Intelligent Media Services:Access TTS models

Python

The following sample code shows how to customize a TTS server:

from aiohttp import web


async def stream_audio(request):
    data = await request.json()
    text = data.get('Text', "")
    token = data.get('Token', None)
    sample_rate = data.get('SampleRate', 48000)
    extend_data = data.get('ExtendData', "")
    print(f"text:{text}, token:{token}, sample_rate:{sample_rate}, extend_data:{extend_data}")
    # Check whether the token is valid.

    response = web.StreamResponse(
        status=200,
        reason='OK',
        headers={'Content-Type': 'audio/mpeg'}
    )

    # Start the response.
    await response.prepare(request)

    # generate_tts_data is a coroutine that is used to generate audio data.
    async for chunk in generate_tts_data(text, sample_rate):
        await response.write(chunk)

    # Return the response.
    await response.write_eof()

    return response


async def generate_tts_data(text: str, sample_rate: int):
    # Call the TTS service to generate the audio data of the corresponding sample rate.
    # The following sample code shows how to read audio data from the file.
    file_path = '/your_dir/sample.pcm'
    with open(file_path, 'rb') as f:
        while True:
            chunk = f.read(4096) # Read 4 KB of data each time.
            if not chunk:
                break
            yield chunk

app = web.Application()
app.add_routes([web.post('/stream-audio', stream_audio)])

if __name__ == '__main__':
    web.run_app(app)