An helm chart for the VLLM project using official images based on the [official documetnation](https://docs.vllm.ai/en/latest/serving/deploying_with_k8s.html}
| Prefix | Option | Value | Description |
|---|---|---|---|
| general | nodeSelector | YAML | YAML object to put under node selector |
| general | tolerations | YAML | YAML object to put under tolerations |
| general | affinity | YAML | YAML object to put under affinity |
| Prefix | Option | Value | Description |
|---|---|---|---|
| vllmConfig | vllmModel | String | The name of the model to use |
| vllmConfig | vllmIsTrustRemoteCode | bool | Enable/Disable trust in external source |
| vllmConfig | vllmIsEnableChunkPrefill | bool | Enable/Disable chunckPrefill |
| vllmConfig | vllmMaxNumBatchedTokens | Int | Number of Batched tokens |
| Prefix | Option | Value | Description |
|---|---|---|---|
| vllmConfig.vllmResources | vllmResourcesGpuEnabled | bool | Enable/Disable GPU |
| vllmConfig.vllmResources | vllmResourcesGpuResourcesNvidiaResourceName | String | The name of the nvidia config option |
| Prefix | Option | Value | Description |
|---|---|---|---|
| vllmConfig.vllmResources.vllmResourcesRequest | vllmResourcesRequestCpu | Mi/int | CPU requests |
| vllmConfig.vllmResources.vllmResourcesRequest | vllmResourcesRequestMem | Size | Memory requests |
| vllmConfig.vllmResources.vllmResourcesRequest | vllmResourcesRequestGpu | int | GPU resources |
| Prefix | Option | Value | Description |
|---|---|---|---|
| vllmConfig.vllmResources.vllmResourcesLimit | vllmResourcesLimitCpu | Mi/int | CPU requests |
| vllmConfig.vllmResources.vllmResourcesLimit | vllmResourcesLimitMem | Size | Memory requests |
| vllmConfig.vllmResources.vllmResourcesLimit | vllmResourcesLimitGpu | int | GPU resources |
| Prefix | Option | Value | Description |
|---|---|---|---|
| hf | hfSecretName | String | The pregenerated secret name for the Hugging face token |
| Prefix | Option | Value | Description |
|---|---|---|---|
| pod | replicaCount | int | The number of replica for the pods |
| pod | imagePullSecrets | List | List of the secrets to use to pull the image |
| pod | nameOverride | String | Override of the deployent name |
| pod | fullnameOverride | String | Override of the deployent name |
| pod | podAnnotations | String | Custom pod annotation |
| pod.podLabels | YAML | Custom labels |
| Prefix | Option | Value | Description |
|---|---|---|---|
| pod.image | repository | String | The image repository source |
| pod.image | pullPolicy | Always/IfNotPresent/Never | When to pull the image |
| pod.image | tag | String | The image tag(default: null) |
| Prefix | Option | Value | Description |
|---|---|---|---|
| pod.security | securityPodSecurityContext | YAML | Custom security context |
| pod.security | securityContext | YAML | Custom security context |
| Prefix | Option | Value | Description |
|---|---|---|---|
| pod.autoscaling | enabled | bool | Enable/Disable HPA(default: false) |
| pod.autoscaling | minReplicas | Int | Minimum number of replicas for autoscaling |
| pod.autoscaling | maxReplicas | Int | Maximum number of replicas for autoscaling |
| pod.autoscaling | targetCPUUtilizationPercentage | String | The maximum CPU utlization to denote scale |
| pod.autoscaling | targetMemoryUtilizationPercentage | String | The maximum memory pressure to denote scale |
| Prefix | Option | Value | Description |
|---|---|---|---|
| storage | volumesAdditional | List | List of additional volumes |
| storage | volumeMounts | List | List of additional mounts |
| Prefix | Option | Value | Description |
|---|---|---|---|
| storage.volumesCache | volumeCacheName | String | The shared cache volume name |
| storage.volumesCache | volumeCacheSize | Size | The Size of the volume for the shared cache volume |
| Prefix | Option | Value | Description |
|---|---|---|---|
| storage.volumesShm | volumeShmSize | Size | The Size of the volume for the Shared memory volume |
| Prefix | Option | Value | Description |
|---|---|---|---|
| serviceAccount | create | bool | Enable/disable the creation of Service Account |
| serviceAccount | automount | bool | Enable/disable automounting the service account |
| serviceAccount | annotations | YAML | Custom annotation object |
| serviceAccount | name | String | Customize the name |
| Prefix | Option | Value | Description |
|---|---|---|---|
| network.service | type | ClusterIP/NodePort/LoadBalancer/ExternalName | K8S service type |
| network.service | port | Int | Internal port of the container(default: 8080) |
| Prefix | Option | Value | Description |
|---|---|---|---|
| network.ingress | enabled | bool | Enable/Disable ingress |
| network.ingress | className | String | Ingress Class name |
| network.ingress | annotations | YAML | Custom ingress annotations |
| network.ingress | hosts | List | List of ingress hosts |
| network.ingress | tls | List | List of secrets to use for certificate |
| Prefix | Option | Value | Description |
|---|---|---|---|
| health | livenessProbe | YAML | YAML object of the liveliness probe check |
| health | readinessProbe | YAML | YAML object of the readiness probe check |