sf job

Job lifecycle: logs, status, stop, pause/resume, batch operations

Inspect

sf job ls                     # Job list (including submission history)
sf job status <JOB_ID>        # Status details
sf job logs [JOB_ID]          # Follow logs; omit the ID to follow the most recent job
sf job logs <JOB_ID> -n 0     # Replay full history then follow (default replays only the last 2000 lines)
sf job samples <JOB_ID>       # Multi-turn conversation traces from the latest validation
sf job samples <JOB_ID> --vidx 2 -n 10   # 3rd validation round, show 10 samples

logs option	Description
-n, --tail	Replay the last N lines of history then follow; -n 0 for everything (default 2000)

samples option	Description
--vidx	Validation round index (default -1 = most recent)
-n, --limit	Number of samples to show (default 6)

Control

sf job stop <JOB_ID>          # Stop (running → terminated)
sf job pause <JOB_ID>         # Pause: keep the checkpoint, resumable
sf job resume <JOB_ID>        # Automatically resume training from the latest checkpoint
sf job delete <JOB_ID>        # Delete the record of a finished job (running jobs must be stopped first)

pause / resume are maintenance-friendly pause semantics: when an administrator drains the cluster in maintenance mode, jobs go through the same path, and after recovery they resume training automatically without losing progress.

Batch

sf job cancel-all             # Stop all of my running / queued jobs
sf job clean                  # Clean up display records of finished jobs
sf job stop-sweep <SWEEP_ID>  # Stop all active jobs of a sweep with one command

sf job

Manage jobs

sf job <subcommand>

Command	Description
sf job cancel-all	Stop every running or queued job of mine
sf job clean	Clean up the display records of finished jobs
sf job delete	Delete one finished job record
sf job logs	Follow a job’s logs
sf job ls	List jobs (including the submission history)
sf job pause	Pause a job, keeping its checkpoint
sf job resume	Resume a paused job
sf job retry	Retry a failed or stopped job
sf job samples	Show one validation round’s traces
sf job status	Show a job’s status
sf job stop	Stop a job (running -> terminated)
sf job stop-sweep	Stop every active job of a sweep

sf job cancel-all

Stop every running or queued job of mine

sf job cancel-all [options]

Option	Description
--yes, -y	Skip the confirmation

sf job clean

Clean up the display records of finished jobs

sf job clean

No options beyond --help.

sf job delete

Delete one finished job record

sf job delete <JOB_ID>

Argument	Description
JOB_ID	Job id (required)

No options beyond --help.

sf job logs

Follow a job’s logs

sf job logs [JOB_ID] [options]

Argument	Description
JOB_ID	Job id (see sf job ls); omitted = the latest

Option	Description
--tail, -n <int>	Replay only the last N log lines before following (default 2000; -n 0 for everything) (default 2000)

sf job ls

List jobs (including the submission history)

sf job ls [options]

Option	Description
--all	Show everything (default: the latest 15)
--exp	Only this experiment (full name or last segment)
--limit, -n <int>	How many to show (ignored with —all) (default 15)

sf job pause

Pause a job, keeping its checkpoint

sf job pause <JOB_ID>

Argument	Description
JOB_ID	Job id (required)

No options beyond --help.

sf job resume

Resume a paused job

sf job resume <JOB_ID>

Argument	Description
JOB_ID	Job id (required)

No options beyond --help.

sf job retry

Retry a failed or stopped job

sf job retry <JOB_ID>

Argument	Description
JOB_ID	Job id (required)

No options beyond --help.

sf job samples

Show one validation round’s traces

sf job samples <JOB_ID> [options]

Argument	Description
JOB_ID	Job id (see sf job ls) (required)

Option	Description
--vidx <int>	Validation round index (default -1 = the latest) (default -1)
--limit, -n <int>	How many samples to show (default 6)

sf job status

Show a job’s status

sf job status <JOB_ID>

Argument	Description
JOB_ID	Job id (required)

No options beyond --help.

sf job stop

Stop a job (running -> terminated)

sf job stop <JOB_ID>

Argument	Description
JOB_ID	Job id (required)

No options beyond --help.

sf job stop-sweep

Stop every active job of a sweep

sf job stop-sweep <SWEEP_ID> [options]

Argument	Description
SWEEP_ID	The sweep id (printed when sf sweep submits) (required)

Option	Description
--yes, -y	Skip the confirmation