Hierarchical Self-Attention Hybrid Sparse Networks for Document Classification

Table 7

Runtime comparison.

IMDb				Yelp
Model	Runtime	Model	Runtime	Model	Runtime	Model	Runtime (s)
GRU	154	LSTM	255	GRU	5472	LSTM	6377
GRU1	117	LSTM1	171	GRU1	3873	LSTM1	5413
GRU2	112	LSTM2	151	GRU2	3720	LSTM2	4945
GRU3	101	LSTM3	126	GRU3	3382	LSTM3	4632