-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtf-example.yaml
99 lines (99 loc) · 2.13 KB
/
tf-example.yaml
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
kind: MLApp
metadata:
name: tfexample
namespace: mapp
labels: # Will be applayed to each resource
testKey: testValue
spec:
tasks:
- name: model
labels:
testModelKey: testModelValue # Will be applayed to each resource
resources:
- name: worker
labels:
testWorkerKey: testWorkerValue # Will be applayed to each resource
replicas: 2
minAvailable: 2
restartPolicy: Never
maxRestartCount: 0
images:
gpu: image-gpu
cpu: image-cpu
command: ['python']
workdir: directory
args: ["--log-dir=$TRAINING_DIR"]
env:
- name: TEST_ENV_V1
value: v1
- name: PYTHONPATH
value: /usr/local
resources:
accelerators:
gpu: 1
dedicated_gpu: true
requests:
cpu: 100mi
memory: 1Gi
limits:
cpu: 100mi
memory: 1Gi
volumes:
- name: lib
- name: training
subPath: build-1
- name: ps
labels:
testPsKey: testPsValue # Will be applayed to each resource
replicas: 1
minAvailable: 1
restartPolicy: Never
maxRestartCount: 0
images:
cpu: image-cpu
command: python
workdir: directory
env:
- name: TEST_ENV_V2
value: v2
volumes:
- name: training
uix:
- name: jupyter
displayName: Jupyter
images:
gpu: image-gpu
cpu: image-cpu
resources:
accelerators:
gpu: 1
requests:
cpu: 100mi
memory: 1Gi
limits:
cpu: 100mi
memory: 1Gi
ports:
- port: 80
targetPort: 8082
protocol: TCP
name: http
volumes:
- name: lib
volumes:
- name: lib
isLibDir: true
isTrainLogDir: false
mountPath: /workspace/lib
subPath: lib
clusterStorage: test
hostPath:
path: /test
- name: training
isLibDir: false
isTrainLogDir: true
mountPath: /workspace/training
subPath: training
clusterStorage: test
hostPath:
path: /test