Run Experiment
Use this only when the spec or command is sufficiently defined.
Procedure
- Confirm the spec maps to real files, backend package, config, data, and mode.
- Confirm expected metrics and artifacts are supported by the current backend.
- Record the resolved command before execution.
- For risky or long jobs, run the smallest smoke path first.
- Execute the command with bounded logging.
- Capture logs, checkpoints, metrics, predictions, and artifact roots.
- Report status: success, failed, timeout, interrupted, or blocked.
Output
- exact command
- resolved config/backend
- artifact paths
- run status
- failure diagnostics if applicable
- next recommended analysis or debug step
Guardrails
- Do not change source code just to make a run work.
- Do not hide failed or partial runs.
- Do not start large training when data paths, compute budget, or user approval are unclear.