Skip to content

Runs and artefacts

Each model run writes artefacts under STORAGE_ROOT/runs/{run_id}/... and exposes them via the API.

Typical cards:

  • Artefacts: browse generated files (plots, CSVs, JSON, models).
  • Logs: stdout/stderr captured during execution.
  • Plots (.png, .jpg, .jpeg) can be viewed inline and downloaded.
  • CSV tables can be opened as a preview table and downloaded.
  • Everything else (models, JSON) is offered as a raw view and/or download.
  • List artefacts: GET /v2/runs/{run_id}/artifacts
  • Download/view artefact bytes: GET /v2/runs/{run_id}/artifacts/{path}
  • Tail logs: GET /v2/runs/{run_id}/logs/tail?stream=stdout|stderr&offset=...