playground.yaml

apiVersion: llmaz.io/v1alpha1
kind: OpenModel
metadata:
  name: qwen2-7b
spec:
  familyName: qwen2
  source:
    # You should replace this with your own oss address following the protocol:
    # oss://<bucket>.<endpoint>/<path-to-your-model>
    uri: oss://llmaz.oss-ap-southeast-1-internal.aliyuncs.com/models/Qwen2-7B
  inferenceConfig:
    flavors:
      - name: t4 # GPU type
        limits:
          nvidia.com/gpu: 1
---
apiVersion: inference.llmaz.io/v1alpha1
kind: Playground
metadata:
  name: qwen2-7b
spec:
  replicas: 1
  modelClaim:
    modelName: qwen2-7b