k8s/roles/vllm/templates/vllm-values.yml.j2

17 lines
454 B
Django/Jinja

replicaCount: {{ vllm_num_replicas }}
image:
repository: {{ vllm_image.repository }}
tag: {{ vllm_image.tag }}
pullPolicy: {{ vllm_image.pull_policy }}
resources:
limits:
nvidia.com/gpu: {{ vllm_resources.limits.nvidia.com/gpu | default('1') }}
requests:
nvidia.com/gpu: {{ vllm_resources.requests.nvidia.com/gpu | default('1') }}
env: {{ vllm_env | default([]) | to_json }}
extraArgs: {{ vllm_extra_args | default([]) | to_json }}