How to enable GPU in localai for KUBERNETES CLUSTER (V1.24)?

[harshsing2891]

LocalAI version:
latest

Environment, CPU architecture, OS, and Version:
kubernetes cluster v1.24

Describe the bug
Not able to GPU for localai using helm chart to deploy in kubernetes cluster

To Reproduce

Expected behavior

Logs

Additional context
I am using helm template of localai

[harshsing2891]

I am using below enviornment varilable in values.yaml of localai helm chart

replicaCount: 1
deployment:
image: quay.io/go-skynet/local-ai:latest
env:
f16: true
gpu_layers: 4
context_size: 1024

[harshsing2891]

I am using text generative feature of localai in kubernetes environment. and I have deployed localai using helm template available in localai repository. Problem is I am getting very delayed response from localai even for very basic question i am asking like "what is ai ?".
I have tried several models like various gguf based models but always getting delayed response.