pandas value_counts tri descendant

df1 = df.groupby('Category')['Name'].value_counts().sort_values(as‌​cending=False)
Vivacious Vulture