Full reference

Training workers and runtime peers

Provision a worker with an explicit model and runtime revision, or register a reachable existing peer. Verify readiness before routing paid work to it.

Updated

Reviewed against Console v0.2.2. Available actions depend on project access and runtime configuration; check your installed CLI version for command differences.

Pair an existing machine or provision a worker

For a machine you own, start with Pair local compute. Settings → Organisation → Compute shows pairing, online state, model compute, and separate sharing controls. GPU provisioning is a different operation that can incur rental charges.

A device record, an online peer, a ready model worker, and a verified inference request are separate milestones. Check each rather than assuming a registered peer can serve every model.

Provision a configured worker

bash
stimulir compute worker <offer-id> \
  --region <offered-region> \
  --ssh-key <registered-key-name> \
  --model-repo <model-repository> \
  --model-slug <runtime-model-slug> \
  --hybrie-ref <reviewed-runtime-revision>

The command provisions, bootstraps and registers a worker. Record the instance and job identifiers, inspect progress, and check the actual model and runtime revision before submitting training.

Understand peer registration

The compute peer registry stores runtime service endpoints. Registration does not prove every model is loaded or every operation is supported. An allocated GPU instance is not automatically a ready inference peer.

Register an existing endpoint

bash
stimulir compute peers add \
  --name <peer-name> \
  --grpc <reachable-grpc-endpoint>

At least one of --grpc or --realtime is required. Use an endpoint reachable from the runtime and keep access controls on the underlying service. Do not disable probing merely to make an unreachable node appear configured.

Inspect or remove peers

bash
stimulir compute peers list --json

Use stimulir compute peers remove <peer-id> to remove registration when appropriate. Removal is separate from terminating the cloud instance that hosts the peer.

Verify artifacts and capacity

Ensure model files, checkpoints and output paths are available to the worker, not merely to your laptop. Confirm memory capacity and supported training family. Inspect job failures before retrying to avoid duplicate provisioning.

Choose the right lifecycle

Use GPU instances for raw capacity, this page for runtime readiness, and Training jobs for execution state. Release capacity after completion and verify that termination succeeded.

Command catalogue

Expand a command for its arguments. These commands follow the released CLI source; check stimulir --version and the command’s --help for your installation. Required flags are marked; some operations also validate combinations of inputs.

stimulir compute peers add

Register a runtime peer (at least one of --grpc / --realtime is required).

--name
Friendly peer name.
--grpc
Peer gRPC URL.
--realtime
Peer realtime URL.
--instance-id
Compute instance id backing the peer.
--no-probe
Register without probing reachability.
--endpoint
HybrIE runtime URL (bypasses the console proxy).
--json
Emit raw JSON output.
stimulir compute peers list

List registered runtime peers.

--endpoint
HybrIE runtime URL (bypasses the console proxy).
--json
Emit raw JSON output.
stimulir compute peers remove

Remove a runtime peer.

peer_id required
Positional argument. Peer id (from stimulir compute peers list).
--yes, -y
Skip the confirmation prompt.
--endpoint
HybrIE runtime URL (bypasses the console proxy).
--json
Emit raw JSON output.
stimulir compute worker

Provision, bootstrap and register an exact managed training worker.

offer_id required
Positional argument. Offer id from stimulir compute offers.
--region required
Region with live capacity.
--ssh-key required
Provider SSH key name (repeatable).
--count
See the installed command’s --help for accepted values.
--provider
See the installed command’s --help for accepted values.
--spot
See the installed command’s --help for accepted values.
--model-repo required
Exact Hugging Face base-model repository.
--model-slug required
HybrIE on-disk family slug.
--hybrie-ref required
Reviewed HybrIE branch or commit to build.
--endpoint
HybrIE runtime URL (bypasses the console proxy).
--json
See the installed command’s --help for accepted values.