"tests/entrypoints/openai/models/test_models.py" did not exist on "f256ebe4df6757d76f1f1642d7e110268a2f8190"
agg.yaml 990 Bytes
Newer Older
1
2
# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0
Alec's avatar
Alec committed
3

4
5
6
apiVersion: nvidia.com/v1alpha1
kind: DynamoGraphDeployment
metadata:
7
  name: vllm-agg
8
9
10
spec:
  services:
    Frontend:
11
      dynamoNamespace: vllm-agg
12
      componentType: frontend
13
14
15
      replicas: 1
      extraPodSpec:
        mainContainer:
16
          image: gitlab-master.nvidia.com:5005/dl/ai-dynamo/dynamo/dynamo:vllm-1029
17
18
    VllmDecodeWorker:
      envFromSecret: hf-token-secret
19
      dynamoNamespace: vllm-agg
20
      componentType: worker
21
      replicas: 1
22
23
24
      resources:
        limits:
          gpu: "1"
25
26
      extraPodSpec:
        mainContainer:
27
          image: gitlab-master.nvidia.com:5005/dl/ai-dynamo/dynamo/dynamo:vllm-1029
Alec's avatar
Alec committed
28
          workingDir: /workspace/components/backends/vllm
29
          command:
30
31
32
            - python3
            - -m
            - dynamo.vllm
33
          args:
34
35
            - --model
            - Qwen/Qwen3-0.6B