Here is the most fresh dataset used for clustering in TOM.It contains time series for 1659 GitHub repos of the metrics.