NAOMI/values_example.yaml at main · copandrej/NAOMI · GitHub

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
# For all values refer to 'helm_charts/SEMR/values.yaml'
# Change <CHANGE-ME> to IP of machine running the k8s cluster
general:
  ip: "<CHANGE-ME>"

# Ray
# Ray can't be disabled currently
rayVersion: "2.54.0"
headGroupSpecs:
  rayHead:
    resources:
      limits:
        cpu: "2"
        memory: "6G"
      requests:
        cpu: "2"
        memory: "6G"

# At least two ray workers required for distributed training
workerGroups:
  amdGroup:
    replicas: 1
    minReplicas: 1
    maxReplicas: 1
    rayStartParams:
      numCpus: "4"
    container:
      resources:
        limits:
          cpu: "4"
          memory: "6G"
        requests:
          cpu: "4"
          memory: "6G"

  # For Raspbery Pis:

  # aarchGroup:
  #   replicas: 1
  #   minReplicas: 1
  #   maxReplicas: 1
  #   rayStartParams:
  #     numCpus: "1"
  #   container:
  #     resources:
  #       limits:
  #         cpu: "1"
  #         memory: "5G"
  #       requests:
  #         cpu: "1"
  #         memory: "5G"

# MLflow
mlflow:
  enabled: true

# MinIO
# MinIO is required by Flyte and Mlflow
minio:
  enabled: true
  # changing credentials will require fixes in other components and workflow examples!
  auth:
    rootUser: "minio"
    rootPassword: "miniostorage"

# Flyte
flyte-binary:
  enabled: true
  configuration:
    storage:
      providerConfig:
        s3:
          endpoint: "http://<CHANGE-ME>:30085"
    inline:
      plugins:
        k8s:
          default-env-vars:
            - FLYTE_AWS_ENDPOINT: "http://<CHANGE-ME>:30085"
            - FLYTE_AWS_ACCESS_KEY_ID: "minio"
            - FLYTE_AWS_SECRET_ACCESS_KEY: "miniostorage"
      task_resources: #adjust this for max flyte task pod resources (this is ceiling for task res. that you specify in example workflows)
        defaults:
          cpu: 2000m
          memory: 4Gi
        limits:
          memory: 4Gi


# NAOMI MCP Server
# When enabled, you can connect AI coding assistants (VS Code Copilot, Claude Code, Cursor, etc.)
# to the cluster via the MCP endpoint: http://<node_ip>:31008/mcp
# MCP Server has generous premisions and can damage the cluster, enable at your own risk
naomiMcp:
  enabled: false

# promethes&grafana
kube-prometheus-stack:
  enabled: true
  grafana:
    grafana.ini:
      auth.anonymous:
        enabled: true