Data with ground truth for validation of spike sorting algorithms. Please see klusta-team/hybrid_analysis - the data is too large to be uploaded onto github.