search

NVIDIA / nim-deploy

A collection of YAML files, Helm Charts, Operator code, and guides to act as an example reference implementation for NVIDIA NIM deployment.
https://build.nvidia.com/
Apache License 2.0
49 stars 17 forks source link

Tests don't work with minimal configuration helm install due to default model.name in values.yaml #14

Open jmineroff opened 2 weeks ago

jmineroff commented 2 weeks ago

I didn't use a custom values file to do the install, but got stuck trying to figure out why the test command was failing. It turns out it was due to the default model.name not matching the name of the default model (llama3-8b-instruct).

Can you either change the default model.name value or add --set model.name='llama3-8b-instruct' to the minimal command?