Unverified Commit cf55e8b8 authored by Hongkuan Zhou's avatar Hongkuan Zhou Committed by GitHub
Browse files

fix: add subcomponenttype to recipes missing this field (#5301)


Signed-off-by: default avatarhongkuanz <hongkuanz@nvidia.com>
parent ec5630ea
......@@ -26,6 +26,7 @@ spec:
decode:
dynamoNamespace: sgl-dsr1-16gpu
componentType: worker
subComponentType: decode
replicas: 1
multinode:
nodeCount: 2
......@@ -71,6 +72,7 @@ spec:
prefill:
dynamoNamespace: sgl-dsr1-16gpu
componentType: worker
subComponentType: prefill
replicas: 1
multinode:
nodeCount: 2
......
......@@ -26,6 +26,7 @@ spec:
decode:
dynamoNamespace: sgl-dsr1-8gpu
componentType: worker
subComponentType: decode
replicas: 1
resources:
limits:
......@@ -68,6 +69,7 @@ spec:
prefill:
dynamoNamespace: sgl-dsr1-8gpu
componentType: worker
subComponentType: prefill
replicas: 1
resources:
limits:
......
......@@ -139,6 +139,7 @@ spec:
prefill:
dynamoNamespace: trtllm-disagg-multinode
componentType: worker
subComponentType: prefill
replicas: 1
# NOTE: Prefill uses 1 node (no multinode section = single node)
# and contributes to ComputeDomain.numNodes (see above)
......@@ -196,6 +197,7 @@ spec:
decode:
dynamoNamespace: trtllm-disagg-multinode
componentType: worker
subComponentType: decode
replicas: 1
volumeMounts:
- name: model-cache
......
......@@ -28,6 +28,7 @@ spec:
decode:
dynamoNamespace: vllm-dsr1
componentType: worker
subComponentType: decode
replicas: 1
multinode:
nodeCount: 2
......@@ -93,6 +94,7 @@ spec:
prefill:
dynamoNamespace: vllm-dsr1
componentType: worker
subComponentType: prefill
replicas: 1
multinode:
nodeCount: 2
......
......@@ -26,6 +26,7 @@ spec:
replicas: 1
VllmPrefillWorker:
componentType: worker
subComponentType: prefill
dynamoNamespace: llama3-70b-disagg-mn
envFromSecret: hf-token-secret
volumeMounts:
......@@ -57,6 +58,7 @@ spec:
gpu: "8"
VllmDecodeWorker:
componentType: worker
subComponentType: decode
dynamoNamespace: llama3-70b-disagg-mn
envFromSecret: hf-token-secret
volumeMounts:
......
......@@ -26,6 +26,7 @@ spec:
replicas: 1
VllmPrefillWorker:
componentType: worker
subComponentType: prefill
dynamoNamespace: llama3-70b-disagg-sn
envFromSecret: hf-token-secret
volumeMounts:
......@@ -69,6 +70,7 @@ spec:
gpu: "2"
VllmDecodeWorker:
componentType: worker
subComponentType: decode
dynamoNamespace: llama3-70b-disagg-sn
envFromSecret: hf-token-secret
volumeMounts:
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment