had to get a bunch of names for this to work since I didn't want to generate these, I wanted these to be the input for the generation of the other attributes. For the names I wanted a diverse mix of countries of origin. Initial google results were mostly from US statistics but I soon found this stackexchange comment and thus used this dataset

While looking through this dataset I found that apart from country and popularity statistics it also had information regarding the "possible" gender of the name which I could also use as part of the input when generating the attributes. Genders were defined from M (male), 1M (male if first part of then ame else mostly male), ?m (mostly male), F (female), 1F (see 1M), ?F (mostly female), ? (unisex)