When using #tabulareditor and serialized folder structure for your #PowerBI dataset, what is a good strategy to utilize a 'local' environment partition that may use different M than the rest of the environments in pipeline? An example would be using table.fromrows to test locally instead of connecting to DW