truss loops command provides subcommands for the Loops deployment lifecycle: pushing a deployment for a base model, viewing runs and samplers, and listing or deploying checkpoints from a run.
push
Provision a Loops session, run, and paired sampler for a base model. If the project already has an active Loops deployment for the base model, the command fails with a validation error.
Arguments
Hugging Face model ID for the base model (for example,
Qwen/Qwen3-8B).Options
Training project ID to associate the deployment with.
Number of data-parallel trainer replicas to provision. The trainer deployment runs this many copies of the base model’s preset node group (for example,
--replicas 4 on a 4-node preset provisions 16 nodes across 4 data-parallel workers). Must be a positive integer; defaults to 1.Name of the remote in
.trussrc to deploy to.Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
deactivate
Shut down a Loops deployment. Saved checkpoints remain accessible after deactivation.
Arguments
ID of the deployment to shut down. Get the deployment ID from
truss loops view.Options
Name of the remote in
.trussrc.-y, --yes
Skip the confirmation prompt.
Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
view
List your Loops deployments by default. Pass --org to list every deployment in your organization, which adds an Owner column to the output. Deployments in terminal states (STOPPED, FAILED) are hidden unless you pass --all.
Options
Name of the remote in
.trussrc.--all
Include deployments in terminal states (STOPPED, FAILED).
--org
List every Loops deployment in your organization (with its owner), not just your own.
Output format: cli-table (default) or json.
Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
--org, an Owner column is included.
logs
Fetch logs from one half of a Loops deployment. A Loops deployment and its sampler have separate log streams, so pass exactly one of --loops-deployment-id or --sampler-deployment-id depending on which side you’re debugging.
Options
Fetch logs from a Loops deployment. The id is the
Deployment ID column in truss loops view.Fetch logs from the sampler’s inference deployment. The id is the
Sampler Deployment ID column in truss loops samplers view. The companion model id is resolved automatically by matching against the caller’s active Loops deployments.--tail
Continue polling for new log lines until the deployment goes inactive (or Ctrl+C).
Remote to use.
Customizes logging.
--non-interactive
Disables interactive prompts, use in CI / automated execution contexts.
truss loops view (Deployment ID) and truss loops samplers view (Sampler Deployment ID).
runs view
List Loops runs visible to the caller. Both filters are optional and can be combined; omit both to list every run.
Options
Filter to a specific run ID.
Filter runs by base model name.
-r, --reverse
Reverse the default order (oldest first) so the most recent run is shown first.
Name of the remote in
.trussrc.Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
samplers view
List Loops samplers visible to the caller.
Options
-r, --reverse
Reverse the default order (oldest first) so the most recent sampler is shown first.
Name of the remote in
.trussrc.Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
checkpoints view
List checkpoints for a Loops run. Identify the run with --run-id, or pass --base-model to pick the most recent run for that base model. The two filters are mutually exclusive.
Options
Loops run ID to list checkpoints for. Mutually exclusive with
--base-model.Base model name. Resolves to the most recent Loops run for that model. Mutually exclusive with
--run-id.Sort checkpoints by checkpoint ID, creation time, size, or type.
Sort order.
Output format.
Name of the remote in
.trussrc.Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
checkpoints deploy
Deploy checkpoints from a Loops run as a vLLM-backed inference deployment. You must pass one of --run-id, --checkpoint-ids, or --config.
Options
Loops run ID. Opens an interactive picker so you can choose checkpoints from the run. Cannot be combined with
--checkpoint-ids.Comma-separated Loops checkpoint IDs (for example,
vL3pQrS8,wK4tUvW9). Bypasses the interactive picker. Use truss loops checkpoints view to find IDs. Cannot be combined with --run-id or --config.Path to a Python file that defines a
DeployCheckpointsConfig. The config must populate checkpoint_details.loops_checkpoint_ids. Cannot be combined with --checkpoint-ids.--dry-run
Render the generated truss config to stdout without deploying.
Name of the remote in
.trussrc.Logging verbosity.
humanfriendly (default) is pretty-printed; INFO, DEBUG, WARNING produce structured logs.--non-interactive
Disable interactive prompts. Use in CI/automated contexts where stdin isn’t a TTY.
Related
- Loops concepts: How sessions, runs, samplers, and checkpoints fit together.
- Loops supported models: Base models you can pass to
truss loops push. - Training SDK reference:
CheckpointListandDeployCheckpointsConfigPython types used with--config.