# Default values for vllm-app.
# This is a YAML-formatted file.
# Declare variables to be passed into your templates.

# 模型配置
model:
  huggingfaceName: "Qwen/Qwen2.5-0.5B-Instruct"  # 用户只需输入这个
  localMountPath: "/Model"                  # PVC 固定挂载路径
  huggingfaceToken: "<your-hf-token>"
  download:
    enabled: false                                  # 启用自动下载
    image: "docker.io/vllm/vllm-openai:latest"  # 包含 huggingface-cli 的镜像

# 功能选择
app: "vllm"

resources:
  gpuLimit: 1
  cpuRequest: 12
  memoryLimit: "16Gi"
  shmSize: "20Gi"

# vLLM 应用配置
vllm:
  image: "docker.io/vllm/vllm-openai:latest"
    #gpuLimit: 2 
    # cpuRequest: 12
    # memoryLimit: "12Gi"
    # shmSize: "15Gi"

llama:
  image: "docker.io/library/one-click:v1"

# lmdeploy  应用配置
lmdeploy:
  image: "docker.io/openmmlab/lmdeploy:latest-cu12"
    #  gpuLimit: 2
    #  cpuRequest: 12
    #  memoryLimit: "12Gi"
    #  shmSize: "15Gi"

# NFS PV/PVC 配置
nfs:
  server: "10.6.80.11"
  path: "/volume1/Dataset/PVStore/lab-data-model-pvc-c0beeab1-6dd5-4c6a-bd2c-6ce9e114c25e/Weight"
  storageClass: "local-path"
  pvSize: "500Gi"
  pvcSize: "50Gi"

# LeaderWorkerSet 配置
replicaCount: 1
workerSize: 2

nodeSelector: {}

tolerations: []

affinity: {}