Dataset validation

Schema and content checks Luna Studio runs on every dataset, and how to fix common errors.

Every dataset added to Luna Studio — uploaded, fetched, generated, or imported — goes through validation. Validation determines whether the dataset can be used in a run.

What Luna Studio checks

Schema checks

• The file parses as CSV or JSONL.
• Required columns are present:
  • Test sets — metric-specific feature columns and label.
  • Training sets — metric-specific feature columns and optionally label when applicable.
• Column types match the metric’s output type (e.g. labels are parseable as Boolean for a Boolean metric).

Content checks

• File encoding (UTF-8 expected).
• Row count > 0.
• Empty rows or empty cells

File checks (uploads and URLs)

• File size within the upload limit.
• For URLs: the URL is reachable; the response content type is appropriate.

Common errors

Error message	Cause	Fix
Missing column "label" in uploaded CSV.	Test set or labelled training set without a label column.	Add the column to your file and re-upload. For unlabelled training logs, use the label-only flow from Step 3.
Missing column "input" in uploaded CSV.	Required input column missing.	Rename your input column to input, or include one.
Could not fetch URL.	The URL was unreachable, returned a non-2xx status, or required auth.	Check the URL is correct; for cloud URLs, confirm the relevant integration has access.
Unsupported file type.	The file isn’t .csv or .jsonl.	Convert the file to one of the supported formats.
Empty dataset.	The file parsed but had zero data rows.	Confirm the file isn’t header-only.
Label values don't match metric output type.	Labels can’t be parsed as the metric’s expected type.	Re-check labels (e.g. for Boolean: only true/false or 1/0).

Where to go next

Add a dataset

Walk through the three sources.

Troubleshooting

Run-time failures and how to recover.