tts_translator.py

# Transcribes user speech to text, and saves it to a file
from pathlib import Path
from dotenv import load_dotenv
from livekit.agents import JobContext, WorkerOptions, cli
from livekit.agents.voice import Agent, AgentSession
from livekit.plugins import rime, elevenlabs, silero
import sys

sys.path.append(str(Path(__file__).parent.parent))
from launch_demos.livekit_plugins_gladia import stt

load_dotenv(dotenv_path=Path(__file__).parent.parent / '.env')

async def entrypoint(ctx: JobContext):
    
    await ctx.connect()
    session = AgentSession()
    
    # Process transcription events - let the agent say what it receives
    @session.on("user_input_transcribed")
    def on_transcript(event):
        # Log the full event object to see all available metadata
        print(f"Transcript event: {event}")
        if event.is_final:
            print(f"Final transcript: {event.transcript}")
            session.say(event.transcript)
    
    await session.start(
        agent=Agent(
            instructions="You are a helpful assistant that speaks what the user says in English.",
            stt=stt.STT(
                languages=["fr", "en"],  # Support French and English input
                code_switching=True,
                sample_rate=16000,
                bit_depth=16,
                channels=1,
                encoding="wav/pcm",
                translation_enabled=True,
                translation_target_languages=["en"],  # Only translate to English
                translation_model="base",
                translation_match_original_utterances=True
            ),
            tts=elevenlabs.TTS(
                model="eleven_multilingual_v2"
            ),
            allow_interruptions=False
        ),
        room=ctx.room
    )

if __name__ == "__main__":
    cli.run_app(WorkerOptions(entrypoint_fnc=entrypoint))