No transcript on live websocket

[Natim]

Hello,

We've been trying to plug Twilio voice API with Gladia for live transcription.
Can you tell us which config we are supposed to use?

We've tried, but we don't receive any transcript.

    {
        "encoding": "wav/ulaw",
        "sample_rate": 8000,
        "bit_depth": 8,
        "channels": 1,
        "language_config": {
            "languages": [],
            "code_switching": true
        }
    }

[Natim]

Refs #65, will try more from there.

[nmorel]

Can you share an id or the ws url you obtain from the POST request ?

[Natim]

Transcription ID: 49c3b9c6-d1f7-439f-a068-1d3d3318209d
WSS: wss://api.gladia.io/v2/live?token=49c3b9c6-d1f7-439f-a068-1d3d3318209d

Transcription ID: fdf0eae1-99ba-4987-a6b2-396af201dd5c
WSS: wss://api.gladia.io/v2/live?token=fdf0eae1-99ba-4987-a6b2-396af201dd5c

[nmorel]

That's what we received for those 2 ids:

You received the transcript "Well" for the 2nd one.
Maybe speak a little more or concat the audio bytes you are sending to check it.

[Natim]

Indeed, the asyncio.sleep(0.1) was the issue.