wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #3611 from pipecat-ai/aleix/aicoustics-example-update

Browse Source 
examples: update 07zd to use vad_analyzer in LLMUserAggregator
This commit is contained in:
Aleix Conchillo Flaqué
2026-01-31 21:02:50 -08:00
committed by GitHub
parent bfd19e867c 675c7c43e3
commit 52cc64019a
1 changed files with 27 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

64
examples/foundational/07zd-interruptible-aicoustics.py
View File

@@ -38,13 +38,6 @@ from pipecat.turns.user_turn_strategies import UserTurnStrategies
load_dotenv(override=True) load_dotenv(override=True)
# Create audio buffer processor so we can hear the audio fitler results.
audiobuffer = AudioBufferProcessor(
num_channels=2, # 1 for mono, 2 for stereo (user left, bot right)
enable_turn_audio=False, # Enable per-turn audio recording
)
def _create_aic_filter() -> AICFilter: def _create_aic_filter() -> AICFilter:
license_key = os.getenv("AICOUSTICS_LICENSE_KEY", "") license_key = os.getenv("AICOUSTICS_LICENSE_KEY", "")
@@ -54,39 +47,29 @@ def _create_aic_filter() -> AICFilter:
) )
aic_filter = _create_aic_filter()
aic_vad_analyzer = aic_filter.create_vad_analyzer(
speech_hold_duration=0.05, minimum_speech_duration=0.0, sensitivity=6.0
)
# We use lambdas to defer transport parameter creation until the transport # We use lambdas to defer transport parameter creation until the transport
# type is selected at runtime. # type is selected at runtime.
transport_params = { transport_params = {
"daily": lambda: ( "daily": lambda: DailyParams(
lambda aic: DailyParams( audio_in_enabled=True,
audio_in_enabled=True, audio_out_enabled=True,
audio_out_enabled=True, audio_in_filter=aic_filter,
vad_analyzer=aic.create_vad_analyzer( ),
speech_hold_duration=0.05, minimum_speech_duration=0.0, sensitivity=6.0 "twilio": lambda: FastAPIWebsocketParams(
), audio_in_enabled=True,
audio_in_filter=aic, audio_out_enabled=True,
) audio_in_filter=aic_filter,
)(_create_aic_filter()), ),
"twilio": lambda: ( "webrtc": lambda: TransportParams(
lambda aic: FastAPIWebsocketParams( audio_in_enabled=True,
audio_in_enabled=True, audio_out_enabled=True,
audio_out_enabled=True, audio_in_filter=aic_filter,
vad_analyzer=aic.create_vad_analyzer( ),
speech_hold_duration=0.05, minimum_speech_duration=0.0, sensitivity=6.0
),
audio_in_filter=aic,
)
)(_create_aic_filter()),
"webrtc": lambda: (
lambda aic: TransportParams(
audio_in_enabled=True,
audio_out_enabled=True,
vad_analyzer=aic.create_vad_analyzer(
speech_hold_duration=0.05, minimum_speech_duration=0.0, sensitivity=6.0
),
audio_in_filter=aic,
)
)(_create_aic_filter()),
} }
@@ -113,12 +96,19 @@ async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
user_aggregator, assistant_aggregator = LLMContextAggregatorPair( user_aggregator, assistant_aggregator = LLMContextAggregatorPair(
context, context,
user_params=LLMUserAggregatorParams( user_params=LLMUserAggregatorParams(
vad_analyzer=aic_vad_analyzer,
user_turn_strategies=UserTurnStrategies( user_turn_strategies=UserTurnStrategies(
stop=[TurnAnalyzerUserTurnStopStrategy(turn_analyzer=LocalSmartTurnAnalyzerV3())] stop=[TurnAnalyzerUserTurnStopStrategy(turn_analyzer=LocalSmartTurnAnalyzerV3())]
), ),
), ),
) )
# Create audio buffer processor so we can hear the audio fitler results.
audiobuffer = AudioBufferProcessor(
num_channels=2, # 1 for mono, 2 for stereo (user left, bot right)
enable_turn_audio=False, # Enable per-turn audio recording
)
pipeline = Pipeline( pipeline = Pipeline(
[ [
transport.input(), # Transport user input transport.input(), # Transport user input