wehub-resource-sync
59 lines
2.0 KiB
YAML
59 lines
2.0 KiB
YAML
# This example demonstrates how to schedule TPU pods using `acceleratorConfig`, i.e.
|
|
# a combination of the chip and underlying topology.
|
|
# See https://cloud.google.com/tpu/docs/supported-tpu-configurations for more details.
|
|
# A unique identifier for the head node and workers of this cluster.
|
|
cluster_name: tputopology
|
|
|
|
max_workers: 2
|
|
|
|
available_node_types:
|
|
ray_head_default:
|
|
min_workers: 0
|
|
max_workers: 0
|
|
resources: {"CPU": 0}
|
|
# Provider-specific config for this node type, e.g. instance type. By default
|
|
# Ray will auto-configure unspecified fields such as subnets and ssh-keys.
|
|
# For more documentation on available fields, see:
|
|
# https://cloud.google.com/compute/docs/reference/rest/v1/instances/insert
|
|
node_config:
|
|
machineType: n1-standard-4
|
|
disks:
|
|
- boot: true
|
|
autoDelete: true
|
|
type: PERSISTENT
|
|
initializeParams:
|
|
diskSizeGb: 50
|
|
# See https://cloud.google.com/compute/docs/images for more images
|
|
sourceImage: projects/ubuntu-os-cloud/global/images/family/ubuntu-2004-lts
|
|
ray_tpu:
|
|
min_workers: 1
|
|
max_workers: 1
|
|
resources: {"TPU": 1} # use TPU custom resource in your code
|
|
node_config:
|
|
# Defining the accelerator configuration, including the:
|
|
# - type (in this case a V4) and
|
|
# - topology 2x2x1 will create a v4-8.
|
|
acceleratorConfig:
|
|
type: V4
|
|
topology: 2x2x1
|
|
runtimeVersion: tpu-vm-v4-base
|
|
|
|
provider:
|
|
type: gcp
|
|
region: us-central2
|
|
availability_zone: us-central2-b
|
|
project_id: null # Replace this with your GCP project ID.
|
|
|
|
initialization_commands:
|
|
- sudo apt-get update
|
|
- sudo apt-get install -y python3-pip python-is-python3
|
|
|
|
setup_commands:
|
|
- pip install 'ray[default]'
|
|
|
|
head_setup_commands:
|
|
- pip install google-api-python-client
|
|
|
|
# Specify the node type of the head node (as configured above).
|
|
head_node_type: ray_head_default
|