added cambai tts integration
This commit is contained in:
Neil Ruaro
207
examples/foundational/07za-interruptible-camb.py
Normal file
207
examples/foundational/07za-interruptible-camb.py
Normal file
@@ -0,0 +1,207 @@
|
||||
#
|
||||
# Copyright (c) 2024–2025, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
|
||||
"""Camb.ai MARS-8 TTS example with interruption handling.
|
||||
|
||||
This example demonstrates:
|
||||
- Basic TTS synthesis with Camb.ai MARS-8
|
||||
- Voice selection
|
||||
- Speed control
|
||||
- Handling interruptions
|
||||
|
||||
Requirements:
|
||||
- CAMB_API_KEY environment variable
|
||||
- OPENAI_API_KEY environment variable (for LLM)
|
||||
- DEEPGRAM_API_KEY environment variable (for STT)
|
||||
|
||||
Usage:
|
||||
export CAMB_API_KEY=your_camb_api_key
|
||||
export OPENAI_API_KEY=your_openai_api_key
|
||||
export DEEPGRAM_API_KEY=your_deepgram_api_key
|
||||
python 07za-interruptible-camb.py --transport daily
|
||||
|
||||
For more information:
|
||||
- Camb.ai API docs: https://camb.mintlify.app/
|
||||
- Pipecat docs: https://docs.pipecat.ai/
|
||||
"""
|
||||
|
||||
import os
|
||||
|
||||
import aiohttp
|
||||
from dotenv import load_dotenv
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.audio.vad.silero import SileroVADAnalyzer
|
||||
from pipecat.audio.vad.vad_analyzer import VADParams
|
||||
from pipecat.frames.frames import LLMRunFrame
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineParams, PipelineTask
|
||||
from pipecat.processors.aggregators.llm_context import LLMContext
|
||||
from pipecat.processors.aggregators.llm_response_universal import (
|
||||
LLMContextAggregatorPair,
|
||||
)
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport
|
||||
from pipecat.services.camb.tts import CambTTSService
|
||||
from pipecat.services.deepgram.stt import DeepgramSTTService
|
||||
from pipecat.services.openai.llm import OpenAILLMService
|
||||
from pipecat.transports.base_transport import BaseTransport, TransportParams
|
||||
from pipecat.transports.daily.transport import DailyParams
|
||||
from pipecat.transports.websocket.fastapi import FastAPIWebsocketParams
|
||||
|
||||
load_dotenv(override=True)
|
||||
|
||||
|
||||
# Transport configuration for different platforms
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=0.2)),
|
||||
),
|
||||
"twilio": lambda: FastAPIWebsocketParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=0.2)),
|
||||
),
|
||||
"webrtc": lambda: TransportParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=0.2)),
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
|
||||
"""Run the bot with Camb.ai TTS.
|
||||
|
||||
Args:
|
||||
transport: The transport to use for audio I/O.
|
||||
runner_args: Runner arguments from the CLI.
|
||||
"""
|
||||
logger.info("Starting Camb.ai TTS bot")
|
||||
|
||||
# Create an HTTP session for the TTS service
|
||||
async with aiohttp.ClientSession() as session:
|
||||
# Initialize Deepgram STT for speech recognition
|
||||
stt = DeepgramSTTService(api_key=os.getenv("DEEPGRAM_API_KEY"))
|
||||
|
||||
# Initialize Camb.ai TTS with MARS-8-flash model (fastest)
|
||||
tts = CambTTSService(
|
||||
api_key=os.getenv("CAMB_API_KEY"),
|
||||
aiohttp_session=session,
|
||||
voice_id=2681, # Attic voice (default)
|
||||
model="mars-8-flash", # Fast inference model
|
||||
params=CambTTSService.InputParams(
|
||||
speed=1.0, # Normal speed (0.5-2.0 range)
|
||||
),
|
||||
)
|
||||
|
||||
# Initialize OpenAI LLM
|
||||
llm = OpenAILLMService(api_key=os.getenv("OPENAI_API_KEY"))
|
||||
|
||||
# System prompt for the assistant
|
||||
messages = [
|
||||
{
|
||||
"role": "system",
|
||||
"content": """You are a helpful voice assistant powered by Camb.ai's MARS-8
|
||||
text-to-speech technology. Your goal is to have natural conversations and demonstrate
|
||||
high-quality speech synthesis. Keep your responses concise and conversational since
|
||||
they will be spoken aloud. Avoid special characters, emojis, or bullet points that
|
||||
can't easily be spoken.""",
|
||||
},
|
||||
]
|
||||
|
||||
# Set up context management
|
||||
context = LLMContext(messages)
|
||||
context_aggregator = LLMContextAggregatorPair(context)
|
||||
|
||||
# Build the pipeline
|
||||
pipeline = Pipeline(
|
||||
[
|
||||
transport.input(), # Transport user input
|
||||
stt, # Speech-to-text
|
||||
context_aggregator.user(), # User context aggregation
|
||||
llm, # Language model
|
||||
tts, # Camb.ai TTS
|
||||
transport.output(), # Transport bot output
|
||||
context_aggregator.assistant(), # Assistant context aggregation
|
||||
]
|
||||
)
|
||||
|
||||
# Create the pipeline task
|
||||
task = PipelineTask(
|
||||
pipeline,
|
||||
params=PipelineParams(
|
||||
enable_metrics=True,
|
||||
enable_usage_metrics=True,
|
||||
),
|
||||
idle_timeout_secs=runner_args.pipeline_idle_timeout_secs,
|
||||
)
|
||||
|
||||
@transport.event_handler("on_client_connected")
|
||||
async def on_client_connected(transport, client):
|
||||
logger.info("Client connected")
|
||||
# Start the conversation with a greeting
|
||||
messages.append(
|
||||
{
|
||||
"role": "system",
|
||||
"content": "Please introduce yourself briefly and ask how you can help.",
|
||||
}
|
||||
)
|
||||
await task.queue_frames([LLMRunFrame()])
|
||||
|
||||
@transport.event_handler("on_client_disconnected")
|
||||
async def on_client_disconnected(transport, client):
|
||||
logger.info("Client disconnected")
|
||||
await task.cancel()
|
||||
|
||||
# Run the pipeline
|
||||
runner = PipelineRunner(handle_sigint=runner_args.handle_sigint)
|
||||
await runner.run(task)
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud.
|
||||
|
||||
Args:
|
||||
runner_args: Arguments passed from the runner.
|
||||
"""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport, runner_args)
|
||||
|
||||
|
||||
async def list_available_voices():
|
||||
"""Helper function to list available Camb.ai voices.
|
||||
|
||||
Run this to see what voices are available for your API key.
|
||||
"""
|
||||
async with aiohttp.ClientSession() as session:
|
||||
voices = await CambTTSService.list_voices(
|
||||
api_key=os.getenv("CAMB_API_KEY"),
|
||||
aiohttp_session=session,
|
||||
)
|
||||
print("\nAvailable Camb.ai voices:")
|
||||
print("-" * 50)
|
||||
for voice in voices:
|
||||
print(f" ID: {voice['id']}, Name: {voice['name']}, Gender: {voice['gender']}")
|
||||
print("-" * 50)
|
||||
print(f"Total: {len(voices)} voices\n")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
import sys
|
||||
|
||||
# If --list-voices flag is passed, list voices and exit
|
||||
if "--list-voices" in sys.argv:
|
||||
import asyncio
|
||||
|
||||
asyncio.run(list_available_voices())
|
||||
else:
|
||||
from pipecat.runner.run import main
|
||||
|
||||
main()
|
||||
181
examples/foundational/test_camb_quick.py
Normal file
181
examples/foundational/test_camb_quick.py
Normal file
@@ -0,0 +1,181 @@
|
||||
#!/usr/bin/env python3
|
||||
"""Quick test script to verify Camb.ai TTS integration works.
|
||||
|
||||
Usage:
|
||||
export CAMB_API_KEY=your_api_key
|
||||
python test_camb_quick.py
|
||||
"""
|
||||
|
||||
import asyncio
|
||||
import os
|
||||
import sys
|
||||
|
||||
# Add the src directory to the path so we can import the module
|
||||
sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", "..", "src"))
|
||||
|
||||
import aiohttp
|
||||
from dotenv import load_dotenv
|
||||
|
||||
load_dotenv()
|
||||
|
||||
|
||||
async def test_list_voices():
|
||||
"""Test listing available voices."""
|
||||
from pipecat.services.camb.tts import CambTTSService
|
||||
|
||||
api_key = os.getenv("CAMB_API_KEY")
|
||||
if not api_key:
|
||||
print("ERROR: CAMB_API_KEY environment variable not set!")
|
||||
return False
|
||||
|
||||
print("\n1. Testing list_voices()...")
|
||||
async with aiohttp.ClientSession() as session:
|
||||
try:
|
||||
voices = await CambTTSService.list_voices(
|
||||
api_key=api_key,
|
||||
aiohttp_session=session,
|
||||
)
|
||||
print(f" SUCCESS: Found {len(voices)} voices")
|
||||
if voices:
|
||||
print(f" First voice: ID={voices[0]['id']}, Name={voices[0]['name']}")
|
||||
return True
|
||||
except Exception as e:
|
||||
print(f" FAILED: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
return False
|
||||
|
||||
|
||||
async def test_tts_synthesis():
|
||||
"""Test basic TTS synthesis."""
|
||||
from pipecat.services.camb.tts import CambTTSService
|
||||
from pipecat.frames.frames import TTSAudioRawFrame, TTSStartedFrame, TTSStoppedFrame, ErrorFrame
|
||||
|
||||
api_key = os.getenv("CAMB_API_KEY")
|
||||
if not api_key:
|
||||
print("ERROR: CAMB_API_KEY environment variable not set!")
|
||||
return False
|
||||
|
||||
print("\n2. Testing TTS synthesis...")
|
||||
async with aiohttp.ClientSession() as session:
|
||||
tts = CambTTSService(
|
||||
api_key=api_key,
|
||||
aiohttp_session=session,
|
||||
voice_id=2681, # Attic voice
|
||||
model="mars-8-flash",
|
||||
)
|
||||
|
||||
# Manually set sample rate (normally done by StartFrame)
|
||||
tts._sample_rate = 24000
|
||||
|
||||
text = "Hello! This is a test of the Camb.ai text to speech integration."
|
||||
print(f" Synthesizing: '{text}'")
|
||||
|
||||
audio_bytes = 0
|
||||
frames_received = []
|
||||
|
||||
try:
|
||||
async for frame in tts.run_tts(text):
|
||||
frames_received.append(type(frame).__name__)
|
||||
if isinstance(frame, TTSAudioRawFrame):
|
||||
audio_bytes += len(frame.audio)
|
||||
elif isinstance(frame, ErrorFrame):
|
||||
print(f" FAILED: {frame.error}")
|
||||
return False
|
||||
|
||||
print(f" Frames received: {frames_received}")
|
||||
print(f" Audio bytes received: {audio_bytes}")
|
||||
|
||||
if audio_bytes > 0:
|
||||
print(" SUCCESS: TTS synthesis works!")
|
||||
|
||||
# Optionally save and play audio
|
||||
save_audio = input("\n Save audio to test_output.wav? (y/n): ").strip().lower()
|
||||
if save_audio == 'y':
|
||||
await save_audio_to_file(tts, text)
|
||||
# Try to play the audio
|
||||
play_audio = input(" Play the audio? (y/n): ").strip().lower()
|
||||
if play_audio == 'y':
|
||||
play_wav_file("test_output.wav")
|
||||
|
||||
return True
|
||||
else:
|
||||
print(" FAILED: No audio received")
|
||||
return False
|
||||
|
||||
except Exception as e:
|
||||
print(f" FAILED: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
return False
|
||||
|
||||
|
||||
async def save_audio_to_file(tts, text):
|
||||
"""Save synthesized audio to a WAV file."""
|
||||
import wave
|
||||
from pipecat.frames.frames import TTSAudioRawFrame
|
||||
|
||||
audio_data = bytearray()
|
||||
async for frame in tts.run_tts(text):
|
||||
if isinstance(frame, TTSAudioRawFrame):
|
||||
audio_data.extend(frame.audio)
|
||||
|
||||
if audio_data:
|
||||
with wave.open("test_output.wav", "wb") as wav_file:
|
||||
wav_file.setnchannels(1) # Mono
|
||||
wav_file.setsampwidth(2) # 16-bit
|
||||
wav_file.setframerate(24000) # 24kHz
|
||||
wav_file.writeframes(bytes(audio_data))
|
||||
print(" Saved to test_output.wav")
|
||||
|
||||
|
||||
def play_wav_file(filepath):
|
||||
"""Play a WAV file using the system's default player."""
|
||||
import subprocess
|
||||
import platform
|
||||
|
||||
system = platform.system()
|
||||
try:
|
||||
if system == "Darwin": # macOS
|
||||
subprocess.run(["afplay", filepath], check=True)
|
||||
elif system == "Linux":
|
||||
subprocess.run(["aplay", filepath], check=True)
|
||||
elif system == "Windows":
|
||||
subprocess.run(["powershell", "-c", f"(New-Object Media.SoundPlayer '{filepath}').PlaySync()"], check=True)
|
||||
else:
|
||||
print(f" Unsupported platform: {system}. Please play {filepath} manually.")
|
||||
except Exception as e:
|
||||
print(f" Could not play audio: {e}")
|
||||
|
||||
|
||||
async def main():
|
||||
print("=" * 50)
|
||||
print("Camb.ai TTS Integration Test")
|
||||
print("=" * 50)
|
||||
|
||||
results = []
|
||||
|
||||
# Test 1: List voices
|
||||
results.append(await test_list_voices())
|
||||
|
||||
# Test 2: TTS synthesis
|
||||
results.append(await test_tts_synthesis())
|
||||
|
||||
# Summary
|
||||
print("\n" + "=" * 50)
|
||||
print("Summary:")
|
||||
print(f" List voices: {'PASS' if results[0] else 'FAIL'}")
|
||||
print(f" TTS synthesis: {'PASS' if results[1] else 'FAIL'}")
|
||||
print("=" * 50)
|
||||
|
||||
if all(results):
|
||||
print("\nAll tests passed!")
|
||||
return 0
|
||||
else:
|
||||
print("\nSome tests failed!")
|
||||
return 1
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
exit_code = asyncio.run(main())
|
||||
sys.exit(exit_code)
|
||||
Reference in New Issue
Block a user