The neural network invents fancy new ingredients

I’m training a neural network to generate recipes based on a database of about 30,000 examples, and although it sometimes comes up with ingredients that sound like a bad idea, sometimes it invents some that you could plausibly ask for at Whole Foods and act all disappointed when they don’t have any:

¼ cup coconut fluff rings
1 teaspoon cooked buster grapes
½ cup shanked whipping peanuts
6 tablespoon lemon turn beans
1 seeds of the chocolate cheese
1 cup milk bat leaves
1 ½ cup sherry stick
¼ line phempherbly ice
½ cup shrimp white pine baking powder
1 teaspoon baking curny sauce
¼ cup milked salt
8 oz canned pomegranate crescents
12 oz can canned and chopped pistachio stock