"vllm/vscode:/vscode.git/clone" did not exist on "d160ae2696a825132adb911ee8604c2d84338357"
agg_router.yaml 1.18 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
# SPDX-License-Identifier: Apache-2.0

apiVersion: nvidia.com/v1alpha1
kind: DynamoGraphDeployment
metadata:
  name: trtllm-agg-router
spec:
  services:
    Frontend:
      dynamoNamespace: trtllm-agg-router
12
      componentType: frontend
13
14
15
      replicas: 1
      extraPodSpec:
        mainContainer:
16
          image: my-registry/tensorrtllm-runtime:my-tag
17
18
19
      envs:
        - name: DYN_ROUTER_MODE
          value: kv
20
21
22
23
24
25
26
27
28
29
    TRTLLMWorker:
      envFromSecret: hf-token-secret
      dynamoNamespace: trtllm-agg-router
      componentType: worker
      replicas: 2
      resources:
        limits:
          gpu: "1"
      extraPodSpec:
        mainContainer:
30
          image: my-registry/tensorrtllm-runtime:my-tag
31
          workingDir: /workspace/
julienmancuso's avatar
julienmancuso committed
32
          command:
33
34
35
          - python3
          - -m
          - dynamo.trtllm
36
          args:
37
38
39
40
41
            - --model-path
            - Qwen/Qwen3-0.6B
            - --served-model-name
            - Qwen/Qwen3-0.6B
            - --extra-engine-args
42
            - ./examples/backends/trtllm/engine_configs/qwen3/agg.yaml
43
            - --publish-events-and-metrics