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

Fix "bot-llm-text" not firing when using AWS Nova Sonic

Browse Source
This commit is contained in:
Paul Kompfner
2026-01-15 14:48:39 -05:00
parent 24082b84f2
commit 7a22d58cf4
1 changed files with 22 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
src/pipecat/services/aws/nova_sonic/llm.py
View File

@@ -38,6 +38,7 @@ from pipecat.frames.frames import (
LLMContextFrame, LLMContextFrame,
LLMFullResponseEndFrame, LLMFullResponseEndFrame,
LLMFullResponseStartFrame, LLMFullResponseStartFrame,
LLMTextFrame,
StartFrame, StartFrame,
TranscriptionFrame, TranscriptionFrame,
TTSAudioRawFrame, TTSAudioRawFrame,
@@ -1077,9 +1078,7 @@ class AWSNovaSonicLLMService(LLMService):
logger.debug(f"Assistant response text added: {text}") logger.debug(f"Assistant response text added: {text}")
# Report the text of the assistant response. # Report the text of the assistant response.
frame = TTSTextFrame(text, aggregated_by=AggregationType.SENTENCE) await self._push_assistant_response_text_frames(text)
frame.includes_inter_frame_spaces = True
await self.push_frame(frame)
# HACK: here we're also buffering the assistant text ourselves as a # HACK: here we're also buffering the assistant text ourselves as a
# backup rather than relying solely on the assistant context aggregator # backup rather than relying solely on the assistant context aggregator
@@ -1112,11 +1111,7 @@ class AWSNovaSonicLLMService(LLMService):
# TTSTextFrame would be ignored otherwise (the interruption frame # TTSTextFrame would be ignored otherwise (the interruption frame
# would have cleared the assistant aggregator state). # would have cleared the assistant aggregator state).
await self.push_frame(LLMFullResponseStartFrame()) await self.push_frame(LLMFullResponseStartFrame())
frame = TTSTextFrame( await self._push_assistant_response_text_frames(self._assistant_text_buffer)
self._assistant_text_buffer, aggregated_by=AggregationType.SENTENCE
)
frame.includes_inter_frame_spaces = True
await self.push_frame(frame)
self._may_need_repush_assistant_text = False self._may_need_repush_assistant_text = False
# Report the end of the assistant response. # Report the end of the assistant response.
@@ -1128,6 +1123,25 @@ class AWSNovaSonicLLMService(LLMService):
# Clear out the buffered assistant text # Clear out the buffered assistant text
self._assistant_text_buffer = "" self._assistant_text_buffer = ""
async def _push_assistant_response_text_frames(self, text: str):
# In a typical "cascade" LLM + TTS setup, LLMTextFrames would not
# proceed beyond the TTS service. Therefore, since a speech-to-speech
# service like Nova Sonic combines both LLM and TTS functionality, you
# would think we wouldn't need to push LLMTextFrames at all. However,
# RTVI relies on LLMTextFrames being pushed to trigger its
# "bot-llm-text" event. So here we push an LLMTextFrame, too, but avoid
# appending it to context to avoid context message duplication.
# Push LLMTextFrame
llm_text_frame = LLMTextFrame(text)
llm_text_frame.append_to_context = False
await self.push_frame(llm_text_frame)
# Push TTSTextFrame
tts_text_frame = TTSTextFrame(text, aggregated_by=AggregationType.SENTENCE)
tts_text_frame.includes_inter_frame_spaces = True
await self.push_frame(tts_text_frame)
# #
# user transcription reporting # user transcription reporting
# #