ASR End of Utterance Detection without Cloud Deployment

What end of utterance detection features are available without using riva-build command.
I justed deployed the Riva Triton Server in docker with the quickstart guide 2.13.0 .

I use the streaming recognition endpoint from my python script and it is working well but I need an appropriate end-of-utterance detection solution.

The Problem: When is the user done talking to some degree of certainty?

I would appreciate some information!

Please provide the following information when requesting support.

Hardware - GPU RTX 4090
Hardware - CPU
Operating System Windows 11
Riva Version 2.13.0
TLT Version (if relevant)
How to reproduce the issue ? (This is for errors. Please share the command and the detailed log here)

I read up more about this and it seems like I need to rebuild the relevant models using the end of utterance specific parameters for prolonging the end of utterance timeframes. For this I seem to need the .riva model files for rebuilding but I cant seem to find them and I dont have .nemo files.


Where could I get the .riva files for rebuilding the rmir files?