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

Autopep linter fixes

Browse Source
This commit is contained in:
Moishe Lettvin
2024-01-25 12:12:46 -05:00
parent f0d9b0613e
commit 5fdda43bed
27 changed files with 147 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
src/dailyai/services/ai_services.py
View File

@@ -65,7 +65,7 @@ class AIService:
raise e
@abstractmethod
async def process_frame(self, frame:QueueFrame) -> AsyncGenerator[QueueFrame, None]:
async def process_frame(self, frame: QueueFrame) -> AsyncGenerator[QueueFrame, None]:
if isinstance(frame, ControlQueueFrame):
yield frame
@@ -75,6 +75,7 @@ class AIService:
if False:
yield QueueFrame()
class LLMService(AIService):
@abstractmethod
async def run_llm_async(self, messages) -> AsyncGenerator[str, None]:
@@ -146,7 +147,7 @@ class ImageGenService(AIService):
# Renders the image. Returns an Image object.
@abstractmethod
async def run_image_gen(self, sentence:str) -> tuple[str, bytes]:
async def run_image_gen(self, sentence: str) -> tuple[str, bytes]:
pass
async def process_frame(self, frame: QueueFrame) -> AsyncGenerator[QueueFrame, None]:
@@ -157,15 +158,16 @@ class ImageGenService(AIService):
(url, image_data) = await self.run_image_gen(frame.text)
yield ImageQueueFrame(url, image_data)
class STTService(AIService):
"""STTService is a base class for speech-to-text services."""
_frame_rate: int
def __init__(self, frame_rate: int = 16000, **kwargs):
super().__init__(**kwargs)
self._frame_rate = frame_rate
@abstractmethod
async def run_stt(self, audio: BinaryIO) -> str:
"""Returns transcript as a string"""
@@ -188,6 +190,7 @@ class STTService(AIService):
text = await self.run_stt(content)
yield TextQueueFrame(text)
@dataclass
class AIServiceConfig:
tts: TTSService

36
src/dailyai/services/azure_ai_services.py
View File

@@ -15,6 +15,7 @@ from PIL import Image
# See .env.example for Azure configuration needed
from azure.cognitiveservices.speech import SpeechSynthesizer, SpeechConfig, ResultReason, CancellationReason
class AzureTTSService(TTSService):
def __init__(self, speech_key=None, speech_region=None):
super().__init__()
@@ -23,18 +24,19 @@ class AzureTTSService(TTSService):
speech_region = speech_region or os.getenv("AZURE_SPEECH_SERVICE_REGION")
self.speech_config = SpeechConfig(subscription=speech_key, region=speech_region)
self.speech_synthesizer = SpeechSynthesizer(speech_config=self.speech_config, audio_config=None)
self.speech_synthesizer = SpeechSynthesizer(
speech_config=self.speech_config, audio_config=None)
async def run_tts(self, sentence) -> AsyncGenerator[bytes, None]:
self.logger.info("Running azure tts")
ssml = "<speak version='1.0' xml:lang='en-US' xmlns='http://www.w3.org/2001/10/synthesis' " \
"xmlns:mstts='http://www.w3.org/2001/mstts'>" \
"<voice name='en-US-SaraNeural'>" \
"<mstts:silence type='Sentenceboundary' value='20ms' />" \
"<mstts:express-as style='lyrical' styledegree='2' role='SeniorFemale'>" \
"<prosody rate='1.05'>" \
f"{sentence}" \
"</prosody></mstts:express-as></voice></speak> "
"xmlns:mstts='http://www.w3.org/2001/mstts'>" \
"<voice name='en-US-SaraNeural'>" \
"<mstts:silence type='Sentenceboundary' value='20ms' />" \
"<mstts:express-as style='lyrical' styledegree='2' role='SeniorFemale'>" \
"<prosody rate='1.05'>" \
f"{sentence}" \
"</prosody></mstts:express-as></voice></speak> "
result = await asyncio.to_thread(self.speech_synthesizer.speak_ssml, (ssml))
self.logger.info("Got azure tts result")
if result.reason == ResultReason.SynthesizingAudioCompleted:
@@ -47,6 +49,7 @@ class AzureTTSService(TTSService):
if cancellation_details.reason == CancellationReason.Error:
self.logger.info("Error details: {}".format(cancellation_details.error_details))
class AzureLLMService(LLMService):
def __init__(self, api_key=None, azure_endpoint=None, api_version=None, model=None):
super().__init__()
@@ -54,11 +57,13 @@ class AzureLLMService(LLMService):
azure_endpoint = azure_endpoint or os.getenv("AZURE_CHATGPT_ENDPOINT")
if not azure_endpoint:
raise Exception("No azure endpoint specified for Azure LLM, please set AZURE_CHATGPT_ENDPOINT in the environment or pass it to the AzureLLMService constructor")
raise Exception(
"No azure endpoint specified for Azure LLM, please set AZURE_CHATGPT_ENDPOINT in the environment or pass it to the AzureLLMService constructor")
model: str | None = model or os.getenv("AZURE_CHATGPT_DEPLOYMENT_ID")
if not model:
raise Exception("No model specified for Azure LLM, please set AZURE_CHATGPT_DEPLOYMENT_ID in the environment or pass it to the AzureLLMService constructor")
raise Exception(
"No model specified for Azure LLM, please set AZURE_CHATGPT_DEPLOYMENT_ID in the environment or pass it to the AzureLLMService constructor")
self.model: str = model
api_version = api_version or "2023-12-01-preview"
@@ -90,9 +95,16 @@ class AzureLLMService(LLMService):
else:
return None
class AzureImageGenServiceREST(ImageGenService):
def __init__(self, image_size:str, api_key=None, azure_endpoint=None, api_version=None, model=None):
def __init__(
self,
image_size: str,
api_key=None,
azure_endpoint=None,
api_version=None,
model=None):
super().__init__(image_size=image_size)
self.api_key = api_key or os.getenv("AZURE_DALLE_KEY")
self.azure_endpoint = azure_endpoint or os.getenv("AZURE_DALLE_ENDPOINT")
@@ -103,7 +115,7 @@ class AzureImageGenServiceREST(ImageGenService):
# TODO hoist the session to app-level
async with aiohttp.ClientSession() as session:
url = f"{self.azure_endpoint}openai/images/generations:submit?api-version={self.api_version}"
headers= { "api-key": self.api_key, "Content-Type": "application/json" }
headers = {"api-key": self.api_key, "Content-Type": "application/json"}
body = {
# Enter your prompt text here
"prompt": sentence,

15
src/dailyai/services/daily_transport_service.py
View File

@@ -30,6 +30,7 @@ from daily import (
VirtualSpeakerDevice,
)
class DailyTransportService(EventHandler):
_daily_initialized = False
_lock = threading.Lock()
@@ -111,7 +112,8 @@ class DailyTransportService(EventHandler):
if self.loop:
asyncio.run_coroutine_threadsafe(handler(*args, **kwargs), self.loop)
else:
raise Exception("No event loop to run coroutine. In order to use async event handlers, you must run the DailyTransportService in an asyncio event loop.")
raise Exception(
"No event loop to run coroutine. In order to use async event handlers, you must run the DailyTransportService in an asyncio event loop.")
else:
handler(*args, **kwargs)
except Exception as e:
@@ -126,8 +128,11 @@ class DailyTransportService(EventHandler):
if event_name not in [method[0] for method in methods]:
raise Exception(f"Event handler {event_name} not found")
if not event_name in self.event_handlers:
self.event_handlers[event_name] = [getattr(self, event_name), types.MethodType(handler, self)]
if event_name not in self.event_handlers:
self.event_handlers[event_name] = [
getattr(
self, event_name), types.MethodType(
handler, self)]
setattr(self, event_name, partial(self.patch_method, event_name))
else:
self.event_handlers[event_name].append(types.MethodType(handler, self))
@@ -317,7 +322,7 @@ class DailyTransportService(EventHandler):
def on_app_message(self, message, sender):
pass
def on_transcription_message(self, message:dict):
def on_transcription_message(self, message: dict):
if self.loop:
participantId = ""
if "participantId" in message:
@@ -360,7 +365,7 @@ class DailyTransportService(EventHandler):
frames_or_frame: QueueFrame | list[QueueFrame] = self.threadsafe_send_queue.get()
if isinstance(frames_or_frame, QueueFrame):
frames: list[QueueFrame] = [frames_or_frame]
elif isinstance(frames_or_frame, list):
elif isinstance(frames_or_frame, list):
frames: list[QueueFrame] = frames_or_frame
else:
raise Exception("Unknown type in output queue")

7
src/dailyai/services/deepgram_ai_services.py
View File

@@ -7,13 +7,14 @@ import requests
from collections.abc import AsyncGenerator
from dailyai.services.ai_services import TTSService
class DeepgramTTSService(TTSService):
def __init__(self, speech_key=None, voice=None):
super().__init__()
self.voice = voice or os.getenv("DEEPGRAM_VOICE") or "alpha-asteria-en-v2"
self.speech_key = speech_key or os.getenv("DEEPGRAM_API_KEY")
def get_mic_sample_rate(self):
return 24000
@@ -22,8 +23,8 @@ class DeepgramTTSService(TTSService):
base_url = "https://api.beta.deepgram.com/v1/speak"
request_url = f"{base_url}?model={self.voice}&encoding=linear16&container=none&sample_rate=16000"
headers = {"authorization": f"token {self.speech_key}"}
body = { "text": sentence }
body = {"text": sentence}
async with aiohttp.ClientSession() as session:
async with session.post(request_url, headers=headers, json=body) as r:
async for data in r.content:
yield data
yield data

6
src/dailyai/services/fal_ai_services.py
View File

@@ -8,6 +8,8 @@ from PIL import Image
from dailyai.services.ai_services import LLMService, TTSService, ImageGenService
# Fal expects FAL_KEY_ID and FAL_KEY_SECRET to be set in the env
class FalImageGenService(ImageGenService):
def __init__(self, image_size):
super().__init__(image_size)
@@ -18,9 +20,9 @@ class FalImageGenService(ImageGenService):
handler = fal.apps.submit(
"110602490-fast-sdxl",
arguments={
"prompt": sentence
"prompt": sentence
},
)
)
print("past fal handler init, about to wait for iter_events...")
for event in handler.iter_events():
if isinstance(event, fal.apps.InProgress):

3
src/dailyai/services/open_ai_services.py
View File

@@ -49,8 +49,9 @@ class OpenAILLMService(LLMService):
else:
return None
class OpenAIImageGenService(ImageGenService):
def __init__(self, image_size:str, api_key=None, model=None):
def __init__(self, image_size: str, api_key=None, model=None):
super().__init__(image_size=image_size)
api_key = api_key or os.getenv("OPEN_AI_KEY")
self.model = model or os.getenv("OPEN_AI_IMAGE_MODEL") or "dall-e-3"

14
src/dailyai/services/to_be_updated/cloudflare_ai_service.py
View File

@@ -4,6 +4,8 @@ from services.ai_service import AIService
# Note that Cloudflare's AI workers are still in beta.
# https://developers.cloudflare.com/workers-ai/
class CloudflareAIService(AIService):
def __init__(self):
super().__init__()
@@ -19,11 +21,11 @@ class CloudflareAIService(AIService):
return response.json()
# https://developers.cloudflare.com/workers-ai/models/llm/
def run_llm(self, messages, latest_user_message=None, stream = True):
def run_llm(self, messages, latest_user_message=None, stream=True):
input = {
"messages": [
{ "role": "system", "content": "You are a friendly assistant" },
{ "role": "user", "content": sentence }
{"role": "system", "content": "You are a friendly assistant"},
{"role": "user", "content": sentence}
]
}
@@ -57,9 +59,9 @@ class CloudflareAIService(AIService):
# https://developers.cloudflare.com/workers-ai/models/embedding/
def run_embeddings(self, texts, size="medium"):
models = {
"small": "@cf/baai/bge-small-en-v1.5", # 384 output dimensions
"medium": "@cf/baai/bge-base-en-v1.5", # 768 output dimensions
"large": "@cf/baai/bge-large-en-v1.5" #1024 output dimensions
"small": "@cf/baai/bge-small-en-v1.5", # 384 output dimensions
"medium": "@cf/baai/bge-base-en-v1.5", # 768 output dimensions
"large": "@cf/baai/bge-large-en-v1.5" # 1024 output dimensions
}
return self.run(models[size], {"text": texts})

3
src/dailyai/services/to_be_updated/deepgram_ai_service.py
View File

@@ -17,7 +17,8 @@ class DeepgramAIService(AIService):
def run_tts(self, sentence):
self.logger.info(f"Running deepgram tts for {sentence}")
base_url = "https://api.beta.deepgram.com/v1/speak"
voice = os.getenv("DEEPGRAM_VOICE") or "alpha-apollo-en-v1" # move this to an environment variable
# move this to an environment variable
voice = os.getenv("DEEPGRAM_VOICE") or "alpha-apollo-en-v1"
request_url = f"{base_url}?model={voice}&encoding=linear16&container=none"
headers = {"authorization": f"token {self.api_key}"}

16
src/dailyai/services/to_be_updated/google_ai_service.py
View File

@@ -2,9 +2,12 @@ from services.ai_service import AIService
import openai
import os
# To use Google Cloud's AI products, you'll need to install Google Cloud CLI and enable the TTS and in your project: https://cloud.google.com/sdk/docs/install
# To use Google Cloud's AI products, you'll need to install Google Cloud
# CLI and enable the TTS and in your project:
# https://cloud.google.com/sdk/docs/install
from google.cloud import texttospeech
class GoogleAIService(AIService):
def __init__(self):
super().__init__()
@@ -15,11 +18,14 @@ class GoogleAIService(AIService):
)
self.audio_config = texttospeech.AudioConfig(
audio_encoding = texttospeech.AudioEncoding.LINEAR16,
sample_rate_hertz = 16000
audio_encoding=texttospeech.AudioEncoding.LINEAR16,
sample_rate_hertz=16000
)
def run_tts(self, sentence):
synthesis_input = texttospeech.SynthesisInput(text = sentence.strip())
result = self.client.synthesize_speech(input=synthesis_input, voice=self.voice, audio_config=self.audio_config)
synthesis_input = texttospeech.SynthesisInput(text=sentence.strip())
result = self.client.synthesize_speech(
input=synthesis_input,
voice=self.voice,
audio_config=self.audio_config)
return result

14
src/dailyai/services/to_be_updated/huggingface_ai_service.py
View File

@@ -1,7 +1,12 @@
from services.ai_service import AIService
from transformers import pipeline
# These functions are just intended for testing, not production use. If you'd like to use HuggingFace, you should use your own models, or do some research into the specific models that will work best for your use case.
# These functions are just intended for testing, not production use. If
# you'd like to use HuggingFace, you should use your own models, or do
# some research into the specific models that will work best for your use
# case.
class HuggingFaceAIService(AIService):
def __init__(self):
super().__init__()
@@ -10,9 +15,12 @@ class HuggingFaceAIService(AIService):
classifier = pipeline("sentiment-analysis")
return classifier(sentence)
# available models at https://huggingface.co/Helsinki-NLP (**not all models use 2-character language codes**)
# available models at https://huggingface.co/Helsinki-NLP (**not all
# models use 2-character language codes**)
def run_text_translation(self, sentence, source_language, target_language):
translator = pipeline(f"translation", model=f"Helsinki-NLP/opus-mt-{source_language}-{target_language}")
translator = pipeline(
f"translation",
model=f"Helsinki-NLP/opus-mt-{source_language}-{target_language}")
return translator(sentence)[0]["translation_text"]

6
src/dailyai/services/to_be_updated/mock_ai_service.py
View File

@@ -4,6 +4,7 @@ import time
from PIL import Image
from services.ai_service import AIService
class MockAIService(AIService):
def __init__(self):
super().__init__()
@@ -20,8 +21,7 @@ class MockAIService(AIService):
time.sleep(1)
return (image_url, image)
def run_llm(self, messages, latest_user_message=None, stream = True):
def run_llm(self, messages, latest_user_message=None, stream=True):
for i in range(5):
time.sleep(1)
yield({"choices": [{"delta": {"content": f"hello {i}!"}}]})
yield ({"choices": [{"delta": {"content": f"hello {i}!"}}]})

11
src/dailyai/services/to_be_updated/playht_ai_service.py
View File

@@ -8,6 +8,7 @@ from pyht.protos.api_pb2 import Format
from services.ai_service import AIService
class PlayHTAIService(AIService):
def __init__(self, **kwargs):
super().__init__(**kwargs)
@@ -23,8 +24,7 @@ class PlayHTAIService(AIService):
voice="s3://voice-cloning-zero-shot/820da3d2-3a3b-42e7-844d-e68db835a206/sarah/manifest.json",
sample_rate=16000,
quality="higher",
format=Format.FORMAT_WAV
)
format=Format.FORMAT_WAV)
def close(self):
super().close()
@@ -43,14 +43,15 @@ class PlayHTAIService(AIService):
fh = io.BytesIO(b)
fh.seek(36)
(data, size) = struct.unpack('<4sI', fh.read(8))
self.logger.info(f"first attempt: data: {data}, size: {hex(size)}, position: {fh.tell()}")
self.logger.info(
f"first attempt: data: {data}, size: {hex(size)}, position: {fh.tell()}")
while data != b'data':
fh.read(size)
(data, size) = struct.unpack('<4sI', fh.read(8))
self.logger.info(f"subsequent data: {data}, size: {hex(size)}, position: {fh.tell()}, data != data: {data != b'data'}")
self.logger.info(
f"subsequent data: {data}, size: {hex(size)}, position: {fh.tell()}, data != data: {data != b'data'}")
self.logger.info("position: ", fh.tell())
in_header = False
else:
if len(chunk):
yield chunk