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

0.5s VAD is interesting

Browse Source
This commit is contained in:
Chad Bailey
2024-02-22 16:14:36 -06:00
parent 18c2b37358
commit 5100a7599b
3 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/dailyai/services/base_transport_service.py
View File

@@ -94,7 +94,7 @@ class BaseTransportService():
self._speaker_sample_rate = kwargs.get("speaker_sample_rate") or 16000 self._speaker_sample_rate = kwargs.get("speaker_sample_rate") or 16000
self._fps = kwargs.get("fps") or 8 self._fps = kwargs.get("fps") or 8
self._vad_start_s = kwargs.get("vad_start_s") or 0.2 self._vad_start_s = kwargs.get("vad_start_s") or 0.2
self._vad_stop_s = kwargs.get("vad_stop_s") or 1.2 self._vad_stop_s = kwargs.get("vad_stop_s") or 0.5
self._context = kwargs.get("context") or [] self._context = kwargs.get("context") or []
self._vad_samples = 1536 self._vad_samples = 1536

2
src/dailyai/services/open_ai_services.py
View File

@@ -10,7 +10,7 @@ from dailyai.services.ai_services import LLMService, ImageGenService
class OpenAILLMService(LLMService): class OpenAILLMService(LLMService):
def __init__(self, *, api_key, model="gpt-4", context): def __init__(self, *, api_key, model="gpt-4-turbo-preview", context):
super().__init__(context) super().__init__(context)
self._model = model self._model = model
self._client = AsyncOpenAI(api_key=api_key) self._client = AsyncOpenAI(api_key=api_key)

2
src/examples/foundational/06c-listen-respond-interruptible-refactor.py
View File

@@ -43,7 +43,7 @@ async def main(room_url: str, token):
# context=context) # context=context)
llm = OpenAILLMService( llm = OpenAILLMService(
context=context, api_key=os.getenv("OPENAI_CHATGPT_API_KEY")) context=context, api_key=os.getenv("OPENAI_CHATGPT_API_KEY"))
llm = GroqLLMService(api_key=os.getenv("GROQ_API_KEY"), context=context) # llm = GroqLLMService(api_key=os.getenv("GROQ_API_KEY"), context=context)
tts = AzureTTSService( tts = AzureTTSService(
api_key=os.getenv("AZURE_SPEECH_API_KEY"), api_key=os.getenv("AZURE_SPEECH_API_KEY"),
region=os.getenv("AZURE_SPEECH_REGION")) region=os.getenv("AZURE_SPEECH_REGION"))