe
fix: AsyncAITTSService wasn't pushing TTSTextFrames
2025-09-17 14:06:41 -07:00
Mark Backman
073b585c52 fix: AsyncAITTSService wasn't pushing TTSTextFrames 2025-09-17 16:54:18 -04:00
Aleix Conchillo Flaqué
81c2e51bec Merge pull request #2669 from pipecat-ai/aleix/interruption-task-frame-wait-fixes
interruption task frame wait fixes
2025-09-17 13:47:57 -07:00
Aleix Conchillo Flaqué
42344125b1 tests: add unit tests for push_interruption_task_frame_and_wait() 2025-09-17 13:38:22 -07:00
Aleix Conchillo Flaqué
db5bcfaa51 FrameProcessor: fix push_interruption_task_frame_and_wait() 2025-09-17 13:38:21 -07:00
kompfner
615239b7d2 Merge pull request #2646 from pipecat-ai/pk/service-switcher-unit-tests
`ServiceSwitcher` unit tests (ended up being much more than that)
2025-09-17 16:30:18 -04:00
Paul Kompfner
27f1e9dd69 Update CHANGELOG with a description of the recently-fixed ServiceSwitcher bugs 2025-09-17 16:27:12 -04:00
Paul Kompfner
bd760deff2 Update comment with more detail for posterity 2025-09-17 16:19:31 -04:00
Paul Kompfner
8bc3c89140 Fix a bug preventing usage of multiple ServiceSwitchers in a pipeline 2025-09-17 16:09:18 -04:00
Paul Kompfner
2cd2567a37 Add a unit tests validating that multiple ServiceSwitchers can be used in the same pipeline (currently failing) 2025-09-17 16:04:30 -04:00
Paul Kompfner
5b55988846 Denote a couple of variables are private with a leading underscore 2025-09-17 15:38:28 -04:00
Paul Kompfner
a12392182c Simplify, undoing the change allowing controlling ServiceSwitcher with immediate frames (SystemFrames). Service switcher frames are ControlFrames, which are easier to reason about. We can always build the immediate option later if needed (i.e. if there's sufficient user pull for it) 2025-09-17 15:35:02 -04:00
Paul Kompfner
b814b70e1e Allow controlling ServiceSwitcher with either immediate frames (SystemFrames) or in-order frames (ControlFrames).
Immediate is the "default", i.e. has the more obvious name (e.g. `ManuallySwitchServiceFrame` v `ManuallySwitchServiceControlFrame`), since that's *probably* what users will want to reach for. Also, the immediate frames are more likely to behave like what we had before the last few commits, where the service switch would always "jump the queue" by having an immediate effect once it hit the `ServiceSwitcher` in the pipeline, jumping ahead of frames in front of it destined for the service.
2025-09-17 15:35:02 -04:00
Paul Kompfner
a1f84e1b50 Remove extraneous unit tests 2025-09-17 15:35:02 -04:00
Paul Kompfner
0839b48da8 Fix an issue where the upstream ServiceSwitcherFilter wouldn't get updated with the current active service 2025-09-17 15:35:02 -04:00
Paul Kompfner
de51637b77 Update ServiceSwitcher so that ServiceSwitcherFrames (which might update the currently active service) are processed and have an effect at the expected time. We should be able to, for example, queue:
- A text frame
- A `ManuallySwitchServiceFrame` (which is a `ServiceSwitcherFrame`)
- Another text frame

And expect that the first text frame be handled by the initially active service and the second text frame be handled by the newly active one.

Previously, the `ManuallySwitchServiceFrame` would have an effect too early, causing both text frames to be handled by the newly active service. Why? Because the frame filtering condition was being updated *directly* by the `ServiceSwitcher`, which is upstream from the services it's switching between. It could therefore update the filters *before* the services received the prior frames.
2025-09-17 15:35:02 -04:00
Paul Kompfner
e1b1dc16ec Add unit tests for ServiceSwitcher 2025-09-17 15:35:02 -04:00
Mark Backman
1fe27eb0a2 Merge pull request #2660 from pipecat-ai/mb/fix-user-idle-processor-cancel-task
fix: clean up how UserIdleProcessor handles return False
2025-09-16 14:48:59 -07:00
Mark Backman
d7e1389497 fix: clean up how UserIdleProcessor handles return False 2025-09-16 17:44:06 -04:00
Aleix Conchillo Flaqué
8c7230aa8f Merge pull request #2668 from pipecat-ai/aleix/livekit-update
livekit package update
2025-09-16 14:43:18 -07:00
Aleix Conchillo Flaqué
2cf71239b0 examples(01b): use TTSSpeakFrame instead of TextFrame 2025-09-16 17:18:45 -04:00
Aleix Conchillo Flaqué
ec2c62e32b pyproject: update to livekit 1.0.13
Fixes #2643
2025-09-16 17:18:44 -04:00
Mark Backman
38ce85e9a0 Merge pull request #2667 from zytegalaxy/mcp-serverparameters-typefix
fix: replace `Tuple` type with `TypeAlias` for server params in MCP client
2025-09-16 14:14:59 -07:00
Mark Backman
2279e5a899 Merge pull request #2663 from pipecat-ai/mb/websockets-15
Add support for websockets 15.0
2025-09-16 14:08:36 -07:00
Mark Backman
cce6eb5d87 Merge pull request #2666 from pipecat-ai/mb/update-38b-local-turn-model
38b: Update bundled ONNX smart-turn model
2025-09-16 14:05:12 -07:00
mehrdad
c2b98ae557 fix(lint): fix space format issue 2025-09-16 13:44:15 -07:00
Filipi da Silva Fuchter
727eb12b16 Merge pull request #2648 from pipecat-ai/filipi/pcc_small_webrtc
Creating SmallWebRTCRequestHandler for managing peer connections.
2025-09-16 16:37:04 -03:00
mehrdad
ba96bd05d3 fix: replace Tuple type with TypeAlias for server params in MCP client 2025-09-16 11:44:25 -07:00
Mark Backman
8ead309f8d 38b: Update bundled ONNX smart-turn model 2025-09-16 13:17:14 -04:00
Mark Backman
fad0e55c64 Add websockets-base optional dependency and use for DRY pyproject.toml 2025-09-16 11:24:38 -04:00
Mark Backman
74b1af56a0 Update uv.lock 2025-09-16 11:21:49 -04:00
Mark Backman
6924850ec4 Add support for websockets 15.0 2025-09-16 11:21:49 -04:00
marcus-daily
dfe7815dc5 Smart Turn v3: removing torch and torchaudio deps 2025-09-16 16:02:41 +01:00
Mark Backman
cca90791c4 Merge pull request #2652 from pipecat-ai/mb/fix-audio-buffer-processor-has-audio
fix: AudioBufferProcessor has_audio returns based on user or bot audi…
2025-09-15 18:43:59 -07:00
Mark Backman
f2a5d408de fix: AudioBufferProcessor has_audio returns based on user or bot audio existing 2025-09-15 21:35:35 -04:00
Aleix Conchillo Flaqué
044c6eba46 Merge pull request #2655 from pipecat-ai/aleix/add-on-pipeline-finalized
PipelineTask: add on_pipeline_finished event
2025-09-15 15:32:04 -07:00
Aleix Conchillo Flaqué
db71089f5e PipelineTask: add on_pipeline_finished event
This deprecates `on_pipeline_stopped`, `on_pipeline_ended` and
`on_pipeline_cancelled`.
2025-09-15 15:28:33 -07:00
Aleix Conchillo Flaqué
f861f5066f Merge pull request #2654 from pipecat-ai/aleix/unify-on-client-disconnected
transports: on_client_disconnected only if remote client disconnects
2025-09-15 15:18:57 -07:00
kompfner
81cede2c60 Merge pull request #2653 from pipecat-ai/pk/llm-context-adapting-tests
`LLMContext`-adapting unit tests
2025-09-15 16:38:46 -04:00
kompfner
7603203230 Merge pull request #2644 from pipecat-ai/pk/run-inference-unit-tests
`run_inference` unit tests
2025-09-15 16:26:10 -04:00
Aleix Conchillo Flaqué
8569b61598 transports: on_client_disconnected only if remote client disconnects 2025-09-15 11:35:40 -07:00
Paul Kompfner
fe42187dc1 Implement LLMService.create_llm_specific_message() so that users don't need to just know what value of llm to provide to the LLMSpecificMessage constructor 2025-09-15 14:15:22 -04:00
Paul Kompfner
999e88c942 Add unit tests for AWSBedrockLLMAdapter.get_llm_invocation_params(), focusing on messages specifically 2025-09-15 12:08:21 -04:00
Paul Kompfner
c04df2f28b Add unit tests for AnthropicLLMAdapter.get_llm_invocation_params(), focusing on messages specifically 2025-09-15 11:55:48 -04:00
Paul Kompfner
100ef0ab5c Add unit tests for GeminiLLMAdapter.get_llm_invocation_params(), focusing on messages specifically 2025-09-15 11:38:23 -04:00
Paul Kompfner
42886d7105 Add unit tests for OpenAILLMAdapter.get_llm_invocation_params(), focusing on messages specifically. Also, fix a bug in OpenAILLMAdapter (found thanks to the unit tests) where it wasn't "unwrapping" LLMSpecificMessages. 2025-09-15 11:17:11 -04:00
Mark Backman
22cbba002a Merge pull request #2651 from pipecat-ai/mb/heygen-bot-speaking-frame
fix: push BotStartedSpeakingFrame in HeyGenVideoService
2025-09-15 08:02:25 -07:00
Filipi Fuchter
0a043154f2 Removing not used import. 2025-09-15 10:46:43 -03:00