Few-NERD
Discover Few-NERD, a detailed dataset for named entity recognition featuring 8 broad categories and 66 detailed entity types. This valuable resource supports supervised and few-shot learning with three benchmark tasks, encompassing 188,200 sentences and around 500,000 entities. Easy BERT integration facilitates advanced training, and regular updates ensure relevance for researchers addressing complex natural language processing problems.
