The raw cloud benchmark runs the model is calibrated on (one row per run), and the fitted curves and sources behind the predictions.
The raw cloud benchmark runs the model is calibrated on (one row per run), and the fitted curves and sources behind the predictions.