| nameOverride: "" | |
| fullnameOverride: "" | |
| h2ogpt: | |
| enabled: true | |
| stack: | |
| # -- Run h2oGPT and vLLM on same pod. | |
| enabled: false | |
| replicaCount: 1 | |
| image: | |
| repository: gcr.io/vorvan/h2oai/h2ogpt-runtime | |
| pullPolicy: IfNotPresent | |
| initImage: | |
| repository: | |
| tag: | |
| pullPolicy: | |
| storage: | |
| size: 128Gi | |
| class: | |
| useEphemeral: true | |
| externalLLM: | |
| enabled: false | |
| secret: | |
| modelLock: | |
| openAIAzure: | |
| enabled: false | |
| openAI: | |
| enabled: False | |
| replicate: | |
| enabled: false | |
| # -- Example configs to use when not using Model Lock and External LLM | |
| # overrideConfig: | |
| # base_model: h2oai/h2ogpt-4096-llama2-7b-chat | |
| # use_safetensors: True | |
| # prompt_type: llama2 | |
| # save_dir: /workspace/save/ | |
| # use_gpu_id: False | |
| # score_model: None | |
| # max_max_new_tokens: 2048 | |
| # max_new_tokens: 1024 | |
| overrideConfig: | |
| service: | |
| type: NodePort | |
| webPort: 80 | |
| gptPort: 8888 | |
| webServiceAnnotations: {} | |
| updateStrategy: | |
| type: RollingUpdate | |
| podSecurityContext: | |
| runAsNonRoot: true | |
| runAsUser: 1000 | |
| runAsGroup: 1000 | |
| fsGroup: 1000 | |
| securityContext: | |
| runAsNonRoot: true | |
| allowPrivilegeEscalation: false | |
| capabilities: | |
| drop: | |
| - ALL | |
| seccompProfile: | |
| type: RuntimeDefault | |
| resources: | |
| nodeSelector: | |
| tolerations: | |
| env: {} | |
| podAnnotations: {} | |
| podLabels: {} | |
| autoscaling: {} | |
| tgi: | |
| enabled: false | |
| replicaCount: 1 | |
| image: | |
| repository: ghcr.io/huggingface/text-generation-inference | |
| tag: 0.9.3 | |
| pullPolicy: IfNotPresent | |
| storage: | |
| size: 512Gi | |
| class: | |
| useEphemeral: true | |
| overrideConfig: | |
| hfSecret: | |
| containerArgs: | |
| service: | |
| type: ClusterIP | |
| port: 8080 | |
| updateStrategy: | |
| type: RollingUpdate | |
| podSecurityContext: | |
| securityContext: | |
| resources: | |
| nodeSelector: | |
| tolerations: | |
| env: {} | |
| podAnnotations: {} | |
| podLabels: {} | |
| autoscaling: {} | |
| vllm: | |
| enabled: false | |
| replicaCount: 1 | |
| image: | |
| repository: gcr.io/vorvan/h2oai/h2ogpt-runtime | |
| pullPolicy: IfNotPresent | |
| imagePullSecrets: | |
| storage: | |
| size: 512Gi | |
| class: | |
| useEphemeral: true | |
| overrideConfig: | |
| containerArgs: | |
| - "--model" | |
| - h2oai/h2ogpt-4096-llama2-7b-chat | |
| - "--tokenizer" | |
| - hf-internal-testing/llama-tokenizer | |
| - "--tensor-parallel-size" | |
| - 2 | |
| - "--seed" | |
| - 1234 | |
| - "--trust-remote-code" | |
| service: | |
| type: ClusterIP | |
| port: 5000 | |
| updateStrategy: | |
| type: RollingUpdate | |
| podSecurityContext: | |
| runAsNonRoot: true | |
| runAsUser: 1000 | |
| runAsGroup: 1000 | |
| fsGroup: 1000 | |
| securityContext: | |
| runAsNonRoot: true | |
| allowPrivilegeEscalation: false | |
| capabilities: | |
| drop: | |
| - ALL | |
| seccompProfile: | |
| env: {} | |
| resources: | |
| nodeSelector: | |
| tolerations: | |
| podAnnotations: {} | |
| podLabels: {} | |
| autoscaling: {} | |