-
Notifications
You must be signed in to change notification settings - Fork 63
Expand file tree
/
Copy pathoutput-pd.yaml
More file actions
215 lines (193 loc) · 5.58 KB
/
Copy pathoutput-pd.yaml
File metadata and controls
215 lines (193 loc) · 5.58 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
# generated by generate-example-output.sh
---
# Source: llm-d-modelservice/templates/serviceaccount.yaml
apiVersion: v1
kind: ServiceAccount
metadata:
name: pd-llm-d-modelservice
labels:
helm.sh/chart: llm-d-modelservice-v0.4.15
app.kubernetes.io/version: "v0.4.0"
app.kubernetes.io/managed-by: Helm
---
# Source: llm-d-modelservice/templates/decode-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: pd-llm-d-modelservice-decode
labels:
helm.sh/chart: llm-d-modelservice-v0.4.15
app.kubernetes.io/version: "v0.4.0"
app.kubernetes.io/managed-by: Helm
spec:
replicas: 1
selector:
matchLabels:
llm-d.ai/inference-serving: "true"
llm-d.ai/model: facebook-opt-125m
llm-d.ai/role: decode
template:
metadata:
labels:
llm-d.ai/inference-serving: "true"
llm-d.ai/model: facebook-opt-125m
llm-d.ai/role: decode
spec:
initContainers:
- name: routing-proxy
args:
- --port=8000
- --vllm-port=8200
- --connector=nixlv2
- --zap-encoder=json
- --zap-log-level=debug
- --secure-proxy=false
image: ghcr.io/llm-d/llm-d-routing-sidecar:latest
imagePullPolicy: Always
ports:
- containerPort: 8000
resources: {}
restartPolicy: Always
securityContext:
allowPrivilegeEscalation: false
runAsNonRoot: true
serviceAccountName: pd-llm-d-modelservice
volumes:
- emptyDir: {}
name: metrics-volume
- name: model-storage
emptyDir:
sizeLimit: 20Gi
containers:
- name: vllm
image: ghcr.io/llm-d/llm-d-cuda:latest
command: ["vllm", "serve"]
args:
- "facebook/opt-125m"
- --port
- "8200"
- --served-model-name
- "facebook/opt-125m"
- --enforce-eager
- --kv-transfer-config
- '{"kv_connector":"NixlConnector", "kv_role":"kv_both"}'
env:
- name: CUDA_VISIBLE_DEVICES
value: "0"
- name: UCX_TLS
value: cuda_ipc,cuda_copy,tcp
- name: VLLM_NIXL_SIDE_CHANNEL_HOST
valueFrom:
fieldRef:
fieldPath: status.podIP
- name: VLLM_NIXL_SIDE_CHANNEL_PORT
value: "5600"
- name: VLLM_LOGGING_LEVEL
value: DEBUG
- name: DP_SIZE
value: "1"
- name: TP_SIZE
value: "1"
- name: DP_SIZE_LOCAL
value: "1"
- name: HF_HOME
value: /model-cache
ports:
- containerPort: 8200
protocol: TCP
- containerPort: 5600
protocol: TCP
resources:
limits:
cpu: "16"
memory: 16Gi
nvidia.com/gpu: "1"
requests:
cpu: "16"
memory: 16Gi
nvidia.com/gpu: "1"
volumeMounts:
- name: model-storage
mountPath: /model-cache
---
# Source: llm-d-modelservice/templates/prefill-deployment.yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: pd-llm-d-modelservice-prefill
labels:
helm.sh/chart: llm-d-modelservice-v0.4.15
app.kubernetes.io/version: "v0.4.0"
app.kubernetes.io/managed-by: Helm
spec:
replicas: 1
selector:
matchLabels:
llm-d.ai/inference-serving: "true"
llm-d.ai/model: facebook-opt-125m
llm-d.ai/role: prefill
template:
metadata:
labels:
llm-d.ai/inference-serving: "true"
llm-d.ai/model: facebook-opt-125m
llm-d.ai/role: prefill
spec:
serviceAccountName: pd-llm-d-modelservice
volumes:
- emptyDir: {}
name: metrics-volume
- name: model-storage
emptyDir:
sizeLimit: 20Gi
containers:
- name: vllm
image: ghcr.io/llm-d/llm-d-cuda:latest
command: ["vllm", "serve"]
args:
- "facebook/opt-125m"
- --port
- "8000"
- --served-model-name
- "facebook/opt-125m"
- --enforce-eager
- --kv-transfer-config
- '{"kv_connector":"NixlConnector", "kv_role":"kv_both"}'
env:
- name: CUDA_VISIBLE_DEVICES
value: "0"
- name: UCX_TLS
value: cuda_ipc,cuda_copy,tcp
- name: VLLM_NIXL_SIDE_CHANNEL_PORT
value: "5600"
- name: VLLM_NIXL_SIDE_CHANNEL_HOST
valueFrom:
fieldRef:
fieldPath: status.podIP
- name: VLLM_LOGGING_LEVEL
value: DEBUG
- name: DP_SIZE
value: "1"
- name: TP_SIZE
value: "1"
- name: DP_SIZE_LOCAL
value: "1"
- name: HF_HOME
value: /model-cache
ports:
- containerPort: 8000
protocol: TCP
- containerPort: 5600
protocol: TCP
resources:
limits:
cpu: "16"
memory: 16Gi
nvidia.com/gpu: "1"
requests:
cpu: "16"
memory: 16Gi
nvidia.com/gpu: "1"
volumeMounts:
- name: model-storage
mountPath: /model-cache