Neural data-to-text generation: A comparison between pipeline and end-to-end architectures

Open paper

Benchmark	Model	Rank	Results
data-to-text-generation-on-webnlg	E2E GRU	#12	BLEU: 57.20
data-to-text-generation-on-webnlg-full	Transformer (Pipeline)	#7	BLEU: 51.68