How can we build a riva model from the Tao Toolkit's .binary file using Kubernetes?

AI & Data Science Deep Learning (Training & Inference) Riva
1

I have the .binary file generated after running the Tao experiments. Now I want to create the Riva model from this file in the Kubernetes environment. So, I wanted to know how we can leverage the riva-build functionality in the Kubernetes environment using the riva-servicemaker container image.

Hardware - GPU - T4
Hardware - CPU - 8 vCPUs → n1-highmem-8 in GCP
Operating System - Container OS
Riva Version - 2.2.1

2

Hi @jimit-modi

Thanks for your interest in Riva,

I will check regarding this request further with the team and provide updates,

Thanks for your patience

3

Hi @jimit-modi

Apologies for the delay

Regarding

  1. tao n-gram tutorial - Please find the below doc on how to deploy it
    Pipeline Configuration — NVIDIA Riva

  2. Please find the below notebook which helps outline servicemaker usage for asr pipelines
    https://catalog.ngc.nvidia.com/orgs/nvidia/teams/tao/resources/speechtotext_notebook/version/v1.3/files/speech-to-text-deployment.ipynb

Please let us know if you need more information

Thanks for patience