Assorted, Archetypal, and Annotated Two Million (3A2M) Cooking Recipe Dataset

The 3A2M dataset is a large collection of culinary recipes labeled in respective categories with extended named entities extracted from recipe descriptions.

BibTex: