88 lines
2.3 KiB
YAML
88 lines
2.3 KiB
YAML
# 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: sglang-disagg
|
|
spec:
|
|
services:
|
|
Frontend:
|
|
dynamoNamespace: sglang-disagg
|
|
componentType: frontend
|
|
replicas: 1
|
|
extraPodSpec:
|
|
mainContainer:
|
|
image: my-registry/sglang-runtime:my-tag
|
|
decode:
|
|
envFromSecret: hf-token-secret
|
|
dynamoNamespace: sglang-disagg
|
|
componentType: worker
|
|
subComponentType: decode
|
|
replicas: 1
|
|
resources:
|
|
limits:
|
|
gpu: "1"
|
|
extraPodSpec:
|
|
mainContainer:
|
|
image: my-registry/sglang-runtime:my-tag
|
|
workingDir: /workspace/examples/backends/sglang
|
|
command:
|
|
- python3
|
|
- -m
|
|
- dynamo.sglang
|
|
args:
|
|
- --model-path
|
|
- Qwen/Qwen3-0.6B
|
|
- --served-model-name
|
|
- Qwen/Qwen3-0.6B
|
|
- --page-size
|
|
- "16"
|
|
- --tp
|
|
- "1"
|
|
- --trust-remote-code
|
|
- --skip-tokenizer-init
|
|
- --disaggregation-mode
|
|
- decode
|
|
- --disaggregation-transfer-backend
|
|
- nixl
|
|
- --disaggregation-bootstrap-port
|
|
- "12345"
|
|
- --host
|
|
- "0.0.0.0"
|
|
prefill:
|
|
envFromSecret: hf-token-secret
|
|
dynamoNamespace: sglang-disagg
|
|
componentType: worker
|
|
subComponentType: prefill
|
|
replicas: 1
|
|
resources:
|
|
limits:
|
|
gpu: "1"
|
|
extraPodSpec:
|
|
mainContainer:
|
|
image: my-registry/sglang-runtime:my-tag
|
|
workingDir: /workspace/examples/backends/sglang
|
|
command:
|
|
- python3
|
|
- -m
|
|
- dynamo.sglang
|
|
args:
|
|
- --model-path
|
|
- Qwen/Qwen3-0.6B
|
|
- --served-model-name
|
|
- Qwen/Qwen3-0.6B
|
|
- --page-size
|
|
- "16"
|
|
- --tp
|
|
- "1"
|
|
- --trust-remote-code
|
|
- --skip-tokenizer-init
|
|
- --disaggregation-mode
|
|
- prefill
|
|
- --disaggregation-transfer-backend
|
|
- nixl
|
|
- --disaggregation-bootstrap-port
|
|
- "12345"
|
|
- --host
|
|
- "0.0.0.0" |