What it does
A Cursor rules file that guides how the assistant sets up AutoML and hyperparameter searches. It starts from a simple baseline and a fixed metric, keeps validation free of leakage, uses time-aware splits for time series, and records seeds, splits and search spaces for each run.
Use cases
- 01Set a baseline model before starting a parameter search
- 02Split time series data without shuffling across time
- 03Log datasets, seeds and search spaces for every run