vllm-helper shard stop
Stop sharded runtimes on cluster nodes
Usage
vllm-helper shard stop --nodes n1,n2 [flags]
Options
--backend
| Type | string |
| Default | all |
backend to stop: all, ray, or mp
--dry-run
| Type | bool |
print commands instead of running them
--head
| Type | string |
head node name (defaults to first --nodes entry)
--nodes
| Type | string |
comma-separated node names
--service-name
| Type | string |
| Default | vllm.service |
service to start with --start-services
--ssh-key
| Type | string |
SSH private key for non-head nodes
--start-services
| Type | bool |
start the normal vLLM service after stopping sharded runtimes