R-Transformer: Recurrent Neural Network Enhanced Transformer

Benchmark Model Rank Results
language-modelling-on-penn-treebank-character-levelR-Transformer#14Bit per Character (BPC): 1.24
sequential-image-classification-on-sequential-mnistR-Transformer#30Unpermuted Accuracy: 99.1%
language-modelling-on-penn-treebank-word-levelR-Transformer#39Test perplexity: 84.38