That error means the backend never received enough logged data to build the experiment. The Console experiment flow expects model inputs, outputs, and predictions to already be logged against the selected dataset. The usual cause is calling create_experiment() before log_model_outputs() finishes, or logging to a different dataset than the one selected in the Console. Fix: confirm the correct dataset is selected in the Console, then re-run log_model_outputs() with your inputs, outputs, and predictions, and only after that retry creating the experiment. If the dataset itself shows as empty in the Console, the upload failed rather than the experiment step.

Context: When I try to create an experiment in the Galileo Console I get "Unable to create experiment. Check that all required data is available." I logged my model outputs, or at least I think I did. What data is the experiment actually missing?

## Matched source
Source: Source: https://docs.galileo.ai/references/faqs/errors#galileo-console-errors
Original query: "Galileo: Unable to create experiment. Check that all required data is available."
Key terms: available, check, create, data, experiment, galileo, required, that, unable