<@UNZB4NW3S> .Yes ray error when I run in flyte. t...
# ask-the-community
k
@Ketan (kumare3) .Yes ray error when I run in flyte. this is a single node
k
k
I am using @task( task_config=RayJobConfig( head_node_config=HeadNodeConfig(ray_start_params={"block": "True"}), worker_node_config=[ WorkerNodeConfig( group_name="ray-group", replicas=5, min_replicas=2, max_replicas=10 ) ], runtime_env={'env_vars': {'__MODIN_AUTOIMPORT_PANDAS__': '1'},"pip": ["numpy", "modin"]}, ) )
message has been deleted
Is it still resource to be increased?? how should i add it here
d
cc @Kevin Su - ray worker resource configuration
k
you could update the worker’s resource by updating the
limit
and
request
in the
@task
.
Copy code
@task(task_config=RayJobConfig(...), requests=Resources(cpu="10"), limits=Resources(gpu="2"))