Feature Needed: Inworld Inference Model Options

[tommymsz006]

Your question

I have been using Inworld inference model with LiveKit JS agent, and it works quite well. I wonder if it's possible to make extra model options available. Currently it only supports the basics, namely language, voice and model. I would like to fine-tune speaking rate, sample rate, temperature and text normalization. Per the below, it seems that these are not open:

agents-js/agents/src/inference/tts.ts

Line 62 in efa8a4b

export interface InworldOptions {}

Can we make these options available as well?